· COLOPHON Title: A National Guideline for the prevention of pressure ulcers Authors: Dimitri (KULeuven), Geert Vanwalleghem ((CNC vzw), Hilde Heyman (WCS), Christian Thyse ((KCE),

2012

KCE REPORT 193S

A NATIONAL GUIDELINEPRESSURE ULCERSAPPENDIX

A NATIONAL GUIDELINE FOR THE PREVENTION OPRESSURE ULCERS

www.kce.fgov.be

FOR THE PREVENTION OF

2012

KCE REPORT 193SGOOD CLINICAL PRACTICE

A NATIONAL GUIDELINEPRESSURE ULCERSAPPENDIX

DIMITRI BEECKMAN, CATHY MATHEÏ, AURÉLIECHRISTIAN THYSE, ADINDA TOPPETS, SABINE STORDEUR, KOEN VA

A NATIONAL GUIDELINE FOR THE PREVENTION OPRESSURE ULCERS

THY MATHEÏ, AURÉLIE VAN LANCKER, SABINE VAN HOUDT, GEERT VANWALLEGHEM, LUC GRYSONE STORDEUR, KOEN VAN DEN HEEDE

www.kce.fgov.be

FOR THE PREVENTION OF

WALLEGHEM, LUC GRYSON, HILDE HEYMAN,

COLOPHON

Title: A National Guideline for the prevention of pressure ulcers

Authors: Dimitri(KULeuven), Geert Vanwalleghem ((CNC vzw), Hilde Heyman (WCS), Christian Thyse ((KCE), Koen Van den Heede (KCE)

Reviewers: Marijke Eyssen (KCE); Dominique Paulus

External experts: Diégo Backaert (Daniëlle Declercq (Institut Jules Bordet); Véronique del Marmolle (Hopital Erasme, ULB); Aurélia Bustillo (HopitalErasme, ULB); Anne Hermand (CPubliqueSchoonhoven (Looverbosch (Zelfstandig Huisarts); Katrien Vanderwee (Waeyenberghe (Home Health Care wound care BVBA); Christiane Vranken (CHU Liège)

Acknowledgements: We thank Liz Avital (NCGC, UK), Katie Jones (NCGC, UK) and Julie Neilson (NCGC, UK) for the collaboration inthe preparation of the evide

External validators: Nicky Cullum (University of Manchester); Siegfried Geens (CEBAM); Philippe Hanson (CHU UCL Mont

Conflict of interest: Dominique Putzeys declare

Diégo Backaert, Hilde Beele, Anne Hermand, Adinda Toppets, Geert Vanwalleghem, Pascal Van Waeyenberghedeclared to have received a fee to lecture or reimbursement for training, travelling or participation to conferencesrelated to the p

Layout: Ine Verhulst

Disclaimer: The external experts were consulted about a (preliminary) version of the scientific report. Theircomments were discussed during meetings. They did not conecessarily agree with its content.

Subsequently, a (final) version was submitted to the validators. The validation of the report resultsfrom a consensus or a voting process between the validators. The validators did not coscientifi

Finally, this report has been approved

Only the KCE is responsible for errors or omissions that could persist. The policy recommendationsare als

A National Guideline for the prevention of pressure ulcers- Supplement

Dimitri Beeckman (UGent), Cathy Matheï (KULeuven), Aurélie Van Lancker (UGent), Sabine Van Houdt(KULeuven), Geert Vanwalleghem (CNC vzw/ WCS/ H.-Hartziekenhuis Roeselare(CNC vzw), Hilde Heyman (WCS), Christian Thyse (AFISCeP.be), Adinda T(KCE), Koen Van den Heede (KCE)

Marijke Eyssen (KCE); Dominique Paulus (KCE)

Diégo Backaert (Thuiszorg Groep Backaert); Hilde Beele (UZ Gent); Lieven Decaevele (OLViëlle Declercq (Institut Jules Bordet); Véronique del Marmolle (Hopital Erasme, ULB); Aurélia Bustillo (Hopital

Erasme, ULB); Anne Hermand (Clinique Universitaire Saint-Luc, Bruxelles); Miguel Lardennois (SPF SantéPublique - FOD Volksgezondheid); Louis Paquay (Wit-Gele Kruis); Dominique Putzeys (CIPIQSchoonhoven (Radboud Universiteit NijmegenEvelien Touriany (Militair Hospitaal Koningin AstridLooverbosch (Zelfstandig Huisarts); Katrien Vanderwee (O.L.V. van Lourdes ziekenhuis WaWaeyenberghe (Home Health Care wound care BVBA); Christiane Vranken (CHU Liège)

We thank Liz Avital (NCGC, UK), Katie Jones (NCGC, UK) and Julie Neilson (NCGC, UK) for the collaboration inthe preparation of the evidence reports

Nicky Cullum (University of Manchester); Siegfried Geens (CEBAM); Philippe Hanson (CHU UCL Mont

Dominique Putzeys declared to have received funding for research related to the prevention o

Diégo Backaert, Hilde Beele, Anne Hermand, Adinda Toppets, Geert Vanwalleghem, Pascal Van Waeyenberghedeclared to have received a fee to lecture or reimbursement for training, travelling or participation to conferencesrelated to the prevention of pressure ulcers.

Ine Verhulst

The external experts were consulted about a (preliminary) version of the scientific report. Theircomments were discussed during meetings. They did not co-author the scientific report anecessarily agree with its content.

Subsequently, a (final) version was submitted to the validators. The validation of the report resultsfrom a consensus or a voting process between the validators. The validators did not coscientific report and did not necessarily all three agree with its content.

Finally, this report has been approved by common assent by the Executive Board.

Only the KCE is responsible for errors or omissions that could persist. The policy recommendationsare also under the full responsibility of the KCE.

ï (KULeuven), Aurélie Van Lancker (UGent), Sabine Van HoudtHartziekenhuis Roeselare-Menen vzw), Luc Gryson

), Adinda Toppets (UZLeuven), Sabine Stordeur

); Hilde Beele (UZ Gent); Lieven Decaevele (OLV-Ziekenhuis, Aalst);iëlle Declercq (Institut Jules Bordet); Véronique del Marmolle (Hopital Erasme, ULB); Aurélia Bustillo (Hopital

Luc, Bruxelles); Miguel Lardennois (SPF SantéGele Kruis); Dominique Putzeys (CIPIQ-s); Lisette

Militair Hospitaal Koningin Astrid); Dirk Van DeO.L.V. van Lourdes ziekenhuis Waregem); Pascal Van

Waeyenberghe (Home Health Care wound care BVBA); Christiane Vranken (CHU Liège)

We thank Liz Avital (NCGC, UK), Katie Jones (NCGC, UK) and Julie Neilson (NCGC, UK) for the collaboration in

Nicky Cullum (University of Manchester); Siegfried Geens (CEBAM); Philippe Hanson (CHU UCL Mont -Godinne)

to have received funding for research related to the prevention of pressure ulcers.

Diégo Backaert, Hilde Beele, Anne Hermand, Adinda Toppets, Geert Vanwalleghem, Pascal Van Waeyenberghedeclared to have received a fee to lecture or reimbursement for training, travelling or participation to conferences

The external experts were consulted about a (preliminary) version of the scientific report. Theirauthor the scientific report and did not

Subsequently, a (final) version was submitted to the validators. The validation of the report resultsfrom a consensus or a voting process between the validators. The validators did not co -author the

c report and did not necessarily all three agree with its content.

by the Executive Board.

Only the KCE is responsible for errors or omissions that could persist. The policy recommendations

Publication date: 11 January 2013

Domain: Good Clinical Practice (GCP)

MeSH: Pressure ulcer ; Practice Guidelines;

NLM Classification: WR 598

Language: English

Format: Adobe

Legal depot: D/201

Copyright: KCE reports are published under a “by/nc/nd” Creative Commons Licencehttp://kce.fgov.be/content/about

How to refer to this document? BeeckmanToppetsSupplementReport

This document is available on the website of the Belgian Health Care Knowledge Centre

11 January 2013

Good Clinical Practice (GCP)

Pressure ulcer ; Practice Guidelines; Prevention and control

WR 598

English

Adobe® PDF™ (A4)

D/2013/10.273/98

KCE reports are published under a “by/nc/nd” Creative Commons Licencehttp://kce.fgov.be/content/about-copyrights-for-kce-reports.

Beeckman D, Matheï C, Van Lancker A, Van Houdt S, VanwalleghemToppets A, Stordeur S, Van den Heede K. A National Guideline for the prevention of pressure ulcers

plement. Good Clinical Practice (GCP). Brussels: Belgian Health Care Knowledge Centre (KCE). 201Reports 193S. D/2013/10.273/98.

This document is available on the website of the Belgian Health Care Knowledge Centre

KCE reports are published under a “by/nc/nd” Creative Commons Licence

, Vanwalleghem G, Gryson L, Heyman H, Thyse C,A National Guideline for the prevention of pressure ulcers -

. Brussels: Belgian Health Care Knowledge Centre (KCE). 2012. KCE

This document is available on the website of the Belgian Health Care Knowledge Centre .

KCE Report 193S

APPENDIX REPORTTABLE OF CONTENTS TABLE OF CONTENTS

LIST OF ABBREVIATION

1.

1.1.

1.2.

1.3.

2.

2.1.

2.2.

2.3.

2.4.

Pressure Ulcer Prevention – Supplement

TABLE OF CONTENTS ................................................................................................

LIST OF ABBREVIATIONS ................................................................................................

1. RISK ASSESSMENT – CLINICAL EFFECTIVENESS................................

1.1. REVIEW PROTOCOL ................................................................

1.2. SEARCH STRATEGY ................................................................

1.2.1. Search filters ................................................................

1.2.2. Selection of articles ................................................................

1.3. CLINICAL EVIDENCE ................................................................

1.3.1. Summary table ................................................................

1.3.2. Clinical evidence GRADE tables................................................................

1.3.3. Forest plots................................................................................................

1.3.4. Evidence tables................................................................

2. RISK ASSESSMENT – PROGNOSTIC................................ ................................

2.1. REVIEW PROTOCOL ................................................................

2.2. SEARCH STRATEGY ................................................................

2.2.1. Search filters ................................................................


2.2.3. List of excluded studies................................................................

2.3. RISK ASSESSMENT TOOLS ................................................................


2.4.1. Search strategy ................................................................

2.4.2. Clinical evidence ................................................................

2.4.3. Summary table ................................................................

2.4.4. Median Area under the ROC curve................................

2.4.5. AUC within studies ................................................................

2.4.6. Predictive ability ................................................................

2.4.7. Quality of the studies................................................................

1

...................................................................... 1

................................................................ 6

......................................................................... 8

........................................................................................... 8

......................................................................................... 10

......................................................................................... 10

............................................................................... 16

......................................................................................... 17

...................................................................................... 17

........................................................... 18

............................................................ 24

..................................................................................... 27

.............................................................. 36

......................................................................................... 36

......................................................................................... 38

......................................................................................... 38

............................................................................... 48

......................................................................... 49

............................................................................. 51

......................................................................................... 53

..................................................................................... 53

................................................................................... 53

...................................................................................... 54

....................................................................................... 60

................................................................................ 65

.................................................................................... 68

............................................................................. 74

2

2.5.

3.

3.1.

3.2.

3.3.

4.

4.1.

4.2.

4.3.


2.5. INCIDENCE AND PREDICTIVE ABILITY OF RISK ASSESSMENT SCALE

2.5.1. Sensitivity and specificity of risk assessment tools ................................

2.5.2. Forest plots area under the receiver operating characteristics curve (AUC)

2.5.3. Forest plots sensitivity and specificity ................................

2.5.4. Clinical evidence tables................................................................

3. SKIN ASSESSMENT – CLINICAL EFFECTIVENESS................................

3.1. REVIEW PROTOCOL ................................................................

3.2. SEARCH STRATEGY ................................................................

3.2.1. Search filters ................................................................



3.3.1. Search strategy ................................................................


3.3.3. Summary table ................................................................


3.3.5. Forest plots................................................................................................

3.3.6. Evidence tables................................................................

4. SKIN ASSESSMENT – PROGNOSTIC................................ ................................

4.1. REVIEW PROTOCOL ................................................................

4.2. SEARCH STRATEGY ................................................................

4.2.1. Search filters ................................................................


4.2.3. List of excluded studies................................................................


4.3.1. Search strategy ................................................................


4.3.3. Summary table ................................................................

4.3.4. Predictive ability ................................................................

4.3.5. Quality of the studies................................................................

4.3.6. Forest plots sensitivity and specificity ................................

KCE Report 193S

ISK ASSESSMENT SCALES .................................. 76

............................................................... 76

under the receiver operating characteristics curve (AUC) ........................ 90

................................................................................... 97

....................................................................... 108

..................................................................... 159

....................................................................................... 159

....................................................................................... 161

....................................................................................... 161

............................................................................. 168

....................................................................................... 169

................................................................................... 169

................................................................................. 169

.................................................................................... 169

......................................................... 170

.......................................................... 172

................................................................................... 173

............................................................ 178

....................................................................................... 178

....................................................................................... 180

....................................................................................... 180

............................................................................. 187

....................................................................... 188

....................................................................................... 189

................................................................................... 189

................................................................................. 189

.................................................................................... 189

.................................................................................. 191

........................................................................... 194

................................................................................. 195

KCE Report 193S

5.

5.1.

5.2.

5.3.

6.

6.1.

6.2.

6.3.

7.

7.1.

7.2.



5. SKIN MASSAGE................................................................ ................................

5.1. REVIEW PROTOCOL ................................................................

5.2. SEARCH STRATEGY ................................................................

5.2.1. Search filters ................................................................


5.2.3. Excluded study................................................................


5.3.1. Search strategy ................................................................

5.3.2. Summary tables ................................................................

5.3.3. Grade evidence profiles ................................................................


5.3.5. Forest plots................................................................................................

6. REPOSITIONING ................................................................ ................................

6.1. REVIEW PROTOCOL ................................................................

6.2. SEARCH STRATEGY ................................................................

6.2.1. Search filters ................................................................


6.2.3. Excluded clinical studies ................................................................


6.3.1. Search strategy ................................................................


6.3.3. Summary table ................................................................

6.3.4. GRADE-tables................................................................

6.3.5. Forest plots................................................................................................


7. RE-DISTRIBUTING DEVICES................................................................

7.1. REVIEW PROTOCOL ................................................................

7.2. SEARCH STRATEGY ................................................................

3

....................................................................... 197

............................................................... 219

....................................................................................... 219

....................................................................................... 221

....................................................................................... 221

............................................................................. 226

.................................................................................... 227

....................................................................................... 227

................................................................................... 227

.................................................................................. 227

...................................................................... 228

....................................................................... 231

.......................................................... 236

.............................................................. 237

....................................................................................... 237

....................................................................................... 240

....................................................................................... 240

............................................................................. 247

..................................................................... 248

....................................................................................... 248

................................................................................... 248

................................................................................. 248

.................................................................................... 249

..................................................................................... 252

.......................................................... 266

....................................................................... 275

.......................................................................... 297

....................................................................................... 297

....................................................................................... 301

4

7.3.

8.

8.1.

8.2.

8.3.

9.

9.1.

9.2.


7.2.1. Search filters ................................................................




7.3.1. Search strategy ................................................................


7.3.3. Glossary of terms ................................................................

7.3.4. Summary of included studies ................................................................

7.4.1. Clinical evidence GRADE-tables................................................................

7.4.2. Forest plots................................................................................................

7.4.3. Alternative foam mattress vs standard foam mattress

7.4.4. Alternating-pressure vs constant low-pressure ................................


8. HEEL ULCER PREVENTION (DEVICES) ................................

8.1. REVIEW PROTOCOL ................................................................

8.2. SEARCH STRATEGY ................................................................

8.2.1. Search filters ................................................................




8.3.1. Summary of included studies ................................................................

8.3.2. Clinical evidence GRADE-tables................................................................

8.3.3. Forest plots................................................................................................


9. NUTRITION AND HYDRATION................................................................

9.1. REVIEW PROTOCOL ................................................................

9.2. SEARCH STRATEGY ................................................................

9.2.1. Search strategy ................................................................

9.2.2. Search filters ................................................................

9.2.3. Flow diagram for article selection................................

KCE Report 193S

....................................................................................... 301

............................................................................. 309

..................................................................... 310

....................................................................................... 312

................................................................................... 312

................................................................................. 312

................................................................................ 313

.............................................................. 316

......................................................... 328

.......................................................... 362

Alternative foam mattress vs standard foam mattress ........................................................ 366

................................................................... 373

....................................................................... 387

........................................................................................ 409

....................................................................................... 409

....................................................................................... 412

....................................................................................... 412

............................................................................. 427

..................................................................... 428

....................................................................................... 430

.............................................................. 430

......................................................... 431

.......................................................... 437

....................................................................... 439

........................................................................ 450

....................................................................................... 450

....................................................................................... 453

................................................................................... 453

....................................................................................... 453

........................................................................................ 461

KCE Report 193S

9.3.

10.

10.1.

11.

12.


9.2.4. Excluded studies ................................................................


9.3.1. Summary table ................................................................


9.3.3. Appendix II: Forest plots ................................................................

9.3.4. Evidence tables................................................................

10. GRADE SYSTEM................................................................ ................................

10.1. DOWN- OR UPGRADING THE EVIDENCE ................................

10.1.1. Risk of bias................................................................................................

10.1.2. Inconsistency................................................................

10.1.3. Indirectness ................................................................................................

10.1.4. Imprecision................................................................................................

10.1.5. Publication Bias................................................................

11. ASSESSMENT OF EXISTING GUIDELINES ................................

12. RECOMMENDATIONS: COMMENTS EXPERT PANEL ................................

REFERENCES ................................................................ ................................

5

................................................................................. 462

....................................................................................... 462

.................................................................................... 463

......................................................... 465

..................................................................... 475

................................................................................... 479

.............................................................. 503

..................................................................................... 503

.......................................................... 504

....................................................................................... 504

......................................................... 504

.......................................................... 504

................................................................................... 506

................................................................................... 506

................................................................. 508

.................................................................. 519

6

LIST OF ABBREVIATIONS


ABBREVIATION DEFINITION

A&E Accident and Emergency

AHCPR Agency of Health Care Policy and Research

AP mattresses Alternating pressure mattresses

BMI Body Mass Index

CI Confidence Interval

CPG Clinical Practice Guideline

DOR Diagnostic Odds Ratio

EPUAP European Pressure Ulcer Advisory Panel

GDG Guideline Development Group

HRQoL Health-Related Quality of Life

HUI The Health Utilities Index

IAD Incontinence Associated Dermatitis

ILD Indentation load deflection

IQR Inter-quartile range

ITT Intention-to-treat

KCE Belgian Healthcare Knowledge Centre

LTCF Long-term care facilities

LR Likelihood ratio

MID Minimal important difference

NBE Non-blanchable erythema

NCGC National Clinical Guideline Centre

NICE National Institute for Health and Clinical Excellence

NPUAP National Pressure Ulcer Advisory Panel

NPV Negative Predictive Value

PICO Population, Intervention, Comparison, Outcome

PPV Positive Predictive Value

PSPS Pressure Sore Prediction Score

KCE Report 193S

Agency of Health Care Policy and Research

National Institute for Health and Clinical Excellence

Intervention, Comparison, Outcome

KCE Report 193S Pressure Ulcer Prevention – Supplement

PU Pressure Ulcer

RAPS Risk Assessment Pressure Sore scale

RCT Randomized Controlled Trial

SF-36 Short form 36 health survey

SS Suriadi and Sanada scale

SD Standard Deviation

TNH-PUPP The Northern Hospital Pressure Ulcer Prevention Plan

WHOQOL-BREF An abbreviated 26 item version of the World Health Organization(WHOQOL-100) instrument

EQ-5D A standardised measure of health status developed by the EuroQol Group

7

The Northern Hospital Pressure Ulcer Prevention Plan

An abbreviated 26 item version of the World Health Organization – Quality of Life

health status developed by the EuroQol Group

8

1. RISK ASSESSMENT – CLINICAL EFFECTIVENE

1.1. Review protocol

Table 1 – Protocol review question

Protocol Risk assessment


Review question What is the clinical effectiveness of risk asses

Population Individuals of all ages

Intervention Clinical judgement based on risk factors

Risk assessment tool (any reported cut

o

o

o

o

o

o

Comparison Each other

No risk assessment

Outcomes Critical outcome for decision

Proportioulcer)

Important outcomes

Patient acceptability Rate of development of pressure ulcers Time to develop new pres Time in hospital or other health care setting (continuous data); Health-related quality of life (continuous data)


CLINICAL EFFECTIVENESS

Risk assessment

Risk assessment

What is the clinical effectiveness of risk assessment tools in the prevention of pressure ulcers?

Individuals of all ages in all settings

Clinical judgement based on risk factors

Risk assessment tool (any reported cut-off score)

Braden,

Norton,

Waterlow,

Cubbin-Jackson,

Braden-Q,

Other scales (e.g. Gosnell scale, Knoll scale, Andersen, Pressure Sore Prediction Score, Risk AssessmentPressure Sore, Douglas, Emina, Glamorgan)

Each other

No risk assessment

Critical outcome for decision-making

Proportion of participants developing new pressure ulcers (dichotomous outcome)(describe different categories of

Important outcomes

Patient acceptability;Rate of development of pressure ulcers;Time to develop new pressure ulcer (time to event data);

n hospital or other health care setting (continuous data);related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes in

KCE Report 193S

sment tools in the prevention of pressure ulcers?

Other scales (e.g. Gosnell scale, Knoll scale, Andersen, Pressure Sore Prediction Score, Risk Assessment

n of participants developing new pressure ulcers (dichotomous outcome)(describe different categories of

(although unlikely to be sensitive enough to detect changes in

KCE Report 193S


pressure ulcer patients, therefore may have to be narratively summarised)

o

o

o

o

o

oo

Study design High quality systemat

Cochrane reviews will be included if tdata such as available case analysis or

Cohort studies will be considered if no RCTs are available.

Exclusion Studies with another population, intervention, comparison or outcome Non-English, non

Search strategy The electronic databases to be searched are:

Medline (OVID interface), Cinahl (EBSCO All years

Review strategy How will individual PICO characteristics be combined across studies)

Population Intervention Comparison Outcomes Blinding Minimum follow up = no minimum Minimum total size = no minimum Use available case analysis for dealing

or if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.

Analysis Strata:


Risk assessment


Short-form health survey (SF36)

Manchester Short Assessment of Quality of Life

EQ-5D

WHOQOL-BREF

Cardiff HRQoL tool

HUIPressure ulcer quality of life (Gorecki)

High quality systematic reviews of RCTs or RCTs only;

Cochrane reviews will be included if they match the inclusion criteria and have appropriate assumptions for missingdata such as available case analysis or intention-to-treat – ITT (with the appropriate assumptions)Cohort studies will be considered if no RCTs are available.

s with another population, intervention, comparison or outcome;English, non-French, non-Dutch language papers.

The electronic databases to be searched are:

Medline (OVID interface), Cinahl (EBSCO-interface), Embase, Library of the CoAll years.

How will individual PICO characteristics be combined across studies) :

Population – any population will be combined except those specified in the strataIntervention – combine same tools only;Comparison – any comparison will be combined;Outcomes – same outcomes will be combined;Blinding – Blinded and unblinded studies will be meta-analysed together;Minimum follow up = no minimum;Minimum total size = no minimum;Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.

9

hey match the inclusion criteria and have appropriate assumptions for missingITT (with the appropriate assumptions);

interface), Embase, Library of the Cochrane Collaboration;

any population will be combined except those specified in the strata;

with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take the

10


The following groups will be

Children (neonates, infants and children) ICU patients Patients with a spinal cord injury Palliative patients

Subgroups:

The following groups will be considered separately as subgroups if data are present and

Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separately Different ulcer locations: sacral, heel and others

1.2. Search strategy

1.2.1. Search filters

Table 2 – Search filters Medline (OVID)

Date 11/9/12

Database Ovid MEDLINE(R) InPresent

Search Strategy 1. Pressure Ulcer/2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage)).ti,ab.4. (bedsore* or bed5. ((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.6. OR/1 – 57. Risk assessment/8. Nursing assessment/9. nursing assess$.tw10. risk assess$.tw11. risk-benefit assess$.tw12. structured assess$.tw13. unstructured assess$.tw14. instrument$.tw


Risk assessment

The following groups will be considered separately if data are present:

Children (neonates, infants and children)ICU patients;

atients with a spinal cord injury;alliative patients.

The following groups will be considered separately as subgroups if data are present and

Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separatelyDifferent ulcer locations: sacral, heel and others

Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations and Ovid MEDLINE(R)

Pressure Ulcer/decubit*.ti,ab.(pressure adj (sore* or ulcer* or damage)).ti,ab.(bedsore* or bed-sore*).ti,ab.

on or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

Risk assessment/Nursing assessment/nursing assess$.twrisk assess$.tw

benefit assess$.twstructured assess$.twunstructured assess$.twinstrument$.tw

KCE Report 193S

The following groups will be considered separately as subgroups if data are present and there is inconsistency:

Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separately )

Indexed Citations and Ovid MEDLINE(R) 1946 to

on or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

913939576283506253

1352115236926960

110828820

439580

9

KCE Report 193S

Date 11/9/12

15. tool$.tw16. scale$.tw17. screening.tw18. Risk factor/19. risk factor$.tw20. risk score$.tw21. assess$ score?.tw22. Judgment/23. clinical judg?ment.tw24. Observation/25. observation?.tw26. OR/7 – 2527. randomized controlled trial.pt.28. controlled clinical trial.pt.29. randomi#ed.ab.30. placebo.ab.31. randomly.ab.32. Clinical Trials as topic/33. trial.ti34. OR/27 – 3335. AND/6, 26, 3436. Limit year: ‘2010


screening.twRisk factor/risk factor$.twrisk score$.twassess$ score?.twJudgment/clinical judg?ment.twObservation/observation?.tw

randomized controlled trial.pt.controlled clinical trial.pt.randomi#ed.ab.placebo.ab.

ly.ab.Clinical Trials as topic/

33AND/6, 26, 34Limit year: ‘2010 – September 2012’ and limit language: ‘English, Dutch, French’

11

September 2012’ and limit language: ‘English, Dutch, French’

165360330115380195289136498760297520

58041858

1157441034267

479442222249333658785134

301653139359184340162333108321824270

31147

12

Table 3 – Search filters Embase

Date 11/9/12

Database Embase

Search Strategy

(attention, for PubMed,check « Details »)

1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (score* OR ulcer* OR damage)):ab,ti4. (bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab5. ((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab6. OR/1 – 57. 'risk assessment'/exp8. ‘risk assessment$’:ti,ab9. ‘assessment$ risk’:ti,ab10. ‘risk-benefit assessment$’:ti,ab11. ‘structured assessment$’:ti,ab12. ‘assessment$ structured’:ti,ab13. ‘unstructured assessment$’:ti,ab14. ‘instrument’/exp15. ‘instrument$’:ti,ab16. ‘tool$’:ti,ab17. ‘scale$’:ti,ab18. ‘screening’:ti,ab19. ‘risk factor’/exp20. ‘risk factor$’:ti,ab21. ‘factor risk$’:ti,ab22. ‘risk score$’:ti,ab23. ‘score$ risk’:ti,ab24. ‘assessment$ score$’:ti,ab25. ‘decision making’/exp26. ‘clinical judg?ment’:ti,ab27. ‘clinical observation’/exp28. ‘nursing assessment’/exp29. ‘nursing assessment’:ti,ab30. ‘observation$’:ti,ab31. OR/7 – 3032. ‘clinical trial’/exp


‘decubitus’/expti,ab

(pressure NEAR/1 (score* OR ulcer* OR damage)):ab,ti(bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

'risk assessment'/exp‘risk assessment$’:ti,ab

sessment$ risk’:ti,abbenefit assessment$’:ti,ab

‘structured assessment$’:ti,ab‘assessment$ structured’:ti,ab‘unstructured assessment$’:ti,ab‘instrument’/exp‘instrument$’:ti,ab

‘scale$’:ti,ab‘screening’:ti,ab‘risk factor’/exprisk factor$’:ti,ab‘factor risk$’:ti,ab‘risk score$’:ti,ab‘score$ risk’:ti,ab‘assessment$ score$’:ti,ab‘decision making’/exp‘clinical judg?ment’:ti,ab‘clinical observation’/exp‘nursing assessment’/exp‘nursing assessment’:ti,ab‘observation$’:ti,ab

‘clinical trial’/exp

KCE Report 193S

((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

694936312357390240

975424648126996

19345851915

66212

60953220029323627285932383549113655

1046423253

112164313

131213399

250213

1909561632721

KCE Report 193S

Date 11/9/12

33. ‘clinical trial (topic)’/exp34. random*:ti,ab35. factorial*:ti,ab36. crossover*:ti,ab OR (cross NEXT/1 over*):ti,ab37. ((doubl* or singl*) NEAR/2 blind*):ti,ab38. (assign* or allocat* or volunteer* or placebo*):ti,ab39. ‘crossover procedure’40. ‘single blind procedure’/exp41. ‘double blind procedure’/exp42. OR/32 – 4143. AND/6, 31, 4244. Limit year: ‘2010

Table 4 – Search filters Cochrane Library

Date 11/9/12

Database The Library of the Cochrane Collabo

Search Strategy


1. “Pressure ulcer” [MeSH]2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulcer* or damage*)):ti,ab,kw4. (bedsore* or bed5. ((friction or shear) near/2 (sore* or ulcer* or damage o6. OR/1 – 57. “risk assessment”[MeSH]8. (risk assess*):ti,ab,kw9. (risk-benefit assess*):ti,ab,kw10. (structured assessment):ti,ab,kw11. (unstructured assess*):ti,ab,kw12. (instrument*):ti,ab,kw13. (tool*):ti,ab,kw14. (scale*):ti,ab,kw15. (screening*):ti,ab,kw


‘clinical trial (topic)’/exprandom*:ti,abfactorial*:ti,abcrossover*:ti,ab OR (cross NEXT/1 over*):ti,ab((doubl* or singl*) NEAR/2 blind*):ti,ab(assign* or allocat* or volunteer* or placebo*):ti,ab‘crossover procedure’/exp‘single blind procedure’/exp‘double blind procedure’/exp

41AND/6, 31, 42Limit year: ‘2010 – September 2012’ and language: ’English, Dutch, French’

The Library of the Cochrane Collaboration

“Pressure ulcer” [MeSH]Decubit*:ti,ab,kw(pressure near/2 (sore* or ulcer* or damage*)):ti,ab,kw(bedsore* or bed-sore*):ti,ab,kw((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or lesion*)):ti,ab,kw

“risk assessment”[MeSH](risk assess*):ti,ab,kw

benefit assess*):ti,ab,kw(structured assessment):ti,ab,kw(unstructured assess*):ti,ab,kw(instrument*):ti,ab,kw(tool*):ti,ab,kw(scale*):ti,ab,kw

creening*):ti,ab,kw

13

71957237330

6132321335454706

129598474964290891214492686

1404456372108

r wound* or injur*or lesion*)):ti,ab,kw

48934983433

3

11155960

23877132

187244

210136381

48545

14

Date 11/9/12

16. “risk factors”[MeSH]17. (risk factor):ti,ab,kw18. (risk score):ti,ab,kw19. (assessment score):ti,ab,kw20. “Judgment”[MeSH]21. “nursing assessment”[MeSH]22. (nurs* assess*):ti,ab,kw23. (clinical judg?ment):ti,ab,kw24. “Observation”[MeSH]25. (observation*):ti,ab,k26. OR/7 – 2527. (Clinical Trial):pt28. (Randomized Controlled Trial):pt29. “clinical trial as topic” [MeSH]30. (trial*):ti,ab,kw31. (randomized or randomised):ti,ab,kw32. (randomly):ti,ab,kw33. (group*):ti,ab,kw34. OR/27 – 3235. AND/6, 26, 3336. Limit year: ‘2010

Table 5 – Search filters CINAHL

Date 12/9/12

Database CINAHL (EBSCO-interface)

Search Strategy


1. MH “Pressure Ulcer”2. bedsore* OR bed3. pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*4. decubit*5. ((friction or shear) and (sore* or ulcer* or damage or wound* or injur6. OR/1 – 5


“risk factors”[MeSH](risk factor):ti,ab,kw(risk score):ti,ab,kw(assessment score):ti,ab,kw“Judgment”[MeSH]“nursing assessment”[MeSH](nurs* assess*):ti,ab,kw(clinical judg?ment):ti,ab,kw“Observation”[MeSH](observation*):ti,ab,kw

(Clinical Trial):pt(Randomized Controlled Trial):pt“clinical trial as topic” [MeSH](trial*):ti,ab,kw(randomized or randomised):ti,ab,kw(randomly):ti,ab,kw(group*):ti,ab,kw

AND/6, 26, 33Limit year: ‘2010 – September 2012’

nterface)

MH “Pressure Ulcer”bedsore* OR bed-sore*pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*

((friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))

KCE Report 193S

120291619028808

626111419

430493

4271260139

2296214541229457631365251548

24837826494785941

273734533623

32350

7749157

8547

487806

KCE Report 193S

Date 12/9/12

7. MH “Risk assessment”8. MH “Nursing assessment”9. “risk assessment$” or “assessment$ risk”10. “nurs$ assessment$ or “assessment$ nurs$”11. “risk-benefit assessment$”12. “structured assessment$” or “assessment$ structured”13. “unstructured assessment14. “instrument$”15. “tool$”16. “scale$”17. “screening”18. MH “risk factor”19. “risk factor$” or “factor$ risk”20. “risk score$” or “score$ risk”21. “assessment score$”22. MH "Decision Making, Clinical"23. MH "Judgment"24. “clinical judg?ment”25. “observation” or “observations”26. OR/7 – 2527. MH "Clinical Trials+"28. “trial*”29. “randomi#ed”30. “randomly”31. “randomized controlled trial”32. PT “randomized controlled trial”33. PT “clinical trial”34. OR/27 – 3335. AND/6, 26, 3436. Limit year: ‘2010


MH “Risk assessment”MH “Nursing assessment”“risk assessment$” or “assessment$ risk”“nurs$ assessment$ or “assessment$ nurs$”

benefit assessment$”“structured assessment$” or “assessment$ structured”“unstructured assessment$”“instrument$”

MH “risk factor”“risk factor$” or “factor$ risk”“risk score$” or “score$ risk”“assessment score$”MH "Decision Making, Clinical"MH "Judgment"“clinical judg?ment”“observation” or “observations”

MH "Clinical Trials+"

“randomi#ed”

“randomized controlled trial”PT “randomized controlled trial”PT “clinical trial”

AND/6, 26, 34Limit year: ‘2010 – September 2012’ and language: ’English, Dutch, French’

15

9407284581371530248

336345139

23113530351

6687531545350113663

931274

131201889277

266802826531075381382016669225374

91441099051404

16944122236

16

1.2.2. Selection of articles

Figure 1 – Flow chart search strategy

Pressure Ulcer Prevention – Supplement KCE Report 193S

KCE Report 193S

1.3. Clinical evidence

1.3.1. Summary table

Table 6 – Summary of included studies

Study Intervention/comparator

Saleh 20091

(1) Braden scale and training

(2) Training only

(3) Clinical judgement

Webster 20112

(1) Waterlow scale

(2) Ramstadius scale



Intervention/comparator Population Outcome

(1) Braden scale and training


Hospitalized patients with apressure ulcer and/or Bradenscale ≤ 18

Incidence ofulcers

(1) Waterlow scale

(2) Ramstadius scale


Hospitalized patients olderthan 18 years with or withouta pressure ulcer

Incidence ofulcers

17

Study length

Incidence of pressure Eight weeks

Incidence of pressure Maximum 98 days

18

1.3.2. Clinical evidence GRADE tables

Table 7 – Braden scale versus clinical judgement

Quality assessment

No of

studies

Design Risk of

bias

Inconsistency Indirectness

Incidence of pressure ulcers – all grades

Saleh

2009

randomised

trials

very

serious1

no serious

inconsistency

no serious

indirectness

1 Sequence allocation and blinding not reported; inadequate allocation concealment (ward allocation); difference at baseline2 Confidence interval crossed one MID


Braden scale versus clinical judgement

Quality assessment No of patients

Indirectness Imprecision Other

considerations

Braden

scale

Clinical

judgement

Relative

no serious

indirectness

serious2

none 16/74

(21.6%)

16/106

(15.1%)

RR 1.43

(0.77 to

15.1%

1 Sequence allocation and blinding not reported; inadequate allocation concealment (ward allocation); difference at baseline not reported; no intention

KCE Report 193S

Effect Quality Importance

Relative

(95%

CI)

Absolute

RR 1.43

(0.77 to

2.68)

65 more

per 1000

(from 35

fewer to

254

more)

VERY

LOW

CRITICALOUTCOME

65 moreper 1000(from 35fewer to

254more)

not reported; no intention-to-treat analysis

KCE Report 193S

Table 8 – Braden scale versus training only

Quality assessment

No of

studies

Design Risk of

bias



Saleh

2009

randomised

trials

very

serious1

no serious

inconsistency

no serious

indirectness

1 Sequence allocation and blinding not reported; inadequate allocation concealment (ward allocation); dif2 Confidence interval crossed both MIDs




considerations

Braden

scale

Training

only

Relative

(95% CI)

no serious

indirectness

very

serious2

none 16/74

(21.6%)

17/76

(22.4%)

RR 0.97

(0.53 t

1.77)

22.4%

1 Sequence allocation and blinding not reported; inadequate allocation concealment (ward allocation); dif ference at baseline not reported; no intention

19


Relative

(95% CI)

Absolute

RR 0.97

(0.53 to

1.77)

7 fewer

per 1000

(from 105

fewer to

172 more)

VERY

LOW

CRITICALOUTCOME

7 fewerper 1000(from 105fewer to

172 more)

ference at baseline not reported; no intention-to-treat analysis

20

Table 9 – Training only versus clinical judgement

Quality assessment

No of

studies

Design Risk of

bias



Saleh

2009

randomised

trials

very

serious1

no serious

inconsistency

no serious

indirec

1 Sequence allocation and blinding not reported; inadequate allocation concealment (ward allocation);2 Confidence interval crossed one MID


Training only versus clinical judgement



considerations

Training

only

Clinical

judgement

Relative

(95%

no serious

indirectness

serious2

none 17/76

(22.4%)

16/106

(15.1%)

RR 1.48

(0.8 to

2.74)

15.1%

inadequate allocation concealment (ward allocation); difference at baseline not reported; no intention

KCE Report 193S


Relative

(95%

CI)

Absolute

RR 1.48

(0.8 to

2.74)

72 more

per 1000

(from 30

fewer to

263 more)

VERY

LOW

CRITICALOUTCOME


263 more)

difference at baseline not reported; no intention-to-treat analysis

KCE Report 193S

Table 10 – Waterlow scale versus clinical judgement

Quality assessment

No of

studies

Design Risk of

bias



Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness

Incidence of pressure ulcers – grade 2

Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness

1Confidence interval crossed both MIDs


Waterlow scale versus clinical judgement



considerations

Waterlow

scale

Clinical

judgement

Relative

(95% CI)

no serious

indirectness

very

serious1

none 31/411

(7.5%)

28/410

(6.8%)

RR 1.1

(0.68 to

1.81)

6.8%

no serious

indirectness

very

serious1

none 10/411

(2.4%)

8/410

(2%)

RR 1.25

(0.5 to

3.13)

2%

21


Relative

(95% CI)

Absolute

RR 1.1

(0.68 to

1.81)

7 more per

1000 (from 22

fewer to 55

more)

LOW

CRITICALOUTCOME

7 more per1000 (from 22

fewer to 55more)

RR 1.25

(0.5 to

3.13)

5 more per

1000 (from 10

fewer to 42

more)

LOW

CRITICALOUTCOME


fewer to 43more)

22

Table 11 – Waterlow scale versus Ramstadius scale

Quality assessment

No of

studies

Design Risk of

bias



Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness


Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness

1 Confidence interval crossed one MID


Waterlow scale versus Ramstadius scale



considerations

Waterlow

scale

Ramstadius

scale

Relative

(95% CI)

no serious

indirectness

serious1 none 31/411

(7.5%)

22/410

(5.4%)

RR 1.41

(0.83 to

2.39)

5.4%

no serious

indirectness

serious1 none 10/411

(2.4%)

4/410

(1%)

RR 2.49

(0.79 to

7.89)

1%

KCE Report 193S


Relative

(95% CI)

Absolute

RR 1.41

(0.83 to

2.39)

22 more per

1000 (from

9 fewer to

75 more)

MODERATE

CRITICALOUTCOME

22 more per1000 (from9 fewer to75 more)

RR 2.49

(0.79 to

7.89)

15 more per

1000 (from

2 fewer to

67 more)

MODERATE

CRITICALOUTCOME


KCE Report 193S

Table 12 – Ramstadius scale versus clinical judgement

Quality assessment

No of

studies

Design Risk of

bias



Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness


Webster

2011

randomised

trials

No

serious

risk of

bias

no serious

inconsistency

no serious

indirectness

1Confidence interval crossed both MIDs


inical judgement



considerations

Ramstadius

scale

Clinical

judgement

Relative

(95% CI)

no serious

indirectness

very

serious1

none 22/410

(5.4%)

28/410

(6.8%)

RR 0.79

(0.46 to

1.35)

6.8%

no serious

indirectness

very

serious1

none 4/410

(1%)

8/410

(2%)

RR 0.5

(0.15 to

1.65)

2%

23


Relative

(95% CI)

Absolute

RR 0.79

(0.46 to

1.35)

14 fewer per

1000 (from 37

fewer to 24

more)

LOW

CRITICALOUTCOME

14 fewer per1000 (from 37

fewer to 24more)

RR 0.5

(0.15 to

1.65)

10 fewer per

1000 (from 17

fewer to 13

more)

LOW

CRITICALOUTCOME


fewer to 13more)

24

1.3.3. Forest plots

Figure 2 – Braden scale versus clinical judgement

Figure 3 – Braden scale versus training only –

Study or Subgroup

Saleh 2009

Total (95% CI)

Total events

Heterogeneity: Not applicable

Test for overall effect: Z = 1.12 (P = 0.26)

Events

16

16

Total

74

74

Events

16

16

Total

106

106

Braden scale Clinical judgement

Study or Subgroup

Saleh 2009

Total (95% CI)

Total events



Events

16

16

Total

74

74

Events

17

17

Total

76

76

Weight

100.0%

100.0%

Braden scale Training only


versus clinical judgement – all stages

all stages

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.43 [0.77, 2.68]

1.43 [0.77, 2.68]

Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favour Braden scale Favour clinical judgement

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.97 [0.53, 1.77]

0.97 [0.53, 1.77]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Braden scale Favours training only

KCE Report 193S

KCE Report 193S

Figure 4 – Training only versus clinical judgement

Figure 5 – Waterlow scale versus clinical judgement

Figure 6 – Waterlow scale versus Ramstadius scale

Study or Subgroup

Saleh 2009

Total (95% CI)

Total events



Events

17

17

Total

76

76

Events

16

16

Total

106

106

Training only Clinical judgement

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

31

31

Total

411

411

Events

28

28

Total

410

410

Waterlow scale Clinical judgement

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

31

31

Total

411

411

Events

22

22

Total

410

410

Waterlow scale Ramstadius scale


Training only versus clinical judgement – all stages

Waterlow scale versus clinical judgement – all stages

Waterlow scale versus Ramstadius scale – all stages

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.48 [0.80, 2.74]

1.48 [0.80, 2.74]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favour training only Favour clinical judgement

Total

410

410

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.10 [0.68, 1.81]

1.10 [0.68, 1.81]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Waterlow scale Favours clinical judgment

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.41 [0.83, 2.39]

1.41 [0.83, 2.39]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Waterlow scale Favours Ramstadius scale

25

26

Figure 7 – Ramstadius scale versus clinical judgement

Figure 8 – Waterlow scale versus clinical judgement

Figure 9 – Waterlow scale versus Ramstadius scale

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

22

22

Total

410

410

Events

28

28

Total

410

410

Ramstadius scale Clinical judgement

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

10

10

Total

411

411

Events

8

8

Total

410

410

Waterlow scale Clinical judgement

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

10

10

Total

411

411

Events

4

4

Total

410

410

Waterlow scale Ramstadius scale


Ramstadius scale versus clinical judgement – all stages

Waterlow scale versus clinical judgement – stage 2

Waterlow scale versus Ramstadius scale – stage 2

Total

410

410

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.79 [0.46, 1.35]

0.79 [0.46, 1.35]

Clinical judgement Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favour Ramstadius scale Favour clinical judgement

Total

410

410

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.25 [0.50, 3.13]

1.25 [0.50, 3.13]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favour Waterlow scale Favour clinical judgement

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

2.49 [0.79, 7.89]

2.49 [0.79, 7.89]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Waterlow scale Favours Ramstadius scale

KCE Report 193S

100

KCE Report 193S

Figure 10 – Ramstadius scale versus clinical judgement

1.3.4. Evidence tables

Table 13 – Saleh 2009

Reference Patient Characteristics

Author and year:Saleh (2009)*

Title: The impact ofpressure ulcer riskassessment onpatient outcomesamong hospitalisedpatients

Journal: Journal ofClinical Nursing, 18;1923-29.

Study type: clusterrandomizedcontrolled trial

Sequence generation:not reported

Patient group:hospitalized patientswith or without a PU

All patients

Randomised N: notreported

Completed N: 256

Drop-outs: not reported

Group 1


Completed N: 74

Dropouts: not reported

Study or Subgroup

Webster 2011

Total (95% CI)

Total events



Events

4

4

Total

410

410

Events

8

8

Total

410

410

Ramstadius scale Clinical judgement


Ramstadius scale versus clinical judgement – stage 2

Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group:hospitalized patients

not

not reported

not

Group 1: assessment of allpatients with the Bradenscale. All nurses received amandatory training on woundcare management, PUprevention and use of theBraden scale.

Group 2: All nurses receiveda mandatory training onwound care management,PU prevention and use of theBraden scale. Use of theBraden scale was notrequired.

Group 3: All nurses receiveda mandatory training onwound care management.

Outcome 1:Incidence of PU

Group 1: 16/74

Group 2: 17/76

Group 3: 16/106

Total

410

410

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.50 [0.15, 1.65]

0.50 [0.15, 1.65]

Clinical judgement Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favour Ramstandius scale Favour clinical judgement

27

Comments

16/74

17/76

16/106

Funding: /

Limitations:sequencegeneration notreported;allocationconcealment notreported; noblinding; no reporton baselinedifferenceregardingpresence of PUon admission; nointention-to-treatanalyses and highdropout

100Favour clinical judgement

28


Allocationconcealment: wardswere unit ofallocation Blinding:not reported

Addressing incompleteoutcome data: nointention-to-treatanalysis. 198patients wereexcluded becausethey were dischargedbefore 8 weeks (totalstudy period)

Statistical analysis:Data were analysedby using descriptiveand inferentialstatisticalprocedures (tests).The inferentialstatistics may beparametric ornonparametric. Chi-square test was usedto test independenceof nominal variables.Student t test forindependent groupsand one way ANOVAwere not used to testdifferences betweenrespectively two or

Age: /

Gender (m/f): /

Group 2


Completed N: 76


Age: /

Gender (m/f): /

Group 3


Completed N: 106


Age: /

Gender (m/f): /

Inclusion criteria:Braden scale ≤ 18 and/or having a PUstage I-IV.

Exclusion criteria:Patients with a PU stageI-IV and a Braden score> 18


Intervention

Comparison

Outcomemeasures

Effect sizes

not

not

Inclusion criteria:≤ 18

and/or having a PU

Exclusion criteria:Patients with a PU stage

IV and a Braden score

Clinical judgement group.

All groups: all patients weremonitored for preventivemeasures and includedfollowing categories:

Protective mattresses suchas the standard hospital bed

mattress (Stryker®, Inc.,Hamilton, ON, Canada),alternating pressure reliefsystem Therakair® (KineticConcepts, Inc., San Antonio,TX, USA), Gen Air 8000®(Genadyne Inc., Great Neck,NY, USA), Atmosair®(Kinetic Concepts, Inc., USA)and gel overlay or airfluidised bed (Clinitron®, Hill-Rom, Inc., Batesville, IN,USA);

Creams and skin barriers;

Vitamin supplements andspecial nutritional formulas;

Patients’ turning (positioning)schedules every two, three tofour, or six hours.

KCE Report 193S

Comments

(discharge beforeend of studyperiod); patientswith PU beforeinterventionincluded.

Very high risk ofbias.

Additionaloutcomes:association wasmeasured withPU incidence.AUC for Bradenscale and clinicaljudgement werereported.

Notes: /

KCE Report 193S


more than twogroups because thedata were notnormally distributed.Mann-Whitney U(MW) test andKruskal-Wallis (KW)test were used to testdifferences betweenrespectively two ormore than twogroups with data thatwere at least ordinal,but not sufficientlynormally distributedto warrant parametrictesting. Logisticregression analysiswas used to producea predictive modelfrom those recordedvariables which arerelated to PUdevelopment. ROCcurve analysis wasused to show theeffects of the Bradenscale compared tonurses' clinicaljudgement in relationto PU development.

Baseline differences:Baseline differences


Intervention

Comparison

Outcomemeasures

Effect sizes

29

Comments

30


for medicaldiagnosis, protectivemeasures, use ofbarrier creams andvitamin therapy.

Study power/samplesize: A priori samplesize calculationindicated a samplesize of 108 patients.Final sample sizewas higher thancalculated.

Setting: Militaryhospital, Riyadh,Saudi Arabia

Length of study: eightweeks

Assessment of PUs:

PU were classifiedaccording to the USAgency for HealthCare Policy andResearch (1992). Atissue viability nursespecialist and twotrained staff nursesassed the wounds.

Multiple ulcers: PU atstart and patientcould havedeveloped a new


Intervention

Comparison

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


ulcer. If the patientdeveloped more thanone PU lesion, onlythe first one wastaken into account.Number of patientswith multiple ulcersnot reported

* The authors were contacted for additional information. This publication is part of a doctoral thesis and can be retrieved on https://www.dora.dmu.ac.uk/handle/2086/

Table 14 – Webster 2011


Author and year:Webster (2011)

Title: Pressure ulcers:effectiveness of risk-assessment tools. Arandomized controlledtrial (the ULCER trial)

Journal: BMJ Quality& Safety, 20 (4); 297-306

Study type:randomized controlledtrial

Sequencegeneration: acomputer-generatedrandomized list was

Patient group:hospitalized patientsolder than 18 years withor without a PU

All patients

Randomised N: 1231

Completed N: 1231

Drop-outs: 0

Group 1

Randomised N: 411

Completed N: 411

Dropouts: 0

Age (mean yrs (SD);


Intervention

Comparison

Outcomemeasures

Effect sizes

al information. This publication is part of a doctoral thesis and can be retrieved on https://www.dora.dmu.ac.uk/handle/2086/

Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group:hospitalized patientsolder than 18 years with

Age (mean yrs (SD);

Group 1: the Waterlow scale

Group 2: the Ramstadiusscale

Group 3: Clinical judgement.

All groups: preventionmeasures were documented.

Outcome 1:Incidence of PU(all stages)

Outcome 2:Incidence of PU(stage I)

Outcome 3:Incidence of PU(stage II)

Group 1: 31/411

Group 2: 22/410

Group 3: 28/410

P value: 0.44

Group 1: 21/411

Group 2: 18/410

Group 3: 20/410

Group 1: 10/411

Group 2: 4/410

Group 3: 8/410

31

Comments

al information. This publication is part of a doctoral thesis and can be retrieved on https://www.dora.dmu.ac.uk/handle/2086/ 4343

Comments

31/411

22/410

28/410

0.44

21/411

18/410

20/410

10/411

4/410

8/410

Funding: /

Limitations: typeof method usedfor allocationconcealment notreported; healthcare professionalnot blinded; finalsample size lowerthan a prioricalculated; noreport on baselinedifferenceregardingpresence of PUon admission;patients with PU

32


used to allocate thepatients. Block andstratifiedrandomizationaccording to type ofpatient(medical/oncological),presence/absence ofPU on admission andmobility status.

Allocationconcealment:allocation wasconcealed; method notreported

Blinding: patient andoutcome assessorwere blinded to groupassignment.

Addressingincomplete outcomedata: Intention-to-treatanalysis. 7 missingdata on comorbidity;247 excluded frommodel because datawas not available.

Statistical analysis:Baseline clinical anddemographiccharacteristics werecompared usingfrequencies or means

range): 62.6 (19.6); 18100

Gender (m/f): 200/211

Ability to turnindependently: 374

Wheelchairdependent: 30

Pressure ulcer onadmission: 25

Length of stay (meandays (SD); range): 8.8(9.5); 1-98

Group 2

Randomised N: 410

Completed N: 410

Dropouts: 0

Age (mean yrs (SD);range): 63.2 (19.2); 1898

Gender (m/f): 205/205




Length of stay (meandays (SD); range): 9.4


Intervention

Comparison

Outcomemeasures

Effect sizes

62.6 (19.6); 18-

Ability to turn

Pressure ulcer on

Length of stay (mean8.8

Age (mean yrs (SD);63.2 (19.2); 18-

Ability to turn

Pressure ulcer on


KCE Report 193S

Comments

beforeinterventionincluded.

Additionaloutcomes:process of carebetween the threegroups weremeasured.Predictor ofpressure injurywere calculated.

Notes: /

KCE Report 193S


and standarddeviations (SD). Theinter-rater agreementwas assessed usingthe percentageagreement betweenraters. For the primaryoutcome, the OR andtheir 95% CIs werecalculated for theproportion of patientswith pressure ulcers ineach group. Logisticregression modelswere used todetermine risk factorsassociated withpatients developing apressure ulcer afteradmission. The initiallogistic regressionmodel incorporated allvariables that weresignificant in theunivariate analyses,and also adjusted forthe treatment group.Using this initial model,the backwardselimination was usedto the select finalmodel. As the vastmajority of inpatientdietician reviews are

(99.9); 1-81

Group 3

Randomised N: 410

Completed N: 410

Dropouts: 0

Age (mean yrs (SD);range): 61.9 (19.0); 19100

Gender (m/f): 214/196




Length of stay (meandays (SD); range): 8.5(8.5); 1-81

Inclusion criteria:admitted through theemergency departmentor any outpatientdepartment

Exclusion criteria:hospital stay < 3 days;hospitalized more than24h before baseline


Intervention

Comparison

Outcomemeasures

Effect sizes

Age (mean yrs (SD);61.9 (19.0); 19-

Ability to turn

Pressure ulcer on


Inclusion criteria:admitted through theemergency departmentor any outpatient

Exclusion criteria:hospital stay < 3 days;hospitalized more than24h before baseline

33

Comments

34


for malnutrition,referral to a dieticianwas used in themodels as a proxy formalnutrition.Regression modelsare adjusted forpotential confoundingof treatment group.

Baseline differences:Statistical differencewas calculated formean hours inemergencydepartment (p=0.56)and average length ofstay (p=0.38).

Study power/samplesize: A priori samplesize calculationindicated a samplesize of 466 patientsper group. Finalsample size lower thancalculated.

Setting: Internalmedicine ward andoncological ward at theRoyal Brisbane andWomen’s Hospital,Australia

Length of study: notreported; length of

assessment


Intervention

Comparison

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


stay: range 1-98 days

Assessment of PUs:

Research assistantstrained in pressureulcer staging asses thewounds using astandardizedassessment method(Black et al. 2007).

Multiple ulcers: notreported


Intervention

Comparison

Outcomemeasures

Effect sizes

35

Comments

36

2. RISK ASSESSMENT – PROGNOSTIC

2.1. Review Protocol



Review question What is the predictive ability of risk assessment tools for pressure ulcer


Risk assessment tool Clinical judgement based on risk factors

Risk assessment tool (any reported cut

o

o

o

o

o

o

Outcomes

Statistical measures

Critical outcomes

Incidence of pressure ulcers (all grades and grades 2

Incidence of pressure ulcers (all grades and grades 2


Area under the ROC (AUC)

Sensitivity for a defined threshold

Specificity for a defined threshold

Study design High quality systematic reviews

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginning of thestudy and with a follow

Exclusion Non-English, non


PROGNOSTIC

Risk assessment

What is the predictive ability of risk assessment tools for pressure ulcer development?

Individuals of all ages in all settings without a pressure ulcer

Clinical judgement based on risk factors

Risk assessment tool (any reported cut-off score):

Braden,

Norton,

Waterlow,

Cubbin-Jackson,

Braden-Q,

Other scales (e.g. Gosnell scale, Knoll scale, Andersen, Pressure Sore Prediction Score, Risk AssessmentPressure Sore, Douglas, Emina, Glamorgan)

outcomes

Incidence of pressure ulcers (all grades and grades 2-4)– up to one week

Incidence of pressure ulcers (all grades and grades 2-4) – up to three months


Area under the ROC (AUC);

Sensitivity for a defined threshold;

Specificity for a defined threshold;

High quality systematic reviews of prospective cohort studies.

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginning of thestudy and with a follow-up in a systematic way during an established period

English, non-French, non-Dutch language papers

KCE Report 193S

?

les (e.g. Gosnell scale, Knoll scale, Andersen, Pressure Sore Prediction Score, Risk Assessment

up to three months

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginning of the

KCE Report 193S




Review strategy How will individual PICO characteris

Population Risk tool Outcomes Minimum follow up = no minimum Minimum total size = no minim

Analysis The following groups will be considered separately if data are present:


The following analys

The AUC and 95used to calculate the median AUC and range.

Three cut(primarily) and specificity


Risk assessment


Medline (OVID interface), Cinahl (EBSCO-interface), Embase, Library of the Cochrane CollaborationAll years

How will individual PICO characteristics be combined across studies)

Population – any population will be combined except those specified in the strataRisk tool – combine same tools onlyOutcomes – same outcomes will be combinedMinimum follow up = no minimumMinimum total size = no minimum


Children (neonates, infants and children);ICU patients;


The following analyses will be performed

The AUC and 95% CI for each scale (within studies and between studies; if data are available) will be extracted andused to calculate the median AUC and range.

Three cut-off scores will be determined for each scale with an acceptable median AUC(primarily) and specificity

37

interface), Embase, Library of the Cochrane Collaboration

any population will be combined except those specified in the strata

% CI for each scale (within studies and between studies; if data are available) will be extracted and

off scores will be determined for each scale with an acceptable median AUC, optimising sensitivity

38




Date 25/7/12


Search Strategy 1. Pressure Ulce2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage)).ti,ab.4. (bedsore* or bed5. ((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.6. OR/1 – 57. Risk assessment/8. Nursing assessment/9. nursing assess$.t10. risk assess$.tw11. risk-benefit assess$.tw12. structured assess$.tw13. unstructured assess$.tw14. instrument$.tw15. tool$.tw16. scale$.tw17. screening.tw18. Risk factor/19. risk factor$.tw20. risk score$.tw21. assess$ score?.tw22. Judgment/23. clinical judg?ment.tw24. Observation/25. observation?.26. OR/7 – 2527. Braden$.tw28. Waterlow.tw



Pressure Ulcer/decubit*.ti,ab.(pressure adj (sore* or ulcer* or damage)).ti,ab.(bedsore* or bed-sore*).ti,ab.((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.


benefit assess$.twstructured assess$.twunstructured assess$.twinstrument$.tw

risk factor$.twrisk score$.twassess$ score?.tw

clinical judg?ment.tw

observation?.tw

KCE Report 193S


((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

913939576283506253

1352115236926960

110828820

439580

9165360330115380195289136498760297520

58041858

1157441034267

4794422222493

327

KCE Report 193S

Date 25/7/12

29. Norton.tw30. OR/27 – 2931. AND/6, 26, 3032. Limit year: ‘2003

Date 25/7/12


Search Strategy 1. Pressure Ulcer/2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage)).ti,ab.4. (bedsore* or bed5. ((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.6. OR/1 – 57. Risk assessment/8. Nursing assessment/9. nursing assess$.tw10. risk assess$.tw11. risk-benefit assess$.tw12. structured assess$.tw13. unstructured assess$.tw14. instrument$.tw15. tool$.tw16. scale$.tw17. screening.tw18. Risk factor/19. risk factor$.tw20. risk score$.tw21. assess$ score?.tw22. Judgment/23. clinical judg?ment.tw24. Observation/


AND/6, 26, 30Limit year: ‘2003 – July 2012’ and limit language: ‘English, Dutch, Flemish, French’


Pressure Ulcer/decubit*.ti,ab.

dj (sore* or ulcer* or damage)).ti,ab.(bedsore* or bed-sore*).ti,ab.((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.


assess$.twstructured assess$.twunstructured assess$.twinstrument$.tw

risk factor$.twrisk score$.twassess$ score?.tw

clinical judg?ment.tw

39

July 2012’ and limit language: ‘English, Dutch, Flemish, French’

164450880422215


((friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

913939576283506253

1352115236926960

110828820

439580

9165360330115380195289136498760297520

58041858

115744103

40

Date 25/7/12

25. observation?.tw26. OR/7 – 2527. Sensitivity and S28. sensitiv:.mp.29. predictive value:.mp.30. accuracy:.tw31. specificit$.mp32. OR/17 – 3133. AND/6, 26, 3234. Limit year: ‘2003


Date 25/7/12

Database Embase

Search Strategy


1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (score* OR ulcer* OR damage)):ab,ti4. (bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab5. ((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab6. OR/1 – 57. 'risk assessment'/exp8. ‘risk assessment$’:ti,ab9. ‘assessment$ risk’:ti,ab10. ‘risk-benefit assessment$’:ti,ab11. ‘structured assessment$’:ti,ab12. ‘assessment$ structured’:ti,ab13. ‘unstructured assessment$’:ti,ab14. ‘instrument’/exp15. ‘instrument$’:ti,ab16. ‘tool$’:ti,ab17. ‘scale$’:ti,ab18. ‘screening’:ti,ab19. ‘risk factor’/exp


observation?.tw

Sensitivity and Specificity/

predictive value:.mp.

specificit$.mp

AND/6, 26, 32Limit year: ‘2003 – July 2012’ and limit language: ‘English, Dutch, Flemish, French’

‘decubitus’/exp


'risk assessment'/expisk assessment$’:ti,ab

‘assessment$ risk’:ti,abbenefit assessment$’:ti,ab


:ti,ab‘risk factor’/exp

KCE Report 193S

July 2012’ and limit language: ‘English, Dutch, Flemish, French’

4267479442

2222493253657

1096653161212194778772727

1747453407220


694936312357390240

975424648126996

19345851915

66212

60953220029323627285932

KCE Report 193S

Date 25/7/12

20. ‘risk factor$’:ti,ab21. ‘factor risk$’:ti,ab22. ‘risk score$’:ti,ab23. ‘score$ risk’:ti,ab24. ‘assessment$ score$’:ti,ab25. ‘decision making’/exp26. ‘clinical judg?ment’:ti,ab27. ‘clinical observation’/exp28. ‘nursing assessment’/exp29. ‘nursing assessment’:30. ‘observation$’:ti,ab31. OR/7 – 3032. ‘braden scale’/exp33. ‘braden$’:ti,ab34. ‘waterlow’:ti,ab35. ‘norton’:ti,ab36. OR/32 – 3537. AND/6, 31, 3638. Limit year: ‘2003

Date 25/7/12

Database Embase

Search Strategy


1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (score* OR ulcer* OR damage)):ab,ti4. (bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab5. ((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab6. OR/1 – 57. 'risk assessment'/exp8. ‘risk assessment$’:ti,ab9. ‘assessment$ risk’:ti,ab10. ‘risk-benefit assessment$’:ti,ab


‘risk factor$’:ti,ab‘factor risk$’:ti,ab‘risk score$’:ti,ab‘score$ risk’:ti,ab‘assessment$ score$’:ti,ab‘decision making’/exp‘clinical judg?ment’:ti,ab‘clinical observation’/exp‘nursing assessment’/exp‘nursing assessment’:ti,ab‘observation$’:ti,ab

‘braden scale’/exp

‘waterlow’:ti,ab

Limit year: ‘2003 – July 2012’ and language: ’English, Dutch, French’

‘decubitus’/exp


'risk assessment'/exp‘risk assessment$’:ti,ab

ssment$ risk’:ti,abbenefit assessment$’:ti,ab

41

383549113655

1046423253

112164313

131213399

250213

1909561632721

3911690

30649813978


694936312357390240

975424648126996

193

42

Date 25/7/12

11. ‘structured assessment$’:ti,ab12. ‘assessment$ structured’:ti,ab13. ‘unstructured assessment$’:ti,ab14. ‘instrument’/exp15. ‘instrument$’:ti,ab16. ‘tool$’:ti,ab17. ‘scale$’:ti,ab18. ‘screening’:ti,ab19. ‘risk factor’/exp20. ‘risk factor$’:ti,ab21. ‘factor risk$’:ti,ab22. ‘risk score$’:ti,ab23. ‘score$ risk’:ti,ab24. ‘assessment$ score$’:ti,ab25. ‘decision making’/exp26. ‘clinical judg?ment’:ti,ab27. ‘clinical observation’/exp28. ‘nursing assessment’/exp29. ‘nursing assessment’:ti,ab30. ‘observation$’:ti,ab31. OR/7 – 3032. ‘sensitivity and Specificity’/exp33. ‘predictive validity’/exp34. ‘Predictive Value’/exp35. ‘sensitive$’:ti,ab36. ‘specificit$’:ti,ab37. ‘accuracy’:ti,ab38. ‘Predictive value’:ti,ab39. ‘predictive validity’:ti,ab40. OR/36 – 4341. AND/10, 35, 4442. Limit year: ‘2003



‘screening’:ti,ab‘risk factor’/exp

sk factor$’:ti,ab‘factor risk$’:ti,ab‘risk score$’:ti,ab‘score$ risk’:ti,ab‘assessment$ score$’:ti,ab‘decision making’/exp‘clinical judg?ment’:ti,ab‘clinical observation’/exp‘nursing assessment’/exp‘nursing assessment’:ti,ab‘observation$’:ti,ab

‘sensitivity and Specificity’/exp‘predictive validity’/exp‘Predictive Value’/exp‘sensitive$’:ti,ab‘specificit$’:ti,ab‘accuracy’:ti,ab‘Predictive value’:ti,ab‘predictive validity’:ti,ab

AND/10, 35, 44Limit year: ‘2003 – July 2012’ and language: ’English, Dutch, French’

KCE Report 193S

45851915

66212

60953220029323627285932383549113655

1046423253

112164313

131213399

250213

1909561632721101683

440419025

4447163

17847455568

3060724733

10573

KCE Report 193S


Date 25/7/12

Database The Library of the Cochrane Collaboration

Search Strategy


1. “Pressure ulcer” [MeSH]2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulc4. (bedsore* or bed5. ((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or lesion*)):ti,ab,kw6. OR/1 – 57. “risk assessment”[MeSH]8. (risk assess*):ti,ab,kw9. (risk-benefit assess*):ti,ab,kw10. (structured assessment):ti,ab,kw11. (unstructured assess*):ti,ab,kw12. (instrument*):ti,ab,kw13. (tool*):ti,ab,kw14. (scale*):ti,ab,kw15. (screening*):ti,ab,kw16. “risk factors”[MeSH]17. (risk factor):ti,ab,kw18. (risk score):ti,ab,kw19. (assessment score):ti,ab,kw20. “Judgment”[MeSH]21. “nursing assessment”[MeSH]22. (nurs* assess*):ti,ab,kw23. (clinical judg?ment):ti,ab,kw24. “Observation”[MeSH]25. (observation*):ti,ab,kw26. OR/7 – 2527. (braden*):ti,ab,kw28. (waterlow):ti,ab,kw29. (Norton):ti,ab,kw30. OR/27 – 2931. AND/6, 26, 3032. Limit year: ‘2003





benefit assess*):ti,ab,kwd assessment):ti,ab,kw

(unstructured assess*):ti,ab,kw(instrument*):ti,ab,kw

(scale*):ti,ab,kw(screening*):ti,ab,kw“risk factors”[MeSH](risk factor):ti,ab,kw(risk score):ti,ab,kw(assessment score):ti,ab,kw“Judgment”[MeSH]

assessment”[MeSH](nurs* assess*):ti,ab,kw(clinical judg?ment):ti,ab,kw“Observation”[MeSH](observation*):ti,ab,kw

(braden*):ti,ab,kw(waterlow):ti,ab,kw(Norton):ti,ab,kw

Limit year: ‘2003 – July 2012’

43

((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or lesion*)):ti,ab,kw

48934983433

3

11155960

23877132

187244

210136381

48545120291619028808

626111419

430493

4271260139

22962145412

271332694321

44

Date 25/7/12


Search Strategy


1. “Pressure ulcer” [MeSH]2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulcer* or damage*)):ti,ab,kw4. (bedsore* or bed5. ((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or lesion*)):ti,ab,kw6. OR/1 – 57. “risk assessmen8. (risk assess*):ti,ab,kw9. (risk-benefit assess*):ti,ab,kw10. (structured assessment):ti,ab,kw11. (unstructured assess*):ti,ab,kw12. (instrument*):ti,ab,kw13. (tool*):ti,ab,kw14. (scale*):ti,ab,kw15. (screening*):ti,ab,kw16. “risk factors”[MeSH]17. (risk factor):ti,ab,kw18. (risk score):ti,ab,kw19. (assessment score):ti,ab,kw20. “Judgment”[MeSH]21. “nursing assessment”[MeSH]22. (nurs* assess*):ti,ab,kw23. (clinical judg?ment):ti,ab,kw24. “Observation”[MeSH]25. (observation*):ti,ab,kw26. OR/7 – 2527. “Sensitivity and Specificity” [MeSH]28. (Sensitive*):ti,ab,kw29. (predictive value*):ti,ab,kw30. (predictive validity):ti,ab,kw31. (accuracy):ti,ab,kw32. (specificit*):ti,ab,kw33. OR/27 – 32





benefit assess*):ti,ab,kw(structured assessment):ti,ab,kw(unstructured assess*):ti,ab,kw(instrument*):ti,ab,kw

(scale*):ti,ab,kw(screening*):ti,ab,kw“risk factors”[MeSH](risk factor):ti,ab,kwrisk score):ti,ab,kw

(assessment score):ti,ab,kw“Judgment”[MeSH]“nursing assessment”[MeSH](nurs* assess*):ti,ab,kw(clinical judg?ment):ti,ab,kw“Observation”[MeSH](observation*):ti,ab,kw

“Sensitivity and Specificity” [MeSH]ab,kw

(predictive value*):ti,ab,kw(predictive validity):ti,ab,kw(accuracy):ti,ab,kw(specificit*):ti,ab,kw

KCE Report 193S

((friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or lesion*)):ti,ab,kw

48934983433

3

11155960

23877132

187244

210136381

48545120291619028808

626111419

430493

4271260139

2296214541213587

86367144312

680612760

KCE Report 193S

Date 25/7/12

34. AND/6, 26, 3335. Limit year: ‘2003

Note


Date 25/7/12


Search Strategy


1. MH “Pressure Ulcer”2. bedsore* OR bed3. pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*4. decubit*5. ((friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))6. OR/1 – 57. MH “Risk assessment”8. MH “Nursing assessment”9. “risk assessment$” or “asse10. “nurs$ assessment$ or “assessment$ nurs$”11. “risk-benefit assessment$”12. “structured assessment$” or “assessment$ structured”13. “unstructured assessment$”14. “instrument$”15. “tool$”16. “scale$”17. “screening”18. MH “risk factor”19. “risk factor$” or “factor$ risk20. “risk score$” or “score$ risk”21. “assessment score$”22. MH "Decision Making, Clinical"23. MH "Judgment"24. “clinical judg?ment”


Limit year: ‘2003 – July 2012’

interface)

MH “Pressure Ulcer”re* OR bed-sore*

pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*



benefit assessment$”“structured assessment$” or “assessment$ structured”“unstructured assessment$”

MH “risk factor”“risk factor$” or “factor$ risk”“risk score$” or “score$ risk”“assessment score$”MH "Decision Making, Clinical"MH "Judgment"“clinical judg?ment”

45

300852311

7732157

8512

482806

9370284261370030197

335335139

2310053012366388528525346913557

917273

13087

46

Date 25/7/12

25. “observation” or “observations”26. OR/7 – 2527. (MH "Braden Scale for Predicting Pressure Sore Risk")28. “Braden$”29. “Waterlow”30. “Norton”31. OR/27 – 3032. AND/6, 26, 3133. Limit year: ‘2003

Note

Date 25/7/12


Search Strategy


1. MH “Pressure Ulcer”2. bedsore* OR bed3. pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*4. decubit*5. ((friction or shear) and (so6. OR/1 – 57. MH “Risk assessment”8. MH “Nursing assessment”9. “risk assessment$” or “assessment$ risk”10. “nurs$ assessment$ or “assessment$ nurs$”11. “risk-benefit assessment$”12. “structured assessment$” or “asses13. “unstructured assessment$”14. “instrument$”15. “tool$”16. “scale$”17. “screening”18. MH “risk factor”


“observation” or “observations”

(MH "Braden Scale for Predicting Pressure Sore Risk")


interface)

MH “Pressure Ulcer”bedsore* OR bed-sore*pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*



benefit assessment$”“structured assessment$” or “assessment$ structured”“unstructured assessment$”

MH “risk factor”

KCE Report 193S

1883273

26519281311

596708115218976806451

7732157

8512

482806

9370284261370030197

335335139

2310053012366388

KCE Report 193S

Date 25/7/12

19. “risk factor$” or “factor$ risk”20. “risk score$” or “score$ risk”21. “assessment score$”22. MH "Decision Making, Clinical"23. MH "Judgment"24. “clinical judg?ment”25. “observation” or “observations”26. OR/7 – 2527. MH "Sensitivity and Specificity"28. MH “predictive validity”29. MH "Predictive Value of Tests”30. “sensitive$”31. “Specificit$”32. “accuracy”33. “Predictive value”34. “predictive validity”35. OR/27 – 3436. AND/6, 26, 3537. Limit year: ‘2003

Note


“risk factor$” or “factor$ risk”“risk score$” or “score$ risk”“assessment score$”MH "Decision Making, Clinical"MH "Judgment"“clinical judg?ment”“observation” or “observations”

MH "Sensitivity and Specificity"MH “predictive validity”MH "Predictive Value of Tests”

“Predictive value”“predictive validity”


47

528525346913557

917273

130871883273

2651928131123888

20561495816366292991549717626

262667412

265146

48


Figure 1 – Flow chart


KCE Report 193S

2.2.3. List of excluded studies

Reference

Anthony 2008

Anthony 2010

Balzer 2007

Bergstrom 2005

Bergquist 2001

Bolton 2007

Bolton 2007

Boyle 2001

Brown 2004

Cowan 2012

Defloor 2004

Fernandes 2008

Franks 2003

Gherghina 2011

Gray 2004

Gunningberg 1999

Hagisawa 1999

Harris 2010

He 2012

Iranmanesh 2012

Kim 2006

Kring 2007

Matuo 2008

Mertens 2008

Mertens 2010


Reason of exclusion

Cross-sectional study

No qualitative systematic review


Letter

Retrospective cohort study

Abstract; insufficient information

No prospective cohort study

Patients with PU at start included


Retrospective study


No predictive validity



Report




Systematic review of a subgroup; reference list screened


Case-control





49

Systematic review of a subgroup; reference list screened

50

Reference

Mitchell 2004

Montague 2009

Nonnemacher 2009

Okuwa 2005

Papanikolaou 2003

Poss 2010

Price 2005

Quesada 2009

Reynolds 2006

Saleh 2009

Schoonhoven 2005

Serpa 2011

Sharp 2006

Stausberg 2011

Stotts 2007

Suddaby 2006

Tannen 2010

Thompson 2005

Walsh 2011

Webster 2007

Willock 2009


Reason of exclusion




No risk assessment tool





No risk assessment tool





Comment


Not only pressure ulcers






KCE Report 193S

KCE Report 193S

2.3. Risk assessment tools

Risk assessment tool Population*

Braden scale

(Bergstrom 1987a3)

Skilled nursing facility patients

Norton scale

(Norton 19624)

Elderly hospitalized patients

Waterlow scale

(Waterlow 19855; revised

Waterlow, 20056)

Hospitalized patients on amedical or surgical ward

Cubbin-Jackson scale

(Cubbin 19917; revised

Jackson 19998)

Intensive care patients


Risk factors

Skilled nursing facility patients Sensory perception (completely limited – no impairment)

Moisture (constantly – rarely)

Activity (bedfast – walks frequently)

Mobility (completely immobile – no limitation)

Nutrition (very poor – excellent)

Friction and shear (problem – no apparent problem)

Elderly hospitalized patients Physical condition (very bad – good)

Mental condition (stupor – alert)

Activity (bedfast – ambulant)

Mobility (immobile – full)

Incontinent (urinary and faecal – not)

Hospitalized patients on aedical or surgical ward

Build/weight for height (average – below average)

Skin type visual risk area (healthy – broken/spots grade 2

Sex (male – female)

Age (14 – 81+)

Continence (complete/catheterised – urinary and faecalincontinence)

Mobility (fully – chair bound)

Malnutrition screening tool (MST) (nutrition score)

Special risk: tissue malnutrition (terminal cachexia, multiple/singleorgan failure, peripheral vascular disease, anaemia, smoking);neurological deficit (diabetes, MS, CVA, motor/sensory,major surgery/trauma (orthopaedic/spinal, on tablemedication (cytotoxic, long term/high dose steroids, antiinflammatory)

Intensive care patients Age

Weight

Skin condition of the whole body

Mental state

51

Scores

mpairment) Score ranges from 6to 23**

Score ranges from 5to 20**

broken/spots grade 2-4)

urinary and faecal

Special risk: tissue malnutrition (terminal cachexia, multiple/singleorgan failure, peripheral vascular disease, anaemia, smoking);neurological deficit (diabetes, MS, CVA, motor/sensory, paraplegia);major surgery/trauma (orthopaedic/spinal, on table ≥2hrs/6hrs); medication (cytotoxic, long term/high dose steroids, anti-

Score ranges from 2to 20+***

Score ranges from10 to 40**

52

Risk assessment tool Population*

Braden-Q scale

(Quigley 19969)

Paediatric patients

* Population for which the scale was originally developed** Lowest score indicates the highest risk*** Lowest score indicates the lowest risk


Risk factors

Mobility

Nutrition

Respiration

Incontinence

Hygiene

Hemodynamic state

Mobility (completely immobile – no limitations)

Activity (bedfast – all patients too young to ambulate or walksfrequently)

Sensory perception (completely limited – no impairment)

Moisture (constantly – rarely)

Friction and shear (problem – no apparent problem)

Nutrition (very poor – excellent)

Tissue perfusion and oxygenation (extremely compromisedexcellent)

Population for which the scale was originally developed

KCE Report 193S

Scores

all patients too young to ambulate or walks

no impairment)

Tissue perfusion and oxygenation (extremely compromised –

Scores ranges from 7to 28**

KCE Report 193S


2.4.1. Search strategy

A systematic review by Pancorbo-Hidalgo et al. (2006)adapted for this review. We updated the review by(2006)

10through a systematic search of multiple electronic databases.

This resulted in 1053 records: 369 in Medline (OVID),in Embase and 30 in the Cochrane Library. 308 duplicates were removed.Based on screening of title and/or abstract 687 records were excluded. 58records were reviewed in detail and an additionalexcluded. The remaining 16 studies were included in this review.additional study was retrieved trough screening of reference lists.

The review by Pancorbo-Hidalgo et al. (2006)10

which five were excluded because they didn’t meet the iour review:

One was excluded as it was a retrospective cohort study


Hidalgo et al. (2006)10

was identified andby Pancorbo-Hidalgo et al.

through a systematic search of multiple electronic databases.in Medline (OVID), 515 in CINAHL, 138

duplicates were removed.records were excluded. 58

and an additional 42 records wereexcluded. The remaining 16 studies were included in this review. One

l study was retrieved trough screening of reference lists.10

included 32 studies, ofwhich five were excluded because they didn’t meet the inclusion criteria of

One was excluded as it was a retrospective cohort study11

;

Another study was removed as it was written in Spanish

Three other studies were exclupressure ulcer at start of the study

The update of the Pancorbo-Hidalgo (2006)articles resulting in a final inclusion of, 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 ,

51 , 52 , 53 , 54 , 55 , 56 , 57

2.4.2. Clinical evidence

The systematic review by Pancorboreference for this review. All studies(2006)

10review, identified through

our review were reviewed in detail. Sensitivity and specificity of each scaleand cut-off score were re-calculated by using the raw data as presented inthe individual studies. Some adjustments were made to the PancorboHidalgo review

10.

53

Another study was removed as it was written in Spanish12

;

hree other studies were excluded as they included patients with apressure ulcer at start of the study

13-15.

Hidalgo (2006)10

review yielded 16 additionalfinal inclusion of 44 studies

3 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24

, 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 ,

Pancorbo-Hidalgo (2006)10

was used as areference for this review. All studies included in the Pancorbo-Hidalgo

through the update and meeting the criteria ofwere reviewed in detail. Sensitivity and specificity of each scale

calculated by using the raw data as presented inthe individual studies. Some adjustments were made to the Pancorbo-

54



Study Risk tool

Outcome

Andersen 198216

Andersen scale

Pressure ulcerdevelopment

Anthony 200317

Waterlow scale


Barnes 199318

Braden scale


Bergstrom 1987a3(1)

(2)

Braden scale


Bergstrom 1987b19

Braden scale


Bergstrom 199820

(1)

(2)

(3)

Braden scale


Braden 199421

Braden scale



Population Length of follow-up Number ofpatients

Pressure ulcer

Patients in an acuteobservation ward

Maximum three months 3398

Pressure ulcer

Hospitalized patients Not reported

(median days in hospitaltwo days for PU freepatients versus 22 daysfor PU patients)

45735

Pressure ulcer

Nursing home patients Maximum two weeks 361

Pressure ulcer

Medical/surgical patients

Medical/surgical patients(unit with higher acuitylevels and longer expectedlength of stay than group 1)

Maximum six weeks

Maximum 12 weeks

99

100

Pressure ulcer

Intensive care patients Maximum two weeks 60

Pressure ulcer

Patients in a tertiary carehospital

Patients in a VeteranMedical Centre

Patients in a skilled nursingfacility

48-72 hours andmaximum 11 days

306

282

255

Pressure ulcer

Patients in a skilled nursingfacility

Maximum four weeks 102

KCE Report 193S

Number ofpatients

Number of events

398 40

45735 203

361 22

100

7

9

24

306

282

255

26

21

61

102 28

KCE Report 193S

Study Risk tool

Outcome

Capobianco 199622

Braden scale


Chan 200923

(1) Braden scale

(2) Modified Braden scale


Compton 200824

Waterlow scale

Pressure ulcer grade II+development

Curley 200325

Braden-Q


de Souza 201026

Braden scale


Edwards 199527

Waterlow scale


Feuchtinger 200728

(1) Braden scale

(2) Modified Norton scale

(3) 4-factor model

PU development

Goodridge 199829

Braden scale


Halfens 200030

(1) Braden scale

(2) Extended Braden



Pressure ulcer

Medical/surgical patients Maximum two weeks 50


ure ulcer

Orthopaedic patients Maximum nine days 197

grade II+

Intensive care patients Maximum 13 days 698

Pressure ulcer

Paediatric intensive carepatients

Maximum 10 days 322

Pressure ulcer

Patients in a long-term carefacility; aged 60 years andolder; a Braden score < 19

Three months 233

Pressure ulcer

Home care patients Eight weeks 31


Intensive care patients(cardiac surgery patients)

Maximum four days 53

Pressure ulcer

Patients from the medicaland geriatric unit of atertiary care hospital andlong-term care facility;aged 65 years and older

Maximum three months 330

(2) Extended Braden

Surgical, neurological,orthopaedic and internal

Not reported 320

55

Number ofpatients

Number of events

14

197 18

698 121

322 86

233 44

2

26

330 32

320 186

56

Study Risk tool

Outcome

scale

Pressure ulcerdevelopment and/or useof preventive measures

Hatanaka 200831

Braden scale


Jalali 200532

(1) Braden scale

(2) Norton scale

(3) Waterlow scale

(4) Gosnell scale


Kim 200933

(1) Braden scale

(2) Cubbin-Jackson scale

(3) Song and Choi scale


Kwong 200534

(1) Braden scale


(3) Norton scale


Langemo 199135

(1)

(2)

Braden scale


Lewicki 200036

Braden scale


Lincoln 198637

Norton scale

Pressure ulcer



ssure ulcerdevelopment and/or useof preventive measures

medicine patients

Pressure ulcer

Bedridden hospitalizedpatients with a respiratorydisorder

Maximum 79 days 149

Pressure ulcer

Neurological, intensivecare, orthopaedic andmedical care patients.

Maximum 14 days 230

Jackson scale

(3) Song and Choi scale

Pressure ulcer

Surgical intensive carepatients

Maximum 90 days 219


Pressure ulcer

Acute care patients Maximum 21 days 429

Pressure ulcer

Hospitalized patients

Patients in a long-term carefacility

Maximum 16 days

Maximum 31 days

74

25

Pressure ulcer

Elective cardiac surgerypatients

Five days 337

Pressure ulcer

Medical/surgical patients;aged 65 years and older

Maximum 26 days 36

KCE Report 193S

Number ofpatients

Number of events

149 38

230 74

219 40

429 9

11

7

337 7

5

KCE Report 193S

Study Risk tool

Outcome

development

Lindgren 200238

Risk AssessmentPressure Sore scale(RAPS)


Lothian 198939

Pressure Sore PredictionScore (PSPS)


Lyder 199940

Braden scale


Ongoma 200641

(1) Sunderland PressureSore Risk Calculator(modified Cubbin-Jacksonscale)



Page 201142

The Northern HospitalPressure UlcerPrevention Plan (TNHPUPP)


Pang 199843

(1) Braden scale

(2) Norton scale

(3) Waterlow scale




Risk AssessmentPressure Sore scale

Pressure ulcer

Acute care patients Maximum 12 weeks 488

Pressure Sore Prediction

Pressure ulcer

Orthopaedic patients Maximum three weeks 1244

Pressure ulcer

Patients from a tertiaryhospital

Not reported 177

(1) Sunderland PressureSore Risk Calculator

Jackson


Pressure ulcer

Intensive care patients Three weeks 66

The Northern HospitalPressure UlcerPrevention Plan (TNH-

Pressure ulcer

Acute care patients Not reported 165

Pressure ulcer

Medical and orthopaedicpatients

Maximum 14 days 106

57

Number ofpatients

Number of events

488 54

1244 53

177 24

25

165 7

106 21

58

Study Risk tool

Outcome

Perneger 200244

(1) Braden scale

(2) Norton scale

(3) Fragmment scale


Ramundo 199545

Braden scale

Pressure ulcer grade IIdevelopment

Salvadalena 199246

(1) Braden scale



Schoonhoven 200247

(1) Braden scale

(2) Norton scale

(3) Waterlow scale

Pressure uldevelopment

Seongsook 200448

(1) Braden scale

(2) Cubbin-Jackson scale

(3) Douglas scale


Serpa 200949

Waterlow scale


Serpa 201158

Braden scale


Smith 198950

(1) Norton scale

(2) Waterlow scale



(3) Fragmment scale

Pressure ulcer

Internal medicine,abdominal surgery,orthopaedic, neurosurgery,intensive care, anddermatological patients

Maximum three weeks 1190

grade II

Home care patients Maximum four weeks 48


Pressure ulcer

Acute medical carepatients

Maximum six months 99

Pressure ulcer

Surgical, internal care,neurological, and geriatricpatients

Maximum 12 weeks 1229

Jackson scale

Pressure ulcer

Internal, surgical andneurological intensive carepatients

Study duration of 1 year 112

Pressure ulcer

Hospitalized patients; aBraden score < 19 and/or aWaterlow score > 15

Maximum six days 98

Pressure ulcer

Intensive care patients; aBraden score < 19

Maximum six days 72

Hospitalized patients Nor reported 101

KCE Report 193S

Number ofpatients

Number of events

1190 170

7

20

1229 135

112 35

7

8

101 30

KCE Report 193S

Study Risk tool

Outcome


Stotts 198851

Norton scale


Suriadi 200652

Braden scale

PU development

Suriadi 200853

Suriadi and Sanada scale(SS)


Towey 198854

Knoll scale


VandenBosch 199655

(1) Braden scale



Wai-Han 199756

(1) Norton scale

(2) Waterlow scale


Weststrate 199857

Waterlow scale

Pressure ulcer grade II+development



Pressure ulcer

Pressure ulcer

Cardiovascular surgery andneurosurgery patients

Maximum three weeks 387

Intensive care patients Maximum 22 days 105

Suriadi and Sanada scale

ressure ulcer

Intensive care patients Not reported 253

Pressure ulcer

Patients in a long-term carefacility; aged 65 years andolder

Fourteen and 38 days 60


Pressure ulcer

General and intensive carepatients and rehabilitationinpatients

Maximum two weeks 103

Pressure ulcer

Geriatric hospitalizedpatients; aged 70 yearsand older

Four weeks 185

grade II+

Surgical intensive carepatients

Maximum of 183 days 594

59

Number ofpatients

Number of events

387 67

105 35

253 47

28

103 29

185 8

594 47

60

2.4.4. Median Area under the ROC curve

Table 3 – Braden scale

Study Risk ofbias

Inconsistency

All populations

9 (Schoonhoven2002; Perneger2002; Seongsook2004**; Suriadi2006; Hatanaka2007; Chan 2009;Kim 2009; de Souza2010**; Serpa2011)

Veryserious

1Serious

2

General population

5 (Schoonhoven2002; Perneger2002; Hatanaka2007; Chan 2009;de Souza 2010**)

Veryserious

1Serious

2


4 (Seongsook2004**; Suriadi2006; Kim 2009;Serpa 2011)

Veryserious

1No seriousinconsistency

* 90.0-100.0: very good discrimination; 80.0-89.0: good** Unclear if patients with a pressure ulcer at start of the study were included.1 All studies had high to very high risks of bias (see quality2 Wide variation in AUC across the studies3 Low event rates (< 100), except for the study of Schoonhoven 2002 and Perneger 2002

Table 4 – Modified Braden scale

Study Risk of bias Inconsistency


Inconsistency Indirectness Imprecision Numberofpatients(range)

Numberof events(range)

MedianAUC (%)(range)

No seriousindirectness

Serious3

72-1229 8-170 74.0

(55.088.0)


Serious3

149-1229

38-170 68.0

(55.081.0)

inconsistencyNo seriousindirectness

Serious3

72-219 8-40 79.0

(71.088.0)

89.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0** Unclear if patients with a pressure ulcer at start of the study were included.1 All studies had high to very high risks of bias (see quality table)

, except for the study of Schoonhoven 2002 and Perneger 2002

Inconsistency Indirectness Imprecision Number Number Median

KCE Report 193S

MedianAUC (%)(range)

Acceptabilityof values*

Quality

(55.0 –88.0)

Fairdiscrimination

VERY LOW

(55.0 –81.0)

Poordiscrimination

VERY LOW

(71.0 –88.0)

Fairdiscrimination

VERY LOW

69.0: poor discrimination;50.0-59.0: fail to discriminate

Median Acceptability Quality

KCE Report 193S

General population

1 (Chan 2009) Serious1

No seriousinconsistency

* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.01 Study had high risks of bias (see quality table)2 Low event rates (< 100)

Table 5 – Braden-Q scale


Paediatric ICU

1 (Curley 2003) Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.01 Study had high risks of bias (see quality table)2 Low event rates (< 100)

Table 6 – Norton scale

Study Risk ofbias

Inconsistency

General population

2 (Schoonhoven2002; Perneger2002)

Serious1

Serious2

* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.01 Studies had high risks of bias (see quality table)


ofpatients

of events AUC

(%) (range)

No serious No seriousindirectness

Serious2

197 18 74.0

(95% CI:63.0

89.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0

Inconsistency Indirectness Imprecision Numberofpatients

Numberof events

MedianAUC

(%) (range)


Serious2

322 86 83.0

(95% CI:76.0



Numberofevents(range)

MedianAUC

(%) (range)


No seriousimprecision

1190-1229

135-170 65.0

(56.074.0)


61

(%) (range)

of values*

74.0

(95% CI:63.0-84.0)

Fairdiscrimination

LOW


Median

(%) (range)


Quality

83.0

(95% CI:6.0-91.0)

Gooddiscrimination

LOW


MedianAUC

3

(%) (range)


Quality

65.0

(56.0 –74.0)

Poordiscrimination

LOW


62

2 Wide variation in AUC across the two studies3 Mean AUC: only two studies

Table 7 – Waterlow scale

Study Risk ofbias

Inconsistency

All populations

4 (Schoonhoven2002; Anthony2003; Compton2008; Serpa 2009)

Veryserious

1Serious

2

General population

3 (Schoonhoven2002; Anthony2003; Serpa 2009)

Veryserious

1Serious

2


1 (Compton 2008) Veryserious


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.01 The studies had high to very high risks of bias (see quality table)2 Wide variation in AUC across the studies3 Very low event rates (< 100) for the study of Serpa 2009. The other studies had an event rate > 100




MedianAUC

(%) (range)


Serious3

98-45735

7-203 60.0

(54.090.0)


Serious3

98-45735

7-203 61.0

(54.090.0)




698 121 59.0

(95% CI:54.0

od discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.01 The studies had high to very high risks of bias (see quality table)

0) for the study of Serpa 2009. The other studies had an event rate > 100

KCE Report 193S

MedianAUC

(%) (range)


Quality

(54.0 –90.0)

Poordiscrimination

VERY LOW

(54.0 –90.0)

Poordiscrimination

VERY LOW

59.0

(95% CI:54.0-65.0)

Faildiscrimination

LOW


KCE Report 193S

Table 8 – Cubbin-Jackson scale



2 (Kim 2009;Seongsook2004**)

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.0** Unclear if patients with a pressure ulcer at start of the study were included.1 The studies had high risks of bias (see quality table)2 Low event rates (< 100); no confidence intervals3 Mean AUC: only two studies

Table 9 – Douglas scale

Study Risk ofbias

Inconsistency


1 (Seongsook2004**)

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination; 70.0** Unclear if patients with a pressure ulcer at start of the study were included.1 Study had high risks of bias (see quality table)2 Low event rates (< 100); no confidence interval




MedianAUC

3

(%) (range)



Serious2

112-219 35-40 87.0

(83.0

d discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0** Unclear if patients with a pressure ulcer at start of the study were included.


Numberof events

MedianAUC

(%) (range)

ensive care patients



Serious2

112 170 79.0

89.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0** Unclear if patients with a pressure ulcer at start of the study were included.

63

Median3

(%) (range)


Quality

87.0

(83.0-90.0)

Gooddiscrimination

LOW


MedianAUC

(%) (range)


Quality

79.0 Fairdiscrimination

LOW


64

Table 10 – Fragmment scale


General population and intensive care patients

1 (Perneger 2002) Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 Study had high risks of bias (see quality table)

Table 11 – Song and Choi scale



1 (Kim 2009) Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 Study had high risks of bias (see quality table)2 Low event rates (< 100); no confidence interval

Table 12 – The Northern Hospital Pressure Ulcer Prevention Plan (TNH



1 (Page 2011) Veryserious


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 Study had very high risks of bias (see quality table)2 Very low event rates (< 100)



Numberof events

MedianAUC

(%) (range)





1190 170 79.0

(95% CI:75.0

.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0


Numberof events

MedianAUC

(%) (range)


Serious2

219 40 89.0


The Northern Hospital Pressure Ulcer Prevention Plan (TNH-PUPP)


Numberof events

MedianAUC

(%) (range)


Serious2

165 7 90.0

(95% CI:82.0


KCE Report 193S

Median

(%) (range)


Quality

79.0

(95% CI:75.0-82.0)

Fairdiscrimination

LOW

.0-59.0: fail to discriminate

Median

(%) (range)


Quality

89.0 Gooddiscrimination

LOW


Median

(%) (range)


Quality

90.0

(95% CI:-99.0)

Perfectdiscrimination

VERY LOW


KCE Report 193S

2.4.5. AUC within studies

Table 13 – Schoonhoven 2002

Study Risk ofbias


General population

Braden scale

Norton scale

Waterlow scale

Serious1



* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had high risks of bias (see quality table)

Table 14 – Perneger 2002

Study Risk ofbias

Inconsistency


Braden scale

Norton scale

Fragmment scale

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had high risks of bias (see quality table)


Indirectness Imprecision Numberofpatients

Numberof events

AUC (%)(95% CI)



1129 135 55.0 (95%CI 49.0

56.0 (95%CI 51.0

61.0 (95%CI 56.0


Indirectness Imprecision Numberofpatients

Number ofevents

AUC (%)(95% CI)




1190 170 74.0 (95%CI 70.0

78.0)

74.0 (95%CI 70.0

78.0)

79.0 (95%CI 75.0

82.0)


65

AUC (%)(95% CI)


Quality

55.0 (95%CI 49.0-

60.0)

56.0 (95%CI 51.0-

61.0)

61.0 (95%CI 56.0-

66.0)

Fail

Fail

Poor

MODERATE


AUC (%)(95% CI)


Quality

74.0 (95%CI 70.0-

78.0)

74.0 (95%CI 70.0-

78.0)

79.0 (95%CI 75.0-

82.0)

Fair

Fair

Fair

MODERATE


66

Table 15 – Seongsook 2004**

Study Risk ofbias

Inconsistency


Braden scale

Cubbin-Jacksonscale

Douglas scale

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;** Unclear if patients with a pressure ulcer at start of the study were included.1 The study had high risks of bias (see quality table)2 Low event rates (< 100); no confidence interval

Table 16 – Chan 2009

Study Risk ofbias

Inconsistency

General population

Braden scale

Modified Bradenscale

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had high risks of bias (see quality table)2 Low event rates (< 100); wide confidence interval



Number ofevents

AUC (%)(95% CI)


Serious2

112 35

.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0** Unclear if patients with a pressure ulcer at start of the study were included.

Indirectness Imprecision Number ofpatients

Number ofevents

AUC (%)(95% CI)


Very serious2

197 18 73.0 (95% CI63.0

68.0 (95% CI51.0

discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.0

KCE Report 193S

AUC (%)(95% CI)


Quality

71.0

83.0

79.0

Fair

Good

Fair

LOW


AUC (%)(95% CI)


Quality

73.0 (95% CI63.0-84.0)

68.0 (95% CI51.0-79.0)

Fair

Poor

VERYLOW


KCE Report 193S

Table 17 – Kim 2009

Study Risk ofbias

Inconsistency


Braden scale

Cubbin-Jacksonscale

Song and Choiscale

Serious1


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had high risks of bias (see quality table)2 Low event rates (< 100); no confidence interval

Table 18 – Serpa 2009

Study Risk ofbias

Inconsistency

General population

Waterlow scale(48 hours)

Waterlow scale(4 days)

Waterlow scale(6 days)

Veryserious


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had very high risks of bias (see quality table)2 Very low event rates (< 100); very wide confidence intervals



Number ofevents

AUC (%)(95% CI)


Serious2

219 40



Number ofevents

AUC (%)(95% CI)


Very serious2

98 7 64.0 (95% CI35.0

59.0 (95% CI34.0

54.0 (95% CI35.0

.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.01 The study had very high risks of bias (see quality table)

ow event rates (< 100); very wide confidence intervals

67

AUC (%)(95% CI)


Quality

88.0

91.0

89.0

Good

Excellent

Good

LOW


AUC (%)(95% CI)


Quality

64.0 (95% CI35.0-93.0)

59.0 (95% CI34.0-83.0)

54.0 (95% CI35.0-74.0)

Poor

Fail

Poor

VERYLOW


68


Study Risk ofbias

Inconsistency


Braden scale(48 hours)

Braden scale (4days)

Braden scale (6days)

Veryserious


* 90.0-100.0: perfect discrimination; 80.0-89.0: good discrimination;1 The study had very high risks of bias (see quality table)2 Very ow event rates (< 100); very wide confidence intervals

2.4.6. Predictive ability


Study

Follow-up < 1 week – all stages – general population

2 (Bergstrom 1998a; Braden 1994)

a


a


a

Follow-up < 1 week – all stages – ICU

1 (Serpa 2011 – 48 hours)

1 (Serpa 2011 – 4 days and 6 days)

1 (Feuchtinger 2007)

Follow-up > 1 week – all stages – general population



Number ofevents

AUC (%)(95% CI)


Very serious2

72 8 79.0 (95% CI29.0

79.0 (95% CI27.0

80.0 (95% CI28.0

.0: good discrimination; 70.0-79.0: fair discrimination; 60.0-69.0: poor discrimination;50.01 The study had very high risks of bias (see quality table)

0); very wide confidence intervals

Cut-off score* Median sensitivity**

(range)

general population

≤ 17 59.0

(50.0-78.0)

≤ 18 70.0

(60.0-88.0)

≤ 19 83.5

(51.0-100.0)

≤ 12 87.5

≤ 13 75.0

(75.0-75.0)

≤ 16 76.9

general population

KCE Report 193S

AUC (%)(95% CI)


Quality

79.0 (95% CI29.0-100.0)

79.0 (95% CI27.0-100.0)

80.0 (95% CI28.0-100.0)

Fair

Fair

Good

VERYLOW


Specificity**‡

(range)

70.5

(52.0-81.0)

58.0

(48.0-81.0)

60.5

(42.0-73.0)

64.1

82.1

(81.3-82.8)

29.6

KCE Report 193S

Study

10 (Bergstrom 1987a; Bergstrom1998

a; Braden 1994

a; Capobianco

1996; Chan 2009; Goodridge 1998;Langemo 1991; Lyder 1999; Pang1998; Salvadalena 1992)


a; Braden 1994

a; Capobianco

1996;Salvadalena 1992)


a; Braden 1994

a; Capobianco

1996; Salvadalena 1992)

Follow-up > 1 week – all stages – ICU

1 (Braden 1994)

2 (Braden 1994a; Seongsook

2004a***)

1 (Braden 1994)

Follow-up > 1 week – stage 2+ – general population

1 (Ramundo 1995)e

1 (Ramundo 1995)e

1 (Ramundo 1995)e

* The reported thresholds are these with the highest values f** Percentage*** Unclear if patients with a pressure ulcer at start of the study were included.‡ Specificity corresponding to the median sensitivitya Study of which sensitivity and specificity are presented.b Sensitivity analysis without studies with < 10 events (Bergstrom 1987a and Langemo 1991) revealed a median sensitivity of 78.6 (r(range: 14.0-100.0)c Sensitivity analysis without study with < 10 events (Bergstrom 1987



(range)

≤ 18 79.5b

(46.2-100.0)

≤ 19 86.3c

(71.4-100.0)

≤ 20 93.2d

(43.2-100.0)

≤ 15 75.0

≤ 16 90.2

(83.3-97.1)

≤ 17 87.5

general population

≤ 17 42.9

≤ 18 100.0

≤ 19 100.0

* The reported thresholds are these with the highest values for median sensitivity and specificity

** Unclear if patients with a pressure ulcer at start of the study were included.

Study of which sensitivity and specificity are presented.itivity analysis without studies with < 10 events (Bergstrom 1987a and Langemo 1991) revealed a median sensitivity of 78.6 (r

c Sensitivity analysis without study with < 10 events (Bergstrom 1987a) revealed a median sensitivity of 85.7 (range: 71.4-100.0) and a corresponding 59.1 (range: 43.0

69

Specificity**‡

(range)

73.6b

(14.0-100.0)

67.5c

(42.9-77.8)

53.5d

(31.6-66.7)

66.7

45.0

(26.0-63.9)

50.0

63.4

34.1

22.0

itivity analysis without studies with < 10 events (Bergstrom 1987a and Langemo 1991) revealed a median sensitivity of 78.6 (r ange: 46.2-90.5) and a corresponding 74.3

100.0) and a corresponding 59.1 (range: 43.0-77.8)

70

d Sensitivity analysis without study with < 10 events (Bergstrom 1987a) revealed a median sensitivity of 91.7 (range: 43.2e The study of Ramundo 1995 had 7 events


Study

Follow-up > 1 week – all stages – paediatric ICU patients

1 (Curley 2003)

1 (Curley 2003)

1 (Curley 2003)

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage‡ Specificity corresponding to the median sensitivity


Study


4 (Kwong 2005; Lincoln 1986; Stotts1998

a***; Wai-Han 1997

a)

1 (Schoonhoven 2002)c

2 (Pang 1998a; Smith 1989

a***)

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage*** Unclear if patients with a pressure ulcer at start of the study were included.‡ Specificity corresponding to the median sensitivitya Study of which sensitivity and specificity are presented.b Sensitivity analysis without studies with < 10 events (Kwong 2005 and Lincoln 1986) revealed a median sensitivit(range: 66.7-94.4)c The study of Schoonhoven 2002 had 135 events


d Sensitivity analysis without study with < 10 events (Bergstrom 1987a) revealed a median sensitivity of 91.7 (range: 43.2 -100.0) and a corresponding 4


paediatric ICU patients

≤ 15 75.6

≤ 16 88.4

≤ 17 91.9



general population

≤ 14 45.7b

(0.0-88.9)

≤ 15 45.9

≤ 16 70.5

(60.0-81.0)


f the study were included.

Study of which sensitivity and specificity are presented.b Sensitivity analysis without studies with < 10 events (Kwong 2005 and Lincoln 1986) revealed a median sensitivit y of 45.7 (range: 16.4

KCE Report 193S

100.0) and a corresponding 40.1 (range: 31.6-66.7)

Specificity**‡

67.8

58.1

44.1

Specificity**‡

80.6b

(61.0-94.4)

60.3

44.9

(31.0-58.8)

y of 45.7 (range: 16.4-75.0) and a corresponding 80.6

KCE Report 193S


Study

Follow-up < 1 week – all stages – general populat

1 (Serpa 2009 – 48 hours)b

1 (Serpa 2009 – 4 days and 6 days)b


3 (Anthony 2003a; Schoonhoven

2002; Wai-Han 1997)

1 (Anthony 2003)d

2 (Pang 1998a; Smith 1989

a***)

Follow-up < 1 week – stage 2+ – ICU

1 (Weststrate 1998)

* The reported thresholds are these with the highest value** Percentage*** Unclear if patients with a pressure ulcer at start of the study were included.‡ Specificity corresponding to the median sensitivitya Study of which sensitivity and specificity are presented.b The study of Serpa 2009 had 7 eventsc Sensitivity analysis with only studies with > 100 events (Antony 2003 and Schoonhoven 2002) revealed a median sensitivity o53.8 (range: 22.4-85.2)d The study of Antony 2003 had 203 events


Study


1 (Seongsook 2004***)

1 (Kim 2009)

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage*** Unclear if patients with a pressure ulcer at start of the study were included.‡ Specificity corresponding to the median sensitivity



general population

≥ 17 71.4

≥ 20 85.7

(85.7-85.7)

general population

≥ 10 87.5c

(82.3-89.6)

≥ 15 48.8

≥ 16 84.3

(73.3-95.2)

≥ 15 80.9



Study of which sensitivity and specificity are presented.

c Sensitivity analysis with only studies with > 100 events (Antony 2003 and Schoonhoven 2002) revealed a median sensitivity o f 86.0 (range: 82.3


≤ 24 88.6

≤ 28 95.0

he highest values for median sensitivity and specificity


71

Specificity**‡

67.0

36.9

(33.0-40.7)

28.2c

(22.4-85.2)

94.4

40.8

(38.0-43.5)

28.5

f 86.0 (range: 82.3-89.6) and a corresponding

Specificity**‡

61.0

81.6

72


Study


1 (Perneger 2002)a

1 (Perneger 2002)a

1 (Perneger 2002)a

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage‡ Specificity corresponding to the median sensitivitya The study of Perneger 2002 had 170 events


Study


1 (Seongsook 2004***)

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage*** Unclear if patients with a pressure ulcer at start of the study were included.‡ Specificity corresponding to the median sensitivity

Table 27 – The Northern Hospital Pressure Ulcer Prevention Pl

Study


1 (Page 2011)a

1 (Page 2011)a

1 (Page 2011)a

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage‡ Specificity corresponding to the median sensitivitya The study of Page 2011 had a 7 events



general population

≤ 1 78.7

≤ 2 76.7

≤ 3 62.1



≤ 18 100.0

lues for median sensitivity and specificity




general population

≥ 2 85.7

≥ 3 71.4

≥ 4 71.4


KCE Report 193S

Specificity**‡

53.5

71.9

85.0

Specificity**‡

18.2

Specificity**‡

62.0

81.0

88.0

KCE Report 193S


Study


1 (Kim 2009)

* The reported thresholds are these with the highest values for media** Percentage‡ Specificity corresponding to the median sensitivity

Table 29 – Suriadi and Sanada scale

Study


1 (Suriadi 2008)

1 (Suriadi 2008)

1 (Suriadi 2008)

* The reported thresholds are these with the highest values for median sensitivity and specificity** Percentage‡ Specificity corresponding to the median sensitivity

Table 30 – Clinical judgement

Study


2 (Salvadalena 1992a; VandenBosch

1996a)

** Percentage‡ Specificity corresponding to the median sensitivitya Study of which sensitivity and specificity are presented.



≤ 21 95.0



≥ 3 97.2

≥ 4 80.6

≥ 5 72.2


Cut-off score Median sensitivity**

Yes/no 50.9

(50.0-51.7)

Study of which sensitivity and specificity are presented.

73

Specificity**‡

69.3

Specificity**‡

53.0

82.9

86.7

Specificity**‡

68.9

(58.1-79.7)

74

2.4.7. Quality of the studies

Table 31 – Quality of the studies

Study Selection bias*

Andersen 1982 Low

Anthony 2003 High

Barnes 1993 High

Bergstrom 1987a Low

Bergstrom 1987b Low

Bergstrom 1998 Low

Braden 1994 Low

Capobianco 1996 Low

Chan 2009 High

Compton 2008 Very high

Curley 2003 Low

de Souza 2010 Very high

Edwards 1995 Low

Feuchtinger 2007 Low

Goodridge 1998 High

Halfens 2000 High

Hatanaka 2008 High

Jalali 2005 High

Kim 2009 High

Kwong 2005 High

Langemo 1991 High

Lewicki 2000 High

Lincoln 1986 High

Lindgren 2002 High

Lothian 1989 Very high


Risk tool bias** Outcome bias***

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High Low

High High

Very high Low

High Low

High Low

High Low

High Low

High Low

High Low

Very high Very high

KCE Report 193S

Analysis bias****

High

High

High

Very high

High

High

High

High

High

High

High

High

Very high

High

High

High

High

High

High

Very high

Very high

Very high

Very high

High

High

KCE Report 193S

Study Selection bias*

Lyder 1999 High

Ongoma 2006 High

Page 2011 Very high

Pang 1998 High

Perneger 2002 High

Ramundo 1995 High

Salvadalena 1992 High

Schoonhoven 2002 High

Seongsook 2004 Very high

Serpa 2009 High

Serpa 2011 High

Smith 1989 Very high

Stotts 1988 Very high

Suriadi 2006 High

Suriadi 2008 High

Towey 1988 Low

VandenBosch 1996 High

Wai-Han 1997 High

Weststrate 1998 High

* inappropriate patient enrolment, inappropriate study design,** unclear definition and measurement of index test, absence*** unclear definition and measurement of reference test,**** no use of time to event analysis, number of events


Risk tool bias** Outcome bias***

High High

Very high Very high

High High

High Low

High Low

High Low

High Low

High Low

High Low

High High

High High

High High

High Low

High Low

High Low

High High

High Low

High High

High High

study design, not representative populationabsence of imputation technique or unclear description of exclusion, inadequate threshold

reference test, inappropriate durationuse of time to event analysis, number of events < 100, reason for missing data not reported

75

Analysis bias****

High

High

Very high

High

High

Very high

High

Low

High

Very high

Very high

High

High

High

High

High

High

High

High

adequate threshold

76

2.5. Incidence and predictive ability of risk assessment scales

2.5.1. Sensitivity and specificity of risk assessment tools


Study Time point

Barnes 1993 2 weeks

Braden 1994 48-72 hours‡

4 weeks

Bergstrom 1987a (a) 6 weeks


Incidence and predictive ability of risk assessment scales

Sensitivity and specificity of risk assessment tools

Incidence* Cut-off score Sensitivity*

6.1 ≤ 16 72.7

NR

27.5

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

2

43.0

46.0

61.0

79.0

93.0

96.0

21.4

32.1

50.0

57.1

78.6

85.7

92.9

7.1 ≤ 9

≤ 10

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

≤ 23

14.3

14.3

28.6

42.9

71.4

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

KCE Report 193S

Sensitivity* Specificity*

72.7 90.6

9.0

43.0

46.0

61.0

79.0

93.0

96.0

21.4

32.1

50.0

57.1

78.6

85.7

92.9

97.0

95.0

84.0

78.0

68.0

51.0

35.0

95.9

94.6

89.2

85.1

74.3

59.3

43.2

14.3

14.3

28.6

42.9

71.4

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

98.9

98.9

98.9

94.6

90.2

88.0

82.6

73.9

65.2

50.0

35.9

0.0

KCE Report 193S

Study Time point

(b) 12 weeks

Bergstrom 1987 2 weeks

Bergstrom 1998 (c) 48-72 hours‡



9.0 ≤ 8

≤ 9

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

≤ 23

11.1

11.1

22.2

44.4

55.6

66.7

77.8

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

40.0 ≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

8.3

8.3

16.7

33.3

58.3

70.8

75.0

83.3

87.5

91.7

91.7

95.8

95.8

100.0

NR ≤ 9

≤ 10

4.0

12.0

77


11.1

11.1

22.2

44.4

55.6

66.7

77.8

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

95.6

91.2

89.0

86.8

83.5

78.0

73.

63.7

60.4

50.5

42.9

31.9

26.4

9.9

0.0

8.3

8.3

16.7

33.3

58.3

70.8

75.0

83.3

87.5

91.7

91.7

95.8

95.8

100.0

100.0

97.2

91.7

88.9

77.8

75.0

66.7

63.9

50.0

38.9

25.0

13.9

5.6

0.0

4.0

12.0

100.0

100.0

78

Study Time point

(d)

11 days

48-72 hours‡



8.5

NR

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

19.0

31.0

38.0

38.0

46.0

58.0

62.0

88.0

100.0

100.0

100.0

7.7

11.5

11.5

15.4

15.4

15.4

23.1

30.8

38.5

46.2

100.0

100.0

100.0

0.0

10.0

10.0

10.0

10.0

20.0

20.0

KCE Report 193S


19.0

31.0

38.0

38.0

46.0

58.0

62.0

88.0

100.0

100.0

100.0

7.7

11.5

11.5

15.4

15.4

15.4

23.1

30.8

38.5

46.2

100.0

100.0

100.0

0.0

10.0

10.0

10.0

10.0

20.0

20.0

99.0

99.0

98.0

95.0

90.0

84.0

76.0

68.0

59.0

40.0

23.0

100.0

100.0

98.9

98.9

97.9

97.1

92.9

88.9

83.9

68.9

58.9

40.0

22.9

99.0

99.0

98.0

98.0

97.0

96.0

94.0

KCE Report 193S

Study Time point

(e)

11 days

48-72 hours‡



7.4

NR

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

30.0

50.0

60.0

80.0

80.0

90.0

0.0

0.0

0.0

0.0

28.6

28.6

52.4

52.4

61.9

71.4

71.4

90.5

90.5

0.0

2.0

2.0

5.0

13.0

23.0

33.0

41.0

56.0

72.0

67.0

83.0

79


30.0

50.0

60.0

80.0

80.0

90.0

0.0

0.0

0.0

0.0

28.6

28.6

52.4

52.4

61.9

71.4

71.4

90.5

90.5

0.0

2.0

2.0

5.0

13.0

23.0

33.0

41.0

56.0

72.0

67.0

83.0

90.0

85.0

81.0

73.0

69.0

41.0

99.2

99.2

99.2

98.1

98.1

96.9

93.9

92.0

87.0

78.9

70.9

50.2

32.2

99.0

99.0

99.0

99.0

99.0

97.0

93.0

88.0

81.0

68.0

48.0

34.0

80

Study Time point

11 days

Capobianco 1996 2 weeks

Chan 2009 9 days

de Souza 2010 (f)

(g)

3 months

3 months

Feuchtinger 2007 4 days



23.9

≤ 21

≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

97.0

0.0

19.7

29.5

8.2

13.1

19.7

31.1

49.2

60.7

80.3

86.9

93.4

98.4

28.0 ≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

28.6

28.6

28.6

35.7

42.9

57.1

71.4

85.7

92.9

9.1 ≤ 16

≤ 17

≤ 18

66.7

72.2

88.9

3.9

3.9

≤ 13

≤ 17

56.8

71.4

62.3 ≤ 9

≤ 10

≤ 11

19.2

23.1

30.8

KCE Report 193S


97.0

0.0

19.7

29.5

8.2

13.1

19.7

31.1

49.2

60.7

80.3

86.9

93.4

98.4

17.0

99.0

99.0

97.9

97.9

97.9

95.9

94.8

90.2

86.1

73.2

57.2

40.2

25.3

28.6

28.6

28.6

35.7

42.9

57.1

71.4

85.7

92.9

97.2

97.2

97.2

94.4

91.7

91.7

83.3

77.8

66.7

66.7

72.2

88.9

64.2

40.8

21.2

56.8

71.4

71.9

75.8

19.2

23.1

30.8

100.0

100.0

100.0

KCE Report 193S

Study Time point

Goodridge 1998 3 months

Halfens 2000 NR

Jalali 2005 14 days

Kim 2009 90 days

Kwong 2005 21 days

Langemo 1991 (h)

(i)

16 days

31 days

Pang 1998 2 weeks

Lyder 1999 (j)

(k)

(l)

NR‡

NR‡

NR‡

Ramundo 1995 4 weeks



≤ 16

≤ 20

76.9

96.2

9.7 ≤ 11

≤ 16

≤ 18

12.5

25.0

50.0

58.1 ≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

≤ 23

1.1

3.2

5.4

11.8

17.7

22.0

32.3

40.9

5

61.3

73.7

78.5

88.2

100.0

9.1 NR 52.7

18.3 ≤ 14 92.5

2.1 ≤ 14 88.9

14.9

28.0

≤ 15

≤ 16

≤ 18

54.5

63.6

57.1

19.8 ≤ 18 90.5

NR

NR

NR

≤ 18

≤ 16

≤ 18

81.0

77.0

90.0

14.6 ≤ 11 14.3

81


76.9

96.2

29.6

3.7

12.5

25.0

50.0

97.3

85.6

52.3

1.1

3.2

5.4

11.8

17.7

22.0

32.3

40.9

1.1

61.3

73.7

78.5

88.2

100.0

100.0

100.0

99.3

97.8

97.0

94.8

91.8

90.3

85.8

79.9

70.1

56.7

42.5

29.9

52.7 100.0

92.5 69.8

88.9 71.9

54.5

63.6

57.1

93.7

87.3

61.1

90.5 62.4

81.0

77.0

90.0

100.0

50.0

14.0

14.3 97.6

82

Study Time point

Salavadalena 1992 6 months

Schoonhoven 2002 12 weeks

Seongsook 2004 NR

Serpa 2011 48 hours

4 days

6 days



≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

14.3

14.3

14.3

14.3

28.6

42.9

100.0

100.0

100.0

100.0

100.0

20.2 ≤ 9

≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

≤ 23

0.0

5.0

5.0

15.0

20.0

30.0

30.0

40.0

45.0

60.0

80.0

85.0

95.0

100.0

100.0

11.0 ≤ 17 43.7

31.3 ≤ 16 97.1

11.1

11.1

11.1

≤ 12

≤ 13

≤ 13

87.5

75.0

75.0

KCE Report 193S


14.3

14.3

14.3

14.3

28.6

42.9

100.0

100.0

100.0

100.0

100.0

95.1

95.1

90.2

82.9

80.5

63.4

34.1

22.0

12.2

4.9

0.0

0.0

5.0

5.0

15.0

20.0

30.0

30.0

40.0

45.0

60.0

80.0

85.0

95.0

100.0

100.0

98.7

97.5

91.1

89.9

86.1

79.

77.2

69.6

63.3

54.4

43.0

31.6

13.9

1.3

0.0

43.7 67.8

97.1 26.0

87.5

75.0

75.0

64.1

81.3

82.8

KCE Report 193S

Study Time point

Suriadi 2006 21 days

VandenBosch 1996 2 weeks

* Percentage‡ No raw data was available to recalculate the sensitivity and specificityNR: not reported(a) ward one in Bergstrom 1987a study; (b) ward two in Bergstrom 1987a study; (c) tertiary care hospitals; (d) veteran medicapatients with a Braden score < 18 on admission; (g) group of patients with a Braden score < 19 on admission; (h) hospitalized patientselders ≥ 75 yrs; (k) black elders < 75 yrs; (l) Latino/Hispanic < 75 yrs

Table 37 – Extended Braden scale

Study Time point

Halfens 2000 NR

* PercentageNR: not reported



33.3 ≤ 14 80.0

28.8 ≤ 17

≤ 18

59.0

NR

raw data was available to recalculate the sensitivity and specificity

(a) ward one in Bergstrom 1987a study; (b) ward two in Bergstrom 1987a study; (c) tertiary care hospitals; (d) veteran medica l centres; (e) skilled nursing facilities;with a Braden score < 18 on admission; (g) group of patients with a Braden score < 19 on admission; (h) hospitalized patients

≥ 75 yrs; (k) black elders < 75 yrs; (l) Latino/Hispanic < 75 yrs


58.1 ≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

≤ 21

≤ 22

≤ 23

≤ 24

≤ 25

≤ 26

0.5

1.6

2.2

3.8

6.5

12.4

17.7

24.2

32.8

40.9

51.1

62.9

73.7

78.5

88.2

100.0

83


80.0 54.3

59.0

NR

NR

79.0

l centres; (e) skilled nursing facilities; (f) group ofwith a Braden score < 18 on admission; (g) group of patients with a Braden score < 19 on admission; (h) hospitalized patients ; (i) long-term care patients; (j) black


0.5

1.6

2.2

3.8

6.5

12.4

17.7

24.2

32.8

40.9

51.1

62.9

73.7

78.5

88.2

100.0

100.0

100.0

100.0

99.3

98.5

97.9

96.3

94.8

91.0

88.8

85.1

79.1

69.4

55.2

42.5

29.1

84

Table 38 – Modified Braden scale

Study Time point

Chan 2009 9 days

Kwong 2005 21 days

* Percentage


Study Time point

Curley 2003 10 days

* Percentage


Study Time point

Kwong 2005 21 days

Lincoln 1986 26 days

Pang 1998 2 weeks


Smith 1989 NR



9.1 ≤ 17

≤ 18

≤ 19

38.9

55.6

88.9

2.1 ≤ 16 88.9


26.7 ≤ 10

≤ 11

≤ 12

≤ 13

≤ 14

≤ 15

≤ 16

≤ 17

≤ 18

≤ 19

≤ 20

3.5

16.3

47.7

67.4

72.1

75.6

88.4

91.9

100.0

100.0

100.0


2.1 ≤ 14 8

13.9 ≤ 14 0.0

19.8 ≤ 16 81.0

11.0 ≤ 15 45.9

29.7 ≤ 16 60.0

KCE Report 193S


38.9

55.6

88.9

79.9

72.6

62.0

88.9 75.0


3.5

16.3

47.7

67.4

72.1

75.6

88.4

91.9

100.0

100.0

100.0

100.0

97.0

92.8

89.0

78.8

67.8

58.1

44.1

30.1

19.9

8.1


8.9 61.0

0.0 93.5

81.0 58.8

45.9 60.3

60.0 31.0

KCE Report 193S

Stotts 1988 3 weeks

Wai-Hang 1997 4 weeks

* PercentageNR: Not reported

Table 41 – Modified Norton scale (ICU)

Study Time point


* Percentage

Table 42 – Modified Norton scale (South African Hospital)

Study Time point

Ongoma 2005 1 week

* Percentage


Study Time point

Anthony 2003 NR

Compton 2008 13 days

Edwards 1995 8 weeks

Jalali 2005 14 days

Pang 1998 2 weeks

Serpa 2009 48 hours

4 days

6 days


17.3 ≤ 14 16.4

4.3 ≤ 14 75.0


62.3 ≤ 19

≤ 21

≤ 23

≤ 25

26.9

34.6

42.3

57.7

fied Norton scale (South African Hospital)


37.9 ≤ 20 92.0


0.4 ≥ 10

≥ 15

≥ 20

82.3

48.8

16.7

17.3 NR 37.2

6.5 NR 100.0

9.1 NR 63.5

19.8 ≥ 16 95.2

7.1

7.1

7.1

≥ 17

≥ 20

≥ 20

71.4

85.7

85.7

85

16.4 94.4

75.0 66.7


26.9

34.6

42.3

57.7

100.0

92.6

88.9

48.1


92.0 29.3


82.3

48.8

16.7

85.2

94.5

98.1

37.2 94.6

100.0 10.3

63.5 83.3

95.2 43.5

71.4

85.7

85.7

67.0

40.7

33.0

86

Study Time point


Smith 1989 NR

Wai-Han 1997 4 weeks

Weststrate 1998 183 days* Percentage; NR: not reported

Table 44 – Andersen scale

Study Time point

Andersen 1982 3 months

* Percentage

Table 45 – Pressure Sore Prediction Score scale (PSPS)

Study Time point

Lothian 1989 3 weeks

* Percentage

Table 46 – Knoll scale

Study Time point

Towey 1988 28 days

* Percentage


Study Time point

Kim 2009 90 days

Seongsook 2004 NR

* Percentage -NR: not reported

Table 48 – Sunderland Pressure Sore Risk Calculator (modified Cubbin

Study Time point

Ongoma 2005 1 week

* Percentage



11.0 ≥ 10 89.6

29.7 ≥ 16 73.3

4.3 ≥ 10 87.5

7.9 ≥ 15 80.9


1.2 ≥ 2 87.5

Pressure Sore Prediction Score scale (PSPS)


4.3 > 6 88.7


46.7 ≥ 12 85.7


18.3 ≤ 28 95.0

31.3 ≤ 24 88.6

Sunderland Pressure Sore Risk Calculator (modified Cubbin-Jackson)


37.9 ≤ 34 80.0

KCE Report 193S


89.6 22.4

73.3 38.0

87.5 28.2

80.9 28.5


87.5 86.7

sitivity* Specificity*

88.7 76.0


85.7 56.3


95.0 81.6

88.6 61.0


80.0 70.7

KCE Report 193S

Table 49 – Risk Assessment Pressure Sore scale (RAPS)

Study Time point

Lindgren 2002 12 weeks

* Percentage


Study Time point

Perneger 2002 3 weeks

* Percentage


Study Time point

Seongsook 2004 NR

* PercentageNR: Not reported


Risk Assessment Pressure Sore scale (RAPS)


11.7 ≤ 31

≤ 32

≤ 33

≤ 34

≤ 35

≤ 36

≤ 37

≤ 38

31.5

33.3

38.9

46.3

50.0

57.4

70.4

77.8


29.9 = 0

≤ 1

≤ 2

≤ 3

≤ 4

≤ 5

≤ 6

≤ 7

≤ 8

91.6

78.7

76.7

62.1

49.7

40.2

27.0

17.7

2.2


31.3 ≤ 18 100.0

87


31.5

33.3

38.9

46.3

50.0

57.4

70.4

77.8

84.6

80.2

75.3

69.4

64.3

57.6

46.5

34.8


91.6

78.7

76.7

62.1

49.7

40.2

27.0

17.7

2.2

34.2

53.5

71.9

85.0

91.0

94.2

97.6

98.9

99.5


100.0 18.2

88

Table 52 – Grosnell scale

Study Time point

Jalali 2005 2 weeks



Study Time point

Kim 2009 90 days

* Percentage

Table 54 – 4-factor model

Study Time point


* Percentage

Table 55 – Suriadi and Sanada scale (SS)

Study Time point

Suriadi 2008 NR




9.1 NR 85.1


18.3 ≤ 21 95.0


62.3 ≥ 2 84.6


28.5 ≥ 0

≥ 2

≥ 3

≥ 4

≥ 5

≥ 6

≥ 7

≥ 9

100.0

97.2

97.2

80.6

72.2

61.1

58.3

6.9

KCE Report 193S


85.1 83.3


95.0 69.3


84.6 29.6


100.0

97.2

97.2

80.6

72.2

61.1

58.3

6.9

0.0

42.0

53.0

82.9

86.7

92.3

95.0

100.0

KCE Report 193S

Table 56 – The Northern Hospital Pressure Ulcer Prevention Plan (TNH

Study Time point

Page 2011 NR


Table 57 – Clinical judgement

Study Time point

Salvadalena 1992 6 months

VandenBosch 1996 2 weeks

* Percentage


The Northern Hospital Pressure Ulcer Prevention Plan (TNH-PUPP)


4.2 ≥ 1

≥ 2

≥ 3

≥ 4

≥ 5

≥ 6

100.0

85.7

71.4

71.4

42.9

57.1


20.2 Yes/no 50.0

28.2 Yes/no 51.7

89


100.0

85.7

71.4

71.4

42.9

57.1

34.2

62.0

81.0

88.0

96.2

99.4


50.0 79.7

51.7 58.1

90

2.5.2. Forest plots area under the receiver operating characteristics curve (AUC)

Figure 2: Braden scale


area under the receiver operating characteristics curve (AUC)

KCE Report 193S

KCE Report 193S

Figure 3 – Modified Braden scale

Pressure Ulcer Prevention – Supplement 91

92

Figure 4 – Braden-Q scale

Figure 5 – Norton scale

Figure 6 – Waterlow scale


KCE Report 193S

Figure 7 – Cubbin-Jackson scale

Pressure Ulcer Prevention – Supplement 93

94

Figure 8 – Douglas scale

Figure 9 – Fragmment scale

Figure 10 – Song and Choi scale


KCE Report 193S

Figure 11 – The Northern Hospital Pressure Ulcer Prevention Plan

Figure 12 – Schoonhoven 2002

Figure 13 – Perneger 2002



95

96

Figure 14 – Seongsook 2004

Figure 15 – Chan 2009

Figure 16 – Kim 2009


KCE Report 193S

Figure 17 – Serpa 2009

Figure 18 – Serpa 2011

2.5.3. Forest plots sensitivity and specificity

Figure 19 – Braden scale cut-off score 17 – follow

No raw data available

Bergstrom 1998 (1): tertiary hospital; Bergstrom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facili


follow-up < 1 week – general population – all grades

Bergstrom 1998 (1): tertiary hospital; Bergstrom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facili ty

97

98









Serpa 2011 1: 4 days; Serpa 2011 2: 6 days

Study

Serpa 2011

TP

7

FP

23

FN

1

TN

41

Sensitivity

0.88 [0.47, 1.00]

Specificity

0.64 [0.51, 0.76]






follow-up 48 hours – ICU – all grades

follow-up 4 and 6 days – ICU – all grades

Specificity

0.64 [0.51, 0.76]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

KCE Report 193S



No raw data for Lyder 1991

Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom 1998 (1): tertiary hospital; Bergstrom 1998 (2): veteLangemo 1991 (2): skilled nursing facility; Lyder 1999 (1): black elders

Study

Feuchtinger 2007

TP

20

FP

19

FN

6

TN

8

Sensitivity

0.77 [0.56, 0.91]

Specificity

0.30 [0.14, 0.50]


follow-up < 1 week – ICU – all grades

follow-up > 1 week – general population – all grades

Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom 1998 (1): tertiary hospital; Bergstrom 1998 (2): vete ran medical centre; Bergstrom 1998 (3): skilled nursing facility;emo 1991 (2): skilled nursing facility; Lyder 1999 (1): black elders ≥ 75 yrs; Lyder 1999 (2): Latino/Hispanic < 75 yrs

Specificity

0.30 [0.14, 0.50]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

99

ran medical centre; Bergstrom 1998 (3): skilled nursing facility;

100


Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom 1998 (1): tertiary hospital; Berg


Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom

Study

Bergstrom 1987a (1)

Bergstrom 1987a (2)

Bergstrom 1998 (1)

Bergstrom 1998 (2)

Bergstrom 1998 (3)

Braden 1994

Capobianco 1996

Salvadalena 1992

TP

7

9

12

15

53

24

12

16

FP

24

52

87

76

83

30

8

45

FN

0

0

14

6

8

4

2

4

TN

68

39

193

185

111

44

28

34

Sensitivity

1.00 [0.59, 1.00]

1.00 [0.66, 1.00]

0.46 [0.27, 0.67]

0.71 [0.48, 0.89]

0.87 [0.76, 0.94]

0.86 [0.67, 0.96]

0.86 [0.57, 0.98]

0.80 [0.56, 0.94]

0.74 [0.64, 0.83]

0.43 [0.33, 0.54]

0.69 [0.63, 0.74]

0.71 [0.65, 0.76]

0.57 [0.50, 0.64]

0.59 [0.47, 0.71]

0.78 [0.61, 0.90]

0.43 [0.32, 0.55]

Study

Bergstrom 1987a (1)

Bergstrom 1987a (2)

Bergstrom 1998 (1)

Bergstrom 1998 (2)

Bergstrom 1998 (3)

Braden 1994

Capobianco 1996

Salvadalena 1992

TP

7

9

17

19

57

26

13

17

FP

32

62

126

130

116

42

12

54

FN

0

0

9

2

4

2

1

3

TN

60

29

154

131

78

32

24

25

Sensitivity

1.00 [0.59, 1.00]

1.00 [0.66, 1.00]

0.65 [0.44, 0.83]

0.90 [0.70, 0.99]

0.93 [0.84, 0.98]

0.93 [0.76, 0.99]

0.93 [0.66, 1.00]

0.85 [0.62, 0.97]

0.65 [0.55, 0.75]

0.32 [0.22, 0.42]

0.55 [0.49, 0.61]

0.50 [0.44, 0.56]

0.40 [0.33, 0.47]

0.43 [0.32, 0.55]

0.67 [0.49, 0.81]

0.32 [0.22, 0.43]



Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom 1998 (1): tertiary hospital; Berg strom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facility


Bergstrom 1987a (1): ward one; Bergstrom 1987a (2): ward two; Bergstrom 1998 (1): tertiary hospital; Bergstrom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facility

Specificity

0.74 [0.64, 0.83]

0.43 [0.33, 0.54]

0.69 [0.63, 0.74]

0.71 [0.65, 0.76]

0.57 [0.50, 0.64]

0.59 [0.47, 0.71]

0.78 [0.61, 0.90]

0.43 [0.32, 0.55]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.65 [0.55, 0.75]

0.32 [0.22, 0.42]

0.55 [0.49, 0.61]

0.50 [0.44, 0.56]

0.40 [0.33, 0.47]

0.43 [0.32, 0.55]

0.67 [0.49, 0.81]

0.32 [0.22, 0.43]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

strom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facility

1998 (1): tertiary hospital; Bergstrom 1998 (2): veteran medical centre; Bergstrom 1998 (3): skilled nursing facility

KCE Report 193S





Study

Braden 1994

TP

9

FP

4

FN

19

TN

70

Sensitivity

0.32 [0.16, 0.52]

Specificity

0.95 [0.87, 0.99]

Study

Braden 1994

Seongsook 2004

TP

14

34

FP

8

57

FN

14

1

TN

66

20

Sensitivity

0.50 [0.31, 0.69]

0.97 [0.85, 1.00]

Specificity

0.89 [0.80, 0.95]

0.26 [0.17, 0.37]

Study

Braden 1994

TP

16

FP

11

FN

12

TN

63

Sensitivity

0.57 [0.37, 0.76]

Specificity

0.85 [0.75, 0.92]

Study

Ramundo 1995

TP

3

FP

15

FN

4

TN

26

Sensitivity

0.43 [0.10, 0.82]

Specificity

0.63 [0.47, 0.78]


follow-up > 1 week – ICU – all grades



follow-up > 1 week – general population – stage 2+

Specificity

0.95 [0.87, 0.99]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.89 [0.80, 0.95]

0.26 [0.17, 0.37]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.85 [0.75, 0.92]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.63 [0.47, 0.78]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

101

102



Figure 34 – Braden-Q scale cut-off score 15 – follow



Study

Ramundo 1995

TP

7

FP

27

FN

0

TN

14

Sensitivity

1.00 [0.59, 1.00]

Specificity

0.34 [0.20, 0.51]

Study

Ramundo 1995

TP

7

FP

32

FN

0

TN

9

Sensitivity

1.00 [0.59, 1.00]

Specificity

0.22 [0.11, 0.38]

Study

Curley 2003

TP

65

FP

76

FN

21

TN

160

Sensitivity

0.76 [0.65, 0.84]

Specificity

0.68 [0.61, 0.74]

Study

Curley 2003

TP

76

FP

99

FN

10

TN

137

Sensitivity

0.88 [0.80, 0.94]

Specificity

0.58 [0.51, 0.64]




follow-up > 1 week – paediatric ICU – all stages



Specificity

0.34 [0.20, 0.51]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.22 [0.11, 0.38]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.68 [0.61, 0.74]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.58 [0.51, 0.64]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

KCE Report 193S

Figure 37 – Norton scale cut-off score 14 – follow



Figure 40 – Waterlow scale cut-off score 17 – follow

Study

Kwong 2005

Lincoln 1986

Stotts 1988

Wai-Han 1997

TP

8

0

11

6

FP

164

2

18

59

FN

1

5

56

2

TN

256

29

302

118

Sensitivity

0.89 [0.52, 1.00]

0.00 [0.00, 0.52]

0.16 [0.08, 0.27]

0.75 [0.35, 0.97]

Specificity

0.61 [0.56, 0.66]

0.94 [0.79, 0.99]

0.94 [0.91, 0.97]

0.67 [0.59, 0.74]

Study

Schoonhoven 2002

TP

62

FP

434

FN

73

TN

660

Sensitivity

0.46 [0.37, 0.55] 0.60 [0.57, 0.63]

Study

Pang 1998

Smith 1989

TP

17

18

FP

35

49

FN

4

12

TN

50

22

Sensitivity

0.81 [0.58, 0.95]

0.60 [0.41, 0.77]

Specificity

0.59 [0.48, 0.69]

0.31 [0.21, 0.43]

Study

Serpa 2009

TP

5

FP

30

FN

2

TN

61

Sensitivity

0.71 [0.29, 0.96]

Specificity

0.67 [0.56, 0.77]


follow-up > 1 week – general population – all stages



follow-up 48 hours – general population – all stages

Specificity

0.61 [0.56, 0.66]

0.94 [0.79, 0.99]

0.94 [0.91, 0.97]

0.67 [0.59, 0.74]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.60 [0.57, 0.63]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.59 [0.48, 0.69]

0.31 [0.21, 0.43]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.67 [0.56, 0.77]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

103

104


Serpa 2009 1: 4 days; Serpa 2009 2: 6 days




Study

Anthony 2003

Schoonhoven 2002

Wai-Han 1997

TP

167

121

7

FP

6757

849

127

FN

36

14

1

TN

38775

245

50

Sensitivity

0.82 [0.76, 0.87]

0.90 [0.83, 0.94]

0.88 [0.47, 1.00]

Study

Anthony 2003

TP

99

FP

2519

FN

104

TN

43013

Sensitivity

0.49 [0.42, 0.56] 0.94 [0.94, 0.95]

Study

Pang 1998

Smith 1989

TP

20

22

FP

48

44

FN

1

8

TN

37

27

Sensitivity

0.95 [0.76, 1.00]

0.73 [0.54, 0.88]

Specificity

0.44 [0.33, 0.55]

0.38 [0.27, 0.50]


follow-up 4 days and 6 days – general population – all stages




Specificity

0.85 [0.85, 0.85]

0.22 [0.20, 0.25]

0.28 [0.22, 0.35]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.94 [0.94, 0.95]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.44 [0.33, 0.55]

0.38 [0.27, 0.50]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

KCE Report 193S


Figure 46 – Cubbin-Jackson scale cut-off score 24

Figure 47 – Cubbin-Jackson scale cut-off score 28

Figure 48 – Fragmment scale cut-off score 1 –


Study

Weststrate 1998

TP

38

FP

391

FN

9

TN

156

Sensitivity

0.81 [0.67, 0.91]

Specificity

0.29 [0.25, 0.33]

Study

Seongsook 2004

TP

31

FP

30

FN

4

TN

47

Sensitivity

0.89 [0.73, 0.97]

Specificity

0.61 [0.49, 0.72]

Study

Kim 2009

TP

38

FP

33

FN

2

TN

146

Sensitivity

0.95 [0.83, 0.99]

Specificity

0.82 [0.75, 0.87]

Study

Perneger 2002

TP

280

FP

388

FN

76

TN

446

Sensitivity

0.79 [0.74, 0.83]

Specificity

0.53 [0.50, 0.57]

Study

Perneger 2002

TP

273

FP

234

FN

83

TN

600

Sensitivity

0.77 [0.72, 0.81]

Specificity

0.72 [0.69, 0.75]


follow-up > 1 week – ICU – stage 2+

off score 24 – follow-up > 1 week – ICU – all stages

re 28 – follow-up > 1 week – ICU – all stages

follow-up > 1 week – general population– all stages


Specificity

0.29 [0.25, 0.33]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.61 [0.49, 0.72]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.82 [0.75, 0.87]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.53 [0.50, 0.57]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.72 [0.69, 0.75]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

105

106


Figure 51 – The Northern Hospital Pressure Ulcer Prevention plan cut



Study

Perneger 2002

TP

221

FP

125

FN

135

TN

709

Sensitivity

0.62 [0.57, 0.67]

Specificity

0.85 [0.82, 0.87]

Study

Page 2011

TP

6

FP

60

FN

1

TN

98

Sensitivity

0.86 [0.42, 1.00]

Specificity

0.62 [0.54, 0.70]

Study

Page 2011

TP

5

FP

30

FN

2

TN

128

Sensitivity

0.71 [0.29, 0.96]

Specificity

0.81 [0.74, 0.87]

Study

Page 2011

TP

5

FP

19

FN

2

TN

139

Sensitivity

0.71 [0.29, 0.96]

Specificity

0.88 [0.82, 0.93]



The Northern Hospital Pressure Ulcer Prevention plan cut-off score 2 – follow-up > 1 week – general population

rn Hospital Pressure Ulcer Prevention plan cut-off score 3 – follow-up > 1 week – general population

The Northern Hospital Pressure Ulcer Prevention plan cut-off score 4 – follow-up > 1 week – general population

Specificity

0.85 [0.82, 0.87]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.62 [0.54, 0.70]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.81 [0.74, 0.87]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.88 [0.82, 0.93]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

general population– all stages



KCE Report 193S

Figure 54 – Douglas scale cut-off score 18 – follow

Figure 55 – Song and Choi scale cut-off score 2

Figure 56 – Suriadi and Sanada scale cut-off score 3


Study

Seongsook 2004

TP

35

FP

63

FN

0

TN

14

Sensitivity

1.00 [0.90, 1.00]

Specificity

0.18 [0.10, 0.29]

Study

Kim 2009

TP

38

FP

55

FN

2

TN

124

Sensitivity

0.95 [0.83, 0.99]

Specificity

0.69 [0.62, 0.76]

Study

Suriadi 2008

TP

70

FP

85

FN

2

TN

96

Sensitivity

0.97 [0.90, 1.00]

Specificity

0.53 [0.45, 0.60]

Study

Suriadi 2008

TP

58

FP

31

FN

14

TN

150

Sensitivity

0.81 [0.70, 0.89]

Specificity

0.83 [0.77, 0.88]


follow-up > 1 week – ICU – all stages




Specificity

0.18 [0.10, 0.29]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.69 [0.62, 0.76]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.53 [0.45, 0.60]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.83 [0.77, 0.88]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

107

108


Figure 59 – Clinical judgement – follow-up > 1 wee

2.5.4. Clinical evidence tables

Table 58 – Pancorbo 2006

Reference Method

Author and year:Pancorbo (2006)

Title: Riskassessment scalesfor pressure ulcerprevention: asystematic review.

Journal: Journal ofAdvanced Nursing,54 (1); 94-110.

Design: systematicreview and metaanalysis

Source of funding:grant from the HealthInstitute Carlos III,Ministry of Health andConsumer (Spain)

Search date: 19662003

Searched databases:

DARE; CINAHL;Medline; Currentcontents clinicalmedicine, social andbehaviour science, lifesciences; indice medico

Study

Suriadi 2008

TP

52

FP

24

FN

20

TN

157

Sensitivity

0.72 [0.60, 0.82]

Specificity

0.87 [0.81, 0.91]

Study

Salvadalena 1992

VandeBosch 1996

TP

10

15

FP

16

30

FN

10

14

TN

63

43

Sensitivity

0.50 [0.27, 0.73]

0.52 [0.33, 0.71]

Specificity

0.80 [0.69, 0.88]

0.59 [0.47, 0.70]



up > 1 week – general population – all stages

Patientcharacteristics

Intervention Results

systematicreview and meta-

Source of funding:m the Health

Institute Carlos III,Ministry of Health and

: 1966-

DARE; CINAHL;Medline; Currentcontents clinicalmedicine, social andbehaviour science, lifesciences; indice medico

Eligibility criteria: alltypes of patients


Hospitalized patients(acute ward, medicalward, surgical ward,orthopaedic ward,internal medicine,geriatric ward,cardiovascularsurgery, neurosurgery,orthopaedic surgery),ICU patients, homecare patients, LTCFpatients, rehabilitationpatients, geriatriccentre

Index test

Braden scale;

Norton scale;

Waterlow scale;

Andersen scale;

Pressure SorePrediction Score;

Knoll scale;

Modified Norton scale;

Emina scale;

Cubbin-Jackson scale;

Risk AssessmentPressure Sore;

Fragmment scale;

Douglas scale;

Clinical judgement

See Appendix

Specificity

0.87 [0.81, 0.91]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.80 [0.69, 0.88]

0.59 [0.47, 0.70]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

Critical appraisal ofreview quality

See Appendix 2.5 The critical assessmentguide developed forclinical practice guide forPU assessment andprevention (Rycroft-Malone & McInness 2002)was used to assess thequality of prospectivecohort studies. Results ofthe assessment of themethodological quality arenot reported.

KCE Report 193S

Reference Method

español; cuiden; centroLatinoamericano y delcaribe de información enCiencias de la Salud;Cochrane Library;EBSCO; ScienceDirect;Springer; InterSciencia;ProQuest; Pascal

Included studydesigns: prospectivecohort studies

Inclusion criteria: thepatients considered hadno PU at the beginningof the study; drop-outrate of patients did notexceed 25 %; studies inFrench, Spanish,English or Portuguese

Number of includedstudies: 32

Table 59 – Anthony 2003


Author and year:Anthony (2003)

Title: A regressionanalysis of theWaterlow score inpressure ulcer riskassessment.

Journal: Clinical

Patient group:hospitalised patients ofall ages

All patients

Included N: 45735

Completed N: 45735



Intervention Results

español; cuiden; centroLatinoamericano y delcaribe de información enCiencias de la Salud;

rane Library;EBSCO; ScienceDirect;Springer; InterSciencia;

Included studyprospective

thepatients considered hadno PU at the beginning

outrate of patients did not

25 %; studies inFrench, Spanish,

Number of included

Reference standard:Pressure ulcerdevelopment

Patient Characteristics Intervention

Comparison

Outcomemeasures

Results

Patient group:patients of

Index test 1: the Waterlowscale

Reference standard:development of pressureulcer stage I or above,according to the Torrancegrading (Torrance, 1983)

Outcome 1:

Incidence of PU

Outcome 2:

Area under theROC

Value: 0.4%

AUC: 0.901

95% CI: 0.883

109

Critical appraisal ofreview quality

Comments

0.883-0.919

Funding: /

Limitations:database cohortstudy; no reporton re-assessment ofindex test; no

110


Rehabilitation, 17(2):216-23.

Study type: Databasecohort study butparticipants followedprospectively

Selection patient:

Hospitalizedpatients. All patientsadmitted between1996 and 2000 with acompatible Waterlowscore on admission.

Index test: Waterlowscale was used toassess PU risk atadmission. Re-assessment unclear.Health professionalwere trained toscreen the patients.

Reference standard:The Torrance scorewas used to gradethe PU. Healthprofessional weretrained to screen thepatients.

Imputation: noimputation, noexclusion

Number of events: 203patients developed

Drop-outs: 0

Group with hospitalacquired PU

Number of patientswith a PU: 203 had noPU on admission; 74had a PU on admission

Age (mean years;median age (IQR);range): 63.24; 64.70(17.22); 0 to > 81

Gender (m/f): 81/122

Days in hospital (meandays; median days(IQR)): 31.98; 22.00(34.50)

Group without hospitalacquired PU

Age (mean years;median age (IQR);range): 41.84; 44.50(28.33); 0 to > 81

Gender (m/f):21732/23800

Days in hospital (meandays; median days(IQR)): 3.40; 2.00(2.00)

Inclusion criteria: notreported



Comparison

Outcomemeasures

Results

Group with hospital

Number of patients203 had no

PU on admission; 74had a PU on admission

Age (mean years;n age (IQR);

63.24; 64.70

Days in hospital (meandays; median days

31.98; 22.00

Group without hospital

Age (mean years;median age (IQR);

41.84; 44.50

Gender (m/f):

Days in hospital (meandays; median days

3.40; 2.00

not

Preventive methods: notreported

Outcome 3:

Sensitivity andspecificityWaterlow scalecut-off 10

Outcome 4:

Sensitivity andspecificityWaterlow scalecut-off 15

Outcome 5:

Sensitivity andspecificityBraden scalecut-off 20

Sensitivity: 82.3%

Specificity: 85.2%

Raw data

Referencestandard

Yes

Indextest

Yes 16

No 36

203

Sensitivity: 48.8%

Specificity: 94.5%

Raw data

Referencestandard

Yes

Indextest

Yes 99

No 104

203

Sensitivity: 16.7%

Specificity: 98.1%

Raw data

Refstandard

Yes

Indextest

Yes 34

No 169

203

KCE Report 193S

Comments

82.3%

: 85.2%

Referencestandard

Yes No

167 6757 6924

36 38775 38811

203 45532 45735

48.8%

: 94.5%

Referencestandard

Yes No

99 2519 2618

104 43013 43117

203 45532 45735

16.7%

: 98.1%

Referencestandard

Yes No

34 846 880

169 44686 44855

203 45532 45735

report onduration offollow-up; noreport onblinding; noimputation, noexclusion; notreported whenpatientsdropped fromthe study; noreport oninclusion andexclusioncriteria; noreport on use ofpreventivemeasures; nosub-analysesaccording topreventivemeasures.

Additionaloutcomes: /

Notes: /

KCE Report 193S


ulcers

Addressing missingdata: not reportedwhen patientsdropped from thestudy

Statistical analysis:An ROC curve is aplot of the truepositive

rate (sensitivity)against the falsepositive rate (1–specificity) for giventhresholds. A systemthat performs as onemight expect wouldshow a differing ratioof sensitivity tospecificity as the

threshold increases.

Setting: the Queen’sHospital in Burton.

Blinding: not reported

Exclusion criteria: notreported



Comparison

Outcomemeasures

Results

not

111

Comments

112

Table 60 – Chan 2009


Author and year: Chan(2009)

Title: Assessingpredictive validity ofthe modified Bradenscale for predictionof pressure ulcer riskof orthopaedicpatients in an acutecare setting.

Journal: Journal ofClinical Nursing, 18:1565-73

Study type:prospective cohortstudy

Selection patient:

Chinese patientsaged 18 or abovewithout a pressureulcer on admission.Recruitment unclear.

Index test: Bradenand modified Bradenwere used to assessPU risk at admission.Researcher, a trainednurse, screened thepatients.

Reference standard:

Patient group:hospitalised patientsaged 18 or above

All patients

Included N: 197

Completed N: 197

Drop-outs: 0

Age (mean years (SD);range): 79.4 (10.88);35-98


Number of patientswith a PU: 18

Number of patientswithout a PU: 179

Inclusion criteria:Chinese;

aged 18 or above;

an expected stay of fivedays or more followingadmission;

not ambulant;

no PU on admission.

Exclusion criteria:none



Comparison

Outcomemeasures

Results

Patient group:patients

Age (mean years (SD);79.4 (10.88);

Number of patients

Number of patients

Inclusion criteria:

an expected stay of fivedays or more following

Exclusion criteria:

Index test 1: the Bradenscale

Index test 2: modifiedBraden scale (Kwong et al.2005)

Reference standard:development of pressureulcer stage I or above,according to the NPUAP(2007) classification.

Preventive methods:preventive nursingintervention were performedbut not described..

Outcome 1:

Incidence of PU

(> 1 week; 9days)

Outcome 2:

Area under theROC

Outcome 3:


Outcome 4:


Value: 9.10%

Value: 0.736

95% CI: 0.632

Sensitivity: 66.7%

Specificity: 64.2%

Raw data

Indextest

Yes

No

Sensitivity: 72.2%

Specificity: 40.8%

Raw data

KCE Report 193S

Comments

: 0.632-0.841

66.7%

: 64.2%

Referencestandard

Yes No

12 64 76

6 115 121

18 179 197

72.2%

: 40.8%

Referencestandard

Funding: /

Limitations:index testmeasured onlyat admission; noreport onblinding ofresearchertoward indextest andreferencestandard; noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; nosub-analysesaccording topreventivemeasures.


Notes: /

KCE Report 193S


skin assessment todetect PUs wereperformed daily.Researcher, a trainednurse, screened thepatients. Patientwere observed untilPU development,discharge, transferor death.Observation periodof maximum 9 days.


Number of events: 18patients developedulcers


Statistical analysis:The receiveroperatingcharacteristic (ROC)curve determined thepredictive validity ofthe Braden andmodified Bradenscales.



Comparison

Outcomemeasures

Results

Outcome 5:


Outcome 6:

Sensitivity andspecificitymodified-Bradenscale cut-off 17

Indextest

Yes

No

Sensitivity: 88.9%

Specificity: 21.2%

Raw data

Indextest

Yes

No

Sensitivity: 38.9%

Specificity: 79.9%

Raw data

Indextest

Yes

No

113

Comments

Yes No

13 106 119

5 73 78

18 179 197

88.9%

: 21.2%

Referencestandard

Yes No

16 141 157

2 38 40

18 179 197

38.9%

: 79.9%

Referencestandard

Yes No

7 36 43

11 143 154

114


Setting: twoorthopaedic wards ofan acute carehospital in HongKong

Blinding: blinding ofresearcher whoassess risk and PUdevelopment notreported. Nursesperformed preventivemeasures withoutknowing the scoresof the Braden andmodified Braden.



Comparison

Outcomemeasures

Results

Outcome 7:


Outcome 8:


Sensitivity: 55.6%

Specificity: 72.6%

Raw data

Indextest

Yes

No

Sensitivity: 88.9%

Specificity: 62.0%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

18 179 197

55.6%

: 72.6%

Referencestandard

Yes No

10 49 59

8 130 138

18 179 197

88.9%

: 62.0%

Referencestandard

Yes No

16 68 84

2 111 113

18 179 197

KCE Report 193S

Table 61 – Compton 2008

Reference PatientCharacteristics

Author and year:Compton (2008)

Title: Pressure ulcerpredictors in ICUpatients: nursing skinassessment versusobjective parameters

Journal: Journal ofWound Care, 17(10):417-24.

Study type: databasecohort but participantswere followedprospectively

Selection patient:

All patients admitted tothe medical ICUbetween April 2001 andDecember 2004.

Index test: Waterlowscore at admission. Theadmitting nursescreened the patients

Reference standard:Occurrence of PU wererecorded during the ICUtreatment (median stay(IQ) before PUoccurrence: 7 (4.13))

Imputation: noimputation, no

Patient group:patients hospitalisedin ICU.

All patients

Included N: 698

Completed N: 698

Drop-outs: 0

Age (median yrs(IQ)): 66 (56, 75, 25)

Gender (m/f):392/306



Number of daysbefore occurrenceof PU (median days(IQ)): 7 (4, 13)

Inclusion criteria:patients admitted tothe ICU for at least72 hours; nopressure ulcer onadmission

Exclusion criteria: /


Intervention

Comparison

Outcomemeasures

Results

Patient group:lised

Age (median yrs

Gender (m/f):

Number of patients

Number of patients

Number of daysbefore occurrence

ays

Inclusion criteria:patients admitted tothe ICU for at least72 hours; nopressure ulcer on

Index test 1: the Waterlowscale

Reference standard:development of pressureulcer stage II or above,according to the NPUAP(1999) classification.

Preventive methods: notreported.

Outcome 1:

Incidence of PU

Outcome 2:

Area under theROC

Value: 17.3%

AUC: 0.59

95% CI: 0.54-0.65

115

Comments

0.65

Funding: /

Limitations:

database cohortstudy; index testonly assessedon admission;no report onmaximumduration offollow-up; noreport onblinding; noimputation, noexclusion; notreported whenpatientsdropped fromthe study; noreport on use ofpreventivemeasures; nosub-analysesaccording topreventivemeasures; cut-off score of 0.5does not exist

Additionaloutcomes:logisticregression of 32

116


exclusion


Addressing missing data:not reported whenpatients dropped fromthe study

Statistical analysis: Thepredictive capacity ofthe logistic regressionfunction was assessedand compared with theWaterlow score bycalculating the areaunder the curve of areceiving-operatorcharacteristics curve.AUC, sensitivityspecificity weredisplayed with 95% CI

Setting: medical ICU ofthe Charité CampusBenjamin FranklinBerlin



Intervention

Comparison

Outcomemeasures

Results

KCE Report 193S

Comments

variables. Fiveparameterswere identifiedas predictorsand sensitivityand specificitywas calculated.

Notes: /

KCE Report 193S

Table 62 – Curley 2003


Author and year:Curley (2003)

Title: Predictingpressure ulcer risk inpediatric patients:the Braden Q Scale

Journal: NursingResearch, 52(1): 22-33.


Selection patient:

PICU patients.Consecutive sample.

Index test: Braden-Qwas used to assessPU risk at enrolment.A trained nursescreened thepatients. Patientswere observed up to3 times a week for 2weeks, then once aweek until discharge(stay: 3 – 12 days).

Reference standard:The skin assessmenttool (Braden &

Patient group:paediatric patientshospitalised in PICU.

All patients

Included N: 322

Completed N: 322

Drop-outs: 0

Age (mean months(SD)): 36 (29)

Gender (m/f): 193/129



Inclusion criteria:bedrest for at least 24hours;

age between 21 daysand 8 years.

Exclusion criteria:patients admitted to thePICU with a pre-existingPU;

intra-cardiac shunting;unrepaired congenitalheart disease


ristics Intervention

Comparison

Outcomemeasures

Results

Patient group:paediatric patients

Age (mean months

Number of patients

Number of patients

Inclusion criteria:bedrest for at least 24

age between 21 days

Exclusion criteria:patients admitted to the

existing

cardiac shunting;unrepaired congenital

Index test 1: the Braden-Qscale (Quigley & Curley,1996)

Reference standard:development of pressureulcer stage II or above,according to the NPUAP(1989) classification.

Preventive methods: notreported.

Outcome 1:

Incidence of PU

(> 1 week; 12days)

Outcome 2:

Area under theROC

Outcome 3:

Sensitivity andspecificityBraden-Q scalecut-off 10

Outcome 4:


Value: 26.71%

AUC: 0.830

95% CI: 0.76-0.91

Sensitivity: 3.5%

Specificity: 100%

Raw data

Indextest

Yes

No

Sensitivity: 16.3%

Specificity: 97.0%

Raw data

117

Comments

26.71%

0.91

3.5%

: 100%

Referencestandard

Yes No

3 0 3

83 236 319

86 236 322

16.3%

: 97.0%

Referencestandard

Yes No

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; noreport onpreventivemeasures; nosub-analysesaccording topreventivemeasures.


Notes: /

118


Bergstorm, 1997)was used to detectthe presence orabsence of PUs.

A trained nursescreened thepatients. Patientswere observed up to3 times a week for 2weeks, then once aweek until discharge(stay: 3 – 12 days).




Statistical analysis:Diagnosticprobabilities

(sensitivity,specificity, positivepredictive value, andnegative predicativevalue) werecalculated over a



Comparison

Outcomemeasures

Results

Outcome 5:


Outcome 6:


Indextest

Yes

No

Sensitivity: 47.7%

Specificity: 92.8%

Raw data

Indextest

Yes

No

Sensitivity: 67.4%

Specificity: 89.0%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

14 7 21

72 229 301

86 236 322

47.7%

: 92.8%

Referencestandard

Yes No

41 17 58

45 219 264

86 236 322

67.4%

: 89.0%

Referencestandard

Yes No

58 26 84

28 210 238

86 236 322

KCE Report 193S


range of possibleBraden Q score.

Receiver operatorcharacteristic (ROC)curve analysis

plotting sensitivityagainst 1 –specificity over therange of

Braden Q scores wasconstructed toconfirm the criticalvalue of the BradenQ Scale. The optimalcutoff point wasdetermined by thatwhich provided highsensitivity andadequate specificity.The likelihood ratio(LR) was measuredto identify the ratio ofthe probabilities thata positive test resultsfrom a patient withpressure ulcers tothat for a patientwithout pressureulcers.

Setting: three PICUsof three differenthospitals in the US



Comparison

Outcomemeasures

Results

Outcome 7:


Outcome 8:


Outcome 9:

Sensitivity andspecificityBraden-Q scale

Sensitivity: 72.1%

Specificity: 78.8%

Raw data

Indextest

Yes

No

Sensitivity: 75.6%

Specificity: 67.8%

Raw data

Indextest

Yes

No

Sensitivity: 88.4%

Specificity: 58.1%

Raw data

119

Comments

72.1%

: 78.8%

Referencestandard

Yes No

62 50 112

24 186 210

86 236 322

75.6%

: 67.8%

Referencestandard

Yes No

65 76 141

21 160 181

86 236 322

88.4%

: 58.1%

Reference

120


Blinding: the twonurses were blindedto other’sassessment. Nurse Irated the Braden Qand nurse II rated theskin assessmenttool.



Comparison

Outcomemeasures

Results

cut-off 16

Outcome 10:


Outcome 11:


Indextest

Yes

No

Sensitivity: 91.9%

Specificity: 44.1%

Raw data

Indextest

Yes

No

Sensitivity: 100.0%

Specificity: 30.1%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

standard

Yes No

76 99 175

10 137 147

86 236 322

91.9%

: 44.1%

Referencestandard

Yes No

79 132 211

7 104 111

86 236 322

100.0%

: 30.1%

Referencestandard

Yes No

86 165 251

0 71 71

KCE Report 193S




Comparison

Outcomemeasures

Results

Outcome 12:


Outcome 13:


Sensitivity: 100.0%

Specificity: 19.9%

Raw data

Indextest

Yes

No

Sensitivity: 100.0%

Specificity: 8.1%

Raw data

Indextest

Yes

No

121

Comments

86 236 322

100.0%

: 19.9%

Referencestandard

Yes No

86 189 275

0 47 47

86 236 322

100.0%

: 8.1%

Referencestandard

Yes No

86 217 303

0 19 19

86 236 322

122

Table 63 – de Souza 2010


Author and year: deSouza (2010)

Title: Predictivevalidity of the Bradenscale for pressureulcer risk in elderlyresidents of long-term care facilities

Journal: Geriatricnursing, 31(2): 95-104.

Study type:prospective cohortstudy (secondaryanalysis)

Selection patient:

Elderly patientsresiding in LTCF witha Braden score < 19.Recruitment strategynot reported.

Index test: Bradenscale was used toassess PU risk every2 days for 3 months.Assessment werecarried out by trainedobservers.

Reference standard:Skin assessment

Patient group: elderlypatients residing inLTCFs.

All patients

Included N: 233

Completed N: 233

Drop-outs: 0

Age (mean years(SD)): 76.6 (9.2)

Gender (m/f): 104/129

Length of stay (meandays (SD); range):3685.37 (4266.4); 123360



Subgroup (Bradenscore < 18)

Included N: 94

Completed N: 94

Drop-outs: 0


Gender (m/f): 35/52


Intervention

Comparison

Outcomemeasures

Results

elderlypatients residing in

Age (mean years

Length of stay (meandays (SD); range):3685.37 (4266.4); 1-

Number of patients

Number of patients

Subgroup (Braden

Age (mean years

Index test 1: the Bradenscale (Braden and Bergstrom1994)

Reference standard:development of pressureulcer grade 1 or above,according to the EPUAP(2008) classification.

Preventive methods:

change of the patient’sposition and minimization

of skin exposure to moisture

Outcome 1:

Incidence of PUin total group

(not reported)

Outcome 2:

Incidence of PUin subgroup

(not reported)

Outcome 3:

Sensitivity andspecificityBraden scalecut-off 17 in totalgroup // lastassessment (3months?)

Outcome 4:

Sensitivity andspecificityBraden scalecut-off 17 insubgroup // last

Value: 18.9%

Value: 39.4%

Sensitivity: 75.0%

Specificity: 75.7%

Raw data

Indextest

Yes

No

Sensitivity: 56.8%

Specificity: 71.9%

Raw data

KCE Report 193S

Comments

75.0%

: 75.7%

Referencestandard

Yes No

33 46 79

11 143 154

44 189 233

56.8%

: 71.9%

Referencestandard

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; noreport onblinding; no sub-analysesaccording topreventivemeasures. Onlypatients with aBraden score <19 wereincluded!Unclear ifpatients with apressure ulcerat start of thestudy wereincluded.

Additionaloutcomes:sensitivity andspecificity onday 0

KCE Report 193S


was performed every2 days for 3 months.Assessment werecarried out by trainedobservers.




Statistical analysis:The predictivevalidity of a test isdetermined by thesensitivity and

specificity of the test.Sensitivity andspecificity can begraphicallyrepresented by thereceiver operatingcharacteristic (ROC)curve that plots thetrue-positive rate(sensitivity) againstthe false-positive

Length of stay (meandays (SD)): 3979.51(5371.3)



Inclusion criteria: aged60 years and older;Braden score < 19;agreement to participate



Intervention

Comparison

Outcomemeasures

Results

Length of stay (meandays (SD)): 3979.51

Number of patients

Number of patients

aged60 years and older;Braden score < 19;agreement to participate

assessment (3months?)

Indextest

Yes

No

123

Comments

Yes No

21 16 37

16 41 57

37 57 94

Notes: /

124


rate (1-specificity).The test isconsidered goodwhen the ROC curvefalls above thediagonal line. Thereis a quantitative andqualitativerelationship betweenthe area under thecurve (AUC) andaccuracy, which maybe classified as

excellent (0.80–0.90),very good (0.70–0.79), good (0.60–0.69), and poor (0.50–0.59). The patientswere assessed for 3consecutive months,and data from thefirst and last (beforeany of theaforementionedoutcomes)assessments wereused for statisticalanalysis.

Setting: 4 LTCFslocated in 3 cities inSouthern MinasGerais, Brazil.

Blinding: no blinding


Intervention

Comparison

Outcomemeasures

Results

KCE Report 193S

Comments

KCE Report 193S

Table 64 – Feuchtinger 2007


Author and year:Feuchtinger (2007)

Title: Pressure ulcerrisk assessmentimmediately aftercardiac surgery--does it make adifference? Acomparison of threepressure ulcer riskassessmentinstruments within acardiac surgerypopulation

Journal: Nursing inCritical Care, 12(1):42-49.


Selection patient:

ICU patientsconsecutivelyrecruited aftercardiac surgery.

Index test: Bradenscale, modified

Patient group: cardiacsurgery ICU patients.

All patients

Included N: 53

Completed N: 53completed assessmenton admission to the ICUand day 1. 36 patientscompleted theassessment after day 2,20 after day 3 and 17after day 4.

Drop-outs: 0 forassessment onadmission to the ICUand day 1. 17 forassessment on day 2,another 16 forassessment on day 3and another 3 forassessment on day 4.

Age (mean years (SD);range): 62 (12.1); 25-83

Gender (m/f): 31/22


Number of patients


Intervention

Comparison

Outcomemeasures

Results

cardiac

53completed assessmenton admission to the ICUand day 1. 36 patientscompleted theassessment after day 2,20 after day 3 and 17

0 forassessment on

ion to the ICUand day 1. 17 forassessment on day 2,another 16 forassessment on day 3and another 3 for

Age (mean years (SD);83

Number of patients

Number of patients

Index test 1: the Bradenscale (Bergstorm et al.1987)

Index test 2: the modifiedNorton scale (Bienstein,1991)

Index test 2: the four-factormodel (Halfens et al. 2000)

Reference standard:development of pressureulcer grade 1 or above,according to the EPUAP(2005a) classification.

Preventive methods:

Not reported

Outcome 1:

Incidence of PU

(1 day)

Outcome 2:

Incidence of PU

(1 week)

Outcome 3:

Sensitivity andspecificityBraden scalecut-off 9 // day 1

Outcome 4:

Sensitivity andspecificityBraden scalecut-off 10 // day

Value: 49%

Value: 62.3%

Sensitivity: 19.2%

Specificity: 100.0%

Raw data

Indextest

Yes

No

Sensitivity: 23.1%

Specificity: 100.0%

Raw data

125

Comments

19.2%

: 100.0%

Referencestandard

Yes No

5 0 5

21 27 48

26 27 53

23.1%

: 100.0%

Referencestandard

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; noreport onblinding; noreport onpreventivemeasures; noreport onstatisticalanalysis; nosub-analysesaccording topreventivemeasures.


Notes: /

126


Norton scale and 4-factor model ofHalfens (2000) wereused to assess PUrisk after surgery andthe four followingdays. Assessmentwere carried out bytrained observers.

Reference standard:Skin assessmentwas performedpreoperative,postoperative andthe four followingdays. Assessmentwere carried out bytrained observers.



Addressing missingdata: 53 patientswere assessedpostoperative and onday 1. 36 patientswere assessed onday 2, 20 on day 3and 14 on day 4.

Statistical analysis:

without a PU: 20

Inclusion criteria:cardiac surgery patientswith a length of stay of≥24h in ICU



Intervention

Comparison

Outcomemeasures

Results

Inclusion criteria:cardiac surgery patientswith a length of stay of

1

Outcome 5:

Sensitivity andspecificityBraden scalecut-off 11 // day1

Outcome 6:


Indextest

Yes

No

Sensitivity: 30.8%

Specificity: 100.0%

Raw data

Indextest

Yes

No

Sensitivity: 76.9%

Specificity: 29.6%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

Yes No

6 0 6

20 27 47

26 27 53

30.8%

: 100.0%

Referencestandard

Yes No

8 0 8

18 27 45

26 27 53

76.9%

9.6%

Referencestandard

Yes No

20 19 39

6 8 14

KCE Report 193S


Not reported

Setting: ICU; nofurther information.



Intervention

Comparison

Outcomemeasures

Results

Outcome 7:


Outcome 8:

Sensitivity andspecificitymodified Nortonscale cut-off 19// day 1

Sensitivity: 96.2%

Specificity: 3.7%

Raw data

Indextest

Yes

No

Sensitivity: 26.9%

Specificity: 100%

Raw data

Indextest

Yes

No

Sensitivity: 34.6%

Specificity: 92.6%

127

Comments

26 27 53

96.2%

: 3.7%

Referencestandard

Yes No

25 26 51

1 1 2

26 27 53

26.9%

: 100%

Referencestandard

Yes No

7 0 7

19 27 46

26 27 53

34.6%

: 92.6%

128



Intervention

Comparison

Outcomemeasures

Results

Outcome 9:


Outcome 10:


Outcome 11:

Sensitivity andspecificitymodified Nortonscale cut-off 25

Raw data

Indextest

Yes

No

Sensitivity: 42.3%

Specificity: 88.9%

Raw data

Indextest

Yes

No

Sensitivity: 57.7%

Specificity: 48.1%

Raw data

KCE Report 193S

Comments

Referencestandard

Yes No

9 2 11

17 25 42

26 27 53

42.3%

: 88.9%

Referencestandard

Yes No

11 3 14

15 24 39

26 27 53

57.7%

: 48.1%

Referencestandard

Yes No

KCE Report 193S


Table 65 – Hatanaka 2007


Author and year:Hatanaka (2007)

Title: A newpredictive indicatorfor development ofpressure ulcers inbedridden patientsbased on common

Patient group:bedridden hospitalizedpatients.

All patients

Included N: 149

Completed N:149


Intervention

Comparison

Outcomemeasures

Results

// day 1

Outcome 12:

Sensitivity andspecificity 4-factor modelcut-off 25 // day1

Indextest

Yes

No

Sensitivity: 84.

Specificity: 29.6%

Raw data

Indextest

Yes

No

Intervention

Comparison

Outcomemeasures

Results

Patient group:hospitalized


Reference standard:development of pressureulcer was defined as morethan grade 1 (closed-persistent erythema)

Outcome 1:

Incidence of PU

(5-79 days)

Outcome 2:

Area under theROC Braden

Value: 25.5%

Value: 0.56

129

Comments

15 14 29

11 13 24

26 27 53

84.6%

: 29.6%

Referencestandard

Yes No

22 19 41

4 8 12

26 27 53

Comments

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatients

130


laboratory testsresults

Journal Journal ofClinical Pathology,61: 514-518.


Selection patient:

Bedridden patientshospitalized for arespiratory disorder.

Recruitment strategynot reported.

Index test: Bradenscale was used toassess PU risk onadmission.

Reference standard:Pressure ulcerdevelopment wasobserved over athree months period,hospital discharge orPU development.


Number of events: 38patients developed

Drop-outs: 0





Inclusion criteria:

Required constantattentive care or need ofa considerable amountof assisted care



Intervention

Comparison

Outcomemeasures

Results

Age (mean years

er of patients

Number of patients

Required constantattentive care or need ofa considerable amount

Preventive methods:

All patients were given astandard pressure-relievingmattress duringhospitalization.

scale

KCE Report 193S

Comments

dropped fromthe study; indextest only onadmission; noreport onblinding; nodescription ofpreventivemeasures; nosub-analysesaccording topreventivemeasures.

Additionaloutcomes:AUC of newindicator basedon laboratoryresults

Notes: /

KCE Report 193S


ulcers


Statistical analysis: Areceiver operatingcharacteristic (ROC)curves analysis wasperformed.

Setting: Onehospital, Nara,Japan.


Table 66 – Jalali 2005


Author and year: Jalali(2005)

Title: Predictingpressure ulcer risk:comparing thepredictive validity of4 scales

Journal Advances inSkin & Wound Care,

Patient group:hospitalized patients.

All patients

Included N: 230

Completed N: 230

Drop-outs: 0

Age (mean years;


Intervention

Comparison

Outcomemeasures

Results

Intervention

Comparison

Outcomemeasures

Results

Patient group:

Age (mean years;

Index test 1: the Bradenscale (Bergstorm et al.1987)

Index test 2: the Nortonscale (Norton, 1962)

Index test 3: the Gosnellscale (Gosnell, 1973)

Index test 4: the Waterlowscale (Waterlow 1985)

Outcome 1:

Incidence of PU

(> 1 week; 2weeks)

Outcome 2:

Area under theROC Bradenscale

Value: 9.10%

Sensitivity: 52.7%

Specificity: 100.0%

Raw data

131

Comments

Comments

52.7%

: 100.0%

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped from

132


18(2): 92-97.


Selection patient:

Patients from aneurology, intensivecare, orthopaedicand medical unit.


Index test: Bradenscale, Norton scale,Gosnell scale andWaterlow scale wereused to assess PUrisk within 48h ofadmission. Patientswere screened bytrained researchstaff.

Reference standard:Skin assessmentwas performed onceevery 24h for amaximum of 14 daysto assess thepresence or absenceof a PU. Patientswere screened bytrained research

range): 60; 21-89

Gender (m/f): 100/130

Number of patientswith a PU:

Stage I: 18

Stage II: 48

Stage III: 8

Pressure ulcerlocation:

Sacrum: 54

Buttocks: 10

Heels: 6

Scapula: 4


Inclusion criteria:

age of 21 years or older;

admitted to the hospitalwithin the past 48h;

expected stay of 14daysor longer;

no PU during initial skinassessment



Intervention

Comparison

Outcomemeasures

Results

Number of patients

Pressure ulcer

Number of patients


admitted to the hospital

expected stay of 14days

no PU during initial skin

Reference standard:development of pressureulcer according to criteria ofBergstorm et al. (1994)

Preventive methods:

Common preventive andnursing measures wererecorded. Outcome 3:

Sensitivity andspecificityNorton scalecut-off notreported

Outcome 4:

Sensitivity andspecificityGrosnell scalecut-off notreported

Indextest

Yes

No

Sensitivity: 48.6%

Specificity: 100.0%

Raw data

Indextest

Yes

No

Sensitivity: 85.1%

Specificity: 83.3%

Raw data

Index Yes

KCE Report 193S

Comments

Referencestandard

Yes No

39 0 39

35 156 191

74 156 230

48.6%

: 100.0%

Referencestandard

Yes No

36 0 36

38 156 194

74 156 230

85.1%

: 83.3%

Referencestandard

Yes No

63 26 89

the study; indextest only within48h ofadmission; noreport onblindingconcerning skinassessment;unclear what ismeant withassessment by4 independentnurses; nodescription ofpreventivemeasures; nosub-analysesaccording topreventivemeasures; noreport onthresholds ofrisk assessmenttools.


Notes: /

KCE Report 193S


staff.




Statistical analysis:Predictive power wasmeasured by theoverallconsiderations ofsensitivity,specificity,

positive predictivevalue, and negativepredictive value.

Setting: threeeducational hospitalsin Kermanshah, Iran.

Blinding: the four riskassessment toolwere assessed byfour independentresearch nurses; noinformation for skinassessment.


Intervention

Comparison

Outcomemeasures

Results

Outcome 5:

Sensitivity andspecificityWaterlow scalecut-off notreported

test No

Sensitivity: 63.5%

Specificity: 83.3%

Raw data

Indextest

Yes

No

133

Comments

11 130 141

74 156 230

63.5%

: 83.3%

Referencestandard

Yes No

47 26 73

27 130 157

74 156 230

134

Table 67 – Kim 2009


Author and year: Kim(2009)

Title: Comparison ofthe predictive validityamong pressureulcer riskassessment scalesfor surgical ICUpatients

Journal AustralianJournal of AdvancedNursing, 26(4): 87-94.

Study type:prospective study

Selection patient:

Patients from asurgical intensivecare unit.


Index test: Bradenscale, Song and Choiscale, Cubbin andJackson scale wereused to assess PUrisk at admission.Patients werescreened by a

Patient group: surgicalICU patients ≥ 16 years.

All patients

Included N: 219

Completed N: 219

Drop-outs: 0

Age (mean years (SD);range): 58.1 (1.2); 1698



Stage I: 15

Stage II: 25


Coccyx: 25

Other: 15


Inclusion criteria:


no existing PU onadmission;


Intervention

Comparison

Outcomemeasures

Results

surgical16 years.

Age (mean years (SD);58.1 (1.2); 16-

Number of patients

Pressure ulcer

ients


no existing PU on


Index test 2: the Song andChoi scale (Song and Choi,1991)

Index test 3: the Cubbin andJackson scale (Cubbin andJackson, 1991)

Reference standard:development of pressureulcer according to criteria ofAHRQ (1994)

Preventive methods:

All patients received ordinarynursing interventions,

especially those related topressure ulcer prevention.

Their position was changedevery two hours and they

were dried, cleaned andfriction/shear managed to

prevent pressure ulcers.

Outcome 1:

Incidence of PU

(> 1 week; 90days)

Outcome 2:

Area under theROC Bradenscale

Outcome 3:

Area under theROC Song andChoi scale

Outcome 4:

Area under theROC Cubbinand Jacksonscale

Outcome 5:


Value: 18.3%

Value: 0.881

Value: 0.890

Value: 0.903

Sensitivity: 92.5%

Specificity: 69.8%

Raw data

KCE Report 193S

Comments

92.5%

: 69.8%

Referencestandard

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; indextest only atadmission;blinding unclear;no sub-analysesaccording topreventivemeasures.


Notes: /

KCE Report 193S


trained researchnurse.

Reference standard:Skin assessmentwas performed dailybetween 10:00 and11:00 am untildischarge (stay: 3-90days). Patients werescreened by atrained researchnurse.




Statistical analysis:The parameters forevaluating thepredictive validity ofeach assessmentscale includedsensitivity,specificity, PVP andPVN. The

admitted to the SICU



Intervention

Comparison

Outcomemeasures

Results

Outcome 6:

Sensitivity andspecificity Songand Choi scalecut-off 21

Outcome 7:

Sensitivity andspecificityCubbin andJackson cut-off28

Indextest

Yes

No

Sensitivity: 95.0%

Specificity: 69.3%

Raw data

Indextest

Yes

No

Sensitivity: 95.0%

Specificity: 81.6%

Raw data

Indextest

Yes

No

135

Comments

Yes No

37 54 91

3 125 128

40 179 219

95.0%

: 69.3%

Referencestandard

Yes No

38 55 93

2 124 126

40 179 219

95.0%

: 81.6%

Referencestandard

Yes No

38 33 71

2 146 148

136


ROC curve showshow the sensitivityproportion (verticalaxis) varies with thefalse‑positiveproportion

(horizontal axis,

1‑specificity) as thedecision criterion isvaried.

Setting: one surgicalICU of a South-Korean hospital.

Blinding: the head-nurse assessed eachscale and skinassessment tool.

Table 68 – Kwong 2005


Author and year:Kwong (2005)

Title: Predictingpressure ulcer riskwith the modifiedBraden, Braden, andNorton scales inacute care hospitalsin Mainland China

Patient group:hospitalized patients ofall ages.

All patients

Included N: 429

Completed N: 429

Drop-outs: 0


Intervention

Comparison

Outcomemeasures

Results

Intervention

Comparison

Outcomemeasures

Results

Patient group:patients of

Index test 1: the Bradenscale (Braden andBergstrom, 1987)

Index test 2: the modifiedBraden scale (Pand andWong, 1998)

Index test 3: the Nortonscale (Norton et al., 1975)

Reference standard:

Outcome 1:

Incidence of PU

(> 1 week; 21days)

Outcome 2:

Sensitivity andspecificity

Value: 2.1%

Sensitivity: 88.9%

Specificity: 71.9%

Raw data

KCE Report 193S

Comments

40 179 219

Comments

88.9%

: 71.9%

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped from

KCE Report 193S


Journal: AppliedNursing Research, 18(2); 122-128.


Selection patient:

Patients from anyward in two acutecare hospitals.


Index test: Bradenscale, modifiedBraden scale andNorton scale wereused to assess PUrisk at admission.Patients werescreened by trainednurses.

Reference standard:Skin assessmentwas performed dailyuntil discharge,transfer or 21-dayhospitalisation.Patients werescreened by trainednurses.

Imputation: no


Gender (m/f): 253/176


Stage I: 8

Stage II: 1


Sacral area: 4

Right iliac region: 2

Abdomen: 1

Left knee: 1

Right ankle: 1


Inclusion criteria:

Free of PU within 24h ofadmission



Intervention

Comparison

Outcomemeasures

Results

Age (mean years (SD);54.07 (16.9); 5-

Number of patients

Pressure ulcer

Number of patients

Free of PU within 24h of

development of pressureulcer according to criteria ofthe NPUAP (1989)

Preventive methods:

Nurses working in the wardrelied on their clinicaljudgment to determine andperform preventive nursinginterventions on the subjects.Preventive measures couldbe: turning every 2h, use ofmaterial to reduce pressure,keeping bed linen clean, dry,and smooth, keeping skinclean and dry, positioning,use of draw sheet for liftingpatients, and massage ofpressure points.

Braden scalecut-off 14

Outcome 3:

Sensitivity andspecificitymodified Bradenscale cut-off 16

Outcome 4:

Sensitivity andspecificityNorton scalecut-off 14

Indextest

Yes

No

Sensitivity: 88.9%

Specificity: 75.0%

Raw data

Indextest

Yes

No

Sensitivity: 88.9%

Specificity: 61.0%

Raw data

137

Comments

Referencestandard

Yes No

8 118 126

1 302 303

9 420 429

88.9%

: 75.0%

Referencestandard

Yes No

8 105 113

1 315 316

9 420 429

88.9%

: 61.0%

Referencestandard

Yes No

the study; indextest only atadmission; noblinding ofscales and skinassessment; nosub-analysesaccording topreventivemeasures.


Notes: Pressureulcers located tothe iliac regionand abdomencould be theresult of medicaldevices.However, this isnot stated in thearticle.

138


imputation, noexclusion



Statistical analysis:not reported

Setting: two acutecare hospitals inMainland China.

Blinding: three nursesform each wardassessed the threescales and skinconditionindependent of eachother. No blindingbetween scale andPU development asone of the threenurses performedthis assessment.


Intervention

Comparison

Outcomemeasures

Results

Indextest

Yes

No

KCE Report 193S

Comments

8 164 172

1 256 257

9 420 429

KCE Report 193S

Table 69 – Lincoln 1986


Author and year:Lincoln (1986)

Title: Use of theNorton PressureSore riskassessment scoringsystem with elderlypatients in acute care

Journal: Journal ofEnterostomyTherapy, 13; 132-138.


Selection patient:

Hospitalizedsurgical-medicalpatients.


Index test: Nortonscale was used toassess PU risk atadmission and every3 days untildischarge or death.Patients werescreened byresearch assistants.

Reference standard:

Patient group:hospitalized medicalsurgical patients aged65 years and older.

All patients

Included N: 50

Completed N: 36

Drop-outs: 14 (stayed 3days or less)


Gender (m/f): 23/27

Length of stay (meandays; range): 7.88; 226

Number of patientswith a PU: 5 of the 36


Primarily on heels andelbows, and one sacrallesion


Inclusion criteria:


Intervention

Comparison

Outcomemeasures

Results

Patient group:hospitalized medical-surgical patients aged

14 (stayed 3

Age (mean years (SD);72.2 (15.8); 65-

Length of stay (mean7.88; 2-

Number of patients

Pressure ulcer

Primarily on heels andelbows, and one sacral

Number of patients

Index test 1: the Nortonscale (assessment onadmission used)

Reference standard:development of pressureulcer according to a 5-pointscale: 0 = no change, 1 =erythema, 2 = superficial skinopening, 3 = a lesionextending into underlyingtissue, 4 = involvement ofmuscle and bone

Preventive methods:

Preventive measures weregiven but not reported.Nurses giving preventionwere unaware of Nortonscore

Outcome 1:

Incidence of PU

(max. 26 days)

Outcome 2:

Sensitivity andspecificityNorton scalecut-off 14

Value: 13.9%

Sensitivity: 0.0%

Specificity: 93.5%

Raw data

Indextest

Yes

No

139

Comments

0.0%

: 93.5%

Referencestandard

Yes No

0 2 2

5 29 34

5 31 36

Funding: theresearch wasfunded by theDean’sResearch fund,Frances PayneBolton School ofNursing, CaseWesternReserveUniversity

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; indextest assessedon admissionused; noblinding of; nosub-analysesaccording topreventivemeasures.

Additional

140


Skin assessmentwas performed atadmission and every3 days untildischarge or death.Patients werescreened byresearch assistants.




Statistical analysis:not reported

Setting: two divisionsin a teaching hospitalin the Midwest.


Age over 65 years;absence of pressuresores on admission



Intervention

Comparison

Outcomemeasures

Results

Age over 65 years;absence of pressure

KCE Report 193S

Comments

outcomes: /

Notes: /

KCE Report 193S

Table 70 – Ongoma 2005


Author and year:Ongoma (2009)

Title: Predictivevalidity of pressurerisk assessmentscales in a privatesector traumaintensive care unit

Journal: SouthernAfrican Journal ofCritical Care, 21 (2);78-86.


Selection patient:

Patients admitted tothe ICU of a privateinstitution.

Purposive sampling;not further specified.

Index test:Sunderland PressureSore Risk Calculator(modified Cubbinand Jackson) and amodified Nortonscale were used toassess PU risk atadmission and on a

Patient group: ICUpatients older than 18years.

All patients

Included N: 66

Completed N: 66completed assessmenton admission and afterone week. 34 patientscompleted theassessment after 2weeks and 17 after 3weeks.

Drop-outs: 0 forassessment onadmission and after oneweek. 32 forassessment on week 2and another 17 forassessment on week 3.

Age (range): 18-65

Gender (m/f): 56/10


Pressure ulcerlocation (total of 44PU):

Heels: 19


Intervention

Comparison

Outcomemeasures

Results

ICUpatients older than 18

66completed assessmenton admission and afterone week. 34 patientscompleted theassessment after 2weeks and 17 after 3

0 forassessment on

fter oneweek. 32 forassessment on week 2and another 17 for

Number of patients

Pressure ulcerlocation (total of 44

Index test 1: the SunderlandPressure Sore RiskCalculator (modified Cubbinand Jackson) (Lowery 1995)

Index test 2: a modifiedNorton scale (hospital SouthAfrica)

Reference standard:development of pressureulcer; criteria not specified

Preventive methods:

Not reported

Outcome 1:

Incidence of PU

(1 week)

Outcome 2:

Sensitivity andspecificitySunderlandPressure SoreRisk Calculatorcut-off 35 //week 1

Outcome 3:

Sensitivity andspecificitymodified Nortonscale cut-off 20 /week 1

Value: 37.9%

Sensitivity: 80.0%

Specificity: 70.7%

Raw data

Indextest

Yes

No

Sensitivity: 92.0%

Specificity: 29.3%

Raw data

Indextest

Yes

No

141

Comments

80.0%

: 70.7%

Referencestandard

Yes No

20 12 32

5 29 34

25 41 66

92.0%

: 29.3%

Referencestandard

Yes No

23 29 52

2 12 14

25 41 66

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; noreport onblinding; unclearwhich is themodified Nortonscale; no reporton criteria of PUclassification norassessment; noreport onpreventivemeasure; noreport no sub-analysesaccording topreventivemeasures.

Additionaloutcomes:sensitivity andspecificity onday 0

142


weekly basis forthree weeks or untildischarge or death.

Reference standard:PU developmentbased on recordreview wasperformed daily.



Addressing missingdata: 66 patientswere assessed onadmission and afterone week. 34patients wereassessed after 2weeks and 17 after 3weeks.

Statistical analysis:Inferential statisticswere used tocompare the totalscores

(predicted risk) withthe outcome(pressure ulcerdevelopment), in

Occiput: 7

Buttocks: 7

Sacrum: 3

Ankles: 2

Knees: 2

Elbows: 1

Ears: 1

Nose: 1

Forehead: 1


Inclusion criteria:

Age between 18 and 65years;

no pressure ulcer onadmission;

total bedrest due toinjuries or medicalinterventions

Exclusion criteria:extensive burns in theback, buttocks and legs


Intervention

Comparison

Outcomemeasures

Results

Number of patients

Age between 18 and 65

no pressure ulcer on

total bedrest due toinjuries or medical

Exclusion criteria:the

KCE Report 193S

Comments

Notes: /

KCE Report 193S


order to determinetheir predictivevalues.

Setting: the ICU of aprivate sector healthcare institution,South Africa.


Table 71 – Page 2011


Author and year: Page(2011)

Title: Developmentand validation ofpressure ulcer riskassessment tool foracute hospitalpatients

Journal: WoundRepair andRegeneration, 19; 31-37.


Selection patient:

Patient group:hospitalized patients.

All patients

Included N: 165

Completed N: 165

Drop-outs: 0

Number of patients >65 years: 107

Gender (m/f): 87/78

Length of stay (meandays (SD)): 14.97(22.29)



Intervention

Comparison

Outcomemeasures

Results

Intervention

Comparison

Outcomemeasures

Results

Patient group:

Number of patients >


Number of patients

Index test 1: The NorthernHospital Pressure UlcerPrevention Plan (TNH-PUPP) (Page 2011)

Reference standard:development of pressureulcer grade 1; not furtherspecified

Preventive methods:

A prevention protocol wasimplemented.

Outcome 1:

Incidence of PU

(not reported)

Outcome 2:

Area under theROC TNH-PUPP

Outcome 3:

Sensitivity andspecificity TNH-PUPP cut-off 1

Value: 4.2%

Value: 0.90

95% CI: 0.82-0.99

Sensitivity: 100.0%

Specificity: 34.2%

Raw data

143

Comments

Comments

0.99

100.0%

: 34.2%

Referencestandard

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; noreport on time ofassessment ofindex test andreferencestandard; notreported whenpatientsdropped fromthe study; noinclusion andexclusion

144


Patients admitted toa general ward,critical care oremergencydepartment of ahospital.


Index test: TheNorthern HospitalPressure UlcerPrevention Plan wasused to assess PUrisk. Patients werescreened by trainednurses.

Reference standard:PU development wasidentified by thenursing staff whoreceived aneducation session of30 minutes.



Addressing missingdata: not reportedwhen patients


Inclusion criteria:

/



Intervention

Comparison

Outcomemeasures

Results

Number of patients

Outcome 4:


Outcome 5:


Indextest

Yes

No

Sensitivity: 85.7%

Specificity: 62.0%

Raw data

Indextest

Yes

No

Sensitivity: 71.4%

Specificity: 81.0%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

Yes No

7 104 111

0 54 54

7 158 165

85.7%

: 62.0%

Referencestandard

Yes No

6 60 66

1 98 99

7 158 165

71.4%

: 81.0%

Referencestandard

Yes No

5 30 35

2 128 130

criteria reported;no report onblinding; noreport on criteriaof PUclassification; noreport no sub-analysesaccording topreventivemeasures.

Additionaloutcomes:

Notes: /

KCE Report 193S


dropped from thestudy

Statistical analysis:The predictiveaccuracy of the TNH-PUPP was measured

by the parametersarea under thereceiver operatingcurve (AUC),sensitivity,specificity, PPV,NPV, Youden Index,and prognosticseparation index. AnAUC of 1 indicates

perfect prediction,whereas 0.5represents theprediction expectedby chance.Sensitivity,specificity, PPV, andNPV values > 0.70are reported to beevidence of highpredictive accuracy.

Setting: the generalwards, critical careand emergencydepartment of anacute, metropolitan,public teaching


Intervention

Comparison

Outcomemeasures

Results

Outcome 6:


Outcome 7:


Outcome 8:


Sensitivity: 71.4%

Specificity: 88.0%

Raw data

Indextest

Yes

No

Sensitivity: 42.9%

Specificity: 96.2%

Raw data

Indextest

Yes

No

Sensitivity: 57.1%

Specificity: 99.4%

145

Comments

7 158 165

71.4%

: 88.0%

Referencestandard

Yes No

5 19 24

2 139 141

7 158 165

42.9%

: 96.2%

Referencestandard

Yes No

3 6 9

4 152 156

7 158 165

57.1%

: 99.4%

146


hospital inMelbourne, Australia.




Author and year:Serpa (2009)

Title: Predictivevalidity of WaterlowScale for pressureulcer developmentrisk in hospitalizedpatients.

Journal: Journal ofWound Ostomy &Continence Nursing,36(6); 640-646.


Patient group:hospitalized patientsolder than 18 years.

All patients

Included N: 98

Completed N: 98

Drop-outs: 0 beforethree consecutiveassessments




Intervention

Comparison

Outcomemeasures

Results

Raw data

Indextest

Yes

No

Intervention

Comparison

Outcomemeasures

Results

Patient group:hospitalized patients

0 beforeve

Age (mean years (SD);71.1 (15.5); 29-

Number of patients

Index test 1: the PortugueseWaterlow scale (Paranhos &Santos, 1999)

Reference standard:development of pressureulcer; not further specified.

Preventive methods:

Not reported

Outcome 1:

Incidence of PU

(< 1 week; 2days)

Outcome 2:

Area under theROC firstassessment(48h)

Outcome 3:

Area under theROC second

Value: 7.1%

Value: 0.64

95% CI: 0.35-0.93

Value: 0.59

95% CI: 0.34-0.83

KCE Report 193S

Comments

Referencestandard

Yes No

4 1 5

3 157 160

7 158 165

Comments

0.93

0.83

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; noreport onblinding; noreport on skinassessment andcriteria ofclassification; noreport onpreventivemeasures; nosub-analyses

KCE Report 193S


(secondary analysis)

Selection patient:

Patients at risk forPU from any ward ina general privatehospital.


Index test:Portuguese Waterlowscale was used toassess PU risk atadmission. Thepatient was assessedfor the first time andthen at 48-hoursintervals as long asthe patient remainedat risk or until PUdevelopment,discharge, transferor death.

Reference standard:PU development; nofurther information.



Stage I: 6

Stage II: 1


Inclusion criteria:

Age equal to 18 years orolder;

absence of PU at firstassessment;hospitalized for aminimum period of 24hours and a maximumperiod of 48 hours atfirst assessment;

a total Braden Scalescore equal to 18 or lessand a Waterlow Scalescore equal to 16 ormore.

Exclusion criteria:

Additional criteria (datafrom another study):patients with chronicrenal failure; patients ondialyse for more than 1month; patients withliver insufficiencyaccompanied withascites.


Intervention

Comparison

Outcomemeasures

Results

Number of patients

Age equal to 18 years or

absence of PU at first

for aminimum period of 24hours and a maximumperiod of 48 hours at

a total Braden Scalescore equal to 18 or lessand a Waterlow Scalescore equal to 16 or

Additional criteria (datafrom another study):

ents with chronicrenal failure; patients ondialyse for more than 1month; patients withliver insufficiencyaccompanied with

assessment (4days)

Outcome 4:

Area under theROC thirdassessment (6days)

Outcome 5:

Sensitivity andspecificityWaterlow scalecut-off 17 // 48h

Outcome 6:

Sensitivity andspecificityWaterlow scalecut-off 20 // 4days

Value: 0.54

95% CI: 0.35-0.74

Sensitivity: 71.4%

Specificity: 67.0%

Raw data

Indextest

Yes

No

Sensitivity: 85.7%

Specificity: 40.7%

Raw data

147

Comments

0.74

71.4%

: 67.0%

Referencestandard

Yes No

5 30 35

2 61 63

7 91 98

85.7%

: 40.7%

Referencestandard

according topreventivemeasures.

Only patients atrisk wereincluded!


Notes: Bradenscale scoreswere alsocollected, but noresults of thesescores werereported.

148



Statistical analysis:The predictivevalidity of theWaterlow Scale forthe development

of PU in hospitalizedpatients wasanalyzed by using 2methods: receiveroperatingcharacteristic (ROC)

curve and likelihoodratio (LR).

Setting: a medium-size general privatehospital in the city ofSão Paulo, Brazil.

Blinding: notreported.


Intervention

Comparison

Outcomemeasures

Results

Outcome 7:

Sensitivity andspecificityWaterlow scalecut-off 20 // 6days

Indextest

Yes

No

Sensitivity: 85.7%

Specificity: 33.0%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

Yes No

6 54 60

1 37 38

7 91 98

85.7%

: 33.0%

Referencestandard

Yes No

6 61 67

1 30 31

7 91 98

KCE Report 193S



Author and year:Serpa (2011)

Title: Predictivevalidity of the Bradenscale for pressureulcer risk on criticalcare patients.

Journal: RevistaLatino-Americana deEnfermagem, 19(1);50-57.

Study type:prospective study(secondary analysis)

Selection patient:

Patients at risk forPU from an ICU.


Index test:Portuguese Bradenscale was used toassess PU risk atadmission. Thepatient was assessedfor the first time andthen at 48-hoursintervals as long asthe patient remained

Patient group: ICUpatients older than 18years.

All patients

Included N: 72

Completed N: 72

Drop-outs: 0 beforethree consecutiveassessments

Age (mean years(SD);): 60.9 (16.5)


Stage I: 3

Stage II: 5


Inclusion criteria:

Admitted to one of thefour ICUs; age equal to18 years or older;

absence of PU at firstassessment;hospitalized for aminimum period of 24hours and a maximumperiod of 48 hours at


Intervention

Comparison

Outcomemeasures

Results


0 beforethree consecutive

Age (mean years

Number of patients

Number of patients

Admitted to one of thefour ICUs; age equal to

absence of PU at first

hospitalized for aminimum period of 24hours and a maximumperiod of 48 hours at

Index test 1: the PortugueseBraden scale (Paranhos &Santos, 1999)

Reference standard:development of pressureulcer; not further specified.

Preventive methods:

Not reported

Outcome 1:

Incidence of PU

(< 1 week; 2days)

Outcome 2:

Area under theROC firstassessment(48h)

Outcome 3:

Area under theROC secondassessment (4days)

Outcome 4:

Area under theROC thirdassessment (6days)

Outcome 5:

Sensitivity andspecificityBraden scalecut-off 12 // 48h

Value: 11.1%

Value: 0.788

95% CI: 0.29-1.00

Value: 0.789

95% CI: 0.28-1.00

Value: 0.800

95% CI: 0.28-1.00

Sensitivity: 87.5%

Specificity: 64.1%

Raw data

149

Comments

1.00

1.00

1.00

87.5%

: 64.1%

Reference

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; noreport onblinding; noreport on skinassessment andcriteria ofclassification; noreport onpreventivemeasures; nosub-analysesaccording topreventivemeasures.

Only patients atrisk wereincluded!


Notes: Bradenscale scoreswere also

150


at risk or until PUdevelopment,discharge, transferor death.

Reference standard:PU development; nofurther information.



Addressing missingdata: patient stayedfor a minimum of 6days.

Statistical analysis:Sensitivity wasdefied as theproportion ofindividuals with apositive test whodevelop a disease,and specificity as theproportion ofindividuals with anegative test who donot develop adisease.

The ROC curve is agraphic plot of true

first assessment;

a total Braden Scalescore equal to 18 orless; informed consent.

Exclusion criteria:

Additional criteria (datafrom another study):patients with chronicrenal failure; patients ondialyse for more than 1month; patients withliver insufficiencyaccompanied withascites.


Intervention

Comparison

Outcomemeasures

Results

a total Braden Scalescore equal to 18 or

Additional criteria (datafrom another study):patients with chronicrenal failure; patients ondialyse for more than 1month; patients withliver insufficiency

mpanied with

Outcome 6:

Sensitivity andspecificityBraden scalecut-off 13 // 4days

Outcome 7:

Sensitivity andspecificityBraden scalecut-off 13 // 6days

Indextest

Yes

No

Sensitivity: 75.0%

Specificity: 81.3%

Raw data

Indextest

Yes

No

Sensitivity: 75.0%

Specificity: 82.8%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

standard

Yes No

7 23 30

1 41 42

8 64 72

75.0%

: 81.3%

Referencestandard

Yes No

6 12 18

2 52 54

8 64 72

75.0%

: 82.8%

Referencestandard

Yes No

6 11 17

2 53 55

collected, but noresults of thesescores werereported.

KCE Report 193S


positive values(sensitivity) on theordinate and falsepositive values (1 –specificity) on theabscissa as afunction of each cut-off point. There is anapproximately linearquantitative-qualitativerelationship betweenthe area under thecurve (AUC) andaccuracy, which canbe classified asfollows: excellent(0.80-0.90), very good(0.70-0.79), good(0.60-0.69), and poor(0.50-0.59)

Setting: four ICUs ofa large, non-profitcharitable generalhospital, Brazil.



Intervention

Comparison

Outcomemeasures

Results

151

Comments

8 64 72

152

Table 74 – Suriadi 2006


Author and year:Suriadi (2006)

Title: A newinstrument forpredicting pressureulcer risk in anintensive care unit.

Journal: Journal ofTissue Viability,16(3); 21-26.


Selection patient:

Patients admitted toan ICU.


Index test: TheBraden scale wasused to assess PUrisk after 24 hours.This assessment wasrepeated three timesa week (stay: 3-22days). Patients werescreened by aresearch assistant.

Reference standard:Skin condition wasassessed daily (stay:

Patient group: ICUpatients of all age.

All patients

Included N: 105

Completed N: 105

Drop-outs: 0

Group PU+


Gender (m/f): 24/11


Stage I: 21

Stage II: 14

PU location:

Sacrum: 28

Heel: 4

Trochanter: 1

Elbow: 2

Vertebrae: 1

Scapula: 1

More than one PU: 3

Group PU-


Intervention

Comparison

Outcomemeasures

Results

ICU

Age (mean years (SD);50.9 (17.0); 17-

Number of patients


Reference standard:development of pressureulcer according to the criteriaof the NPUAP classification(Burd et al., 1992).

Preventive methods:

Not reported

Outcome 1:

Incidence of PU

(> 1 week; 22days)

Outcome 2:

Area under theROC

Outcome 3:


Value: 33.3%

Value: 0.770

95% CI: 0.70-0.89

Sensitivity: 80.0%

Specificity: 54.3%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

0.89

80.0%

: 54.3%

Referencestandard

Yes No

28 32 60

7 38 45

35 70 105

Funding: /

Limitations: noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; noreport onpreventivemeasures; nosub-analysesaccording topreventivemeasures.


Notes: /

KCE Report 193S


3-22 days) by theprimary researcher.





In the statisticalmethods, diagnosticprobabilities(sensitivity,specificity, positivepredictive value(PPV), and negativepredictive value(NPV)) werecalculated.

In this study we alsoevaluated thelikelihood ratio (LR)for this tools.

A receiver-operatingcharacteristic (ROC)curve plot of thesensitivity versus 1-

Age PU- (mean years(SD); range): 47.5(17.6); 17-82

Gender (m/f): 48/22


Inclusion criteria:

Free of pressure ulcer;bedfast;

could not walk.

Exclusion criteria:

Physically incapable ofparticipating;

refusal


Intervention

Comparison

Outcomemeasures

Results

(mean years47.5

Number of patients

ee of pressure ulcer;

Physically incapable of

153

Comments

154


specificity over therange of the Bradenscale scoresconfirmed the cut-offvalue of theinstrument

Setting: an intensivecare unit withinPontianak PublicHospital, Sei Jawi inWest Kalimantan,Indonesia

Blinding: The Bradenscale was used by aresearch assistantand the skincondition wasassessed by theprimary researcher.

Table 75 – Suriadi 2008


Author and year:Suriadi (2008)

Title: Development ofa new riskassessment scale forpredicting pressureulcers in an intensivecare unit.

Patient group: ICUpatients older than 18yrs.

All patients

Included N: 253

Completed N: 253


Intervention

Comparison

Outcomemeasures

Results

Intervention

Comparison

Outcomemeasures

Results


Index test 1: the Suriadi andSanada scale

Reference standard:development of pressureulcer according to the criteriaof the NPUAP classification(Ayello et al. 2003).

Outcome 1:

Cumulativeincidence of PU

Outcome 2:

Incidencedensity of PU

Unit 1: 27%

Unit 2: 31.6%

Total: 28.5%

Unit 1: 0.060/100

Unit 2: 0.059/100 person days

KCE Report 193S

Comments

Comments

0.060/100 person days

0.059/100 person days

Funding: /

Limitations:only part indextest repeated;end ofobservation PUdevelopment not

KCE Report 193S


Journal: BritishAssociation ofCritical Care Nurses,13(1); 34-43.


Selection patient:

Patients admitted toan ICU.

Patients wereselected by theresearcher.

Index test: The SS(Suriadi and Sanada)scale was used toassess PU risk within24 hours. Bodytemperature wasrepeated once a day.Patients werescreened by aresearch assistant.

Reference standard:Skin condition wasassessed daily by aresearcher.


Drop-outs: 0

ICU 1

Included N: 174

Completed N: 174

Drop-outs: 0




Stage I: 20

Stage II: 22

Stage III: 5

Stage IV: 1

One patient had morethan one PU

PU location:

Sacrum: 44

Heel: 2

Trochanter: 1

Malleolus: 1

ICU 2

Included N: 79

Completed N: 79

Drop-outs: 0


Intervention

Comparison

Outcomemeasures

Results

Age (mean years

Number of patients

patient had more

Preventive methods:

Not reported Outcome 3:

Area under theROC

Outcome 4:

Sensitivity andspecificity SSscale cut-off 0

Outcome 5:


Value: 0.888

95% CI: 0.84-0.93

Sensitivity: 100.0%

Specificity: 0.0%

Raw data

Indextest

Yes

No

Sensitivity: 97.2%

Specificity: 42.0%

Raw data

Indextest

Yes

No

155

Comments

0.93

100.0%

: 0.0%

Referencestandard

Yes No

72 181 253

0 0 0

72 181 253

97.2%

: 42.0%

Referencestandard

Yes No

70 105 175

2 76 78

72 181 253

reported; noimputation, noexclusion; lowevent rate; notreported whenpatientsdropped fromthe study; noreport onpreventivemeasures; nosub-analysesaccording topreventivemeasures; noreport onwithdrawal.


Notes: /

156





To evaluate theaccuracy of the S.S.scale, diagnostic

probabilities[sensitivity,specificity, positivepredictive

value (PPV), negativepredictive value(NPV) and thelikelihood ratio (LR)]were calculated forthe range of the S.S.score. Area underthe curve (AUC) ofthe ROC wascalculated to assessthe overall validity ofthe scale

Incidence density iscomputed as thenumber of personsdeveloping new


Gender (m/f): 54/25


Stage I: 12

Stage II: 13

PU location:

Sacrum: 25

Inclusion criteria:

Aged 18 yrs or more;admitted to the ICU atleast 24h beforeenrolment; bedfast; noexisting PU at time ofenrolment; ability to giveinformed consent;Indonesian origin.

Exclusion criteria:

Active skin disease;previous enrolment inthe study; physicallyincapable ofparticipating; length ofstay < 72 h after initialdata collection.


Intervention

Comparison

Outcomemeasures

Results

Age (mean years

Number of patients

Aged 18 yrs or more;admitted to the ICU atleast 24h beforeenrolment; bedfast; noexisting PU at time ofenrolment; ability to giveinformed consent;

Active skin disease;previous enrolment inthe study; physicallyincapable ofparticipating; length ofstay < 72 h after initial

Outcome 6:


Outcome 7:


Outcome 8:


Sensitivity: 97.2%

Specificity: 53.0%

Raw data

Indextest

Yes

No

Sensitivity: 80.6%

Specificity: 82.9%

Raw data

Indextest

Yes

No

Sensitivity: 72.2%

Specificity: 86.7%

Raw data

KCE Report 193S

Comments

97.2%

: 53.0%

Referencestandard

Yes No

70 85 155

2 96 98

72 181 253

80.6%

: 82.9%

Referencestandard

Yes No

58 31 89

14 150 164

72 181 253

72.2%

: 86.7%

Reference

KCE Report 193S


pressure ulcers(numerator) dividedby the total person-days [sum of all thedays over whicheach patientparticipated in thestudy (denominator)]

Setting: twointensive care unitsof two hospitals inPontianak, Indonesia

Blinding: two nursesbeing assessorsused their assignedscale toindependently

assess the patients.


Intervention

Comparison

Outcomemeasures

Results

Outcome 9:


Outcome 10:


Indextest

Yes

No

Sensitivity: 61.1%

Specificity: 92.3%

Raw data

Indextest

Yes

No

Sensitivity: 58.3%

Specificity: 95.0%

Raw data

Index Yes

157

Comments

standard

Yes No

52 24 76

20 157 177

72 181 253

61.1%

: 92.3%

Referencestandard

Yes No

44 14 58

28 167 195

72 181 253

58.3%

: 95.0%

Referencestandard

Yes No

42 9 51

158



Intervention

Comparison

Outcomemeasures

Results

Outcome 11:


test No

Sensitivity: 6.9%

Specificity: 100.0%

Raw data

Indextest

Yes

No

KCE Report 193S

Comments

30 172 202

72 181 253

6.9%

: 100.0%

Referencestandard

Yes No

5 0 5

67 181 248

72 181 253

KCE Report 193S

3. SKIN ASSESSMENT – CLINICAL



Protocol Skin assessment

Review question What is the clinical effectiveness of skin assessment methods in the prevention of pressure ulcers?

Population Individuals of all

Intervention Diascopy: finger method and transparent disk

Measuring of skin temperature


No skin assessment

Other

Outcomes Critical outcome for decision

Proportion of participants developing new pressure ulcers (dichotomulcer)

Important outcomes

Rate of development of pressure ulcers Time to develop new pressure ulcer (time to event data) Time in hospital (continuous data) Patient acceptability Health-related quality of lif


o

o

o

o

o

o

o

Study design High quality systematic reviews of RCTs or RCTs only


CLINICAL EFFECTIVENESS

Skin assessment

What is the clinical effectiveness of skin assessment methods in the prevention of pressure ulcers?

Individuals of all in all settings

Diascopy: finger method and transparent disk

Measuring of skin temperature

Each other

No skin assessment

Critical outcome for decision-making

Proportion of participants developing new pressure ulcers (dichotomous outcome)

Important outcomes

Rate of development of pressure ulcersTime to develop new pressure ulcer (time to event data)Time in hospital (continuous data)Patient acceptability

related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes inpressure ulcer patients, therefore may have to be narratively summarised)



EQ-5D

WHOQOL-BREF

Cardiff HRQoL tool

HUI

Pressure ulcer quality of life (Gorecki)

High quality systematic reviews of RCTs or RCTs only

159

What is the clinical effectiveness of skin assessment methods in the prevention of pressure ulcers?

ous outcome) (describe different categories of

(although unlikely to be sensitive enough to detect changes in

160


Cochrane reviews will be included if they match the inclusion criteria and have appropriate assumptions for missingdata such a


Exclusion Studies with another population, intervention, comparison or outcome. Non-English, non


Medline (OVID interface), Cinahl (EBSCO

Review strategy How will individual PICO characteristics be co

Population: any population will be combined except those specified in the strata Intervention Comparison Outcomes Blinding Minimum follow up = no minimum

Minimum total size = no minimum Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groups

or if the missiauthor’s data.

Analysis Strata:



Subgroups:


Different categories of pressure ulcer (from category 2 upwards where

Different ulcer locations: sacral, heel and others


Skin assessment

Cochrane reviews will be included if they match the inclusion criteria and have appropriate assumptions for missingdata such as available case analysis or Intention to treat (with the appropriate assumptions).


Studies with another population, intervention, comparison or outcome.English, non-French, non-Dutch language papers


Medline (OVID interface), Cinahl (EBSCO-interface), Embase, Library of the Cochrane Collaboration (All years)

How will individual PICO characteristics be combined across studies)

Population: any population will be combined except those specified in the strataIntervention – any intervention will be combinedComparison – any comparison will be combinedOutcomes – same outcomes will be combinedBlinding – Blinded and not-blinded studies will be meta-analysed togetherMinimum follow up = no minimum

Minimum total size = no minimumUse available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.


Children (neonates, infants and children)ients;



Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separately

Different ulcer locations: sacral, heel and others

KCE Report 193S

Cochrane reviews will be included if they match the inclusion criteria and have appropriate assumptions for missing(with the appropriate assumptions).

interface), Embase, Library of the Cochrane Collaboration (All years)

Population: any population will be combined except those specified in the strata

Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsng data is higher than the event rate, if cannot work out the available case analysis will take the


outcomes are reported separately

KCE Report 193S

3.2. Search Strategy



Date 30/08/2012

Database Ovid MEDLINE(R) In-Process & Other Non

Search Strategy 1. Pressure ulcer.sh2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage)).ti,ab.4. (bedsore* or bed-5. ((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,6. OR/1 – 57. finger method.tw8. transparent disk*.tw9. diascopy.tw10. ultrasonograph*.tw11. ultrasonography.sh12. ultrasonics.sh13. ultrasound.tw14. durometer.tw15. durometry.tw16. elastometer.tw17. haptic finger.tw18. digital imag*.tw19. digital colo?r imag*.tw20. spectrometer.tw21. multispectral imag*.tw22. multiwavelength imag*.tw23. clinical assessment.tw24. transcutaneous oximetry.tw25. Blood Gas Monitoring, Transcutaneous.sh26. tympanic thermometer*.tw27. Doppler blood flowmetry.tw28. laser Doppler imag*.tw29. Minimum Data Set.tw or MDS.tw or RAI.tw


Process & Other Non-Indexed Citations and Ovid MEDLINE(R) 1946 to Present

Pressure ulcer.sh

(pressure adj (sore* or ulcer* or damage)).ti,ab.-sore*).ti,ab.

((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,

transparent disk*.tw

ultrasonograph*.twultrasonography.sh

digital colo?r imag*.tw

al imag*.twmultiwavelength imag*.twclinical assessment.twtranscutaneous oximetry.twBlood Gas Monitoring, Transcutaneous.shtympanic thermometer*.twDoppler blood flowmetry.twlaser Doppler imag*.twMinimum Data Set.tw or MDS.tw or RAI.tw

161

INE(R) 1946 to Present

((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti, ab.

911839486254506654

1381813

729

698726051119248

13440195

924

19221

4718194

3758

14292137

201214441

484

162

Date 30/08/2012

30. skin assess*.tw31. skin inspect*.tw32. skin exam*.tw33. skin eval*.tw34. skin observ*.tw35. skin risk assess*.tw36. skin status.tw37. skin condition.tw38. judgment.sh39. clinical judgment.tw40. skin temperature.tw41. skin temperature.sh42. OR/7 – 4243. AND/6, 4344. Limit language: ‘English, Dutch, Flemish, Fre


skin risk assess*.tw

clinical judgment.twskin temperature.twskin temperature.sh

Limit language: ‘English, Dutch, Flemish, French’

KCE Report 193S

1162015346

5956156

568

122111538

288953448195

313866665619

KCE Report 193S


Date 30/08/2012

Database Embase

Search Strategy


1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (4. (bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab5. ((moist* or friction or shear) near/2 (sore* or ulcer* or dam6. OR/1 – 57. ‘finger method’:ti,ab8. ‘transparent disk’:ti,ab9. ‘diascopy’:ti,ab10. ultrasonograph*:ti,ab11. ‘echography’/exp12. ‘Doppler echography’/exp13. ‘color ultrasound flowmetry’/exp14. ‘ultrasound’/exp15. ‘ultrasound’:ti,ab16. ‘durometer’:ti,ab17. ‘durometry’:ti,ab18. ‘elastometer’:ti,ab19. ‘haptic finger’:ti,ab20. (digital NEXT/1 imag*):ti,ab21. (‘digital colo?r’ NEXT/1 imag*):ti,ab22. ‘spectrometer’:ti,ab23. ‘mass spectrometer’/exp24. (multispectral NEXT/1 imag*):ti,ab25. (multiwavelength NEXT/1 imag*):ti,ab26. ‘clinical assessment’:ti,ab27. ‘clinical assessment’/exp28. ‘transcutaneous oximetry’:ti,ab29. ‘transcutaneous oxygen monitoring’/exp30. ‘thermographic scanner’:ti,ab31. (tympanic NEXT/1 thermometer*):ti,ab32. ‘tympanic thermometer’/exp33. ‘Doppler blood flowmetry’:ti,ab


NEAR/1 (sore* OR ulcer* OR damage)):ab,ti(bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

‘finger method’:ti,ab‘transparent disk’:ti,ab

ultrasonograph*:ti,ab‘echography’/exp‘Doppler echography’/exp‘color ultrasound flowmetry’/exp

‘elastometer’:ti,ab‘haptic finger’:ti,ab(digital NEXT/1 imag*):ti,ab(‘digital colo?r’ NEXT/1 imag*):ti,ab‘spectrometer’:ti,ab‘mass spectrometer’/exp(multispectral NEXT/1 imag*):ti,ab(multiwavelength NEXT/1 imag*):ti,ab‘clinical assessment’:ti,ab‘clinical assessment’/exp‘transcutaneous oximetry’:ti,ab‘transcutaneous oxygen monitoring’/exp‘thermographic scanner’:ti,ab(tympanic NEXT/1 thermometer*):ti,ab‘tympanic thermometer’/exp‘Doppler blood flowmetry’:ti,ab

163

age or wound* or injur* or lesion*)):ti,ab

1335554597477741812

1826320

640

88906447225266801994885221

179253110

932

110513

1419687

3983379

71865949109

1642268

716863

164

Date 30/08/2012

34. ‘Doppler flowmetry’/exp35. ‘blood flowmetry’/exp36. (‘laser Doppler’ NEXT/1 imag*):ti,ab

37. ‘laser Doppler flowmetry’/exp38. ‘Minimum Data Set’:ti,ab or ‘MDS’:ti,ab or ‘RAI’:ti,ab39. (skin NEXT/1 assess*):ti,ab40. (skin NEXT/1 inspect*):ti,ab41. (skin NEXT/1 exam*):ti,ab42. (skin NEXT/1 eval*):ti,ab43. (skin NEXT/1 observ*):ti,ab44. (‘skin risk’ NEXT/1 assess*):ti,ab45. ‘skin status’:ti,ab46. ‘skin condition’:ti,ab47. ‘clinical judgment’:ti,ab48. ‘clinical observation’/exp49. ‘skin temperature’:ti,ab50. ‘skin temperature’/exp51. OR/7 – 5052. AND/6, 5153. Limit language: ’English, Dutch, French’ and limited to embase


ler flowmetry’/exp‘blood flowmetry’/exp(‘laser Doppler’ NEXT/1 imag*):ti,ab

‘laser Doppler flowmetry’/exp‘Minimum Data Set’:ti,ab or ‘MDS’:ti,ab or ‘RAI’:ti,ab(skin NEXT/1 assess*):ti,ab(skin NEXT/1 inspect*):ti,ab(skin NEXT/1 exam*):ti,ab

/1 eval*):ti,ab(skin NEXT/1 observ*):ti,ab(‘skin risk’ NEXT/1 assess*):ti,ab

‘skin condition’:ti,ab‘clinical judgment’:ti,ab‘clinical observation’/exp‘skin temperature’:ti,ab‘skin temperature’/exp

e: ’English, Dutch, French’ and limited to embase

KCE Report 193S

6123546

2867604

783116943

21349

8758785

996

17683472

1603565249393

7437751195794

KCE Report 193S


Date 30/08/2012


Search Strategy

(attention, for PubMed, check« Details »):ti,ab,kw

1. MeSH descriptor “Pressure ulcer” explod2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw4. (bedsore* or bed5. ((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

lesion*)):ti,ab,kw6. OR/1 – 57. (finger method):ti,ab8. (transparent disk*):ti,ab,kw9. (diascopy):ti,ab,kw10. (ultrasonograph*):ti,ab,kw11. MeSH descriptor “ultrasonography” explode all trees12. MeSH descriptor “ultrasonics” explode all trees13. (ultrasound):ti,ab,kw14. (durometer) :ti,ab,kw15. (durometry) :ti,ab,kw16. (elastomet17. (haptic finger) :ti,ab,kw18. (digital imag*):ti,ab,kw19. (digital colo?r imag*):ti,ab,kw20. (spectrometer):ti,ab,kw21. (multispectral imag*):ti,ab,kw22. (multiwavelength imag*):ti,ab,kw23. (clinical assessment):ti,ab,kw24. (transcutaneous oximetry):ti,ab,kw25. MeSH descriptor “Blood Gas Monitoring, Transcutaneous” explode all trees26. (thermographic scanner):ti,ab,kw27. (tympanic thermometer*):ti,ab,kw28. (Doppler blood flowmetry):ti,ab,kw29. (laser Doppler imag*):ti,ab,kw30. (Minimum Data Set):ti,ab,kw or (MDS) :ti,ab,kw or (RAI):31. (skin assess*):ti,ab,kw32. (skin inspect*):ti,ab,kw



MeSH descriptor “Pressure ulcer” explode all treesDecubit*:ti,ab,kw(pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw(bedsore* or bed-sore*):ti,ab,kw((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*orlesion*)):ti,ab,kw

5(finger method):ti,ab,kw(transparent disk*):ti,ab,kw(diascopy):ti,ab,kw(ultrasonograph*):ti,ab,kwMeSH descriptor “ultrasonography” explode all trees

eSH descriptor “ultrasonics” explode all trees(ultrasound):ti,ab,kw(durometer) :ti,ab,kw(durometry) :ti,ab,kw(elastometer) :ti,ab,kw(haptic finger) :ti,ab,kw(digital imag*):ti,ab,kw(digital colo?r imag*):ti,ab,kw(spectrometer):ti,ab,kw(multispectral imag*):ti,ab,kw(multiwavelength imag*):ti,ab,kw(clinical assessment):ti,ab,kw(transcutaneous oximetry):ti,ab,kw

descriptor “Blood Gas Monitoring, Transcutaneous” explode all trees(thermographic scanner):ti,ab,kw(tympanic thermometer*):ti,ab,kw(Doppler blood flowmetry):ti,ab,kw(laser Doppler imag*):ti,ab,kw(Minimum Data Set):ti,ab,kw or (MDS) :ti,ab,kw or (RAI):ti,ab,kw(skin assess*):ti,ab,kw(skin inspect*):ti,ab,kw

165

((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

4903498343363

1168940

20

93656678230

66266109

70624

11440

1872249

173

136

808150655

166

Date 30/08/2012

33. (skin exam*):ti,ab,kw34. (skin eval*):ti,ab,kw35. (skin observ*):ti,ab,kw36. (skin risk assess*):ti,ab,kw37. (skin status):ti,ab,kw38. (skin condition*):ti,ab,kw39. Mesh descriptor “Judgment” explode all trees40. (clinical judgement):ti,ab,kw41. (Skin temperature):ti,ab,kw42. Mesh descriptor “skin temperature” explode all trees43. OR/7 – 4244. AND/6, 43


Date 30/08/2012

Database CINAHL

Search Strategy

(attention, forPubMed, check« Details »)

1. MH “Pressure Ulcer”2. Decubit*3. Pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 dam4. Bedsore* OR bed-sore*5. ((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))6. OR/1 – 57. finger method8. transparent disk*9. diascopy10. ultrasonograph*11. MH ultrasonography12. MH ultrasonics13. Ultrasound14. durometer15. durometry16. elastometer


(skin exam*):ti,ab,kw(skin eval*):ti,ab,kw(skin observ*):ti,ab,kw(skin risk assess*):ti,ab,kw(skin status):ti,ab,kw(skin condition*):ti,ab,kwMesh descriptor “Judgment” explode all trees

ical judgement):ti,ab,kw(Skin temperature):ti,ab,kwMesh descriptor “skin temperature” explode all trees

42AND/6, 43

MH “Pressure Ulcer”

Pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*sore*

((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))

MH ultrasonography

KCE Report 193S

460257

188651093358675637

1635430405

1837708

46802203


7748487

8540157

1424

987630

42

279837546669

1128617

2

KCE Report 193S

Date 30/08/2012

17. haptic finger18. digital imag*19. digital color imag* or digital colour imag*20. spectrometer21. multispectral imag*22. multiwavelength imag*23. clinical assessment24. transcutaneous oximetry25. MH Blood Gas Monitoring, Transcutaneous26. thermographic scanner27. tympanic thermometer*28. Doppler blood flowmetry29. laser Doppler imag*30. “Minimum Data Set” or MDS or RAI31. skin assess*32. skin inspect*33. skin exam*34. skin eval*35. skin observ*36. skin risk assess*37. skin status38. skin condition39. MH judgment40. clinical judgment41. skin temperature42. MH skin temperature43. OR/7 – 4244. AND/6, 4345. Limit language=’English, Dutch, French’


digital color imag* or digital colour imag*

multiwavelength imag*

transcutaneous oximetryMH Blood Gas Monitoring, Transcutaneousthermographic scanner

ometer*Doppler blood flowmetry

“Minimum Data Set” or MDS or RAI

re

Limit language=’English, Dutch, French’

167

01

230126

11490

5900643

2963

1396297

290877057

51076028680

127350

1889875

1041703

10102317781631

168



Potentially relevant citations identified:

Potentially relevant citations identified afterduplicates removal:

Additionally potentially relevant citations(hand searching): 0

Relevant studies:

Studies retrieved for more detailed evaluation:


Potentially relevant citations identified: 3247

y relevant citations identified afterduplicates removal: 2576

Additionally potentially relevant citations

Based on title and abstract evaluation,citations excluded: 2575

Relevant studies: 1

Based on full text evaluation, studiesexcluded: 0

Studies retrieved for more detailed evaluation: 1

KCE Report 193S

Based on title and abstract evaluation,

Based on full text evaluation, studies

KCE Report 193S



The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in CinahlEmbase, and 203 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted in 2 576 recordtitle and abstract 2 575 records were excluded. Reasons for exclusion were listed. The fullthis review.


One randomized controlled trial of Vanderwee (2007) was included in this reviewprofiles. The forest plot and the study evidence table are presented in respectivelymedians and interquartile ranges, medians were used as a surrogate for means and standard deviations were estimated as 80% of the interqu




Vanderwee 200759

(1) Daily skin assessment withtransparent disk.measures were started whennon-blanchable erythema (appeared.

(2) Braden scoreassessment with transparentdisk. Preventive measures werestarted if the Braden score was<17 or NBE appeared.

Patients received preventivemeasures according to the samepressure redistribution protocol.

The patients were randomized toeither the Polyéthylènemattress (TempurLexington, Kentucky USA), or tothe Alternating pressure airmattress (Alpha-


The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in CinahlEmbase, and 203 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted in 2 576 recordtitle and abstract 2 575 records were excluded. Reasons for exclusion were listed. The full text of the remaining record was reviewed in detail and included in

One randomized controlled trial of Vanderwee (2007) was included in this review59

Evidence from this study is summarised in the clinical GRADE evidenceprofiles. The forest plot and the study evidence table are presented in respectively Appendix 3.3.5. and Appendix 3.3.6.

interquartile ranges, medians were used as a surrogate for means and standard deviations were estimated as 80% of the interqu


skin assessment withtransparent disk. Preventivemeasures were started when

blanchable erythema (NBE)

(2) Braden score and daily skinassessment with transparent

. Preventive measures werestarted if the Braden score was<17 or NBE appeared.

Patients received preventivemeasures according to the samepressure redistribution protocol.

The patients were randomized toPolyéthylène-uréthane(Tempur-World Inc,

Lexington, Kentucky USA), or toAlternating pressure air

-XCell, Huntleigh

Patients with an expectedhospitalization of at least threedays admitted between May2000 and March 2002 in 14surgery, internal medicine andgeriatric wards of six Belgianhospitals

Incidence of PU2-4) per 1 000 days(95% CI)

Time (days) todevelopment of PU(grades 2-4)

169

The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in Cinahl (EBSCO interface), 794 inEmbase, and 203 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted in 2 576 record s. Based on the screening of

text of the remaining record was reviewed in detail and included in

Evidence from this study is summarised in the clinical GRADE evidenceAppendix 3.3.5. and Appendix 3.3.6. In case that data were reported as

interquartile ranges, medians were used as a surrogate for means and standard deviations were estimated as 80% of the interqu artile range.

Study length

Incidence of PU (grades000 days

Time (days) todevelopment of PU

The study was carried outbetween May 2000 andMarch 2002. Each nursingunit took part in the study forthe duration of five months.

170


Healthcare, UK). On thePolyéthylène-uréthane mattresspatients were turned every fourhours, as proved to bin an earlier study (Defloor et al.2005). On thepressure air mattressstandardized position changeswere carried out.


Table 7 – Skin assessment with transparent disk versus skin assessment with transparent disk and Braden scale for the prevention ofulcers development

Quality assessment

No of

studies

Design Risk of

bias


Incidence of pressure ulcers (grades 2-4) per 1000 days

1

(Vanderwee

2007)

randomised

trials

serious1 no serious

inconsistency

no serious

indirectness



Healthcare, UK). On theuréthane mattress,

patients were turned every fourhours, as proved to be indicatedin an earlier study (Defloor et al.2005). On the Alternatingpressure air mattress, nostandardized position changes

Skin assessment with transparent disk versus skin assessment with transparent disk and Braden scale for the prevention of

sessment No of patients


considerations

Skin

assessment

with

transparent

disk for

non-

blanchable

erythema

(NBE)

skin

assessment

with

transparent

disk

combined

with the

Braden

scale

Relative

(95%

CI)

no serious

indirectness

no serious

imprecision

none 56/826

(6.8%)

53/791

(6.7%)

RR 1.01

(0.7 to

1.45)

KCE Report 193S

Study length

Skin assessment with transparent disk versus skin assessment with transparent disk and Braden scale for the prevention of pressure


Relative

(95%

CI)

Absolute

RR 1.01

(0.7 to

1.45)

1 more

per 1000

(from 20

fewer to

30 more)

MODERATE

CRITICAL

KCE Report 193S

1No blinding, randomisation tables used, concealment of allocation sequence, no drop

Table 8 – Time to development of pressure ulcersdisk and Braden scale

Quality assessment

No of

studies

Design Risk of

bias


Time (days) to develop pressure ulcers (grade 2-4) (Better indicated by

1

(Vanderwee

2007)

randomised

trials

Serious1 no serious

inconsistency

no serious

indirectness

1No blinding, randomisation tables used, concealment of allo


6.7%

No blinding, randomisation tables used, concealment of allocation sequence, no drop-outs

pressure ulcers (grade 2-4) for skin assessment with transparent disk versus skin assessment with transparent



considerations

Skin

assessment

with

transparent

disk for

non-

blanchable

erythema

(NBE)

skin

assessment

with

transparent

disk

combined

with the

Braden

scale

Relative

(95%

CI)

4) (Better indicated by higher values)

no serious

indirectness

no serious

imprecision

none 4 (SD 2.4) 8 (SD 6.4) -

No blinding, randomisation tables used, concealment of allocation sequence, no drop-outs

171

1 more

per 1000

(from 20

fewer to

30 more)

with transparent disk versus skin assessment with transparent


Relative

(95%

CI)

Absolute

- MD 4

lower

(4.48 to

3.52

lower)

MODERATE

IMPORTANT

172

3.3.5. Forest plots

Figure 2 – Skin assessment with transparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control)ulcer (grades 2-4) development

Figure 3 – Skin assessment with transparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control)PU (grades 2-4)


Skin assessment with transparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control)

sparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control)

KCE Report 193S

Skin assessment with transparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control) – for pressure

sparent disk (NBE) versus skin assessment with transparent disk and Braden scale (control) – time to develop

KCE Report 193S


Table 9 – Vanderwee 2007


Author and year:

Vanderwee, 2007

Title:

Non-blanchableerythema as anindicator for the needfor PU prevention: arandomized-controlledtrial

Journal:

Journal of ClinicalNursing, 2007;16: 325–335

Study type:

RCT

Sequence generation:based on randomizationtables generated withthe software packageSPSS 10Allocation:Serially numbered,closed envelopes weremade available for eachparticipating nursingunit. Each time a patientwas admitted theenvelope with the

Patient group:

Patients with an expectedhospitalization of atdays admitted between May 2000and March 2002 in 14 surgery,internal medicine and geriatricwards of six Belgian hospitals

All patients

Randomized N: 1 617

Completed N: 1 617

Drop-outs: 0

Group 1

Randomized N: 826

Completed N: 826

Dropouts: 0

Age: (median and interqurange)

78 (70-86)

Gender (m/f): 332/494

Other relevant patientcharacteristics:

Braden score on admission(median and interquartile range):

19 (16-21)



Comparison

Outcome measures Effect sizes

Patients with an expectedhospitalization of at least three

etween May 2000in 14 surgery,

internal medicine and geriatricsix Belgian hospitals

617

(median and interquartile

332/494

Other relevant patient


Group 1 (NBE):Daily skinassessment withtransparent disk.Preventive measureswere started whenNBE appeared. Thepatient continued tobe observed daily.When the NBEdisappeared, themeasures werediscontinued andrestarted only if theNBE reappeared.

Group 2 (Control):Braden score anddaily skinassessment withtransparent disk.Preventive measureswere started if theBraden score was<17 or NBEappeared. If theBraden score was 17or higher, the patientwas scored again onthe Braden scalethree days later.

Outcome 1:

Incidence of PU(grades 2-4) per1 000 days (95% CI)

Outcome 2:

Time (days) todevelop PU (grades2-4)

Mean (IQR)

Group 1:

4.5 (3.3

Group 2:

4.2 (3.0

Risk Ratio

95% CI:

P value:

Fisher exact test,p>0.99

Group 1:

4 (2-5)

Group 2:

8 (4-16)

Mean d(95% CI):

- 4 (- 4.48;

P value:

Mann-test, p=0.001

173

ct sizes Comments

Group 1:

(3.3-5.7%)

Group 2:

(3.0-5.3%)

Risk Ratio: 1.01

95% CI: 0.7-1.45

P value:

Fisher exact test,p>0.99

Group 1:

5)

Group 2:

16)

Mean difference(95% CI):

4.48;- 3.52)

P value:

-Whitney Utest, p=0.001

Funding:

This study wassupported by a grantfrom the GhentUniversity and fromHuntleigh Healthcare

Limitations:

No blinding

Additionaloutcomes:In the group usingAlternating pressureair mattress, theincidence of pressureulcers (grades 2–4)was lower, but notsignificantly differentin the NBE group(14.5%) comparedwith the control group(20.5%) (Fisher’sexact test, P=0.42).In the group usingPolyéthylène-uréthane mattress,the difference in theincidence of pressure

174


lowest number wasopened. The envelopecontained the patient’sadmission form onwhich the assignment ofthe patient wasindicated, by means ofa flow chart. The flowchart indicated whetherthe patient belonged tothe control group or theNBE group, andwhether to use aPolyéthylène-uréthanemattress or anAlternating pressure airmattress if pressureredistribution wasneeded.

Blinding:

No blinding (for practicaland ethical reasons)

Addressingincomplete outcomedata:

No incomplete outcomedata

Statistical analysis:The Mann–Whitney U-test was used forcontinuous variablesthat were not distributednormally. The Fisher’s

Group 2

Randomized N: 791

Completed N: 791

Dropouts: 0

Age: (median and interquartilerange)

79 (71-85)

Gender (m/f): 289/502


19 (17-21)

Inclusion criteria:

Hospitalization of at least 3 days

Exclusion criteria:

-grade 2 pressure ulcer (abrasionor blister), grades 3 (superficialulcer) and 4 (deep ulcer) onadmission

-age younger than 18

-bodyweight of over 140 kg

-contra-indication for turningbecause of medical reasons



Comparison


(median and interquartile

289/502


Hospitalization of at least 3 days

grade 2 pressure ulcer (abrasionor blister), grades 3 (superficial

nd 4 (deep ulcer) on

bodyweight of over 140 kg

indication for turningof medical reasons

Pressure points wereobserved daily

Both groups:Patients receivedpreventive measuresaccording to thesame pressureredistributionprotocol. It consistedof pressureredistribution whilesitting up and while inbed. During sitting inan (arm)chair, an aircushion (Airtech_,Huntleigh Healthcare,UK) was used for allpatients and they hadto stand up every twohours, alone or withsome help. If theback of the armchaircould be tiltedbackwards, thepatient’s legs wereput on a footrest. Ifthe back of thearmchair could not betilted backwards, thepatient’s feet wereplaced on the floor.

The patients were

KCE Report 193S

ct sizes Comments

ulcers (grades 2–4)approachedsignificance (Fisher’sexact test, P =0.052),the incidence beinglower in the controlgroup (14.2%) thanin the NBE group(25.8%). In theintervention group,16% of patientsreceived preventivemeasures, in thecontrol group 32%(Fisher’s exact test,P < 0.001).The sensitivity of therisk assessmentmethod used in thecontrol group was81.1% and thespecificity 71.8%.The sensitivity ofNBE as a method forassigning preventivemeasures was 46.6%and the specificity86.8%.The time whenprevention startedwas not significantlydifferent in the twogroups (Mann–Whitney U = 479, P =

KCE Report 193S


exact test was used forcategorical variables. AKaplan–Meier survivalanalysis was performedto evaluate the effect ofthe risk assessmentmethod on theincidence of PU (grade2 or higher). Allanalyses were carriedout with the softwarepackage SPSS 10. Avalue of P < 0.05 wasconsidered statisticallysignificant.

Baseline differences:The random assignmentproduced comparableintervention and controlgroups with regard toage, gender, Bradenscore on admission,medical specialty andprimary diagnosis.

Study power/samplesize:

Based on a PU (grade 2or higher) incidence of6%, a sample size wascalculated of 1 624patients (812 in eachgroup) to detect adifference of 3% in the



Comparison


randomized to eitherthe Polyéthylène-uréthane mattress(Tempur_-World Inc,Lexington, KentuckyUSA), or to theAlternating pressureair mattress (Alpha-XCell, HuntleighHealthcare, UK). Onthe Polyéthylène-uréthane mattress,patients were turnedevery four hours, asproved to beindicated in an earlierstudy (Defloor et al.2005). On theAlternating pressureair mattress, nostandardized positionchanges were carriedout.

In the interventiongroup 66 patientsreceived pressureredistribution byPolyéthylène-uréthane mattressand 62 by Alternatingpressure airmattress.

175

ct sizes Comments

0.28). The separateanalyses for thePolyéthylène-uréthane mattressgroup and theAlternating pressureair mattress groupdid not reveal asignificant differenceeither. Adjusted forthe preventionprotocols, theKaplan–Meiersurvival analysisrevealed a significantdifference betweencontrol and NBEgroups (Log-ranktest=7.18, d.f.=1,p=0.007).

Notes: any note thereviewer thinks maybe important

176


incidence of PUbetween the NBE andcontrol group (= 0.05;power = 80%).Setting:14 surgery, internalmedicine and geriatricwards of six Belgianhospitals

Length of study:

The study was carriedout between May 2000and March 2002. Eachnursing unit took part inthe study for theduration of five months.

Assessment of PUs:

In the NBE group and inthe control group, theskin was examined atall pressure points, bynursing staff onadmission and thendaily during the morningshift. The observedpressure points werethe sacrum, heels, hips,ankles, shoulder,elbows, ears and knees.PU were classifiedaccording to the fourgrades of the EuropeanPressure Ulcer Advisory



Comparison


In the control group134 patients receivedpressureredistribution byPolyéthylène-uréthane mattressand 117 byAlternating pressureair mattress.

KCE Report 193S

ct sizes Comments

KCE Report 193S


Panel. A patient wasconsidered to have apressure ulcer when apressure ulcer grades2–4 were observed.

A transparent pressuredisk with a size of 5 cmby 5 cm, was used todistinguish betweenblanchable (BE) andnonblanchableerythema (NBE). Thenurse pressed thetransparent disk on theerythema. If theerythema blanched, itwas defined as BE. Ifthe erythema remainedwhile pressing, it wasdefined as NBEMultiple ulcers:

Unit of analysis wasnumber of patientsdeveloping PU



Comparison


177

ct sizes Comments

178

4. SKIN ASSESSMENT – PROGNOSTIC




Review question What is the predictive ability of skin assessmentPopulation Individuals of all agesSkin assessment method Structured, systematic skin assessment methods/tools:

Ultrasonography

Ultrasound

Durometer/durometry


Elastometer

Haptic finger

Multispectral imaging device

Multiwavelength imaging

Multispectral images

Digital color images

Clinical assessment

Transcutaneous oximetry

Termographic scanner

Tympanic thermometers (to measure skin temperature)

Doppler blood flowmetry

Laser

Doppler imagin

Outcomes Patient outcomes


PROGNOSTIC

Skin assessment

What is the predictive ability of skin assessment methods for pressure ulcer risk?


Structured, systematic skin assessment methods/tools:

Ultrasonography

Ultrasound

Durometer/durometry


Elastometer

Haptic finger

Multispectral imaging device

Multiwavelength imaging

Multispectral images

Digital color images

Clinical assessment

Transcutaneous oximetry

Termographic scanner

Tympanic thermometers (to measure skin temperature)

Doppler blood flowmetry

oppler imaging

Patient outcomes

KCE Report 193S

KCE Report 193S



Incidence of pressure ulcers (all grades and grades 2 Incidence of pressure ulcers (all grades and grades 2

Statistical measures Sensitivity Specificity Area under the ROC (AUC) (for skin temperature) Diagnostic Odds Ratio

Study design High quality systematic reviews of prospective cohort studies.

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginningthe study and with a follow

Exclusion Non-English, nonSearch strategy The electronic databases to be searched are:


Review strategy How will individual PICO characteristics be combined across studies) Population Skin assessment method Outcomes Minimum follow up = no minimum. Minimum total size = no minimum

Analysis The following groups will be considered separately if data are present: ICU patients Spinal cord injury pa Palliative care patients Paediatric patients Adults (if not in other previous subgroup)

Following analyses will be performed Sensitivity and specificity analyses Diagnostic odds ratios will be meta


Skin assessment

Incidence of pressure ulcers (all grades and grades 2-4) – up to one weekIncidence of pressure ulcers (all grades and grades 2-4) – up to three months

Statistical measuresSensitivitySpecificity

ea under the ROC (AUC) (for skin temperature)Diagnostic Odds Ratio

High quality systematic reviews of prospective cohort studies.

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginningthe study and with a follow-up in a systematic way during an established period

English, non-French, non-Dutch language papersThe electronic databases to be searched are:

Medline (OVID interface), Cinahl (EBSCO-interface), Embase, Library of the Cochrane CollaborationAll years

How will individual PICO characteristics be combined across studies)Population – any population will be combined except those specified in the strataSkin assessment method – only same methods will be combinedOutcomes – same outcomes will be combinedMinimum follow up = no minimum.Minimum total size = no minimum

The following groups will be considered separately if data are present:ICU patientsSpinal cord injury patientsPalliative care patientsPaediatric patientsAdults (if not in other previous subgroup)

Following analyses will be performedSensitivity and specificity analysesDiagnostic odds ratios will be meta-analysed

179

up to three months

Prospective cohort studies in which the patients considered had not developed pressure ulcers at the beginning ofup in a systematic way during an established period

ce), Embase, Library of the Cochrane Collaboration

any population will be combined except those specified in the strata

180




Date 30/08/2012


Search Strategy 1. Pressure ulcer.sh2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage)).ti,ab4. (bedsore* or bed-5. ((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.6. OR/1 – 57. finger method.tw8. transparent disk*.tw9. diascopy.tw10. ultrasonograph*.tw11. ultrasonography.sh12. ultrasonics.sh13. ultrasound.tw14. durometer.tw15. durometry.tw16. elastometer.tw17. haptic finger.tw18. digital imag*.tw19. digital colo?r imag*.tw20. spectrometer.tw21. multispectral imag*.tw22. multiwavelength imag*.tw23. clinical assessment.tw24. transcutaneous oximetry.tw25. Blood Gas Monitoring, Transcutaneous.26. tympanic thermometer*.tw27. Doppler blood flowmetry.tw28. laser Doppler imag*.tw


Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations and Ovid MEDLINE(R) 1946 to

Pressure ulcer.sh

(pressure adj (sore* or ulcer* or damage)).ti,ab.-sore*).ti,ab.

((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

transparent disk*.tw

ultrasonograph*.twultrasonography.sh

digital colo?r imag*.tw

multispectral imag*.twmultiwavelength imag*.twclinical assessment.twtranscutaneous oximetry.twBlood Gas Monitoring, Transcutaneous.shtympanic thermometer*.twDoppler blood flowmetry.twlaser Doppler imag*.tw

KCE Report 193S



911839486254506654

1381813

729

698726051119248

13440195

924

19221

4718194

3758

14292137

201214441

KCE Report 193S

Date 30/08/2012

29. Minimum Data Set.tw or MDS.tw or RAI.tw30. skin assess*.tw31. skin inspect*.tw32. skin exam*.tw33. skin eval*.tw34. skin observ*.tw35. skin risk assess*.tw36. skin status.tw37. skin condition.tw38. judgment.sh39. clinical judgment.tw40. skin temperature.tw41. skin temperature.sh42. OR/7 – 4143. AND/6, 4244. Limit language: ‘English, Dutch, Flemish, French’


Date 30/08/2012

Database Embase

Search Strategy


1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (4. (bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab5. ((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab6. OR/1 – 57. ‘finger method’:ti,ab8. ‘transparent disk’:ti,ab9. ‘diascopy’:ti,ab10. ultrasonograph*:ti,ab11. ‘echography’/exp12. ‘Doppler echography’/exp


Minimum Data Set.tw or MDS.tw or RAI.tw

skin risk assess*.tw

clinical judgment.twskin temperature.twskin temperature.sh

Limit language: ‘English, Dutch, Flemish, French’

‘decubitus’/exp

NEAR/1 (sore* OR ulcer* OR damage)):ab,ti(bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

‘finger method’:ti,ab‘transparent disk’:ti,ab

raph*:ti,ab‘echography’/exp‘Doppler echography’/exp

181

48411620

15346

5956156

568

122111538

288953448195

313866665619

((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

1335554597477741812

1826320

640

88906447225

182

Date 30/08/2012

13. ‘color ultrasound flowmetry’/exp14. ‘ultrasound’/exp15. ‘ultrasound’:ti,ab16. ‘durometer’:ti,ab17. ‘durometry’:ti,ab18. ‘elastometer’:ti,ab19. ‘haptic finger’:ti,ab20. (digital NEXT/1 imag*):ti,ab21. (‘digital colo?r’ NEXT/1 i22. ‘spectrometer’:ti,ab23. ‘mass spectrometer’/exp24. (multispectral NEXT/1 imag*):ti,ab25. (multiwavelength NEXT/1 imag*):ti,ab26. ‘clinical assessment’:ti,ab27. ‘clinical assessment’/exp28. ‘transcutaneous oximetry’:ti,ab29. ‘transcutaneous oxygen monitoring’/exp30. ‘thermographic scanner’:ti,ab31. (tympanic NEXT/1 thermometer*):ti,ab32. ‘tympanic thermometer’/exp33. ‘Doppler blood flowmetry’:ti,ab34. ‘Doppler flowmetry’/exp35. ‘blood flowmetry’/exp36. (‘laser Doppler’ NEXT/1 imag*):ti,ab37. ‘laser Doppler flowmetry’/exp38. ‘Minimum Data Set’:39. (skin NEXT/1 assess*):ti,ab40. (skin NEXT/1 inspect*):ti,ab41. (skin NEXT/1 exam*):ti,ab42. (skin NEXT/1 eval*):ti,ab43. (skin NEXT/1 observ*):ti,ab44. (‘skin risk’ NEXT/1 assess*):ti,ab45. ‘skin status’:ti,ab46. ‘skin condition’:ti,ab47. ‘clinical judgment’:ti,ab48. ‘clinical observation’/exp


‘color ultrasound flowmetry’/exp

‘elastometer’:ti,ab‘haptic finger’:ti,ab(digital NEXT/1 imag*):ti,ab(‘digital colo?r’ NEXT/1 imag*):ti,ab‘spectrometer’:ti,ab‘mass spectrometer’/exp(multispectral NEXT/1 imag*):ti,ab(multiwavelength NEXT/1 imag*):ti,ab‘clinical assessment’:ti,ab‘clinical assessment’/exp‘transcutaneous oximetry’:ti,ab‘transcutaneous oxygen monitoring’/exphermographic scanner’:ti,ab

(tympanic NEXT/1 thermometer*):ti,ab‘tympanic thermometer’/exp‘Doppler blood flowmetry’:ti,ab‘Doppler flowmetry’/exp‘blood flowmetry’/exp(‘laser Doppler’ NEXT/1 imag*):ti,ab‘laser Doppler flowmetry’/exp‘Minimum Data Set’:ti,ab or ‘MDS’:ti,ab or ‘RAI’:ti,ab(skin NEXT/1 assess*):ti,ab(skin NEXT/1 inspect*):ti,ab(skin NEXT/1 exam*):ti,ab(skin NEXT/1 eval*):ti,ab(skin NEXT/1 observ*):ti,ab(‘skin risk’ NEXT/1 assess*):ti,ab

‘skin condition’:ti,abical judgment’:ti,ab

‘clinical observation’/exp

KCE Report 193S

266801994885221

179253110

932

110513

1419687

3983379

71865949109

1642268

71686361

235462867604

783116943

21349

8758785

996

17683472

KCE Report 193S

Date 30/08/2012

49. ‘skin temperature’:ti,ab50. ‘skin temperature’/exp51. OR/7 – 5052. AND/6, 5153. Limit language: ’English, Dutch, French’ and limited to embase


Date 30/08/2012

Database The Library of the Cochra

Search Strategy


1. MeSH descriptor “Pressure ulcer” explode all trees2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw4. (bedsore* or bed5. ((moist* or

lesion*)):ti,ab,kw6. OR/1 – 57. (finger method):ti,ab,kw8. (transparent disk*):ti,ab,kw9. (diascopy):ti,ab,kw10. (ultrasonograph*):ti,ab,kw11. MeSH descriptor “ultrasonography” explode all tree12. MeSH descriptor “ultrasonics” explode all trees13. (ultrasound):ti,ab,kw14. (durometer) :ti,ab,kw15. (durometry) :ti,ab,kw16. (elastometer) :ti,ab,kw17. (haptic finger) :ti,ab,kw18. (digital imag*):ti,ab,kw19. (digital colo?r imag*):ti,ab,kw20. (spectrometer):ti,ab,kw21. (multispe22. (multiwavelength imag*):ti,ab,kw


‘skin temperature’:ti,ab‘skin temperature’/exp

Limit language: ’English, Dutch, French’ and limited to embase


MeSH descriptor “Pressure ulcer” explode all treesDecubit*:ti,ab,kw(pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw(bedsore* or bed-sore*):ti,ab,kw((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*orlesion*)):ti,ab,kw

5(finger method):ti,ab,kw(transparent disk*):ti,ab,kw(diascopy):ti,ab,kw(ultrasonograph*):ti,ab,kwMeSH descriptor “ultrasonography” explode all trees

eSH descriptor “ultrasonics” explode all trees(ultrasound):ti,ab,kw(durometer) :ti,ab,kw(durometry) :ti,ab,kw(elastometer) :ti,ab,kw(haptic finger) :ti,ab,kw(digital imag*):ti,ab,kw(digital colo?r imag*):ti,ab,kw(spectrometer):ti,ab,kw(multispectral imag*):ti,ab,kw(multiwavelength imag*):ti,ab,kw

183

1603565249393

7437751195794

friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

4903498343363

1168940

20

93656678230

66266109

70624

11440

184

Date 30/08/2012

23. (clinical assessment):ti,ab,kw24. (transcutaneous oximetry):ti,ab,kw25. MeSH descriptor “Blood Gas Monitoring, Transcutaneous” explode all trees26. (thermographic scanner):ti,ab,kw27. (tympanic thermometer*):ti,ab28. (Doppler blood flowmetry):ti,ab,kw29. (laser Doppler imag*):ti,ab,kw30. (Minimum Data Set):ti,ab,kw or (MDS) :ti,ab,kw or (RAI):ti,ab,kw31. (skin assess*):ti,ab,kw32. (skin inspect*):ti,ab,kw33. (skin exam*):ti,ab,kw34. (skin eval*):ti,ab,kw35. (skin observ*):ti,ab,kw36. (skin risk assess*):ti,ab,kw37. (skin status):ti,ab,kw38. (skin condition*):ti,ab,kw39. Mesh descriptor “Judgment” explode all trees40. (clinical judgement):ti,ab,kw41. (Skin temperature):ti,ab,kw42. Mesh descriptor “skin temperature” explode all trees43. OR/7 – 4244. AND/6, 43


(clinical assessment):ti,ab,kw(transcutaneous oximetry):ti,ab,kwMeSH descriptor “Blood Gas Monitoring, Transcutaneous” explode all trees(thermographic scanner):ti,ab,kw(tympanic thermometer*):ti,ab,kw(Doppler blood flowmetry):ti,ab,kw(laser Doppler imag*):ti,ab,kw(Minimum Data Set):ti,ab,kw or (MDS) :ti,ab,kw or (RAI):ti,ab,kw(skin assess*):ti,ab,kw(skin inspect*):ti,ab,kw(skin exam*):ti,ab,kw(skin eval*):ti,ab,kw(skin observ*):ti,ab,kw

in risk assess*):ti,ab,kw(skin status):ti,ab,kw(skin condition*):ti,ab,kwMesh descriptor “Judgment” explode all trees(clinical judgement):ti,ab,kw(Skin temperature):ti,ab,kwMesh descriptor “skin temperature” explode all trees

42AND/6, 43

KCE Report 193S

1872249

173

136

808150655

460257

188651093358675637

1635430405

1837708

46802203

KCE Report 193S


Date 30/08/2012

Database CINAHL

Search Strategy


1. MH “Pressure Ulcer”2. Decubit*3. Pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*4. Bedsore* OR bed-sore*5. ((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))6. OR/1 – 57. finger method8. transparent disk*9. diascopy10. ultrasonograph*11. MH ultrasonography12. MH ultrasonics13. Ultrasound14. durometer15. durometry16. elastometer17. haptic finger18. digital imag*19. digital color imag* or digital colour imag*20. spectrometer21. multispectral imag*22. multiwavelength imag*23. clinical assessment24. transcutaneous oximetry25. MH Blood Gas Monitoring, Transcutaneous26. thermographic scanner27. tympanic thermometer*28. Doppler blood flowmetry29. laser Doppler imag*30. “Minimum Data Set” or MDS or RAI31. skin assess*32. skin inspect*33. skin exam*


MH “Pressure Ulcer”



MH ultrasonography

digital color imag* or digital colour imag*

imag*

transcutaneous oximetryMH Blood Gas Monitoring, Transcutaneousthermographic scannertympanic thermometer*Doppler blood flowmetry

“Minimum Data Set” or MDS or RAI

185


7748487

8540157

1424

987630

42

279837546669

1128617

201

230126

11490

5900643

2963

1396297

290877057

186

Date 30/08/2012

34. skin eval*35. skin observ*36. skin risk assess*37. skin status38. skin condition39. MH judgment40. clinical judgment41. skin temperature42. MH skin temperature43. OR/7 – 4244. AND/6, 4345. Limit language=’English, Dutch, French’


MH skin temperature

Limit language=’English, Dutch, French’

KCE Report 193S

51076028680

127350

1889875

1041703

10102317781631

KCE Report 193S



Potentially relevant citations identified:

Potentially relevant citations identified afterduplicates removal:

Relevant studies:

Studies retrieved for more detailed evaluation:14

Additionally potentially relevant citations(hand searching): 0


Potentially relevant citations identified: 3247

Potentially relevant citations identified afterates removal: 2576

Based on title and abstract evaluation,citations excluded: 2562

Based on full text evaluation, studiesexcluded: 10

Relevant studies: 4

Studies retrieved for more detailed evaluation:

Additionally potentially relevant citations

187

Based on title and abstract evaluation,

Based on full text evaluation, studies

188

4.2.3. List of excluded studies

Reference

Bates-jensen 2007

Bates-jensen 2009

Bates-jensen 2010

Judy 2011

Guihan 2012

Rapp 2006

Rapp 2009

Stordeur 1998

Vanderwee 2006

Vanderwee 2007


Reason of exclusion

Outcome: no sensitivity and specificity reported and impossible to calculatefrom the available data


Design: Abstract describing study protocol


Population – all had pressure ulcers and were hospitalized for care of thePUs

Dissertation: same data as in article “Rapp, M. P., Bergstrom, N., & Padhye,N. S. (2009). Contribution of skin temperature regularity to the risk ofdeveloping pressure ulcers in nursing facility residents. 22, 506

Outcome; no sensitivity and specificity reported and impossible to calculatefrom the available data

Intervention: risk assessment scales

Intervention: assessment of interrater reliability

Included in the clinical effectiveness review

KCE Report 193S

Outcome: no sensitivity and specificity reported and impossible to calculate


n: Abstract describing study protocol


all had pressure ulcers and were hospitalized for care of the

Rapp, M. P., Bergstrom, N., & Padhye,N. S. (2009). Contribution of skin temperature regularity to the risk ofdeveloping pressure ulcers in nursing facility residents. 22, 506-513.”

d specificity reported and impossible to calculate

Intervention: assessment of interrater reliability

Included in the clinical effectiveness review

KCE Report 193S



The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in CinahlEmbase, and 203 in the Library of the Cochrane Collaboration. Duplititle and abstract 2 562 records were excluded. Reasons for exclusion were listed. The full text of the remaining 14this review, 10 studies were excluded. Four studies


Four studies24 , 60 , 61 , 62

were included in this review. Sensitivity and specificity were re



Study Skin assessment

Outcome

Compton 200824

Subjective nursing skin assessmon admission

Occurrence of PU development(grades 2European Pressure Ulcer AdvisoryPanel classification system in thecourse of ICU treatment

Konishi 200860

Presence of blanchable erythemaassessed by finger test

Occurrence of PU developmentaccording toUlcer Advisory Panel classification

Newman 198161

Thermographyanomaly (an area of the skin at least1°C warmer than the surro


The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in CinahlEmbase, and 203 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted in 2 576 records. Based on the screening oftitle and abstract 2 562 records were excluded. Reasons for exclusion were listed. The full text of the remaining 14 studies

studies were included in this review.

were included in this review. Sensitivity and specificity were re-calculated by using the raw data as presented in the individual studies.

Skin assessment test

Outcome

Population

Subjective nursing skin assessmenton admission

Occurrence of PU development(grades 2-4) according to theEuropean Pressure Ulcer AdvisoryPanel classification system in thecourse of ICU treatment

ICU patients

Presence of blanchable erythemad by finger test

Occurrence of PU developmentaccording to the National PressureUlcer Advisory Panel classification


Thermography: presence of thermalanomaly (an area of the skin at least1°C warmer than the surrounding


189

The systematic search through multiple electronic databases resulted in 3 247 records: 619 in Medline (Ovid), 1 631 in Cinahl (EBSCO interface), 794 incate records were excluded, which resulted in 2 576 records. Based on the screening of

studies was reviewed in detail. Based on

calculated by using the raw data as presented in the individual studies.

Length of follow-up

Not reported

Not reported

Not reported

190

Study Skin assessment

Outcome

skin).

Development of skin breakdown inthe buttock region within 10 days ofadmission was reported by thenursing staff and photographed.Redness alone, however marked orpersistent, was not categorized as apressure sore.

Nixon 200762

Skin assessment accordingclassification scale adapted frominternational classi(AHCPR (Agency for Health CarePolicy and Research) 1992; EPUAP,1999)

Occurrence PU development (grades2-4) accordingscale adapted fromclassification scales (AHCPR(Agency for Health Care Policy andResearch) 1992; EPUAP, 1999)


Skin assessment test

Outcome

Population

Development of skin breakdown inthe buttock region within 10 days ofadmission was reported by thenursing staff and photographed.Redness alone, however marked orpersistent, was not categorized as apressure sore.

kin assessment according to theclassification scale adapted frominternational classification scales(AHCPR (Agency for Health Care

y and Research) 1992; EPUAP,

Occurrence PU development (grades4) according the classification

scale adapted from internationalfication scales (AHCPR

(Agency for Health Care Policy andResearch) 1992; EPUAP, 1999)

Surgical in-patients

KCE Report 193S

Length of follow-up

Not reported

KCE Report 193S

4.3.4. Predictive ability

Table 7 – Subjective nursing assessment of moist skin as a predictor for the development ofto the European Pressure Ulcer Advisory Panel classification system

Study Incidence Sensitivity

1 (Compton 2008) / 76%

(67-

Table 8 – Subjective nursing assessment of oedematous skin as a predictor for the development ofaccording to the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 64%

(54-

Table 9 – Subjective nursing assessment of mottled skin as a pto the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 33%

(25-

Table 10 – Subjective nursing assessment of livid skin as a predictor for the development ofto the European Pressure Ulcer Advisory Panel classification system

Study Incidence Sensi

1 (Compton 2008) / 31%

(23-

Table 11 – Subjective nursing assessment of centralised circulation as a predictor for the development ofdevelopment according to the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 71%

(62-


Subjective nursing assessment of moist skin as a predictor for the development of pressure ulcers (grades 2to the European Pressure Ulcer Advisory Panel classification system

Sensitivity Specificity DOR

76%

-83%)

65%

(61-69%)

5.9

(3.84-9.03)

Subjective nursing assessment of oedematous skin as a predictor for the development of pressure ulcersto the European Pressure Ulcer Advisory Panel classification system


64%

-72%)

77%

(73-80%)

5.7

(4.05-8.11)

Subjective nursing assessment of mottled skin as a predictor for the development of pressure ulcers (grades 2to the European Pressure Ulcer Advisory Panel classification system


33%

-42%)

92%

(89-94%)

5.4

(4.21-7.03)

Subjective nursing assessment of livid skin as a predictor for the development of pressure ulcers (grades 2to the European Pressure Ulcer Advisory Panel classification system


31%

-40%)

92%

(89-94%)

5.0

(3.92-6.5)

Subjective nursing assessment of centralised circulation as a predictor for the development ofng to the European Pressure Ulcer Advisory Panel classification system


71%

-79%)

70%

(66-74%)

5.8

(3.95-8.61)

191

(grades 2-4) development according

PPV NPV

31% 93%

pressure ulcers (grades 2-4) development

PPV NPV

36% 91%


PPV NPV

45% 87%


PPV NPV

44% 86%

Subjective nursing assessment of centralised circulation as a predictor for the development of pressure ulcers (grades 2-4)

PPV NPV

33% 92%

192

Table 12 – Subjective nursing assessment of cyanosis as a prto the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 45%

(36-

Table 13 – Subjective nursing assessment of reddened skin as a predictor for the development ofaccording to the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 69%

(60-

Table 14 – Subjective nursing assessment of hyperaemicaccording to the European Pressure Ulcer Advisory Panel classification system


1 (Compton 2008) / 21%

(15-

Table 15 – Presence of blanchable erythema assessed byaccording to the European Pressure Ulcer Advisory Panel classification system


1 (Konishi 2008) / 75%

(35-


Subjective nursing assessment of cyanosis as a predictor for the development of pressure ulcers (grades 2to the European Pressure Ulcer Advisory Panel classification system


45%

-55%)

81%

(77-84%)

3.5

(2.63-4.64)

Subjective nursing assessment of reddened skin as a predictor for the development of pressure ulcersaccording to the European Pressure Ulcer Advisory Panel classification system

sitivity Specificity DOR

69%

-77%)

70%

(66-74%)

5.1

(3.54-7.47)

e nursing assessment of hyperaemic skin as a predictor for the development of pressure ulcersto the European Pressure Ulcer Advisory Panel classification system


21%

-30%)

91%

(89-93%)

2.9

(2.28-3.65)

Presence of blanchable erythema assessed by the finger test as a predictor for the development ofaccording to the European Pressure Ulcer Advisory Panel classification system


75%

-97%)

77%

(71-82%)

9.9

(1.94-50.49)

KCE Report 193S


PPV NPV

33% 88%


PPV NPV

33% 91%


PPV NPV

34% 85%

as a predictor for the development of pressure ulcers development

PPV NPV

10% 99%

KCE Report 193S

Table 16 – Presence of blanchable erythema assessed bydevelopment according to the European Pressure Ulcer Advisory Panel classification system


1 (Konishi 2008) / 75%

(19-

1 (Nixon 2007) / 75

(19-

Table 17 – Thermography (presence of thermal anomalydevelopment of skin breakdown


1 (Newman 1981) / 100%

(54-100%)

Table 18 – Presence of non-blanchable erythema assessed by the finger testdevelopment according to the European Pressure Ulcer Advisory Panel classification system


1 (Nixon 2007) / 73

(45-


Presence of blanchable erythema assessed by the finger test as a predictor for the development ofdevelopment according to the European Pressure Ulcer Advisory Panel classification system


75%

-99%)

76%

(70-81%)

9.4

(0.94-94.58)

75%

-99%)

10%

(4-20%)

0.33

(0.03-3.27)

Thermography (presence of thermal anomaly – an area of the skin at least 1°C warmer than the surrounding skin) as a predictor for the


100%

100%)

74%

(63-83%)

36.7

(1.41-952.24)

able erythema assessed by the finger test as a predictor for the development ofdevelopment according to the European Pressure Ulcer Advisory Panel classification system


73%

-92%)

74%

(64-83%)

8.0

(2.53-25.26)

193

as a predictor for the development of pressure ulcers (grades 2-4)

PPV NPV

5% 99%

5% 86%

of the skin at least 1°C warmer than the surrounding skin) as a predictor for the

PPV NPV

21% 100%

as a predictor for the development of pressure ulcers (grades 2-4)

PPV NPV

34% 94%

194

4.3.5. Quality of the studies

Table 19 – Quality of the studies

Study Selection bias

Compton 2008 Very high1

Konishi 2008 High2

Newman 1981 High2

Nixon 2007 Very high2

1Consecutive patient enrolment, database cohort but participants followed prospectively, unclear validation method

2Consecutive patient enrolment, prospective cohort study, unclear validation method

3selected patient enrolment, prospective cohort study, unclear validation method

4Unclear definition and measurement of prognostic factors, prognostic factors were dichotomised,

threshold5Duration was unclear

6Uncertain if duration was appropriate

7incidence data only

8Incidence data only, inadequate number of events (<100 events)

9incidence data only, inadequate number of events (<100 events), unclear how they dealt with missing data


Risk tool bias Outcome bias

High4

Low

Low High5

Low Low

Low High6

Consecutive patient enrolment, database cohort but participants followed prospectively, unclear validation methodment, prospective cohort study, unclear validation method

selected patient enrolment, prospective cohort study, unclear validation methodUnclear definition and measurement of prognostic factors, prognostic factors were dichotomised, use of imputation technique or clear description of exclusion, adequate

Incidence data only, inadequate number of events (<100 events)vents (<100 events), unclear how they dealt with missing data

KCE Report 193S

Analysis bias

High7

High8

High8

High8

echnique or clear description of exclusion, adequate

KCE Report 193S

4.3.6. Forest plots sensitivity and specificity

Figure 2 – Subjective nursing assessment of moist skin

Figure 3 – Subjective nursing assessment of oedematous

Figure 4 – Subjective nursing assessment of mottled

Figure 5 – Subjective nursing assessment of centralised circulation

Figure 6 – Subjective nursing assessment of livid

Study

Compton 2008

TP

92

FP

202

FN

29

TN

375

Sensitivity

0.76 [0.67, 0.83] 0.65 [0.61, 0.69]

Study

Compton 2008

TP

77

FP

135

FN

44

TN

442

Sensitivity

0.64 [0.54, 0.72] 0.77 [0.73, 0.80]

Study

Compton 2008

TP

40

FP

48

FN

81

TN

529

Sensitivity

0.33 [0.25, 0.42] 0.92 [0.89, 0.94]

Study

Compton 2008

TP

86

FP

171

FN

35

TN

406

Sensitivity

0.71 [0.62, 0.79] 0.70 [0.66, 0.74]

Study

Compton 2008

TP

38

FP

48

FN

83

TN

529

Sensitivity

0.31 [0.23, 0.40] 0.92 [0.89, 0.94]


Subjective nursing assessment of moist skin – ICU- grades 2-4

assessment of oedematous skin – ICU- grades 2-4

tive nursing assessment of mottled skin – ICU- grades 2-4

tive nursing assessment of centralised circulation – ICU- grades 2-4

assessment of livid skin – ICU- grades 2-4

Specificity

0.65 [0.61, 0.69]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.77 [0.73, 0.80]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.92 [0.89, 0.94]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.70 [0.66, 0.74]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.92 [0.89, 0.94]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

195

196

Figure 7 – Subjective nursing assessment of cyanosis

Figure 8 – Subjective nursing assessment of reddened

Figure 9 – Subjective nursing assessment of hyperaemic

Figure 10 – Presence of blanchable erythema assesse

Figure 11 – Presence of thermal anomaly (an area of the skin at least 1°C warmer than the surrounding skininpatients- all grades

Study

Compton 2008

TP

77

FP

135

FN

44

TN

442

Sensitivity

0.64 [0.54, 0.72] 0.77 [0.73, 0.80]

Study

Compton 2008

TP

83

FP

172

FN

38

TN

405

Sensitivity

0.69 [0.60, 0.77] 0.70 [0.66, 0.74]

Study

Compton 2008

TP

26

FP

50

FN

95

TN

527

Sensitivity

0.21 [0.15, 0.30] 0.91 [0.89, 0.93]

Study

Konishi 2008

TP

6

FP

56

FN

2

TN

185

Sensitivity

0.75 [0.35, 0.97]

Specificity

0.77 [0.71, 0.82]

Study

Newman 1981

TP

6

FP

22

FN

0

TN

63

Sensitivity

1.00 [0.54, 1.00]

Specificity

0.74 [0.63, 0.83]


assessment of cyanosis – ICU- grades 2-4

assessment of reddened skin – ICU- grades 2-4

assessment of hyperaemic skin – ICU- grades 2-4

ythema assessed by finger test – hospitalized patients- all grades

an area of the skin at least 1°C warmer than the surrounding skin

Specificity

0.77 [0.73, 0.80]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.70 [0.66, 0.74]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.91 [0.89, 0.93]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.77 [0.71, 0.82]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.74 [0.63, 0.83]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

KCE Report 193S

an area of the skin at least 1°C warmer than the surrounding skin) – follow-up 10 days- geriatric

KCE Report 193S

Figure 12 – Presence of blanchable erythema assesse

Figure 13 – Presence of non-blanchable erythema assesse


Table 20 – Compton 2008


Author and year:

Compton, 2008

Title:

Pressure ulcerpredictors in ICUpatients: nursingskin assessmentversus objectiveparameters

Journal:

Journal of Wound

Patient group:

ICU patients

All patients:

713

Included N:

698

Completed N:

698

Study

Konishi 2008

Nixon 2006

TP

3

3

FP

59

55

FN

1

1

TN

186

6

Sensitivity

0.75 [0.19, 0.99]

0.75 [0.19, 0.99]

Specificity

0.76 [0.70, 0.81]

0.10 [0.04, 0.20]

Study

Nixon 2006

TP

11

FP

21

FN

4

TN

61

Sensitivity

0.73 [0.45, 0.92]

Specificity

0.74 [0.64, 0.83]


f blanchable erythema assessed by finger test – hospitalized patients – grades 2-4

blanchable erythema assessed by finger test – surgical inpatients – grades 2-4

Predictive test

Outcome

Outcomemeasures

Effect sizes

Index test 1:

Nursing skin assessment onadmission

Predictive test 2:

Outcome:

Occurrence of PU (grades 2-4) development according tothe European Pressure UlcerAdvisory Panel classificationsystem in the course of ICUtreatment

Outcome 1:

Sensitivity andspecificity (95%CI)

moist skin

Sensitivity: 76% (67

Specificity: 65% (61

Raw data

Referencestandard

Yes

Indextest

Yes 92

No 29

121

5.9 (3.84-9.03)

Specificity

0.76 [0.70, 0.81]

0.10 [0.04, 0.20]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

Specificity

0.74 [0.64, 0.83]

Sensitivity

0 0.2 0.4 0.6 0.8 1

Specificity

0 0.2 0.4 0.6 0.8 1

197

Comments

% (67-83%)

: 65% (61-69%)

Referencestandard

Yes No

202 294

375 404

121 577 698

Funding:

Supported by aresearch grantof the Robert-Bosch-Stiftung,Stuttgart,Germany

Limitations:

index testmeasured onlyat admission; noreport onblinding of

198


Care, 2008; 17 (10):417-24

Study type:

Prospective cohortstudy

Selection patient:

All patients admittedat ICU between April2001 and December2004 without a PU atadmission andremaining at least 72h were eligible forthe study

Index test:

Subjective nursingassessment of theskin condition onadmission includingthe presence ofmoist skin,oedematous skin,mottled skin, lividskin, centralisedcirculation, cyanosis,reddened skin andhyperaemic skin.

Drop-outs:

0

Age (median years,quartiles):

66 ((56, 75, 25)

Gender (m/f):

392/306


121

Number of patientswithout a PU:

577

Inclusion criteria:

ICU patient

No PU on admission

Exclusion criteria:

Stay in the ICU lessthan 72 h


Predictive test

Outcome

Outcomemeasures

Effect sizes

Preventive methods:

Not reported

DOR 95%CI

Outcome 2:


oedematousskin

DOR 95%CI

Outcome 3:


mottled skin



Raw data

Referencestandard

Yes

Indextest

Yes 77

No 44

121

5.7 (4.05-8.11)



Raw data

Referencestandard

Yes

Index Yes 40

KCE Report 193S

Comments

4% (54-72%)

: 77% (73-80%)

Referencestandard

Yes No

135 212

442 486

121 577 698

% (25-42%)

: 92% (89-94%)

Referencestandard

Yes No

48 88

researchertoward indextest andreferencestandard;unclear ifuninterpretableresults werefound; noinformationabout preventivemeasures;

no sub-analysesaccording topreventivemeasures.

Additionaloutcomes:

With univariateanalysismeasuresrelating to organdysfunction,circulatoryimpairment andsepsis showedsignificantassociation withthe occurrenceof PU. Multiple

KCE Report 193S


Reference standard:

Occurrence of PU(grades 2-4) duringcourse of ICUtreatment.

PU were defined andgraded according tothe EuropeanPressure UlcerAdvisory Panelclassificationsystem.

Addressing missingdata:

To control formissing data, valuesof the continuousmonitoring andlaboratory variableswere recorded intothe point score usedin the acutephysiology score(APS) of theAPACHE II severity-of-disease scoringsystem, where 0 to 4points are assigned


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR 95%CI

Outcome 4:


livid skin

test No 81

121

5.4 (4.21-7.03)

Sensitivity: 31%


Raw data

Referencestandard

Yes

Indextest

Yes 38

No 83

121

199

Comments

529 610

121 577 698

% (23-40%)

: 92% (89-94%)

Referencestandard

Yes No

48 86

529 612

121 577 698

regressionanalysis showedsubjectivenursing skinassessment tooutweigh theseparameters asPU predictors.

A risk functioncomprised of 5skin-related andgender yieldedan overallcorrect PUpredictionproportion of84.6%. ROCanalysis showedan AUC of 0.82(0.79-0.86)compared withan AUC of 0.59(0.54-0.65)obtained withthe Waterlowscale onadmission.Results werevalidated in 392patients treatedin the same ICUbetween

200


according to theextent of deviationfrom thephysiological range.Therefore, onlymonitoring andlaboratory variablesused in the APSscore were enteredin the logisticregression model.


Continuous data aredisplayed as medianand quartiles andwere comparedbetween groupsusing Mann-WhitneyU testing.

Dichotomousparameters aredisplayed asabsolute numbersand percentages andwere comparedbetween groupsusing the chi-squaretest or the Fisher’sexact test. A two-sided p value < 0.05


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR 95%CI

Outcome 5:


centralisedcirculation

5.0 (3.92-6.5)



Raw data

Referencestandard

Yes

Indextest

Yes 86

No 35

121

KCE Report 193S

Comments

71% (62-79%)

: 70% (66-74%)

Referencestandard

Yes No

171 257

406 441

121 577 698

January 2005and May 2006,yielding an AUCof 0.8 (0.73-0.86) comparedwith 0.58 (0.50-0.66) with theWaterlow scale.

Notes:

KCE Report 193S


was consideredsignificant.

Multiple stepwiseregression analysiswas used to analyzewhich of theexaminedparameters predictPU risk in critically illpatients.

The predictivecapacity of thelogistic regressionfunction wasassessed andcompared with theWaterlow scale bycalculating the areaunder the curve(AUC) of a receiver-operatorcharacteristic (ROC)curve. AUCs,sensitivities andspecificities aredisplayed with 95%confidence intervals.

Setting:

Intensive Care Unit,Charité Campus


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR 95%CI

Outcome 6:


Cyanosis

DOR 95%CI

Outcome 7:

Sensitivity and

Specificity (95%CI)

reddened skin

5.8 (3.95-8.61)



Raw data

Referencestandard

Yes

Indextest

Yes 55

No 66

121

3.5 (2.63-4.64)



Raw data

Referencestandard

Yes

Indextest

Yes 83

No 38

201

Comments

45% (36-55%)

: 81% (77-84%)

Referencestandard

Yes No

111 166

466 532

121 577 698

69% (60-77%)

: 70% (66-74%)

Referencestandard

Yes No

172 255

405 443

202


Benjamin Franklin,Berlin, Germany

Blinding: no details


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR 95%CI

Outcome 8:


Hyperaemic skin

DOR 95%CI

121

5.1 (3.54-7.47)



Raw data

Referencestandard

Yes

Indextest

Yes 26

No 95

121

2.9 (2.28-3.65)

KCE Report 193S

Comments

121 577 698

21% (15-30%)

: 91% (89-93%)

Referencestandard

Yes No

50 76

527 622

121 577 698

KCE Report 193S

Table 21 – Konishi 2008


Author and year:

Konishi, 2008

Title:

A prospective studyof

blanchable erythemaamong universityhospital patients

Journal:

International WoundJournal, 2008; 5(3):470-5.

Study type:


Selection patient:

Subjects consistedof patients who wereadmitted to 6 wardsin a universityhospital with 832beds betweenFebruary and April

Patient group:

Patients admitted in auniversity hospital freeof PU and spendingmost of the day in bed.

All patients:

493

Included N:

249

Completed N:

249

Drop-outs:

0

Age (mean years (SD);range):

not reported

Gender (m/f):

not reported


Predictive test

Outcome

Outcomemeasures

Effect sizes

Index test 1:

Presence of blanchableerythema assessed bypressing firmly on the skinwith a finger and by lookingfor blanching followed byprompt return of color to thearea after lifting the finger

Reference standard:

Occurrence of PUdevelopment according tothe National Pressure UlcerAdvisory Panel classification

Preventive methods:

Not reported

Outcome 1:


of blanchableerythema as apredictor for PUdevelopment

DOR 95%CI

Outcome 2:


of blanchableerythema as apredictor for PU(grades 2-4)development

Sensitivity: 75%


Raw data

Referencestandard

Yes

Indextest

Yes 6

No 2

8

9.9 (1.94-50.49)



Raw data

Referencestandard

Yes

Indextest

Yes 3

No 1

4

203

Comments

(35-97%)

: 77% (71-82%)

Referencestandard

Yes No

56 62

185 187

241 249

% (19-99%)

: 76% (70-81%)

ferencestandard

Yes No

59 62

186 187

245 249

Funding:

None reported

Limitations:

No informationabout time offollow-up; noreport onblinding ofresearchertoward indextest andreferencestandard;unclear ifuninterpretableresults werefound; noinformationabout preventivemeasures;


Additional

204


2005. Six wards wereICU, surgicalrecovery room,gastroenterologicalsurgery andmedicine, internalmedicine andcardiovascular andrespiratory surgery.These were selected,as three had thehighest percentagesof bedriddenpatients, and theother three had thelowest percentages.All subjects wererequired to be free ofpressure ulcers atthe beginning of thestudy and spentmost of the day inbed.

Index test:

Daily assessment ofthe presence ofblanchable erythema.

To assess forblanchability,researchers pressed


8 (for all stages of PUdevelopment)

4 (for PU (grades 2-4)development)


241

Inclusion criteria:

Admission in one of the6 participating wards

Free of PU

Bedridden

Exclusion criteria:

none


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR 95%CI 9.4 (0.94-94.58)

KCE Report 193S

Comments

outcomes:

Identification offactorsassociated withthe deteriorationof blanchableerythema.

The number ofpatients whohad a risk underthe item‘pressure’,which is one ofthe triggeringfactors in thescale forpredictingpressure ulcerdevelopment,was significantlyhigher in thedeterioratedgroup (chi-squared=4.277,p= 0.039).

Inadequatemaintenance ofsupport surfaceswas observed inall six patients inthe deteriorated

KCE Report 193S


firmly on the skinwith a finger andlifted the finger andlooked for blanching(sudden whitening ofthe skin), followed byprompt return ofcolor to the area.

Reference standard:

Occurrence of PUassessed by dailyinspection.

Pressure ulcers weredefined by using theNational PressureUlcer Advisory Panelclassification

Addressing missingdata:

No details


To compare eachparameter betweenthe healed and thedeteriorated groups,the chi-squared test


Predictive test

Outcome

Outcomemeasures

Effect sizes

205

Comments

Group (chi-squared =0.228,p= 0.015).

Notes:

206


and Mann-Whitney Utest were performedusing SPSS II forWindows forstatistical analysis. P< 0.05 wasconsideredstatisticallysignificant.

The probability ofblanchable erythemaresulting in pressureulcer developmentwas calculated interms of sensitivity,specificity andpositive likelihoodratio and diagnosticaccuracy wasexamined. In thestatistical methods,diagnosticprobabilities(sensitivity,specificity andpositive likelihoodratio) werecalculated.

Setting:

Six wards in a


Predictive test

Outcome

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


university hospitalwith

832 beds, Ishikawa,Japan.

Blinding:

No details

Table 22 – Newman 1981


Author and year:Newman 1981

Title:

Thermography as apredictor of sacralpressure sores

Journal:

Age and Ageing, 1981;10: 14-8.

Study type:


Patient group:

155 newly admitted ina 12-week periodwithout pressurelesions

64 patients were notincluded because:

- could not bescreened within24 h (N=29)

- too ill toparticipate(N=11)

- refusal (N=11)- miscellaneous

(N=13)


Predictive test

Outcome

Outcomemeasures

Effect sizes

Predictive test

Outcome

Outcomemeasures

Effect sizes

Index test 1:

Thermography: presence ofthermal anomaly (an area ofthe skin at least 1°C warmerthan the surrounding skin).

Reference standard:

Visual inspection

Preventive methods:

No details

Outcome 1:

Proportion ofpatientsdevelopingpressure soresin the sacralregion within 10days afteradmission

Outcome 2:


Thermalanomaly

Value:

6/85 (7%)

Sensitivity:

100% (54-100%)

Specificity:

74% (63-83%)

Raw data

207

Comments

Comments

Funding:

None reported

Limitations:

index testmeasured onlyat admission; noreport onblinding ofresearchertoward indextest andreferencestandard;unclear ifuninterpretable

208


Selection patient:

New admissions to thegeriatric assessmentunit at the SouthernGeneral Hospital,Glasgow, over a 12-week period withunmarked skin wereinvited to participate inthe study

Index test:

Thermography with aprototype, low cost,portable, heat- sensitivethermograph wasperformed within 24 hafter admission.

Patients lay on one sidefor 10 to 15 minuteswith the buttocksexposed to allow skintemperature to stabilize.The ward temperaturewas maintainedbetween 21 and 26°C;relative humidity wasseldom below 40% orabove 60%. The camera

All patients

Included N:

91

Completed N:

91

Drop-outs:

0

Age (mean years(SD); range):

No details

Gender (m/f):

No details


6


85


Predictive test

Outcome

Outcomemeasures

Effect sizes

DOR (95% CI)

Referencestandard

Yes

Indextest

Yes 6

No 0

6

36.7 (1.41-952.24)

KCE Report 193S

Comments

Referencestandard

Yes No

22 28

63 63

85 91

952.24)

results werefound; noinformationabout preventivemeasures;


Additionaloutcomes:

Patients withlow Nortonscores onadmissiondeveloped morefrequently skinbreaks withinthe subsequent10 days thanthose with highscores. Two ofthe 58 controlpatients (4%)developed soreswithin a week ofadmission.

Notes:

KCE Report 193S


was positioned assquare as possible tothe sacrum, ischiumand hip. A smallreflective marker stuckon to the patientsimplified focusing.Thermal images(thermograms) wererecorded on video-tape;the patient was thenturned, and theprocedure, includingstabilization, wasrepeated for the otherbuttock. During thesubsequent 4 weeks,patients admitted weresimilarly examined, butthermography was notcarried out. This controlwas established todetermine whether thethermographicexamination by itselfhad led to any changein the reportedincidence of pressuresores.

Reference

Inclusion criteria:

New admission

Unmarked skin

Exclusion criteria:

Pressure lesion onadmission


Predictive test

Outcome

Outcomemeasures

Effect sizes

209

Comments

210


standard:

Development of skinbreakdown in thebuttock region within 10days of admission wasreported by the nursingstaff and photographed.Redness alone,however marked orpersistent, was notcategorized as apressure sore.

Addressing missing data:

No details


Only descriptive data

Setting:

Geriatric assessmentunit at the Southerngeneral Hospital,Glasgow, Scotland

Blinding:

No details


Predictive test

Outcome

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S

Table 23 – Nixon 2006


Author and year:

Nixon 2006

Title:

Skin alterations ofintact skin and riskfactors associated withpressure ulcerdevelopment in surgicalpatients: a cohort study

Journal:

International Journal ofNursing Studies, 2006;44: 655-663

Study type:


Selection patient:

Surgical in-patientsadmitted to St. James’sUniversity Hospital,Leeds betweenSeptember 1998 andMay 1999.

Patient group:

Surgical in-patients

All patients:

109

Included N:

109

Completed N:

97

Drop-outs:

12

Incomplete follow-upresulted fromcancelled electivesurgery and earlydischarge (N=4),patient request todiscontinue (N=4)and

presence of pressureulcer at baselineassessment (N=4)

Age (median years,


Predictive test

Outcome

Outcomemeasures

Effect sizes

Index test 1:

skin assessment accordingthe classification scaleadapted from internationalclassification scales (AHCPR(Agency for Health CarePolicy and Research) 1992;EPUAP, 1999)

Index test 2:

Reference standard:

Occurrence of stage 2+ PUdevelopment according theclassification scale adaptedfrom internationalclassification scales (AHCPR(Agency for Health CarePolicy and Research) 1992;EPUAP, 1999)

Preventive methods:

Not reported

Outcome 1:


of blanchableerythma as apredictor for PUdevelopment(grades 2-4)

DOR 95%CI

Outcome 2:


of non-blanchableerythma as apredictor for PU


Specificity: 10% (4

Raw data (Grade 1a vs Grade0 erythema)

Referencestandard

Yes

Indextest

Yes 3

No 1

4

0.33 (0.03-3.27)



Raw data (Grade 1b and 1b+vs Grade 1a and Grade 0)

Refestandard

Yes

211

Comments

75% (19-99%)

: 10% (4-20%)

(Grade 1a vs Grade

Referencestandard

Yes No

55 58

6 7

61 65

73% (45-92%)

: 74% (64-83%)

(Grade 1b and 1b+vs Grade 1a and Grade 0)

Referencestandard

Yes No

Funding:

Jane Nixon hasbeenreimbursed forattendingconferences,has been paidspeakers feesand receivedresearchfunding fromHuntleighHealthcare Ltd.

Funding awardsfrom the TissueViability Society

TrainingFellowship (UK)and the Smithand NephewFoundationNursingResearchFellowship weremade to JaneNixon. Theseorganizationspeer reviewedthe grant

application and

212


Index test:

The classification scaleused was adapted frominternationalclassification scales,(AHCPR (Agency forHealth Care Policy andResearch) 1992;EPUAP, 1999) in orderto meet practical datacollection requirementsfor the purpose ofresearch. Specifically,Grade 0 (no skinchanges) was includedto clearly distinguishskin assessment ofnormal skin frommissing data.

In addition, alterationsto intact skin wereclassified as blanching(1a), non-blanching (1b)and non-blanching withother skin changesincluding, localinduration, oedema,pain, warmth ordiscoloration (1b+).

quartiles):

75 (55-95)

Gender (m/f):

38/59


15


82

Inclusion criteria:

(a) Scheduled forelective majorgeneral or vascularsurgery OR acuteorthopaedic, vascularand general surgicaladmission.

(b) Aged 55 years orover on day ofsurgery.

(c) Expected lengthof stay of 5 or moredays.


Predictive test

Outcome

Outcomemeasures

Effect sizes

development(grades 2-4)

DOR 95%CI

I

Indextest

Yes 11

No 4

15

8.00 (2.53-25.26)

KCE Report 193S

Comments

21 32

61 65

82 97

received areport of thefindings.

Limitations:

no report onblinding ofresearchertoward indextest andreferencestandard;unclear ifuninterpretableresults werefound; noinformationabout preventivemeasures;


Additionaloutcomes:

There wassignificantlyincreased oddsof pressure

KCE Report 193S


Reference standard:

The classification scaleused was adapted frominternationalclassification scales,(AHCPR (Agency forHealth Care Policy andResearch) 1992;EPUAP, 1999) in orderto meet practical datacollection requirementsfor the purpose ofresearch. Thedependent outcomevariable ‘pressure ulcer’was defined as a skinarea assessed as>=Grade 2, that is, asuperficial skinbreak/blister or worse.

Grade 5 (black eschar)was included as aseparate grade untilwound debridementenabled classificationby tissue layer.

Addressing missing data:

Variables wereexcluded from furtheranalysis if the p value

Exclusion criteria:

(a) General surgerysub-specialtiesincluding liver,urology and breastsurgery.

(b) Dark skinpigmentation whichprecluded reliableidentification of skinerythema.

(c) Skin conditionsover the sacrum,buttocks or heelswhich precludedreliable identificationof pressure inducedskin erythema.


Predictive test

Outcome

Outcomemeasures

Effect sizes

213

Comments

ulcerdevelopmentassociated withnon-blanchingerythema

(7.98, p = 0.002)and non-blanchingerythema withother skinchanges (9.17,p = 0.035).Logisticregressionmodelingidentified non-blanchingerythema, pre-operativealbumin, weightloss, and intra-operativeminimumdiastolic blood

pressure, asindependentpredictors ofGrade>=2pressure ulcerdevelopment.

Notes:

214


was >=0.2 (Altman,1991) or >=25% of datawas missing.

Missing values werereplaced by imputeddata.


A chi-square test wasused to compare theproportions of patientsclassified as havingGrade 0, Grade 1a,Grade 1b and Grade1b+ on any skin sitepreceding pressureulcer development. Skinchanges precedingpressure ulcerdevelopment were alsoclassified by Grade,independently for eachsite, and the differencein frequency ofpressure ulcersbetween Gradesexamined usingFisher’s exact test.

To identify whichclinical signs oferythema were


Predictive test

Outcome

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


predictive of skin loss,the odds of pressureulcer development forGrade 0, Grade 1a, 1band 1b+ were examinedusing single factorlogistic regression.

To identify variableswhich independentlyare predictive of>=Grade 2 pressureulcer development, therelationship betweenrisk factors andpressure ulcerdevelopment wasexplored using a threestage process forpatients who werepressure ulcer free atbaseline. The ‘worst’skin grade recorded atany time and on anysite during hospital stayor preceding pressureulcer development wasused to categorise skinalteration as a riskfactor. Univariateanalysis used singlefactor logisticregression with a binary


Predictive test

Outcome

Outcomemeasures

Effect sizes

215

Comments

216


response of pressureulcer or no pressureulcer.

Correlations betweenvariables were thenexamined usingPearson’s correlationcoefficient forcontinuous data orSpearman’s rankcorrelation for orderedcategorical data. Wherevariables werecorrelated with acorrelation coefficientof 40.7 and anassociated p-value of0.01 (Fielding et al.,1992), one waseliminated from furtherconsideration.

The final candidatevariables were enteredinto a logisticregression model usingforward stepwiseselection. The p valuedetermined entry(<0.25) and removal(40.9). The variablesidentified by the


Predictive test

Outcome

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


forward stepwiseselection were thenused as the basicmodel for furtherlogistic regressionanalysis. Correlatedvariables were droppedand addedsystematically in orderto determine the finalmodel in which eachvariable independentlypredicted subsequentpressure ulcerdevelopment asassessed by the size ofthe p value.

The model wasdetermined only frompatients with completedata for all candidatevariables. Therefore,when the final set ofvariables was obtainedthe model was refittedwith only those finalvariables in the model

statement.

Analyses were carriedout using the StataStatistical Software


Predictive test

Outcome

Outcomemeasures

Effect sizes

217

Comments

218


package.

Setting:

St. James UniversityHospital Leeds



Predictive test

Outcome

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S

5. SKIN MASSAGE

5.1. Review Protocol


Review question: What is the clinical effectiveness of skin massage in the prevention of pressure ulcers


Intervention Skin massage (method, frequency)

Comparison No skin massage Other preventive methods

Outcomes Critical outcome


Skin damage

Important outcomes:

Patient acceptability

Rate of development of pressure ulcers

Time to develop new pressure ulcer (time to event data)

Time in hospital or other healthcare settings (continuous data)

Healthpressure ulcer patients, therefore may have to be narratively summarised)

o

o

o

o

o

o


What is the clinical effectiveness of skin massage in the prevention of pressure ulcers ?

Individuals of all ages

Skin massage (method, frequency)

No skin massageOther preventive methods

Critical outcomes for decision-making:


Skin damage

Important outcomes:




ime in hospital or other healthcare settings (continuous data)

Health-related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes inpressure ulcer patients, therefore may have to be narratively summarised)

o Short-form health survey (SF36)

o Manchester Short Assessment of Quality of Life

o EQ-5D

o WHO-Quality of life BREF

o Cardiff HRQoL tool

o HUI

219


related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes inpressure ulcer patients, therefore may have to be narratively summarised)

220


o

Study design High quality systematic reviews of RCTs and/or RCTs only. Cochrane reviews

missing data such as available case analysis or ITT (with the appropriate assumptions) Cohort studies will be considered if no RCTs are available.

Exclusion Studie Non-English, non



Review strategy How will individual PICO characteristics be combined across studies in a metareviews)

Population Intervention

combined for meta Comparison Outcomes

separately from other side effects Blinding Unit of analysis Minimum duration of treatment = no minimum. Minimum follow up = no minimum. Minimum total sample size = no minimum. Use available case analysis for dealing with missing data if there is a 10% differential

groups or if the missing data is higher than the event rate, if cannot work out the available case analysis willtake the author’s data.

Analysis The following groups will be considered separately if data are present:

Children and ad



o Pressure ulcer quality of life (Gorecki)

High quality systematic reviews of RCTs and/or RCTs only.Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions formissing data such as available case analysis or ITT (with the appropriate assumptions)Cohort studies will be considered if no RCTs are available.

Studies with another population, intervention, comparison or outcome.English, non-French, non-Dutch language papers


Medline (OVID interface), Cinahl (EBSCO-interface), Embase, Library of the CAll years

How will individual PICO characteristics be combined across studies in a meta

Population – any population will be combined for meta-analysis except combination of childrenIntervention – different types of methods will be combined for meta-combined for meta-analysis; different types of frequency will be combined for metaComparison – any comparison which fits the inclusion criteria will be metaOutcomes – same outcomes will be combined for meta-analysis; single side effects will be metaseparately from other side effectsBlinding – Blinded and unblinded studies will be meta-analysed together.

of analysis – patients, individual pressure ulcersMinimum duration of treatment = no minimum.Minimum follow up = no minimum.Minimum total sample size = no minimum.Use available case analysis for dealing with missing data if there is a 10% differentialgroups or if the missing data is higher than the event rate, if cannot work out the available case analysis willtake the author’s data.


Children and adults (neonates, infants, children).

KCE Report 193S

will be included if they match our inclusion criteria and have appropriate assumptions formissing data such as available case analysis or ITT (with the appropriate assumptions)

interface), Embase, Library of the Cochrane Collaboration

How will individual PICO characteristics be combined across studies in a meta -analysis (for intervention

analysis except combination of children and adults.-analysis; different products will be

analysis; different types of frequency will be combined for meta-analysis.-analysed

analysis; single side effects will be meta-analysed

Use available case analysis for dealing with missing data if there is a 10% differential or higher between thegroups or if the missing data is higher than the event rate, if cannot work out the available case analysis will

KCE Report 193S


The following groups will be considered separately as subgroups if data are present:

Different categories of pressure ulcers Different locations of pressu




Date 05/06/2012

Database Ovid MEDLINE® In-Process & Other Non

Search Strategy 1. Pressure ulcer.sh2. decubit*.ti,ab.3. (pressure adj (sore* or ulcer* or damage*)).ti,ab.4. (bedsore* or bed-sore*).ti,ab.5. ((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.6. OR/1 – 57. Massage.sh8. Massage*.tw9. Rub*.tw10. Emollients.sh11. Emollient*.tw12. Moistur*.tw13. skin care.sh14. skin care.tw or care skin.tw15. OR/7 – 1416. randomized controlled trial.pt.17. controlled clinical trial.pt.18. randomi#ed.ab.19. placebo.ab.20. randomly.ab.21. exp Clinical Trials as topic




Different categories of pressure ulcers(from category 2 upwards where outcomes are reported separately)Different locations of pressure ulcers: sacral, heel and others.

Process & Other Non-Indexed Citations and Ovid MEDLINE(R) 1946 to

(pressure adj (sore* or ulcer* or damage*)).ti,ab.sore*).ti,ab.


skin care.tw or care skin.tw

randomized controlled trial.pt.controlled clinical trial.pt.

Clinical Trials as topic/

221


(from category 2 upwards where outcomes are reported separately) ;

present


901237845916479601

1323343036112

455091201868

1125239961547

7056132764984127

277597131376167239

222

Date 05/06/2012

22. trial.ti23. OR/16 – 2224. AND/6, 15, 2325. Limit language: ‘English, Dutch, Flemish, French’

Note


Date 28/06/2012

Database Embase

Search Strategy


1. ‘decubitus’/exp2. decubit*:ti,ab3. (pressure NEAR/1 (score* OR ulcer* OR4. (bed NEAR/2 sore*):ab,ti OR bedsore*5. ((moist* or friction or shear) near/2 (sore* or ulcer* or damag6. OR/1 – 57. ‘massage’/exp8. massage*:ti,ab9. rub*:ti,ab10. 'emollient agent'/exp11. emollient*:ti,ab12. moistur*:ti,ab13. ‘skin care’/exp14. ‘skin care’:ti,ab or ‘care skin’:ti,ab15. OR/7– 1416. ‘clinical trial’/exp17. ‘clinical trial (topic)’/exp18. random*:ti,ab19. factorial*:ti,ab20. crossover*:ti,ab OR (cross21. ((doubl* or singl*) NEAR/2


t language: ‘English, Dutch, Flemish, French’

(score* OR ulcer* OR damage)):ab,ti(bed NEAR/2 sore*):ab,ti OR bedsore*:ti,ab((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab

‘skin care’:ti,ab or ‘care skin’:ti,ab

xp

cross NEXT/1 over*):ti,ab((doubl* or singl*) NEAR/2 blind*):ti,ab

KCE Report 193S

255547100097825108

10098

e or wound* or injur* or lesion*)):ti,ab

1316854054764736797

1761685528558

6049431741510

1783364272473

10094191258738567

7362011942162987

KCE Report 193S

Date 28/06/2012

22. (assign* or allocat* or volunteer* or placebo*):ti,ab23. ‘crossover procedure’/exp24. ‘single blind procedure’/exp25. ‘double blind procedure’/26. OR/16 – 2527. AND/6, 15, 2628. Limit language: ’English, Dutch, French’

Note


(assign* or allocat* or volunteer* or placebo*):ti,ab‘crossover procedure’/exp‘single blind procedure’/exp‘double blind procedure’/exp

Limit language: ’English, Dutch, French’ AND exclude medline

223

1442215725553322515382

1084181736560

14640

224

Table 4 – Search filters Cochrane

Date 5/06/2012


Search Strategy


1. MeSH descriptor “Pressure ulcer” explode all trees2. Decubit*:ti,ab,kw3. (pressure near/2 (sore* or ulcer* or damage*)4. (bedsore* or bed5. ((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

lesion*)):ti,ab,kw6. OR/1 – 57. MeSH descriptor “massage” explode all trees8. (massage*):ti9. (rub*):ti,ab,kw10. (emollient*):ti,ab,kw11. MeSH descriptor “emollients” explode all trees12. (moistur*):ti,ab,kw13. (skin care):ti,ab,kw14. MeSH descriptor “skin care” explode all trees15. OR/7 – 1416. “Clinical Trial”:pt17. “Randomized Cont18. MeSH descript19. (trial*):ti,ab,kw20. (randomized or randomised21. (randomly):ti,ab,kw22. (group*):ti,ab,kw23. OR/16 – 2224. AND/6, 15, 23

Note



MeSH descriptor “Pressure ulcer” explode all trees:ti,ab,kw

re near/2 (sore* or ulcer* or damage*)):ti,ab,kw(bedsore* or bed-sore*):ti,ab,kw((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

):ti,ab,kw

MeSH descriptor “massage” explode all trees):ti,ab,kw

):ti,ab,kw(emollient*):ti,ab,kwMeSH descriptor “emollients” explode all trees(moistur*):ti,ab,kw(skin care):ti,ab,kwMeSH descriptor “skin care” explode all trees

inical Trial”:pt“Randomized Controlled Trial”:ptMeSH descriptor “clinical trial as topic” explode all trees

):ti,ab,kw(randomized or randomised):ti,ab,kw(randomly):ti,ab,kw

):ti,ab,kw

15, 23

KCE Report 193S

((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur*or

4873498293363

1163590

13551322437258636

1566265

505429449331310750972

24719826385185743

273083532112

109

KCE Report 193S


Date 05/06/2012


Search Strategy


34. MH “Pressure Ulcer”35. Decubit*36. Pressure n1 sore* OR pressure n1 ulcer* O37. Bedsore* OR bed-sore*38. ((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))39. OR/1 – 540. MH "Massage"41. “massage*”42. “rub*”43. MH "Emollients+"44. "Emollient*”45. “moistur*”46. MH "Skin Care"47. “skin care” or “care skin”48. OR/7 – 1449. MH "Clinical Trials+"50. “trial*”51. “randomi#ed”52. “randomly”53. “randomized controlled trial”54. PT “randomized controlled trial”55. PT “clinical trial”56. OR/16 – 2257. AND/6, 15, 2358. Limit language=’English, Dutch, French’

Note


interface)



skin”

“randomized controlled trial”PT “randomized controlled trial”

Limit language=’English, Dutch, French’ AND exclude medline records

225

7641480

8402156

1399

9715553773544095780679750

37734884

172651053651349916463224832

88579758

51022165760

7823

226


Figure 1 – Flowchart search strategy


KCE Report 193S

5.2.3. Excluded study

Study

Arashi, M., Sugama, J., Sanada, H., Konya, C., Okuwa, M., Nakagami, G. et al.healing of Stage I pressure ulcers in older adult patients.



The systematic search through multiple electronic databases resulted in 270 records: 98 in Medline (Ovid), 23 in Cinahl (EBSCOand 109 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted inabstracts another 191 records were excluded. Reasons for exclusion are listed in figure 1. The full text of the remaining 2 records wasBased on this review, 1 record was excluded. Reason for exclusion was listed. This resulted in retaining 1Dutch nursing homes.

63

5.3.2. Summary tables

Table 6 – Summary table

Study ID Intervention/ comparator

Duimel-Peeters 200763

Massage with indifferent creamversus massage with dimethylsulfoxide (DMSO) cream versusno massage


Reason for exclusion

, Sanada, H., Konya, C., Okuwa, M., Nakagami, G. et al. (2010). Vibration therapy accelerateshealing of Stage I pressure ulcers in older adult patients. Advances in Skin & Wound Care.23(7):321-7.

Non

he systematic search through multiple electronic databases resulted in 270 records: 98 in Medline (Ovid), 23 in Cinahl (EBSCOand 109 in the Library of the Cochrane Collaboration. Duplicate records were excluded, which resulted in 193 records. Based on the screening of title

another 191 records were excluded. Reasons for exclusion are listed in figure 1. The full text of the remaining 2 records wason for exclusion was listed. This resulted in retaining 1 clustered cross

Intervention/ comparator Population Outcome

Massage with indifferent creamwith dimethyl

sulfoxide (DMSO) cream versus

Residents of 8 Dutch nursinghomes

Incidence of pressure ulcers

227


Non-randomized trial

he systematic search through multiple electronic databases resulted in 270 records: 98 in Medline (Ovid), 23 in Cinahl (EBSCO interface), 40 in Embase,193 records. Based on the screening of titles and

another 191 records were excluded. Reasons for exclusion are listed in figure 1. The full text of the remaining 2 records was reviewed in detail.clustered cross-over randomized trial performed in

Length of study

Incidence of pressure ulcers 4 weeks of treatmentfollowed by a wash-outperiod of 2 weeks andanother 4 weeks oftreatment

228

5.3.3. Grade evidence profiles

Table 7 – Clinical GRADE evidence profile: Massage with indifferent creamprevention of pressure ulcers

Quality assessment

No of

studies

Design Risk of

bias


Incidence of PU (follow-up 4 weeks; assessed with the four-grade system of the EPUAP using a

1

Duimel-

Peeters,

2007

randomised

trials

very

serious1

no serious

inconsistency

no serious

indirectness

1 No details of allocation concealment.It was not clear whether the outcome assessors were blinded2 Confidence interval crossed both MID points.


inical GRADE evidence profile: Massage with indifferent cream (Vaseline) + position change versus position change only for the

No of patients


considerations

Massage with

indifferent cream +

position change

Position

change

only

Relative

(95%

CI)

grade system of the EPUAP using a transparent disk)

no serious

indirectness²

very

serious2

none 13/31

(41.9%)

7/18

(38.9%)

RR 1.22

(0.61 to

2.41)

38.9%

1 No details of allocation concealment.It was not clear whether the outcome assessors were blinded

KCE Report 193S

+ position change versus position change only for the


Relative

(95%

CI)

Absolute

RR 1.22

(0.61 to

2.41)

86 more per 1000 (from

152 fewer to 548 more)

VERY

LOW

CRITICAL

86 more per 1000 (from152 fewer to 548 more)

KCE Report 193S

Table 8 – Clinical GRADE evidence profile: Massage with DMSO creampressure ulcers

Quality assessment

No of

studies

Design Risk

of

bias

Inconsisten

cy

Indirectnes

s

Incidence of PU (follow-up 4 weeks; assessed with the four-grade system of the EPUAP using a transparent disk)

1 Duimel-

Peeters,

2007

randomis

ed trials

very

serio

us1

no serious

inconsistenc

y

no serious

indirectness

1 No details of allocation concealment. It was not clear whether the outcome assessors were blinded2 Confidence interval crossed one MID point.


l GRADE evidence profile: Massage with DMSO cream + position change versus position change

No of patients

Imprecisi

on

Other

consideratio

ns

Massage with DMSO

cream + position change

Position

change only

Relative

(95% CI)


serious2 none 18/29

(62.1%)

7/18

(38.9%)

RR 1.85

(0.87 to

2.99

38.9%

1 No details of allocation concealment. It was not clear whether the outcome assessors were blinded

229

position change only for the prevention of

Effect Qual

ity

Importan

ce

Relative

(95% CI)

Absolute

1.85

(0.87 to

2.99)

331 more per 1000 (from

51 fewer to 774 more)

VER

Y

LOW

CRITICA

L


230

Table 9 – Clinical GRADE evidence profile: Massage with DMSO creamposition change for the prevention of pressure ulcers

Quality assessment

No of

studies

Design Risk of

bias


Incidence of PU (follow-up 4 weeks; assessed with the four-grade system of the EPUAP using a transparent di

1 Duimel-

Peeters,

2007

randomised

trials

very

serious1

no serious

inconsistency

no serious

indirectness

1 No details of allocation concealment. It was not clear whether the outcome assessors were blinded2 The confidence interval crossed one MID point.


linical GRADE evidence profile: Massage with DMSO cream + position change versus massage with indifferent creamfor the prevention of pressure ulcers

No of patients


considerations

Massage

with DMSO

cream +

position

change

Massage with

indifferent

cream +

Position

change

Relative

(95% CI)


no serious

indirectness

very

serious2

none 18/29

(62.1%)

13/31

(41.9%)

RR

1.43(0.

to 2.

38.9%

1 No details of allocation concealment. It was not clear whether the outcome assessors were blinded

KCE Report 193S

versus massage with indifferent cream (Vaseline) +


Relative

(95% CI)

Absolute

RR

(0.79

to 2.14)

180 more

per 1000

(from 88

fewer to 478

more)

VERY

LOW

CRITICAL

167 moreper 1000(from 82

fewer to 443more)

KCE Report 193S



Author and year:

Duimel-Peeters, 2007

Title:

The effectiveness ofmassage with andwithout dimethylsulfoxide in preventingpressure ulcers: Arandomized double-blind cross-over clinicaltrial in patients prone topressure ulcers.

Journal:

International Journal ofNursing Studies, 2007;44: 1285-95.

Study type:

Multicentric randomizeddouble-blinded cross-over trial

Sequence generation:Throwing a dice

Allocation:not reported

Blinding: Not reported

Addressingincomplete outcomedata:

Not reported

Patient group:

Residents of 8 Dutchnursing homes

All patients

Randomised N: 79

Completed N:

Period 1: 78

Period 2: 64

Drop-outs:

Period 1: 1

Period 2: 15

Some participantsdecided not toparticipate any longer

Some health careworkers got tired ofapplying the treatmentas accurately aspossible

Gender: 69.6% female

Age: mean 81.3, SD9.76, range 45-97

Group 1 (period 1)

Randomised N: 31

Completed N: 31


Intervention

Comparison


Group 1:

A 2–3-min massage ofthe coccyx, both heelsand ankles with anindifferent cream(Cremor vaselinicetomacrogolis

FNA; ‘Vaseline’). Thisprocedure wasrepeated every 6 h for 4weeks

Group 2: A 2–3-minmassage of the coccyx,both heels and ankleswith a dimethylsulfoxide (DMSO)cream (5%), Thisprocedure wasrepeated every 6 h for 4weeks

Group 3: positionchange only

All groups:

30° position change

Outcome 1:

Incidence of PU(%/period of 4 weeks) inperiod 1

Incidence of PU(%/period of 4 weeks) inperiod 2

Group 1:

Period 1: 13/31 (41.9%)

Period 2: 3/22 (13.6%)

Group 2:

Period 1: 18/29 (62.1%)

Period 2: 3/25 (12.0%)

Group 3:

Period 1: 7/18 (38.9%)

Period 2: 1/17 (5.9%)

Period 1

P value=0.189

Period 2

P value=0.726

Period 1:

Treatment 1

OR: 1.135

95% CI:

P value: 0.834

Treatment 2:

OR: 2.571

95% CI:

P value: 0.126

231

Comments

13/31 (41.9%)

Period 2: 3/22 (13.6%)

Period 1: 18/29 (62.1%)

Period 2: 3/25 (12.0%)

Period 1: 7/18 (38.9%)

Period 2: 1/17 (5.9%)

P value=0.189

726

0.834

0.126

Funding:

none reported

Limitations:

Underpowered

Randomization processby throwing a dice for 2of the 3 interventions.

Unclear allocationconcealment

Not clear whetheroutcome assessorswere blinded

Relatively high dropoutrate in period 2

Additional outcomes:

KM survival curves.

Massaging with theindifferent cream or onlychanging of positionsseemed to result inbetter pressure ulcerfree prognosis thanbeing massaged withthe DMSO cream. Astimes goes on, thedashed and bold curvesappear to grow further

232


Statistical analysis:Differences incharacteristics betweenpatients in the varioustreatment groups weretested for each periodwith Chi-square testsfor categorical data andt-tests for continuousdata. Mann–Whitneyand Kruskal–Wallistests were usedbecause of non-normality of somevariables.

Frequency tables forthe outcome variablewere constructed foreach treatment period.

Logistic regression wasused to examine theresults of eachtreatment in terms ofpressure ulcerprevention.

To correct for

possible confoundingvariables, the followingcovariates were added(together andseparately): length,

Dropouts: 0

Age: not reported

Gender (m/f): notreported

Other relevant patientcharacteristics: none

Group 2 (period 1)

Randomised N: 29

Completed N: 29

Dropouts: 0

Age: not reported

Gender (m/f):

Not reported


Group 3 (period 1)

Randomised N: 18

Completed N: 18

Dropouts: 0

Age: not reported

Gender (m/f):

Not reported



Intervention

Comparison


Period 2:

Treatment 1

OR: 2.526

95% CI:

P value: 0.441

Treatment 2:

OR: 2.182

95% CI:

P value: 0.516

KCE Report 193S

Comments

0.441

0.516

apart (until day 18),suggesting that thebeneficial effects of onlychanging positionrelative to massagingwith a DMSO-cream

increase as treatmentcontinued for a longerperiod. However,beyond day 18, thethree treatments tendedto have the sameeffects.

Notes: none

KCE Report 193S


weight, body massindex (BMI), length ofstay on the ward (inmonths), age, sex,incontinence level, typeof pressure-relievingcushions used and useof other preventivemethods. Non-significant covariateswere removed usingbackward deletion.

Kaplan–Meier curveswere constructed toobtain a clearerrepresentation of thesurvival prognosis foreach treatment.

Baseline differences:Patients were notsignificantly differentacross periods withrespect to age, sex,length, weight, BMI,length of stay on theward, incontinencelevel, type of pressure-relieving cushions usedand use of otherpreventive methods.

Study power/samplesize:

Group 1 (period 2)

Randomised N: 28

Completed N: 22

Dropouts: 6

Age: not reported

Gender (m/f): notreported


Group 2 (period 2)

Randomised N: 27

Completed N: 25

Dropouts: 2

Age: not reported

Gender (m/f):

Not reported


Group 3 (period 2)

Randomised N: 24

Completed N: 17

Dropouts: 7

Age: not reported

Gender (m/f):


Intervention

Comparison


233

Comments

234


No a priori sample sizecalculation

Setting:

Dutch nursing homes

Length of study:

4 weeks in period 1

4 weeks in period 2

2 weeks wash-outperiod between periods1 and 2

Assessment of PUs:

Braden scale to assessPU risk (cutoff point of20)

PU were gradedaccording to the four-grade system of theEuropean PressureUlcer Advisory Panelusing a transparentdisk. Because of thereversibility of grade Iulcers, these ulcerswere only recorded aspressure ulcers if theywere still present after 4h and if two externalobservers confirmed thenurse’s rating of gradeI. A transparent disk

Not reported


Inclusion criteria:

1) have a light skincolour, 2) have residedin a long-stay ward of anursing home for more

than two months

3) rest on an anti-pressure ulcer mattress(i.e. poly urethanemattress or equivalent),

4) be willing to giveinformed consent orhave this provided bytheir relative/legalrepresentative

5) to be at high risk ofdeveloping pressureulcers according to theBraden scale using acut-off point of 20.

Exclusion criteria:

1) already being treatedwith massage foranother medical


Intervention

Comparison


KCE Report 193S

Comments

KCE Report 193S


with a diameter of 6.5cm was used to assesslocal redness. Thisinvolved first releasingthe pressure on thebody part, for exampleby changing thepatient’s position. If thelocal redness persistedafter 10min, whenpushing the convex lensagainst the skin, thegrade 1 pressure ulcerwas confirmed.

Multiple ulcers:

The outcome variabledevelopment of PU ornot regardless of thenumber of PU

indication (and it wasnot possible to end thistreatment)

2) undergoing surgeryin the near future or hadundergone surgery lessthan two weekspreviously

3) had pressure ulcersalready present at thecoccyx, heels or ankles(the only places thatwere massaged in thisresearch

4) expected to haveshort length of stay

5) a short lifeexpectancy (<10months).


Intervention

Comparison


235

Comments

236

5.3.5. Forest plots

Figure 2 – Incidence of pressure ulcers for comparison: m

Figure 3 – Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versus position change only

Study or Subgroup

Duimel-Peeters 2007 (1)

Total (95% CI)

Total events



Events

13

13

Total

31

31

Events

7

7

Total

18

18

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.08 [0.53, 2.20]

1.08 [0.53, 2.20]

Experimental Control Risk Ratio

(1) Period 1

Study or Subgroup


Total (95% CI)

Total events



Events

18

18

Total

29

29

Events

7

7

Total

18

18

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.60 [0.84, 3.04]

1.60 [0.84, 3.04]


(1) Period 1


Incidence of pressure ulcers for comparison: massage with indifferent cream + position change versus p

Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versus position change only

M-H, Fixed, 95% CI

1.08 [0.53, 2.20]

1.08 [0.53, 2.20]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours massage Favours standard only

M-H, Fixed, 95% CI

1.60 [0.84, 3.04]

1.60 [0.84, 3.04]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours massage with DMSO Favours standard only

Version 11/14/2012 3:08 PM

with indifferent cream + position change versus position change only

Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versus position change only

KCE Report 193S

Figure 4 – Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versuposition change.

6. REPOSITIONING



Protocol Repositioning

Review question How and at what frequency should repositioning be undertaken for the prevention of pressure ulc

Population Individuals

Intervention Repositioning

Frequency of repositioning

Different positions (e

Devices included for repositioning:

Profiling bed

Tilt in spac

Comparison No repositioning

Different frequencies of repositioning

Different positions for repositioning

Outcomes Critical outcomes for decision

Study or Subgroup


Total (95% CI)

Total events



Events

18

18

Total

29

29

Events

13

13

Total

31

31

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.48 [0.90, 2.45]

1.48 [0.90, 2.45]


(1) Period 1


Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versu

Repositioning

How and at what frequency should repositioning be undertaken for the prevention of pressure ulc

of all ages in all settings

Repositioning technique

Frequency of repositioning

Different positions (e.g. 90-degree lateral rotation, 30 degree tilt)

Devices included for repositioning:

Profiling bed

Tilt in space chairs

No repositioning

Different frequencies of repositioning

Different positions for repositioning


M-H, Fixed, 95% CI

1.48 [0.90, 2.45]

1.48 [0.90, 2.45]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours massage + DMSO Favours massage

237

Incidence of pressure ulcers for comparison: massage with DMSO cream + position change versu s massage with indifferent cream +

How and at what frequency should repositioning be undertaken for the prevention of pressure ulc ers?

238


Proportion of participants developing new pressure ulcers (ulcer)

Important outcomes:




Time in hospital

Healthpressure ulcer patients, therefore may have to be narratively summarised

o Short


o EQ

o WHOQOL


o HUI


Study design High quality s

Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missindata such as available case analysis or ITT (with the appropriate assumptions)


Exclusion Studies with outcomes that do not involve pressure ulcers

Abstracts unless no RCTs are found

Non-Englis

Search strategy The databases to be searched are:

Medline, Embase, Cinahl, the Cochrane Library.

All years.

Studies will be restricted to English language only


Repositioning

Proportion of participants developing new pressure ulcers (dichotomous outcome)(describe

Important outcomes:




Time in hospital or other healthcare setting (continuous data)

Health-related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes inpressure ulcer patients, therefore may have to be narratively summarised



EQ-5D

WHOQOL BREF

Cardiff HRQoL tool

HUI


High quality systematic reviews of RCTs and/or RCTs only.

Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missindata such as available case analysis or ITT (with the appropriate assumptions)


Studies with outcomes that do not involve pressure ulcers


English language papers

The databases to be searched are:


All years.


Version 11/14/2012 3:08 PM

dichotomous outcome)(describe different categories of

ty of life (continuous data) (although unlikely to be sensitive enough to detect changes in

Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missin gdata such as available case analysis or ITT (with the appropriate assumptions)

KCE Report 193S


The review strategy How will individual PICO characteristics be combined i

Population

Intervention

Outcomes

Study designstudies will be meta

Unit of analysisunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe differentcategories of ulcer).

Describe which support surfaces a

Minimum duration of treatment = no minimumimprovements.

Minimum follow up = no minimum.

Minimum total sample size = no minimum.

Use available case analysis for dealing with missing data ifor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.

MIDs: 0.75 to 1.25 for dichotomous variables and 0.5 x standard dev

Analysis Strata – where included studies are split up at outset as separate reviews (dissimilar groups andconfident that the intervention will work very differently in the two (or more) strata.recommendations on these.

The following groups will be considered separately as strata if data are present:

Children (neonates, infants, children) and adults

People with

Self repositi

People with sensory impairment


Repositioning

How will individual PICO characteristics be combined in a meta-analysis?:

Population – any population will be combined for meta-analysis except for different strata.

ervention – different types of frequency will be meta-analysed, different positions will be meta

Outcomes – single side effects

tudy design – randomised and quasi-randomised studies will be meta-analysed together. Blinded and unblindedstudies will be meta-analysed together. Crossover trials will be meta-analysed together with parallel trials

Unit of analysis – patients, clusters (hospital wards), individual pressure ulcersunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe differentcategories of ulcer).

Describe which support surfaces are used.

Minimum duration of treatment = no minimum, but would expect at least a fortnight before they showimprovements.



Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data..

MIDs: 0.75 to 1.25 for dichotomous variables and 0.5 x standard deviation for continuous variables.

where included studies are split up at outset as separate reviews (dissimilar groups andconfident that the intervention will work very differently in the two (or more) strata.recommendations on these.



People with neurological impairment or spinal cord damage or injury

Self repositioning versus manual repositioning versus repositioning by a device


239

analysis?:

for different strata.

analysed, different positions will be meta-analysed.

analysed together. Blinded and unblindedanalysed together with parallel trials

(hospital wards), individual pressure ulcers – for those where patients are theunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different

, but would expect at least a fortnight before they show

there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take the

iation for continuous variables.

where included studies are split up at outset as separate reviews (dissimilar groups and we need to beconfident that the intervention will work very differently in the two (or more) strata. The GDG will make separate


oning versus manual repositioning versus repositioning by a device

240


Subgroup analysison the basis of pre

The following groups will be considered separately as subgroups (if there is heterogeneity):

different risk stratification

different clinical populations

Notes Where have said ‘describe’ or ‘descriptive’ this will be noted in the summary table.

6.2. search Strategy


Table 2 – Search filters in OVID Medline

Search strategy Repositioning

Date 27th Mar 2012

Database Medline-Ovid

Search strategy 1 pressure ulcer/

2 decubit*.ti,ab.

3 (pressure adj (sore* or ulcer* or damage)).ti,ab.

4 (bedsore* or bed-sore*).ti,ab.

5 (incontinen* adj2 dermatitis).ti,ab.

6 ((moist* or friction or shear) adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

7 or/1-6

8 limit 7 to english language

9 exp posture/

10 exp patient positioning

11 "moving and lifting patients"/

12 (re-position* or reposition*).ti,ab.

13 (mobilis* or mobiliz*).ti,ab.


Repositioning

Subgroup analysis – combining all the studies together initially and then looking at any inconsistency between studiesbasis of pre-defined subgroups.



different clinical populations

Where have said ‘describe’ or ‘descriptive’ this will be noted in the summary table.

(pressure adj (sore* or ulcer* or damage)).ti,ab.

sore*).ti,ab.

(incontinen* adj2 dermatitis).ti,ab.


limit 7 to english language

exp patient positioning/

"moving and lifting patients"/

position* or reposition*).ti,ab.

(mobilis* or mobiliz*).ti,ab.

Version 11/14/2012 3:08 PM

combining all the studies together initially and then looking at any inconsistency between studies



Results


8802

3830

5969

494

49

614

13334

10621

56402

689

160

9322

53398

KCE Report 193S


14 (turn* adj5 (patient* or interval* or frequen*)).ti,ab.

15 or/9-14

16 8 and 15

17 randomized controlled trial.pt.

18 controlled clinical trial.p

19 randomi#ed.ab.

20 placebo.ab.

21 drug therapy.fs.

22 randomly.ab.

23 trial.ab.

24 groups.ab.

25 or/17-24

26 Clinical Trials as topic.sh.

27 trial.ti.

28 or/17-20,22,26-27

29 Meta-Analysis/

30 Meta-Analysis as Topic/

31 (meta analy* or metanaly* or me

32 ((systematic* or evidence*) adj2 (review* or overview*)).ti,ab.

33 (reference list* or bibliograph* or hand search* or manual search* or relevant journals).ab.

34 (search strategy or search criteria or systematic search or study selecti

35 (search* adj4 literature).ab.

36 (medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl orscience citation index or bids or cancerlit).ab.

37 cochrane.jw.

38 or/29-37

39 28 or 38

40 16 and 39

41 letter/


(turn* adj5 (patient* or interval* or frequen*)).ti,ab.

randomized controlled trial.pt.

controlled clinical trial.pt.

Clinical Trials as topic.sh.

Analysis as Topic/

(meta analy* or metanaly* or metaanaly*).ti,ab.

((systematic* or evidence*) adj2 (review* or overview*)).ti,ab.

(reference list* or bibliograph* or hand search* or manual search* or relevant journals).ab.

(search strategy or search criteria or systematic search or study selection or data extraction).ab.

(search* adj4 literature).ab.

(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl orscience citation index or bids or cancerlit).ab.

241

Results

on or data extraction).ab.

(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or

6913

125576

1219

317876

83345

278870

131961

1492734

171910

241007

1128651

2856539

157206

99919

777187

31028

11703

40322

48100

19344

20846

18952

59408

7644

140085

877198

154

242


42 editorial/

43 news/

44 exp historical article/

45 Anecdotes as Topic/

46 comment/

47 case report/

48 (letter or comment*).ti.

49 or/41-48

50 randomized controlled trial/ or random*.ti,ab.

51 49 not 50

52 animals/ not humans/

53 exp Animals, Laboratory/

54 exp Animal Experimentation/

55 exp Models, Animal/

56 exp Rodentia/

57 (rat or rats or mouse or mice).ti.

58 or/51-57

59 40 not 58


(letter or comment*).ti.

randomized controlled trial/ or random*.ti,ab.

animals/ not humans/

Animals, Laboratory/

exp Animal Experimentation/

(rat or rats or mouse or mice).ti.

Version 11/14/2012 3:08 PM

Results

745026

297341

142441

300400

4103

484718

1546366

82026

2995614

661256

2980844

3553260

655846

5130

358217

2423006

1019659

7043844

154

KCE Report 193S

Table 3 – Search filters in Embase


Date 27th Mar 2012

Database Embase

Search strategy 1 decubitus/

2 decubit*.ti,ab.


4 (bedsore* or bed-sore*).t



7 or/1-6


9 exp position/

10 patient positioning/

11 patient lifting/

12 mobilization/

13 (re-position* or reposition*).ti,ab.

14 (mobilis* or mobiliz*).ti,ab.

15 (turn* adj5 (patient* or interval* or frequen*)).ti,ab.

16 or/9-15

17 8 and 16

18 random*.ti,ab.

19 factorial*.ti,ab.

20 (crossover* or cross over*).ti,ab.

21 ((doubl$ or singl$) adj blind$).ti,ab.

22 (assign* or allocat* or volunteer* or placebo*).ti,ab.

23 crossover procedure/

24 double blind procedure/

25 single blind procedure/



sore*).ti,ab.




position* or reposition*).ti,ab.

(mobilis* or mobiliz*).ti,ab.

(turn* adj5 (patient* or interval* or frequen*)).ti,ab.

(crossover* or cross over*).ti,ab.

singl$) adj blind$).ti,ab.

(assign* or allocat* or volunteer* or placebo*).ti,ab.

crossover procedure/

double blind procedure/

single blind procedure/

243

Results


12141

4617

6831

631

737

51

16424

12654

83846

10602

92

12892

10628

61541

8307

178268

1711

677053

17713

57802

130691

526036

31644

102550

14668

244


26 randomized controlled trial/

27 or/18-26

28 systematic review/

29 meta-analysis/

30 (meta analy* or metanaly* or metaanaly*).ti,ab.

31 ((systematic or evidence) adj2 (review* or overview*)).ti,ab.


33 (search strategy or search criteria or systematic


35 (medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl orscience citation index or bids or cancerlit).ab.

36 ((pool* or combined) adj2 (data or trials or studies or results)).ab.

37 cochrane.jw.

38 or/28-37

39 letter.pt. or letter/

40 note.pt.

41 editorial.pt.

42 case report/ or case study/


44 or/39-43

45 randomized controlled trial/ or random*.ti,ab

46 44 not 45

47 animal/ not human/

48 nonhuman/

49 exp Animal Experiment/

50 exp experimental animal/

51 animal model/

52 exp Rodent/



randomized controlled trial/

analy* or metanaly* or metaanaly*).ti,ab.

((systematic or evidence) adj2 (review* or overview*)).ti,ab.




(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl orscience citation index or bids or cancerlit).ab.

ined) adj2 (data or trials or studies or results)).ab.

case report/ or case study/



exp Animal Experiment/

exp experimental animal/


Version 11/14/2012 3:08 PM

Results

search or study selection or data extraction).ab.

(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or

295607

1123098

46604

58505

51361

54624

23210

24799

22510

70467

29479

11040

210071

755399

462400

389343

1772519

132536

3256668

752555

3232924

1267429

3772499

1486602

366425

619749

2423085

KCE Report 193S


54 or/46-53

55 27 or 38

56 17 and 55

57 56 not 54

Table 4 – Search filters in CINAHL


Date 27th Mar 2012

Database CINAHL

Search strategy S19 s17 not s18

S18 PT anecdote or PT audiovisualbrief item or PT cartoon or PT commentary or PT computer program or PT editorial or PT games or PT glossary orPT historical material or PT interview or PT letter or PT listservs or PTpamphlet or PT pamphlet chapter or PT pictorial or PT poetry or PT proceedings or PT “questions and answers” orPT response or PT software or PT teaching materials or PT website

S17 S7 and S15 Limiters –

Search modes – Boolean/Phrase

S16 S7 and S15

S15 S8 or S9 or S10 or S11 or S12 or S13 or S14

S14 turn* N5 frequen*

S13 turn* N5 interval*

S12 turn* N5 patient*

S11 mobilis* or mobiliz*

S10 re-position* or repositio

S9 (MH "Posture+")

S8 (MH "Patient Positioning+")

S7 S1 or S2 or S3 or S4 or S5 or S6


PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PTbrief item or PT cartoon or PT commentary or PT computer program or PT editorial or PT games or PT glossary or

or PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PTpamphlet or PT pamphlet chapter or PT pictorial or PT poetry or PT proceedings or PT “questions and answers” orPT response or PT software or PT teaching materials or PT website

– English Language; Exclude MEDLINE records

Boolean/Phrase

S8 or S9 or S10 or S11 or S12 or S13 or S14

position* or reposition*

(MH "Patient Positioning+")

S1 or S2 or S3 or S4 or S5 or S6

245

Results

1073325

8598716

1271496

298

286

Results

or PT bibliography or PT biography or PT book or PT book review or PTbrief item or PT cartoon or PT commentary or PT computer program or PT editorial or PT games or PT glossary or

masters thesis or PT obituary or PTpamphlet or PT pamphlet chapter or PT pictorial or PT poetry or PT proceedings or PT “questions and answers” or

156

979974

243

734

20951

112

29

1043

4063

975

9597

5903

9430

246


S6 ((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*))

S5 incontinen* n2 dermatitis

S4 bedsore* OR bed-sore*

S3 pressure n1 sore* OR pressure n1 ulcer* OR pressure n1 damage*

S2 decubit*

S1 (MH "Pressure Ulcer")

Table 5 – Search filters in Cochrane


Date 27th Mar 2012

Database Cochrane

Search strategy #1 MeSH descriptor Pressure Ulcer explode all trees

#2 decubit*:ti,ab,kw

#3 (pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw

#4 (bedsore* or bed-sore*):ti,ab,kw

#5 (incontinen* near/2 dermatitis):ti,ab,kw

#6 ((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab,kw

#7 (#1 OR #2 OR #3 OR #4 OR #5 OR #6)

#8 MeSH descriptor Posture explode all trees

#9 MeSH descriptor Patient Positi

#10 MeSH descriptor Moving and Lifting Patients explode all trees

#11 (mobilis* or mobiliz*):ti,ab,kw

#12 (re-position* or reposition*):ti,ab,kw

#13 (turn* near/5 (patient* or interval* or frequen*)):ti,ab,kw

#14 (#8 OR #9 OR #10 OR

#15 (#7 AND #14)



incontinen* n2 dermatitis

sore*


(MH "Pressure Ulcer")

MeSH descriptor Pressure Ulcer explode all trees

(pressure near/2 (sore* or ulcer* or damage)):ti,ab,kw

sore*):ti,ab,kw

* near/2 dermatitis):ti,ab,kw

((moist* or friction or shear) near/2 (sore* or ulcer* or damage or wound* or injur* or lesion*)):ti,ab,kw

(#1 OR #2 OR #3 OR #4 OR #5 OR #6)

MeSH descriptor Posture explode all trees

MeSH descriptor Patient Positioning explode all trees

MeSH descriptor Moving and Lifting Patients explode all trees

(mobilis* or mobiliz*):ti,ab,kw

position* or reposition*):ti,ab,kw

(turn* near/5 (patient* or interval* or frequen*)):ti,ab,kw

(#8 OR #9 OR #10 OR #11 OR #12 OR #13)

Version 11/14/2012 3:08 PM

Results

((moist* or friction or shear) and (sore* or ulcer* or damage or wound* or injur* or lesion*)) 1336

65

152

8135

467

7399

Results


480

341

818

32

10

62

1151

3009

39

8

2525

413

477

6328

138

KCE Report 193S


Figure 1 – Flow diagram of clinical article selection

Titles and abstractsidentified, n = 472

Full copiesretrieved andassessed foreligibility, n = 8

Publicationsincluded in review,n = 7 + 2 (fromCochrane Review)

Excluded, n = 1


Flow diagram of clinical article selection

Titles and abstracts472

Excluded, n = 464

247

248

6.2.3. Excluded clinical studies

Table 6 – Studies excluded from the clinical review

Reference

Defloor 2005



Nine studies were included in the review.64

,65

,66

,67,


Evidence from these are summarised in the clinical GRADE evidence profile. We searched for randomised trials assessing effectthe prevention of pressure ulcers in patients of all ages in any setting.

Seven randomised trials (three cluster randomised trials

Included population varied from geriatric patients to critically ill infants and children, all assessed in different inpatiengeriatric patients with a mean age of 80 years, one trial included acute inpatient with a mean age of 70 years and the sixthchildren (Table 1). Two studies were of turning tables, were included in thethey were deemed more relevant to the repositioning review than the devices for prevention review.

Studies looked at different reposition techniques applied at different time intervals. For this purpose of this review, the trials have been grouped andanalysed in five different comparisons:

o Repositioning (frequent turning with or without the use of pressure reducing mattress) versus no repos

o Different frequencies of repositioning –64

o Different positions for repositioning – 30° tilt position versus 90° lateral and supine positionhead and back) versus standard care (supine position)

o Different positions for repositioning – prone/semi recum

Turning tables for repositioning –105,106

Trials reported the incidence of pressure ulcers (proportion of participants devedevelopment’ and patient tolerability.

Included studies had varying time periods (ranging from one night to 5 weeks). Cluster randomised trials and trials includingseparately.


d from the clinical review


Sub-population ofincluded trial (Defloor2005B)

,68,69,70 , 71 , 72

Evidence from these are summarised in the clinical GRADE evidence profile. We searched for randomised trials assessing effectthe prevention of pressure ulcers in patients of all ages in any setting.

(three cluster randomised trials64

,65

,66

and six parallel RCTs67,68,69,70 , 71 , 72

were identified.

Included population varied from geriatric patients to critically ill infants and children, all assessed in different inpatiengeriatric patients with a mean age of 80 years, one trial included acute inpatient with a mean age of 70 years and the sixth

). Two studies were of turning tables, were included in the Cochrane Review Support surfaces for the prevention of pressure ulcersthey were deemed more relevant to the repositioning review than the devices for prevention review.

techniques applied at different time intervals. For this purpose of this review, the trials have been grouped and

Repositioning (frequent turning with or without the use of pressure reducing mattress) versus no repos itioning (standard care without turning).64,66,68

30° tilt position versus 90° lateral and supine position65,72

and semi recumbhead and back) versus standard care (supine position)

69

prone/semi recumbent positioning versus control supine positioning67

rials reported the incidence of pressure ulcers (proportion of participants deve loping pressure ulcers, Grades I

Included studies had varying time periods (ranging from one night to 5 weeks). Cluster randomised trials and trials including

Version 11/14/2012 3:08 PM

Evidence from these are summarised in the clinical GRADE evidence profile. We searched for randomised trials assessing effect iveness of repositioning for

were identified.

Included population varied from geriatric patients to critically ill infants and children, all assessed in different inpatien t hospital settings. Four trials includedgeriatric patients with a mean age of 80 years, one trial included acute inpatient with a mean age of 70 years and the sixth trial included infants and

Cochrane Review Support surfaces for the prevention of pressure ulcers73

,

techniques applied at different time intervals. For this purpose of this review, the trials have been grouped and

itioning (standard care without turning).64

and semi recumbent position (i.e., 45° position of the

loping pressure ulcers, Grades I-IV)65

,66

,72

, the ‘time to pressure ulcer

Included studies had varying time periods (ranging from one night to 5 weeks). Cluster randomised trials and trials including children67

have been analysed

KCE Report 193S


Table 7 – Summary of studies included in the review


Defloor 200564

2, 3 hourly turning scheme on astandard institutional mattress and 4,6-hourly turning scheme on apressure reducing mattress.

The turning schemes consistealternating a semi-recumbentposition with a lateral position.

Standard care involving preventivenursing care based on clinicaljudgement of the nurses. Preventivemeasures used were watermattresses, alternating mattresses,sheepskins and gel cushiPreventive care did not includeturning.

Fineman 200667

Prone positioning: a 2-hr cyclicrotation from full prone to rightlateral/prone to full prone to leftlateral/prone and then to full prone.

Supine positioning.

All patients were maintained onstandard hospital beds. Individuallysized head, chest, pelvic, distalfemoral and lower limb cushionswere created using pressurerelieving material.


Summary of studies included in the review

Population Outcomes

2, 3 hourly turning scheme on astandard institutional mattress and 4,

hourly turning scheme on a

The turning schemes consisted inrecumbent

position with a lateral position.

Standard care involving preventivenursing care based on clinicaljudgement of the nurses. Preventivemeasures used were watermattresses, alternating mattresses,sheepskins and gel cushions.Preventive care did not include

Geriatric nursing home patients.Mean age: 84.4 (SD 8.33) years, Themean Braden score was 13.2 (SD2.36) and the mean Norton scorewas 10.0 (SD 1.96). Patients wereconsidered to be at risk to developpressure ulcers.

Development of Nonerythema: redness which cannot bepressed away with the thumb andwhich lasts longer than I day(GRADE I in the Agency of HealthCare Policy and Research (AHCPR).

Development of pressure ulcerlesion: blistering, spressure ulcer (grades II, III and IV inthe AHCPR classification).

hr cyclicrotation from full prone to right

to full prone to leftlateral/prone and then to full prone.

All patients were maintained onstandard hospital beds. Individuallysized head, chest, pelvic, distalfemoral and lower limb cushionswere created using pressure-

One hundred and two paediatricpatients with acute lung injury.

Proportion of people that developedstage II or greater pressure ulcers.

249

Study length

Development of Non-blanchableerythema: redness which cannot bepressed away with the thumb andwhich lasts longer than I day(GRADE I in the Agency of HealthCare Policy and Research (AHCPR).

Development of pressure ulcerlesion: blistering, superficial or deeppressure ulcer (grades II, III and IV inthe AHCPR classification).

4 weeks.

Proportion of people that developedstage II or greater pressure ulcers.

28-days

250


Gentilello 198870

Kinetic treatment table (rotatesthrough an arc of 124 degrees every7 minutes) vs conventional beds(patients turned in conventionalfashion every 2 hours)

Moore 201165

Repositioning by using the 30(left side, back, right side, back)every three hours during the night.

Repositioning every six hours atnight, using 90° lateral rotation.

Both groups were nursed during theday according to planned care.Pressure redistribution devices incurrent use on the bed and on thechair was continued. Patiepositions were altered every 2hours.

Smith 199068

Small shift in body (adjusting theposition of a limb or body part byplacing a small rolled towel todesignated areas). Shifts werecompleted in less than one minute.Sites for placement of rolled towelwere under each arm, shoulder, hip,and leg.

Both groups received normal, routinecare and were turned every twohours.

Summer 198971

Kinetic treatment table vs routine 2hourly turning ICU conventionalbeds


Population Outcomes

Kinetic treatment table (rotatesthrough an arc of 124 degrees every7 minutes) vs conventional beds(patients turned in conventional

Critically ill patients in surgical ICUimmobilised because of head injury,spinal injuries or traction

Incidence of pulmonarycomplications; incidence of pressureulcers

Repositioning by using the 30° tilt(left side, back, right side, back)

hours during the night.

Repositioning every six hours atlateral rotation.

Both groups were nursed during theday according to planned care.Pressure redistribution devices incurrent use on the bed and on thechair was continued. Patients’positions were altered every 2-3

Participants from 12 long-term careof the older person hospital settings.Seventy-nine percent were women.Eighty-seven per cent were chair-fastand 77% had very limited activity.Participants were at risk ofdeveloping pressure ulcers (using theBraden pressure ulcer riskassessment scale).

Proportion of people developingpressure ulcers (Grades I

Time to pressure ulcer development.

Small shift in body (adjusting theposition of a limb or body part byplacing a small rolled towel todesignated areas). Shifts werecompleted in less than one minute.

t of rolled towelwere under each arm, shoulder, hip,

Both groups received normal, routinecare and were turned every two

Elderly patients. Participants rangedin age from 65 years to 91 years witha mean age of 80.55. Fourteenparticipants were women and fivewere men.

Proportion of people developingpressure ulcers (Grades II andhigher)

Kinetic treatment table vs routine 2-hourly turning ICU conventional

Patients admitted to ICU Incidence of pressure ulcers

Version 11/14/2012 3:08 PM

Study length

Incidence of pulmonarycomplications; incidence of pressure

Duration offollow-upunclear

Proportion of people developingpressure ulcers (Grades I – IV).

Time to pressure ulcer development.

4 weeks

Proportion of people developingpressure ulcers (Grades II and

2 weeks

Incidence of pressure ulcers Duration offollow-upunclear

KCE Report 193S


Vanderwee 200766

4 hours in a semi-recumbent 30°position and 2 hours in a lateralposition 30°.

Repositioning was the same asabove but with equal time intervals of4 hours in lateral 30° as in semirecumbent 30° position.

Patients in both groups were lying ona visco-elastic foam overlaymattress.

VanNieuwenhoven2006

69

Semi recumbent position. Aim was toachieve 45° position of the head andback. The 45° position was notachieved for 85% of the study time,and these patients more frequentlychanged position than supinepositioned patients.

Standard care (supine position)

Young 200472

30° tilt position during the night.

90° side-lying position during thenight.


Population Outcomes

recumbent 30°position and 2 hours in a lateral

ame asabove but with equal time intervals of4 hours in lateral 30° as in semi-

Patients in both groups were lying onelastic foam overlay

Geriatric nursing home patients.Mean age: 84.4 (SD 8.33) years, themean Braden score was 13.2 (SD2.36) and the mean Norton scorewas 10.0 (SD 1.96).

Proportion of people developingpressure ulcers (Grades II andhigher).

Time to developing pressure ulcers.

Semi recumbent position. Aim was toachieve 45° position of the head andback. The 45° position was notachieved for 85% of the study time,and these patients more frequently

sition than supine

Standard care (supine position)

221 adult patients admitted to fourICUs in three university hospitals inthe Netherlands. 112 randomised tosemi recumbent positioning and 109to supine positioning. Mean age of63.9 years

Proportion of patients developingulcer (Grades I-IV)

30° tilt position during the night.

lying position during the

Acute inpatient in a district generalhospital. Mean age of 70.3 years.Patients were at risk of developingpressure ulcers (confirmed by aWaterlow risk assessment scoreabove ten).

Proportion of people developingpressure ulcers (Grade 1: nonblanching erythema).

Patient tolerability.

251

Study length

Proportion of people developingpressure ulcers (Grades II and

Time to developing pressure ulcers.

5 weeks

Proportion of patients developingIV)

7 days

Proportion of people developingpressure ulcers (Grade 1: non-blanching erythema).

Patient tolerability.

One night

252

6.3.4. GRADE-tables

Frequencies of repositioning

Table 8 – Clinical evidence profile: Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standardcare without turning).

Quality assessment

No ofstudies

Design Riskof

bias


Proportion of people developing pressure ulcer (Grade I: Non-

1

Defloor(2005)

randomised trials

Veryserious1

no seriousinconsistency

no seriousindirectness


1

Defloor(2005)

randomised trials

Veryserious1




1

Defloor(2005)

randomised trials

Veryserious1




1

Deflo

randomised trials

Veryseriou




vidence profile: Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard

No of patients

Indirectness Imprecision Otherconsiderations

Repositioning Norepositioning

Relative(95% CI)

-blanching Erythema) – 2-h turning scheme on a standard institutional mattress ( follow


serious2 none 30/63(47.6%)

220/511(43.1%)

RR 1.11(0.84 to

1.46)(from 69 fewer to 198

43.1%(from 69 fewer to 198

-blanching Erythema) – 3-h turning scheme on a standard institutional mattress (follow



220/511(43.1%)

RR 1.04(0.77 to



-blanching Erythema) – 4-h turning + pressure reducing mattress (follow-up 4 weeks)


Veryserious3

none 28/66(42.4%)

220/511(43.1%)

RR 0.99(0.73 to

1.33)

4 fewer per 1000 (from

43.1% 4 fewer per 1000 (fro

-blanching Erythema) – 6-h turning + pressure reducing mattress (follow-up 4 weeks)

ousindirectness

serious2 none 29/63(46%)

220/511(43.1%)

RR 1.07(0.8 to (from 86 fewer to 181

Version 11/14/2012 3:08 PM

vidence profile: Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard


Absolute

follow-up 4 weeks)

47 more per 1000(from 69 fewer to 198

more)

VERYLOW

CRITICAL


more)

h turning scheme on a standard institutional mattress (follow-up 4 weeks)


more)

VERYLOW

CRITICAL


more)

4 fewer per 1000 (from116 fewer to 142

more)

VERYLOW

CRITICAL

4 fewer per 1000 (from116 fewer to 142

more)


VERY

CRITICAL

KCE Report 193S

Quality assessment

No ofstudies

Design Riskof

bias


or(2005)

s1

Proportion of people developing pressure ulcer (Grade II and higher)

1

Defloor(2005)

randomised trials

Veryserious1




1

Defloor(2005)

randomised trials

Veryserious1




1

Defloor(2005)

randomised trials

Veryserious1




1

Defloor(2005)

randomised trials

Veryserious1




No of patients


Repositioning Norepositioning

Relative(95% CI)

1.42)


Proportion of people developing pressure ulcer (Grade II and higher) – 2-h turning scheme on a standard institutional mattress (follow-up 4 weeks)


no seriousimprecision

none 9/63(14.3%)

220/511(43.1%)

RR 0.33(0.18 to



Proportion of people developing pressure ulcer (Grade II and higher) – 3-h turning scheme on a standard institutional mattress (follow-up 4 weeks)



220/511(43.1%)

RR 0.56(0.35 to



Proportion of people developing pressure ulcer (Grade II and higher) – 4-h turning + pressure reducing mattress (follow-up 4 weeks)



none 2/66(3%)

220/511(43.1%)

RR 0.07(0.02 to


43.1%(from 31

Proportion of people developing pressure ulcer (Grade II and higher) – 6-h turning + pressure reducing mattress (follow-up 4 weeks)

no seriouss


none 10/63(15.9%)

220/511(43.1%)

RR 0.37(0.21 to



253


Absolute

more) LOW


more)

288 fewer per 1000(from 168 fewer to 353

fewer)

LOW

CRITICAL


fewer)


fewer)

VERYLOW

CRITICAL


fewer)


fewer)

LOW

CRITICAL


fewer)


fewer)

LOW

CRITICAL


fewer)

254

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for the experimental group.2 Confidence interval crossed one MID point (0.75 to 1.25 for dichotomous outcomes)3 Confidence interval crossed both ends of MID point (0.75 to 1.25 for dichotomous outcomes)

Table 9 – Clinical evidence profile: Different frequencies of repositioning: 2standard institutional mattress.

Quality assessment

No ofstudies

Design Risk ofbias


Proportion of people developing pressure ulcer (Grade I: Non-up 4 weeks)

1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)


1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used was

2 Confidence interval crossed one MID point (0.75 to 1.25 for dichotomous outcomes)3 Confidence interval crossed both ends of MID point (0.75 to 1.25 for dichotomous outcomes)

ifferent frequencies of repositioning: 2-h turning on a standard institutional mattress versus 3

No of patients

Indirectness Imprecision

Otherconsiderations

2-h turning 3-h turning Relative(95% CI)

-blanching Erythema) – 2-h turning on a standard institutional mattress versus 3- h turning


veryserious2

none 30/63(47.6%)

26/58(44.8%)

RR 1.06(0.72 to1.56)

44.8%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealme

2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)

Version 11/14/2012 3:08 PM

ar allocation concealment, mattress used was not

h turning on a standard institutional mattress versus 3-h turning on a


Absolute

h turning on a standard institutional mattress (follow-

27 more per 1000(from 126 fewer to

251 more)

VERYLOW

CRITICAL


251 more)


KCE Report 193S

Table 10 – Different frequencies of repositioning: 2

Quality assessment

No ofstudies

Design Risk ofbias


Proportion of people developing pressure ulcer (Grade I: Nonweeks)

1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.; 2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)


Different frequencies of repositioning: 2-h turning on a standard institutional mattress versus 4-h turning+ pressure reducing mattress

No of patients


Otherconsiderations

2-h turning 4-h turning+pressurereducingmattress

Relative(95% CI)

(Grade I: Non-blanching Erythema) – 2-h turning on a standard institutional mattress versus 4- h turning+ pressure reducing mattress (follow


Serious2 none 30/63(47.6%)

28/66(42.4%)

RR 1.12(0.77 to1.64)

42.4%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used wa2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)

255

h turning+ pressure reducing mattress


Absolute

h turning+ pressure reducing mattress (follow-up 4


272 more)

VERYLOW

CRITICAL


271 more)


256


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryerious1





Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.; 2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)


Different frequencies of repositioning: 2-h turning on a standard institutional mattress versus 6-h turning+ pressure r

No of patients


Otherconsiderations


Relative(95% CI)



veryserious2

none 30/63(47.6%)

29/63(46%)

RR 1.03(0.71 to

1.5)

46%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used was


of repositioning: 3-h turning on a standard institutional mattress versus 4-h turning+ pressure red

No of patients


Otherconsiderations


Relative(95% CI)



veryserious2

none 26/58(44.8%)

28/66(42.4%)

RR 1.06(0.71 to

1.58)

42.4%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used w2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)

Version 11/14/2012 3:08 PM



Relative(95% CI)

Absolute


RR 1.03(0.71 to


230 more)

VERYLOW

CRITICAL


230 more)




Relative(95% CI)

Absolute


RR 1.06(0.71 to


246 more)

VERYLOW

CRITICAL


246 more)


KCE Report 193S


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1





Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Serious1


no seriouindirectness

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)


Different frequencies of repositioning: 3-h turning on a standard institutional mattress versus 6-h turning+ pressure

No of patients


Otherconsiderations


Relative(95% CI)



veryserious2

none 26/58(44.8%)

29/63(46%)

RR 0.97(0.66 to1.44)

46%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used wa


Different frequencies of repositioning: 4-h turning+ pressure reducing mattress versus 6-h turning+ pressure reducing ma

No of patients


Otherconsiderations

4-h turning+pressurereducingmattress


Relative(95% CI)

-blanching Erythema) – 4-h versus 6-h turning+ pressure reducing mattress (follow- up 4 weeks)


veryserious2

none 28/66(42.4%)

29/63(46%)

RR 0.92(0.63 to1.36)

46%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment.erval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes) ;

257



Absolute


14 fewer per 1000(from 157 fewer to

203 more)

VERYLOW

CRITICAL


202 more)




Relative(95% CI)

Absolute

up 4 weeks)

RR 0.92(0.63 to


166 more)

VERYLOW

CRITICAL


166 more)

ar allocation concealment.

258

Table 15 – Different frequencies of repositioning: turning 2

Quality assessment

No ofstudies

Design Risk ofbias



1

Vanderwee(2007)

randomised trials

veryserious1



Time to develop a pressure ulcer

1

Vanderwee(2007)

randomised trials

veryserious1



1 Blinding, intention to treat analysis and allocation concealment not reported. Sample size lower than the desired (calculatends of MID point (0.75 to 1.25 for dichotomous outcomes)


Different frequencies of repositioning: turning 2-h in a lateral and 4-h in a supine position versus repositioning 4

No of patients


Otherconsiderations

2-h in alateral and 4-h in a supineposition

4-hrlyturning

Relative(95% CI)

(Grade II and higher) – Turning with unequal time intervals (follow-up 5 weeks)


veryserious2

none 20/122(16.4%)

24/113(21.2%)

RR 0.77(0.45 to1.32)

21.2%


N/A Very serious3 - - -

1 Blinding, intention to treat analysis and allocation concealment not reported. Sample size lower than the desired (calculat ed) needed..25 for dichotomous outcomes); 3 No data could be analysed in Revman.

Version 11/14/2012 3:08 PM

h in a supine position versus repositioning 4-hrly


Absolute


68 more)

VERYLOW

CRITICAL


68 more)

Log rank test 1.18(d.f 0.1), p=0.28

VERYLOW

IMPORTANT

ed) needed. ; 2 Confidence interval crosses both

KCE Report 193S

Table 16 – Different frequencies of repositioning: unscheduled small shifts in body position versus 2

Quality assessment

No ofstudies

Design Risk ofbias



1

Smith(1990)

randomised trials

veryserious1



1 Blinding, intention to treat analysis and allocation concealment not reported. Sample size lower than the desired (calculatbetween control and experimental greater than 10%)2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes)

Table 17 – Different frequencies of repositioning: 2mattress

Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1





Different frequencies of repositioning: unscheduled small shifts in body position versus 2 -hr turning

No of patients


Otherconsiderations

unscheduledsmall shifts

2-hrlyturning

Relative(95% CI)

ressure ulcer (Grade II and higher) – Unscheduled (small) shifts in body positions (follow-up 2 weeks)


veryserious2

none 1/9(11.1%)

1/10(10%)

RR 1.11(0.08 to15.28)

10%

1 Blinding, intention to treat analysis and allocation concealment not reported. Sample size lower than the desired (calculat ed) needed, high rate of drop outs (difference


Different frequencies of repositioning: 2-h turning on a standard institutional mattress versus 3-h t

No of patients


Otherconsiderations

2-h turning 3-h turning Relative(95% CI)

ressure ulcer (Grade II and higher) – 2-h turning on a standard institutional mattress versus 3-h turning on a standard institutional mattress


veryserious2

none 9/63(14.3%)

14/58(24.1%)

RR 0.59(0.28 to1.26)

24.1%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment. Mattress used w


259

hr turning


Absolute


1000 more)

VERYLOW

CRITICAL


1000 more)

ed) needed, high rate of drop outs (difference

h turning on a standard institutional


Absolute

on a standard institutional mattress (follow-up 4 weeks)


63 more)

VERYLOW

CRITICAL


63 more)

ar allocation concealment. Mattress used was not

260


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.; 2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.; 2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes


Different frequencies of repositioning: 2-h turning on a standard institutional mattress versus 4-h turning+ pressure

No of patients


Otherconsiderations


Relative(95% CI)

ressure ulcer (Grade II and higher) – 2-h turning on a standard institutional mattress versus 4-h turning+ pressure reducing mattress (follow



2/66(3%)

RR 4.71(1.06 to20.98)

3%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used was n2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)

Different frequencies of repositioning: 2-h turning on a standard institutional mattress versus 6-h turning

No of patients


Otherconsiderations


Relative(95% CI)



veryserious2

none 9/63(14.3%)

10/63(15.9%)

RR 0.9(0.39 to2.06)

15.9%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used w2 Confidence interval crosses both ends of MID point (0.75 to 1.25 for dichotomous outcomes

Version 11/14/2012 3:08 PM



Absolute

h turning+ pressure reducing mattress (follow-up 4 weeks)

112 more per1000 (from 2 more

to 605 more)

VERYLOW

CRITICAL

111 more per1000 (from 2 more

to 599 more)


h turning + pressure reducing mattress


Absolute



168 more)

VERYLOW

CRITICAL


169 more)


KCE Report 193S


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryserious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, unclethe same for both groups.2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

Veryerious1





Different frequencies of repositioning: 3-h turning on a standard institutional mattress versus 4-h turning+ pressure r

No of patients


Otherconsiderations


Relative(95% CI)




2/66(3%)

RR 7.97(1.89 to33.59)

3%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used was n

2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)

Different frequencies of repositioning: 3-h turning on a standard institutional mattress versus 6-h turning+ pressure redu

No of patients


Otherconsiderations


Relative(95% CI)



veryserious2

none 14/58(24.1%)

10/63(15.9%)

RR 1.52(0.73 to3.15)

15.9%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment, mattress used wa


261



Absolute



more to 988 more)

VERYLOW

CRITICAL


more to 978 more)




Absolute



341 more)

VERYLOW

CRITICAL


342 more)


262


Quality assessment

No ofstudies

Design Risk ofbias



1

Defloor(2005)

randomised trials

serious1



1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle2 Confidence interval crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)


Different frequencies of repositioning: 4-h turning+ pressure reducing mattress versus 6-h turning+ pressure reducing

No of patients


Otherconsiderations



Relative(95% CI)

ressure ulcer (Grade II and higher) – 4-h versus 6-h turning+ pressure reducing mattress (follow-up 4 weeks)

no serioustness

Serious2 none 2/66(3%)

10/63(15.9%)

RR 0.19(0.04 to0.84)

15.9%

1 Incomplete data for 3 patients though authors claim that analysis including these patients did not change the result, uncle ar allocation concealment.val crosses one end of MID point (0.75 to 1.25 for dichotomous outcomes)

Version 11/14/2012 3:08 PM



Absolute

129 fewer per1000 (from 25fewer to 152

fewer)

LOW

CRITICAL


fewer)

ar allocation concealment.

KCE Report 193S

Table 23 – Clinical evidence profile: Different positions for repositioning

Quality assessment

No ofstudie

s

Design Risk ofbias


Proportion of people developing pressure ulcer (Grades I – IV)

1

Moore(2011)

randomised trials

veryserious1



Proportion of people developing pressure ulcer (Grade I: non-

1

Young(2004)

randomised trials

serious3



Mean time to pressure ulcer development

1Moore(2011)

randomised trials

veryserious1



Tolerability

1

Young(2004)

randomised trials

serious3



1 Blinding not reported, sample size was lower than the desired (calculated) power needed.2 Confidence interval crossed one MID point (0.75 to 1.25 for dichotomous outcomes)3 Small sample size4 Confidence interval crosses both ends of MID (0.75 to 1.25 for dichotomous outcomes)5 Details could not be analysed in Revman.6 Details only given for one arm of the trial.


Clinical evidence profile: Different positions for repositioning – 30° tilt position versus 90° lateral and supine position (control)

No of patients


Otherconsideratio

ns

30° tilt position 90° lateralandsupineposition

Relative(95% CI)

IV) – 30 degree tilt 3 hourly- (cluster) (follow-up 4 weeks)



13/114(11.4%)

RR 0.27(0.08 to0.91)

11.4%

-blanching erythema) – 30 degree tilt – (follow-up 1 night)


veryserious4

none 3/23(13%)

2/23(8.7%)

RR 1.5(0.28 to8.16)

8.7%


N/A Very serious5 26 days (range 3days)

17 days(range 24

days)

-


N/A very serious6 5/23 (22%) - -

1 Blinding not reported, sample size was lower than the desired (calculated) power needed.MID point (0.75 to 1.25 for dichotomous outcomes)

4 Confidence interval crosses both ends of MID (0.75 to 1.25 for dichotomous outcomes)

263

30° tilt position versus 90° lateral and supine position (control)


Absolute


105 fewer)

VERYLOW

CRITICAL


105 fewer)


623 more)

VERYLOW

CRITICAL


623 more)

- VERYLOW

IMPORTANT

- VERYLOW

IMPORTANT

264

Table 24 – Clinical evidence profile: Different positions for repositioningstandard care (supine position)

Quality assessment

No ofstudie

s

Design Risk ofbias


Proportion of people developing pressure ulcers (Grade I-IV) –

1

VanNieuwenhoven(2006)

randomised trials

noseriousrisk ofbias



1 Confidence interval crossed both ends of MID points (0.75 to 1.25 for dichotomous outcomes).

Comparison between kinetic beds and conventional beds


Clinical evidence profile: Different positions for repositioning – semi recumbent position (45° position of the head and back) versus

No of patients


Otherconsiderations

Semi recumbentposition (45 degreeposition of the head

and back)

Supine position Relative(95% CI)

– semi recumbent position (45° position of the head and back) (follow-up 7 days)


veryserious1

none 31/112(27.7%)

30/109(27.5%)

RR 1.01(0.66 to1.54)

27.5%

1 Confidence interval crossed both ends of MID points (0.75 to 1.25 for dichotomous outcomes).

tic beds and conventional beds

Version 11/14/2012 3:08 PM

semi recumbent position (45° position of the head and back) versus


Relative(95% CI)

Absolute

RR 1.01.66 to

1.54)


LOW

CRITICAL


KCE Report 193S

Table 25 – Kinetic treatment table vs standard care for pressure ulcer prevention

Quality assessment

No of studies Design Risk ofbias

Inconsistency

Pressure ulcer incidence

2 Gentilello(1988)Summer (1989)

randomisedtrials

veryserious1


Time in hospital (days)

1 Summer (1989) randomisedtrials

veryserious1


1 Unclear allocation concealment and blinding (Gentilello 1988, Summer 1989) and unclear addressing of incomplete outcome datbaseline (Summer 1989).; 2 Confidence interval crossed both MIDs.4 Patients in Summer (1989) randomised only obtunded or unconscious patients (although this was not the initial intention) anfrom head injury, spinal injuries or traction. Most patients would not be able to reposition themselves so the two studies were meta

Table 26 – Critically ill infants and children: different positions for repositioning

Quality assessment

No ofstudies

Design Riskof

bias


Proportion of people developing (Grade II and higher) – Prone positioning (2 hour cyclic rotation) (follow

1

Fineman(2006)

randomised trials

serious1



1 Blinding of any kind not reported2 Confidence interval crosses both ends of MID (0.75 to 1.25 for dichotomous outcomes)


Kinetic treatment table vs standard care for pressure ulcer prevention


Inconsistency Indirectness Imprecision Otherconsiderations

Kinetictreatment

table

Standardcare

Relative(95% CI)



veryserious2

none 9/70(12.9%)

10/81(12.3%)

RR 1.23(0.57 to



none very serious3 6.7 days 11.6 days

1 Unclear allocation concealment and blinding (Gentilello 1988, Summer 1989) and unclear addressing of incomplete outcome dat a. (Gentilello 1988). Unclear if similar atrval crossed both MIDs.; 3 Not enough data for analysis in Revman.

4 Patients in Summer (1989) randomised only obtunded or unconscious patients (although this was not the initial intention) an d Gentillello (1988) included patients immobilisedury, spinal injuries or traction. Most patients would not be able to reposition themselves so the two studies were meta -analysed together.

Critically ill infants and children: different positions for repositioning – prone positioning versus control supine positioning (control)

No of patients


Otherconsiderations

Prone positioning Supinepositioning

Relative(95% CI)

Prone positioning (2 hour cyclic rotation) (follow-up 28 days)


veryserious2

none 10/51(19.6%)

8/51(15.7%)

RR 1.25(0.54 to

2.91)15.7%

.75 to 1.25 for dichotomous outcomes)

265


Relative(95% CI)

Absolute

RR 1.23(0.57 to2.65)


VERYLOW

Critical

- - VERYLOW

Important

a. (Gentilello 1988). Unclear if similar at

d Gentillello (1988) included patients immobilisedanalysed together.

ontrol supine positioning (control)


Absolute


VERYLOW

CRITICAL


266

6.3.5. Forest plots

Figure 2 – Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turnblanching erythema (Grade I pressure ulcer)

Study or Subgroup

1.1.1 2-h turning scheme

Defloor 2005Subtotal (95% CI)

Total events





Total events



1.1.3 4-h turning+mattress


Total events





Total events



Test for subgroup differences: Chi² = 0.33, df = 3 (P = 0.95), I² = 0%

Events

30

30

26

26

28

28

29

29

Total

6363

5858

6666

6363

Events

220

220

220

220

220

220

220

220

Total

511511

511511

511511

511511

Repositioning No repositioning


Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turn


Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

1.11 [0.84, 1.46]1.11 [0.84, 1.46]

1.04 [0.77, 1.41]1.04 [0.77, 1.41]

0.99 [0.73, 1.33]0.99 [0.73, 1.33]

1.07 [0.80, 1.42]1.07 [0.80, 1.42]


M-H, Fixed, 95% CI

0.05 0.2 1 5 20Repositioning No repositioning

Version 11/14/2012 3:08 PM

Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turn ing): Non-

KCE Report 193S

Figure 3 – Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turnPressure ulcers (Grades II – IV)

Study or Subgroup



Total events





Total events





Total events





Total events



Test for subgroup differences: Chi² = 8.63, df = 3 (P = 0.03), I² = 65.2%

Events

9

9

14

14

2

2

10

10

Total

6363

5858

6666

6363

Events

220

220

220

220

220

220

220

220

Total

511511

511511

511511

511511

Repositioning No repositioning


Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turn


Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

0.33 [0.18, 0.61]0.33 [0.18, 0.61]

0.56 [0.35, 0.89]0.56 [0.35, 0.89]

0.07 [0.02, 0.28]0.07 [0.02, 0.28]

0.37 [0.21, 0.66]0.37 [0.21, 0.66]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Repositioning No repositioning

267

Repositioning (Frequent turning or the use of pressure reducing mattress) versus no repositioning (standard care without turn ing):

268

Figure 4 – Different frequencies of repositioninginstitutional mattress: Non-blanching erythema (Grade I pressure ulcer)

Figure 5 – Different frequencies of repositioningreducing mattress: non-blanching erythema (Grade I pressure ulcer)

Study or Subgroup

3.8.1 2-h versus 3-h turning


Total events



Test for subgroup differences: Not applicable

Events

30

30

Total

6363

Events

26

26

Total

5858

Weight

100.0%100.0%

2h turning 3h turning

Study or Subgroup

3.9.2 2-h versus 4-h turning+ pressure reducing mattress


Total events




Events

30

30

Total

6363

Events

28

28

Total

6666

Weight

100.0%100.0%

2h turning 4h turning+mattress


Different frequencies of repositioning – 2-hour turning on a standard institutional mattress versus 3blanching erythema (Grade I pressure ulcer)


Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.06 [0.72, 1.56]1.06 [0.72, 1.56]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours 2h turning Favours 3h turning

Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.12 [0.77, 1.64]1.12 [0.77, 1.64]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours 2h turning Favours 4h turning+mattre

Version 11/14/2012 3:08 PM

titutional mattress versus 3-hour turning on a standard

hour turning on a standard institutional mattress versus 4-hour turning scheme + pressure

KCE Report 193S

Figure 6 – Different frequencies of repositioningreducing mattress: non-blanching erythema (Grade I pressure ulcer)

Figure 7 – Different frequencies of repositioningreducing mattress: Non-blanching erythema (Grade I pre

Figure 8 – Different frequencies of repositioningreducing mattress: Non-blanching erythema (Grade I pressure ulcer)

Study or Subgroup



Total events




Events

30

30

Total

6363

Events

29

29

Total

6363

Weight

100.0%100.0%


Study or Subgroup



Total events




Events

26

26

Total

5858

Events

28

28

Total

6666

Weight

100.0%100.0%


Study or Subgroup



Total events




Events

26

26

Total

5858

Events

29

29

Total

6363

Weight

100.0%100.0%






Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.03 [0.71, 1.50]1.03 [0.71, 1.50]


M-H, Fixed, 95% CI


Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.06 [0.71, 1.58]1.06 [0.71, 1.58]


M-H, Fixed, 95% CI


Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.97 [0.66, 1.44]0.97 [0.66, 1.44]


M-H, Fixed, 95% CI


269




270

Figure 9 – Different frequencies of repositioningreducing mattress: Non-blanching erythema (Grade I pressure ulcer).

Figure 10 – Different frequencies of repositioningpressure ulcers (Grade II and higher).

Figure 11 – Different frequencies of repositioning(Grade II and higher).

Study or Subgroup



Total events




Events

28

28

Total

6666

Events

29

29

Total

6363

Weight

100.0%100.0%

4h turning+mattress 6h turning+mattress

Study or Subgroup

3.13.1 Turning with unequal time intervals

Vanderwee 2007Subtotal (95% CI)

Total events




Events

20

20

Total

122122

Events

24

24

Total

113113

2-h in a lateral and 4-h in a supine 4-hrly turning

Study or Subgroup

3.14.2 Unscheduled (small) shifts in body positions

Smith 1990Subtotal (95% CI)

Total events




Events

1

1

Total

99

Events

1

1

Total

1010

Small shifts in positions 2hrly turn


frequencies of repositioning – 4-hour turning scheme + pressure reducing mattress versus 6blanching erythema (Grade I pressure ulcer).

Different frequencies of repositioning – turning 2-h in a lateral and 4-h in a supine position versus repositioning 4

Different frequencies of repositioning – unscheduled small shifts in body position versus 2-hrly turning: incidenc

Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.92 [0.63, 1.36]0.92 [0.63, 1.36]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours 4h turning+mattre Favours 6h turning+mattre

Total

113113

Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.77 [0.45, 1.32]0.77 [0.45, 1.32]

4-hrly turning Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours 2-h + 4hrly turn Favours 4hrly turning

Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.11 [0.08, 15.28]1.11 [0.08, 15.28]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours Small shifts Favours 2hrly turning

Version 11/14/2012 3:08 PM

hour turning scheme + pressure reducing mattress versus 6-hour turning scheme + pressure

h in a supine position versus repositioning 4-hrly: incidence of

hrly turning: incidence of pressure ulcers

KCE Report 193S

Figure 12 – Different frequencies of repositioningpressure ulcers (Grade II and higher).

Figure 13 – Different frequencies of repositioningreducing mattress: incidence of pressure ulcers (Grade II and higher).

Study or Subgroup

3.15.3 2-h versus 3-h turning


Total events




Events

9

9

Total

6363

Events

14

14

Total

5858

Weight

100.0%100.0%

2-h turning 3-h turning

Study or Subgroup



Total events




Events

9

9

Total

6363

Events

2

2

Total

6666

Weight

100.0%100.0%

2-h turning 4-h turning+mattress


Different frequencies of repositioning – 2-hour turning on a standard institutional mattress versus 3

erent frequencies of repositioning – 2-hour turning on a standard institutional mattress versus 4reducing mattress: incidence of pressure ulcers (Grade II and higher).

Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.59 [0.28, 1.26]0.59 [0.28, 1.26]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours 2-h turning Favours 3-h turning

Weight

100.0%100.0%

M-H, Fixed, 95% CI

4.71 [1.06, 20.98]4.71 [1.06, 20.98]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours 2-h turning Favours 4-h turning+mattr

271

g on a standard institutional mattress versus 3-hour turning scheme: incidence of


272




Study or Subgroup



Total events




Events

9

9

Total

6363

Events

10

10

Total

6363

Weight

100.0%100.0%

2-h turning 6-h turning + mattress

Study or Subgroup



Total events




Events

14

14

Total

5858

Events

2

2

Total

6666

Weight

100.0%100.0%


Study or Subgroup



Total events




Events

14

14

Total

5858

Events

10

10

Total

6363

Weight

100.0%100.0%



Different frequencies of repositioning – 2-hour turning on a standard institutional mattress versus 6reducing mattress: incidence of pressure ulcers (Grade II and higher).



Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.90 [0.39, 2.06]0.90 [0.39, 2.06]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours 2-h turning Favours 6-h turning+mattr

Weight

100.0%100.0%

M-H, Fixed, 95% CI

7.97 [1.89, 33.59]7.97 [1.89, 33.59]


M-H, Fixed, 95% CI


Weight

100.0%100.0%

M-H, Fixed, 95% CI

1.52 [0.73, 3.15]1.52 [0.73, 3.15]


M-H, Fixed, 95% CI


Version 11/14/2012 3:08 PM


tutional mattress versus 4-hour turning scheme + pressure


KCE Report 193S


Figure 18 – Different positions for repositioning

Study or Subgroup



Total events




Events

2

2

Total

6666

Events

10

10

Total

6363

Weight

100.0%100.0%

4h turning+mattress 6h turning+mattress

Study or Subgroup

2.3.1 30 degree tilt - all stages (cluster)

Moore 2011Subtotal (95% CI)

Total events



2.3.2 30 degree tilt - erythema (non-cluster)

Young 2004Subtotal (95% CI)

Total events



Events

3

3

3

3

Total

9999

2323

Events

13

13

2

2

Total

114114

2323

30 degree tilt position 90 degree positions


Different frequencies of repositioning – 4-hour turning scheme + pressure reducing mattress versus 6: incidence of pressure ulcers (Grade II and higher).

Different positions for repositioning – 30° tilt position versus 90° lateral and supine position: incidence of pressure ulcer (Grade I

Weight

100.0%100.0%

M-H, Fixed, 95% CI

0.19 [0.04, 0.84]0.19 [0.04, 0.84]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours 4h turning+mattre Favours 6h turning+mattre

Total

114114

2323

Weight

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

0.27 [0.08, 0.91]0.27 [0.08, 0.91]

1.50 [0.28, 8.16]1.50 [0.28, 8.16]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours 30 degree tilt Favours 90 degree

273

hour turning scheme + pressure reducing mattress versus 6-hour turning scheme + pressure

30° tilt position versus 90° lateral and supine position: incidence of pressure ulcer (Grade I – IV).

274

Figure 19 – Different positions for repositioningposition): incidence of pressure ulcer (Grade I-

Figure 20 – Critically ill infants and children: different positions for repositioningulcer (Grade II and higher)

Figure 21 – Kinetic treatment table vs standard care

Study or Subgroup

van Nieuwenhoven 2006

Total (95% CI)

Total events



Events

31

31

Total

112

112

Events

30

30

Total

109

109

Weight

100.0%

100.0%

Semi recumbent position Supine position

Study or Subgroup

Fineman 2006

Total (95% CI)

Total events



Events

10

10

Total

51

51

Events

8

8

Total

51

51

Weight

100.0%

100.0%

Prone positioning Supine positioning

Study or Subgroup

Gentilello 1988

Summer 1989

Total (95% CI)

Total events

Heterogeneity: Chi² = 0.35, df = 1 (P = 0.55); I² = 0%


Events

8

1

9

Total

27

43

70

Events

10

0

10

Total

38

43

81

Weight

94.3%

5.7%

100.0%

KTT Standard


positioning – semi recumbent position (45° position of the head and back) versus standard care (supine-IV).

Critically ill infants and children: different positions for repositioning – prone positioning versus control supine positioning. Pressure

Kinetic treatment table vs standard care

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.01 [0.66, 1.54]

1.01 [0.66, 1.54]


M-H, Fixed, 95% CI

0.05 0.2 1 5 20Favours semi recumbent pt Favours supine position

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.25 [0.54, 2.91]

1.25 [0.54, 2.91]


M-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours prone positioning Favours supine positionin

Weight

94.3%

5.7%

100.0%

M-H, Fixed, 95% CI

1.13 [0.51, 2.48]

3.00 [0.13, 71.65]

1.23 [0.57, 2.65]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours KTT Favours Std

Version 11/14/2012 3:08 PM

semi recumbent position (45° position of the head and back) versus standard care (supine

ositioning versus control supine positioning. Pressure

KCE Report 193S


Table 27 – FINEMAN 2006


Author and year:Fineman 2006

Title: Pronepositioning can besafely performed incritically ill infantsand children

Journal: PaediatricCritical CareMedicine

Sequencegeneration:Randomisation doneusing a permutedblock sizes

Allocationconcealment: Eachcentre receivedserially numbered,opaque, sealedenvelopes containingstudy assignments

Blinding: notreported

Addressingincomplete outcomedata: not reported

Analysis: Analysiswere carried out on

Patient group: Onehundred and twopaediatric patients withacute lung injury.

All patients

Randomised N: 102

Completed N: 98

Drop-outs: 4

Group 1

Randomised N: 51

Completed N: 47

Dropouts: 4

Group 2

Randomised N: 51

Completed N: 51

Dropouts: none

Inclusion criteria:Paediatric patients (2wks to 18 yrs) who wereintubated andmechanically ventilatedwith a PaO2/FIO2 ratio


Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: Onehundred and twopaediatric patients with

Inclusion criteria:Paediatric patients (2wks to 18 yrs) who wereintubated andmechanically ventilatedwith a PaO2/FIO2 ratio

Group 1: Prone positioning: a2-hr cyclic rotation from fullprone to right lateral/prone tofull prone to left lateral/proneand then to full prone. Pronepositioning continued eachday during the acute phaseof their Acute Lung Injuryillness for a maximum of 7days of treatment.Infants/toddlers were liftedup, turned 45°, and turnedprone on their cushions.School-aged and adolescentpatients were turned usingthe mummy technique.During each turn, thepatient’s head was kept inalignment with the body,avoiding hyperextension.

Group 2: Supine positioning

All patients were maintainedon standard hospital beds.Individually sized head,chest, pelvic, distal femoraland lower limb cushionswere created using pressure-relieving material.

Outcome 1:Adverse event(proportion ofparticipants thatdeveloped stageII or greaterpressure ulcers)

Group 1: 10/51 (19.60%)

Group 2: 8/51

275

Comments

Group 1: 10/51 (19.60%)

Group 2: 8/51 (15.69%)

Funding: notreported.

Limitations:

Blinding foroutcomeassessors notreported

Additionaloutcomes:

276


an intention-to-treatbasis

Statistical analysis:Wilcoxon’s rank-sumtest or Fisher’s exacttest, as appropriate,to compare proneand supine groups intheir baselinecharacteristics andoutcomes that werecalculated on a perpatient basis.

Baseline differences:There were nosignificantdifferences betweenthe prone and supinegroups

Study power/samplesize: Study powernot reported.

Setting: Sevenpaediatric intensivecare units thatparticipate in thePaediatric AcuteLung Injury andSepsis Investigators(PALISI) Network inthe United States

Length of study: 28

of ≤300, bilaterallypulmonary infiltrates,and no clinical evidenceof left atrial hypertension

Exclusion criteria: <2wks of age (newbornphysiology), <42 wkspost conceptual age(considered preterm),were unable to toleratea position change(persistent hypotension,cerebral hypertension),had respiratory failurefrom cardiac disease,had hypoxemia withoutbilateral infiltrates, hadreceived a bone marrowor lung transplant, weresupported onextracorporealmembrane oxygenation,had a nonpulmonarycondition that could beexacerbated by theprone position, or hadparticipated in otherclinical trials within thepreceding 30 days.


Intervention

Comparison

Outcomemeasures

Effect sizes

300, bilaterallypulmonary infiltrates,and no clinical evidenceof left atrial hypertension

Exclusion criteria: <2wks of age (newbornphysiology), <42 wkspost conceptual age(considered preterm),were unable to toleratea position change

potension,cerebral hypertension),had respiratory failurefrom cardiac disease,had hypoxemia withoutbilateral infiltrates, hadreceived a bone marrowor lung transplant, weresupported on

membrane oxygenation,had a nonpulmonary

that could beexacerbated by theprone position, or hadparticipated in otherclinical trials within the

KCE Report 193S

Comments

KCE Report 193S


days

Assessment of PUs:Not reported

Multiple ulcers: Notreported

Table 28 – DEFLOOR 2005B


Author and year:Defloor 2005B

Title: The effect ofvariouscombinations ofturning and pressurereducing devices onthe incidence ofpressure ulcers

Journal: InternationalJournal of NursingStudies

Sequencegeneration: clusterrandomisation doneusing a permutedblock sizes. Clusterrandomisation usingcomputerisedrandomisation

Patient group: 838geriatric nursing homepatients. Mean age:84.4 (SD 8.33) years,The mean Braden scorewas 13.2 (SD 2.36) andthe mean Norton scorewas 10.0 (SD 1.96).

All patients

Randomised N: 838

Completed N: 761

Drop-outs: 77

Group 1

Randomised N: 65

Completed N: 63

Dropouts: 2 (1 died and1 transferred to hospital)


Intervention

Comparison

Outcomemeasures

Effect sizes

Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: 838nursing home

patients. Mean age:84.4 (SD 8.33) years,The mean Braden scorewas 13.2 (SD 2.36) andthe mean Norton score

s: 2 (1 died and1 transferred to hospital)

Group 1: 2-hour turningscheme on a standardinstitutional mattress

Group 2: 3-hour turningscheme on a standardinstitutional mattress

Group 3: 4-hour turningscheme + pressure reducingmattress

Group 4: 6-hour turningscheme + pressure reducingmattress.

The turning schemesconsisted in alternating asemi-Fowler position with alateral position.

Group 5: Standard careinvolving preventive nursingcare based on clinicaljudgement of the nurses.

Outcome 1:Development ofNon-blanchableerythema:redness whichcannot bepressed awaywith the thumband which lastslonger than I day(GRADE I in theAgency of HealthCare Policy andResearch(AHCPR)

Group 1: 30/63 (47.6%)

Group 2: 26/58 (44.8%)

Group 3: 28/66 (42.4%)

Group 4: 29/63 (46.0%)

Group 5: 220/511 (43.0%)

Outcome 2:Development ofpressure ulcerlesion: blistering,superficial ordeep pressure

Group 1: 9/63 (14.3%)

Group 2: 14/58 (24.1%)

Group 3: 2/66 (3%)

Group 4: 10/63 (15.9%)

Group 5: 102/511 (20%)

277

Comments

Comments

Group 1: 30/63 (47.6%)

Group 2: 26/58 (44.8%)

Group 3: 28/66 (42.4%)

Group 4: 29/63 (46.0%)

Group 5: 220/511 (43.0%)


Limitations:Intention-To-Treatanalysis notreported.

Additionaloutcomes:

Group 1: 9/63 (14.3%)

Group 2: 14/58 (24.1%)

Group 3: 2/66 (3%)

p 4: 10/63 (15.9%)

Group 5: 102/511 (20%)

278


tables.

Allocationconcealment: Sealedenvelope containingall room numbers ina random order.

Blinding: Outcomeassessors blinded

Addressingincomplete outcomedata: Gave details ofwhat happened todrop outs and data ofavailable patients

Analysis: notreported

Statistical analysis:The incidence ofpressure ulcerlesions in relation tothe different turningschemes wasvisualized usingsurvival curvesestimated accordingto the Kaplan-Meiermethod

Baseline differences:No significantdifferences betweenthe group

Study power/sample

Group 2

Randomised N: 65

Completed N: 58

Dropouts: 7 (5transferred to hospitaland 2 missing data)

Group 3

Randomised N: 67

Completed N: 66

Dropouts: 1 (missingdata)

Group 4

Randomised N: 65

Completed N: 63

Dropouts: 2 (2 died)

Group 5

Randomised N: 576

Completed N: 511

Dropouts: 65 (20 died,24 transferred tohospital and 21 missingdata)


Intervention

Comparison

Outcomemeasures

Effect sizes

Dropouts: 7 (5transferred to hospital

Dropouts: 1 (missing

Dropouts: 65 (20 died,24 transferred tohospital and 21 missing

Nurses did not use apressure ulcer riskassessment scale and werenot familiar with those scales.Preventive care did notinclude turning.

ulcer (grades II, IIIand IV in theAHCPRclassification)

KCE Report 193S

Comments

KCE Report 193S


size: Power analysiswas performed usingthe national Belgianpressure ulcerprevalence figures.Desired power of80% and asignificance level of0.05, a sample of 60in each group wasdeemed sufficient.

Setting: Elevengeriatric nursinghomes in Flanders(Belgium)

Length of study: 4-week study period

Assessment of PUs:not reported

Multiple ulcers: N/A

Inclusion criteria: ABraden score of lessthan 17 or a Nortonscore of less than 12;informed consent ofpatient/family

Exclusion criteria: noreported


Intervention

Comparison

Outcomemeasures

Effect sizes

Inclusion criteria: ABraden score of lessthan 17 or a Nortonscore of less than 12;informed consent of

Exclusion criteria: no

279

Comments

280

Table 29 – SMITH 1990


Author and year:Smith 1990

Title: Preventingpressure ulcers ininstitutionalizedelders: assessing theeffects of small,unscheduled shiftsin body position

Journal: Decubitus

Sequencegeneration:Participants wererandomly assignedto the treatment orcontrol group bydrawing names froma hat.

Allocationconcealment

Blinding: Notreported

Addressingincomplete outcomedata: Provideddetails to missingdata and usedavailable patients


Patient group:Participants ranged inage from 65 years to 91years with a mean ageof 80.55. Fourteenparticipants werewomen and five weremen. Elderly patients:

All patients

Randomised N: 26

Completed N: 19

Drop-outs: 7

Group 1

Randomised N: 14

Completed N: 9

Dropouts: 5 (3 found tohave pressure ulcerbefore study and 2missing data)

Group 2

Randomised N: 12

Completed N: 10

Dropouts: 2 (1 found tohave pressure ulcerbefore study and


Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group:Participants ranged inage from 65 years to 91

ith a mean ageof 80.55. Fourteenparticipants werewomen and five were

Dropouts: 5 (3 found tohave pressure ulcer

nd 2

Dropouts: 2 (1 found tohave pressure ulcerbefore study and

Group 1: Small shift in body(adjusting the position of alimb or body part by placing asmall rolled towel todesignated areas). A handtowel was used because itwas efficient, convenient,and an existing resource.Shifts were completed in lessthan one minute. Sites forplacement of rolled towelwere under each arm,shoulder, hip, and leg.

Group 2: Turning every twohours.

Both groups received normal,routine care and were turnedevery two hours.

Outcome 1:Development ofpressure ulcer.

Throughout the second weekof the study, one subject ineach of the two groupsdeveloped a pressure ulcerwhich healed by the end ofthe study.

The mean post test Nortonscores for the experimentalgroup decreased to 9.44,while the conincreased to 12.5. There wasno difference between posttest scores for the twogroups.

KCE Report 193S

Comments

Throughout the second weekof the study, one subject ineach of the two groupsdeveloped a pressure ulcerwhich healed by the end of

The mean post test Nortonscores for the experimentalgroup decreased to 9.44,while the control groupincreased to 12.5. There wasno difference between post-test scores for the two


Limitations:

Allocationconcealment notreported.

Intention-To-Treatanalysis notreported.

Blinding notreported.

High rate of dropouts (differencebetween controland experimentalgreater than10%).

Small samplesize.

Clinicallyexperimentalgroup were moreat risk.

Narrative report ofeffect sizes wasgiven.

Additional

KCE Report 193S



Baseline differences:No significantdifferences betweenthe group

Study power/samplesize: not reported

Setting: Participantswere drawn from asingle, skilled, 100-bed long –term carefacility in a largeMidwesternmetropolitan city.


Assessment of PUs:When a pressureulcer was found, itwas measured usinga Medirule.Information on theprogression ofpressure ulcerformation, chartinformation, andobservationspertinent to the studywere kept in a diary.

Multiple ulcers: nodetails

1missing data)

Inclusion criteria:Patients who received a14 or below on theNorton scale and were65 years or older.

Exclusion criteria: Nodetails provided


Intervention

Comparison

Outcomemeasures

Effect sizes

Inclusion criteria:Patients who received a14 or below on theNorton scale and were

iteria: No

281

Comments

outcomes:

282

Table 30 – VANDERWEE 2007


Author and year:Vanderwee 2007

Title: Effectivenessof turning withunequal timeintervals on theincidence ofpressure ulcerlesions.

Journal: JANOriginal Research

Sequencegeneration:Randomisation doneat ward level usingrandomisation listsgenerated with thesoftware packageSPSS 12.

Allocationconcealment: Notreported


Addressingincomplete outcomedata: None reported.No loss to follow up.

Analysis: no detailsprovided.

Patient group: 838geriatric nursing homepatients. Mean age:84.4 (SD 8.33) years,The mean Braden scorewas 13.2 (SD 2.36) andthe mean Norton scorewas 10.0 (SD 1.96).

All patients

Randomised N: 235

Completed N: 235

Drop-outs: not reported

Group 1

Randomised N: 122

Completed N: 122


Group 2

Randomised N: 113

Completed N: 113



Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: 838geriatric nursing homepatients. Mean age:

(SD 8.33) years,The mean Braden scorewas 13.2 (SD 2.36) andthe mean Norton score

Group 1: 4 hours in a semi-Fowler 30° position and 2hours in a lateral position30°. The semi-Fowlerposition consisted of a 30°elevation of the head endand the foot end of the bed.In a lateral position, theposition, the patient wasrotated 30°, with their backsupported with an ordinarypillow.

Group 2: Repositioning wasthe same as above but withequal time intervals of 4hours in lateral 30° as insemi-Fowler 30° position.

Patients in both groups werelying on a visco-elastic foamoverlay mattress

Outcome 1:Incidence ofpressure ulcer(proportion ofpatientsdeveloping ulcer)

Group 1: 20/122 (16.4%)

Group 2: 24/113 (21.2%)

Outcome 2: Theseverity ofpressure ulcerlesion

The majority of patients inthe experimental group(17/122; 13.9%) and thecontrol group (22/113;19.5%) developed a grade 2pressure ulcer. Threepatients (2.5%) in theexperimental group and two(1.8%) in the control grouphad a grade 3ulcer. No statisticallysignificant difference in theseverity of pressure ulcer.

Outcome 3:Location ofpressure ulcerlesion

Group 1: 13 patients (10.7%)developed a pressure ulcerat the sacral area; 7 patients(5.7%) on the heels orankles.

Group 2: 20 patients (17.7%)had a pressure ulcer on thesacrum and four (3.5%) onthe heels or ankles.Difference between the twogroups was not statisticallysignificant.

KCE Report 193S

Comments

Group 1: 20/122 (16.4%)

Group 2: 24/113 (21.2%)


Limitations:Intention-To-Treatanalysis notreported.

Blinding notreported.

Allocationconcealment notmentioned.

Sample size waslower than thedesired powerneeded.

Results should beinterpreted withcaution.

Additionaloutcomes:

The majority of patients inthe experimental group(17/122; 13.9%) and thecontrol group (22/113;19.5%) developed a grade 2pressure ulcer. Threepatients (2.5%) in theexperimental group and two(1.8%) in the control grouphad a grade 3 or 4 pressureulcer. No statisticallysignificant difference in theseverity of pressure ulcer.

Group 1: 13 patients (10.7%)developed a pressure ulcerat the sacral area; 7 patients(5.7%) on the heels or

Group 2: 20 patients (17.7%)had a pressure ulcer on thesacrum and four (3.5%) onthe heels or ankles.Difference between the twogroups was not statistically

KCE Report 193S


Statistical analysis:Data were analysedusing the softwarepackage SPSSversion 12.0.

Baseline differences:The two groups werecomparable withrespect to baselineand mobilitycharacteristics.

Study power/samplesize: Sample size forthe trial wascalculated based onan incidence ofpressure ulcerlesions (grade 2 orhigher) in nursinghomes of 17% (todetect a difference of0.05; power = 80%).In order to detect adifference of 10% inthe pressure ulcerincidence betweenthe groups, 148patients per groupwould have to beincluded in the trial.

Setting: 84 wards of16 Belgian elder carenursing homes

Inclusion criteria:Patients were eligible forthe study if they had nopressure ulcer lesion(grades 2, 3 or 4)(EPUAP 1999) at thestart of the study, if theycould be repositioned,and if they are expectedto stay for >3 days in thenursing home.

Exclusion criteria:


Intervention

Comparison

Outcomemeasures

Effect sizes

Inclusion criteria:Patients were eligible forthe study if they had nopressure ulcer lesion(grades 2, 3 or 4)(EPUAP 1999) at thestart of the study, if theycould be repositioned,

y are expectedto stay for >3 days in the

Outcome 4: Timeto developingpressure ulcer(analysed using aKaplan-Meiersurvival analysis)

No statistically significantdifference between the twoturning protocols (Log Ranktest = 1.18, d.f. = .1, p =0.28). To account for thedelay in which a pressureulcer becomes visible on theskin surface, the survivalanalysis wastarting from day 4. Nostatistically significantdifference was found (LogRank test = 1.04, d.f. = 1; P =0.31)

283

Comments

No statistically significantdifference between the twoturning protocols (Log Ranktest = 1.18, d.f. = .1, p =0.28). To account for thedelay in which a pressureulcer becomes visible on theskin surface, the survivalanalysis was repeatedstarting from day 4. Nostatistically significantdifference was found (LogRank test = 1.04, d.f. = 1; P =

284



Assessment of PUs:Occurrence ofpressure ulcers wasassessed daily bythe nursing staff. Theskin was observed atall the pressurearrears. Pressureulcer categorizedaccording to theEPUAP-classificationsystem

Multiple ulcers: nonereported

Table 31 – MOORE 2011


Author and year:Moore 2011

Title: A randomisedcontrolled clinicaltrial of repositioning,using the 30° tilt, forthe prevention ofpressure ulcers

Journal: Journal ofClinical Nursing

Patient group: 213participants enrolled intostudy, 114 assigned tothe control arm and 99enrolled in theexperimental arm.Seventy-nine percentwere women, with 53%aged between 81-90years, 13% agedbetween 91-100 years.


Intervention

Comparison

Outcomemeasures

Effect sizes

Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: 213participants enrolled intostudy, 114 assigned tothe control arm and 99enrolled in theexperimental arm.

nine percentith 53%

90years, 13% aged

100 years.

Group 1: repositioning by theclinical staff, using the 30° tilt(left side, back, right side,back) every three hoursduring the night.

Group 2: Repositioning everysix hours at night, using 90°lateral rotation. Night timewas taken to mean betweenthe hours of 8pm-8 am. No


Group 1: 3/99 (3%)

Group 2: 13/114 (11%)

Outcome 2: Timeto pressure ulcerdevelopment

Group 1: Mean 26 days(range 3 days).

Group 2: Mean 17 days(range 24 days)

KCE Report 193S

Comments

Comments

Group 1: 3/99 (3%)

Group 2: 13/114 (11%)

Funding: HealthResearch Boardof Ireland ClinicalNursing andMidwiferyResearchFellowship.

Limitations:

Blinding not

Group 1: Mean 26 days(range 3 days).

Group 2: Mean 17 days(range 24 days)

KCE Report 193S


Sequencegeneration: Clusterrandomisation usingcomputerisedrandomisation

Allocationconcealment:Achieved throughuse of distancerandomisation:statistician, notresearcher controlledrandomisationsequence.


Addressingincomplete outcomedata: None reported.No loss to follow upreported.

Analysis: no details

Statistical analysis:Data were analysedusing SPSS version13 on an intention totreat (ITT) basis.Differences betweenthe two arms of thestudy assessedusing the chi-squared test.Multiple regression

Eighty-seven per cent ofthe participants werechair-fast and 77% hadvery limited activity

All patients

Randomised N: 213

Completed N: 213

Drop-outs: Nonereported

Group 1

Randomised N: 99

Completed N: 99

Dropouts: Nonereported

Group 2

Randomised N: 114

Completed N: 114

Dropouts: none reported

Inclusion criteria: An inpatient in a long termcare of the older personhospital; >65 years; at


Intervention

Comparison

Outcomemeasures

Effect sizes

seven per cent ofthe participants were

fast and 77% had

outs: None

Dropouts: None


Inclusion criteria: An in-patient in a long termcare of the older personhospital; >65 years; at

further manipulation ofpatient care was undertaken.

Both groups were nursedduring the day according toplanned care. Pressureredistribution devises incurrent use on the bed andon the chair was continued.Patients’ positions werealtered every 2-3 hours.


Ninety-four percent ofpressure was located on thesacrum/buttocks. One waslocated on the knee, with nopressure ulcer on the heels.

Sixteen pressure ulcersdeveloped during the studyperiod, seven classified asgrade 1 (6 in control group; 1in the experimental group).Nine classified as grade 2 (7in control group; 2 in theexperimental group).

285

Comments

four percent ofpressure was located on thesacrum/buttocks. One waslocated on the knee, with nopressure ulcer on the heels.

Sixteen pressure ulcersdeveloped during the studyperiod, seven classified asgrade 1 (6 in control group; 1

erimental group).Nine classified as grade 2 (7in control group; 2 in theexperimental group).

reported.

Sample size waslower than thedesired powerneeded.


Additionaloutcomes:

286


analysis wasconducted todetermine which riskfactors reflectedpressure ulcer risk.

Baseline differences:No statisticaldifference betweenthe groups for age,sex and Bradenactivity scores. Astatisticallysignificantassociation wasnoted for Bradenmobility scores, withmore of theexperimental groupnoted to be bed fast.

Study power/samplesize: Sample sizewas determined onthe basis of anexpected incidenceof 15% in the controlgroup and a 90%power to detect areduction in pressureulcer incidence from15-10%. The samplesize required wastwo groups of 398participants.

risk of pressure ulcerdevelopment; nopressure ulcer at thetime of recruitment tothe study; no medicalcondition that wouldpreclude the use ofrepositioning; consent toparticipate in the study.

Exclusion criteria: Notreported


Intervention

Comparison

Outcomemeasures

Effect sizes

risk of pressure ulcerdevelopment; no

sure ulcer at thetime of recruitment tothe study; no medicalcondition that wouldpreclude the use ofrepositioning; consent to

Exclusion criteria: Not

KCE Report 193S

Comments

KCE Report 193S


Setting: Participantswere selected from12 long-term care ofthe older personhospital settings inthe Republic ofIreland


Assessment of PUs:Patients’ skin wasassessed at eachturning episode. Ifany changes in skinintegrity were noted,the researcher wasinformed. The skinwas then assessedby the assigned keystaff member, theclinical nursemanager and theresearcher.Agreement wasachieved bycomparing theparticipants’ skincondition to theimages on theEPUAP gradingsystem.

Multiple ulcers: nonereported


Intervention

Comparison

Outcomemeasures

Effect sizes

287

Comments

288

Table 32 – YOUNG 2004


Author and year:Young 2004

Title: The 30° tiltposition vs the 90°lateral and supinepositions in reducingthe incidence of non-blanching erythemain a hospitalinpatient population:a randomisedcontrolled trial.

Journal: Journal ofTissue Viability.

Sequencegeneration:Randomisation wasbased on blockallocation

Allocationconcealment:Sequential openingof sealed opaqueenvelopes.

Blinding: Researcherwas unaware ofwhich method ofrepositioning hadbeen used.

Addressingincomplete outcome

Patient group: 46participants with 23randomised to theexperimental arm and23 to the control arm ofthe study. Mean age of70.3 years

All patients

Randomised N: 46

Completed N: 46

Drop-outs: Nonereported

Group 1

Randomised N: 23

Completed N: 23

Dropouts: Nonereported

Group 2

Randomised N: 23

Completed N: 23


Inclusion criteria:Elderly, at risk ofdeveloping pressure


Intervention

Comparison

Outcomemeasures

Effect sizes

46participants with 23randomised to theexperimental arm and23 to the control arm ofthe study. Mean age of

outs: None

Dropouts: None


Inclusion criteria:Elderly, at risk ofdeveloping pressure

Group 1: 30° tilt positionduring the night.

Group 2: 90° side-lyingposition during the night.


Group 1: 3/23 (13%)

Group 2: 2/23 (


Group 1: one (4%) over thesacrum, 2 (9%) developedtwo discrete areas ofdamage (one on the lefttrochanter and heel, and theother on the right trochanterand heel).

Group 2: 2 (9%) developedpressure damage at thesacrum.

Outcome 3:

Patientacceptability

Group 1: 5/23 (22%) wereunable to tolerateintervention

Group 2: None reported forthe control group

KCE Report 193S

Comments

Group 1: 3/23 (13%)

Group 2: 2/23 (9%)

Funding: Notreported

Limitations:

Study lacksgeneralisability(small samplesize; one nightstudy).


Additionalcomment:

Among thesubjects whocompleted thestudy, theexperimentalintervention (30°tilt repositioning)was difficult toimplement for 20subjects (87%),whereas only fivesubjects (22%) inthe control group(90° side-lyingposition)experienced

Group 1: one (4%) over thesacrum, 2 (9%) developedtwo discrete areas ofdamage (one on the lefttrochanter and heel, and the

the right trochanter

Group 2: 2 (9%) developedpressure damage at the

Group 1: 5/23 (22%) wereunable to tolerate

Group 2: None reported forthe control group

KCE Report 193S


data: None reported.No loss to follow upreported.

Statistical analysis:Statisticalcomparisons weremade on anintention-to-treatbasis. Primaryoutcome analysedusing Fisher’s exacttest

Baseline differences:Groups were similarwith respect toidentified variables

Study power/samplesize: Eighty per centpower of detecting adifference,significant at a 5%level, 46 subjectswere recruited intothe study

Setting: Acuteinpatient districtgeneral hospital

Length of study: Onenight

Assessment of PUs:Non-blanchingerythema was used

ulcers (confirmed by aWaterlow riskassessment score ofabove ten), able to lie30° tilt position, hadgiven informed consent

Exclusion criteria: Notreported


Intervention

Comparison

Outcomemeasures

Effect sizes

ulcers (confirmed by aWaterlow riskassessment score ofabove ten), able to lie

tilt position, had

Exclusion criteria: Not

289

Comments

difficulty withrepositioning.

Reported reasonsfor difficulty withrepositioningincludes: inabilityto get into andstay in position,joint stiffness,pain, anxiety.

290


as a definition ofpressure damage.This is ascertainedby applying lightfinger pressure toany reddened areas.If the area does notblanch underexertion then tissuedamage is said tohave occurred.


Table 33 – VAN NIEUWENHOVEN 2006


Author and year: VanNieuwenhoven 2006

Title: Feasibility andeffects of the semirecumbent positionto prevent ventilator-associatedpneumonia.

Journal: Critical Caremedical journal.

Sequencegeneration: Patientswere randomly

Patient group: 221participants with 112randomised to theexperimental arm and109 to the control arm ofthe study. Mean age of63.9 years

All patients

Randomised N: 221

Completed N: Not clear

Drop-outs: Not clear


Intervention

Comparison

Outcomemeasures

Effect sizes

Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: 221participants with 112randomised to theexperimental arm and109 to the control arm ofthe study. Mean age of

Group 1: Semi recumbentposition. Aim was to achieve45° position of the head andback. The 45° position wasnot achieved for 85% of thestudy time, and thesepatients more frequentlychanged position than supinepositioned patients.

Group 2: Standard care(supine position)


Group 1: 31/112 (28%)

Group 2: 33/109 (9%)

.

KCE Report 193S

Comments

Comments

Group 1: 31/112 (28%)

Group 2: 33/109 (9%)

Funding: Notreported

Limitations:

Additionaloutcomes:

KCE Report 193S


assigned on a one toone allocation basis.

Allocationconcealment:Closed, nontransparent,numberedenvelopes.

Blinding:Investigatorsremained blinded forthe results of interimanalysis

Addressingincomplete outcomedata: None reported.

Statistical analysis:Power calculationwas carried out.Study did notachieve estimatedsample calculated.Intention to treatanalysis done.

Baseline differences:Groups were similarwith respect toidentified variables

Study power/samplesize: an expectedtotal of 252 patientswould be needed to

Group 1

Randomised N: 112

Completed N: Not clear

Dropouts: Not clear

Group 2

Randomised N: 109

Completed N: not clear

Dropouts: not clear

Inclusion criteria: Adultpatients intubated within24hrs of ICU admissionand had an expectedduration of ventilation ofat least 48hrs.

Exclusion criteria: Ifpatients wereundergoing selectivedecontamination of theirdigestive tract or if theycould not be randomisedto one or two positions.


Intervention

Comparison

Outcomemeasures

Effect sizes

Inclusion criteria: Adultpatients intubated within24hrs of ICU admissionand had an expectedduration of ventilation of

Exclusion criteria: Ifpatients wereundergoing selectivedecontamination of theirdigestive tract or if theycould not be randomised

291

Comments

292


reject the nullhypothesis and anexpected totalsample size of 176patients would beneeded to accept thehypothesis.

Setting: Adultspatients admitted tofour ICUs in threeuniversity hospitalsin the Netherlands.

Length of study: 7days

Assessment of PUs:Pressure soredevelopment wasstaged daily byresearch nursesaccording to the fourstages described bythe NationalPressure UlcerAdvisory Panelsystem (stage I-IV)



Intervention

Comparison

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S

Table 34 – GENTILELLO1988


Author and year:Gentilello 1988

Title: Effect of arotating bed on theincidence ofpulmonarycomplications incritically ill patients

Journal: Critical CareMedicine 1988, 16(8),783-786.

Study type: RCT

Sequencegeneration:randomisationperformed bydrawing a card

Allocationconcealment: notreported

Blinding: study onlyreported that thephysician in chargeof interpreting x-rayswas blinded totreatment allocation.

Addressingincomplete outcomedata: noreasons/numbers forattrition/exclusions

Patient group: criticallyill patients in surgicalICU immobilisedbecause of head injury,spinal injuries oftraction.

All patients

Randomised N: 65

Completed N: 64

Drop-outs: 1 withdrew,not included in analysis

Group 1

Randomised N:27

Completed N: unclear

Dropouts: unclear

Sex (% male): 74.1

Age: 34.8 (s.d 20.6)years

Injury of spinal cord (%):14.8

Group 2

Randomised N: 38


Dropouts: unclear


Intervention

Comparison

Outcomemeasures

Effect sizes

callyill patients in surgicalICU immobilisedbecause of head injury,spinal injuries of

outs: 1 withdrew,

Age: 34.8 (s.d 20.6)

Injury of spinal cord (%):

Group 1: Kinetic treatmenttable (rotates through arc of124

oevery 7 minutes).

Nurses left bed rotatingexcept when vital signsrecorded and treatmentsgiven. IF there were seriouscomplications due to thetable they were moved to aconventional bed.

Group 2: Conventional bed.Patients turned in usual wayevery 2 hours. Patients whodeveloped a chest infectionwhich was thought due topositioning were moved tothe kinetic treatment table.

Outcome 1:Incidence ofpressure ulcer s

Group 1: 30%

Group 2: 26%

.

293

Comments

Group 1: 30%

Group 2: 26%

Funding: KineticCocnepts.

Limitations: Unclearallocationconcealment andblinding andaddressing ofincomplete outcomedata.

Additionaloutcomes: thetrial was notprimarily apressure ulcertrial and theprimary outcomewas incidence ofpulmonarycomplications

294


reported.

Statistical analysis:Incidence of PUs byZ-statistic.

Baseline differences:similar for mostdemographicvariables. Theconventional bedgroup had higherincidence of smoking

Study power/samplesize: no a priorisample sizecalculation but smallsample size.

Setting: a surgicalICU

Length of study:follow-up unclear.

Assessment of PUs:evaluated daily, nodetails of method.


Sex (% male): 76.3

Age: 35.1 (s.d 15.4)years

Injury of spinal cord (%):10.5

Inclusion criteria:patients withorthopaedic injuriesrequiring traction, headinjuries or spinal injuries

Exclusion criteria: notreported see above forinclusion criteria


Intervention

Comparison

Outcomemeasures

Effect sizes

Age: 35.1 (s.d 15.4)

Injury of spinal cord (%):

Inclusion criteria:ith

orthopaedic injuriesrequiring traction, headinjuries or spinal injuries

Exclusion criteria: notreported see above for

KCE Report 193S

Comments

KCE Report 193S

Table 35 – SUMMER1989


Author and year:Summer 1989

Title: Continuousmechanical turningof intensive care unitpatients shortestlength of stay insome diagnostic-related groups

Journal: Journal ofCritical Care 1989, 4,45-53.

Sequencegeneration: randomsequences of letterscorresponding to thetreatment groups

Allocationconcealment: notreported

Blinding: the studynurse collectingAPACHE score datawas not involved inpatient managementof triage decisions,but there is noindication thatoutcome assessorswere blinded.

Addressing

Patient group: patientsadmitted to the ICU indiagnostic groups –sepsis-sepsissyndrome/pneumonia;respiratory failure; drugoverdose; metaboliccoma;stroke/neuromusculardisease; adultrespiratory distresssyndrome

All patients

Randomised N: 86

Completed N: 83

Drop-outs: 3 lost tofollow-up

Groupings:

Sepsis n=30

COPD/asthma n=16

Overdose n=11

Metabolic coma n=12

Stroke/neuromuscularn=14

Group 1

Randomised N:43


Intervention

Comparison

Outcomemeasures

Effect sizes

Patient group: patientsadmitted to the ICU in

–

respiratory failure; drugoverdose; metabolic

disease; adultrespiratory distress

outs: 3 lost to

Group 1: Kinetic treatmenttable (7 feet x 3 feet padded,vinyl-covered platform oncentral rotating pivot whichturns through an arc every1.7 seconds). Reported to beof value in respiratory failure.

Group 2: Routine 2-hourlyturning on conventional beds

Outcome 1:Incidence ofpressure ulcer

Group 1: 1/43 (small facialulcer)

Group 2: 0/43

295

Comments

Group 1: 1/43 (small facial

Group 2: 0/43

Funding: notreported

Limitations:Unclear allocationconcealment andblinding. Unclearif similar atbaseline Patientsrandomised onlyobtunded orunconsciouspatients (althoughthis was not theinitial intention)

Additionaloutcomes:

296


incomplete outcomedata: no

Statistical analysis:covariance analysis.

Baseline differences:comparable forAPACHE score,condition of pressurearea at baseline notdiscussed.

Study power/samplesize: no a priorisample sizecalculation but smallsample size

Setting: ICU

Length of study:follow-up unclear

Assessment of PUs:APACHE-II scores



Dropouts: unclear

Group 2

Randomised N: 43


Dropouts: unclear

Inclusion criteria: mostpatients admitted to theICU: sepsis-sepsissyndrome orpneumonia; respiratoryfailure secondary tochronic obstructivepulmonary disease orasthma; drug overdose;diabetic ketoacidosis orother metabolic coma(uremia, hepaticencephalopathy); strokeor neuromusculardisease; adultrespiratory distresssyndrome.

Exclusion criteria: notreported but see abovefor inclusion criteria


Intervention

Comparison

Outcomemeasures

Effect sizes

Inclusion criteria: mostpatients admitted to the

sepsissyndrome orpneumonia; respiratoryfailure secondary tochronic obstructivepulmonary disease orasthma; drug overdose;diabetic ketoacidosis orother metabolic coma(uremia, hepaticencephalopathy); strokeor neuromusculardisease; adultespiratory distress

Exclusion criteria: notreported but see above

KCE Report 193S

Comments

KCE Report 193S

7. RE-DISTRIBUTING DEVICES



Protocol Re-distributing devices

Review question What are the most clinically effective pressure re


Intervention Mattresses/overlays

Standard foam mattresses (needs to be identified)

Alternative

Specialised foam mattresses

Gel-filled mattresses/ overlays

Fibre-filled mattresses/ overlays

Air-filled mattresses/ overlays

Water-

Bead-filled mattresses/ overlays

AP mattresses/ overlays (air

Low-air

Operating

Sheepskins (synthetic/natural)

Beds

Air-fluid

Low-air

Air flotation beds

Bead-filled


DISTRIBUTING DEVICES

distributing devices

What are the most clinically effective pressure re-distributing devices for the prevention of pressure ulcers?


Mattresses/overlays

Standard foam mattresses (needs to be identified)

Alternative foam mattresses/ overlays (e.g. convoluted foam, cubed foam)

Specialised foam mattresses

filled mattresses/ overlays


led mattresses/ overlays

-filled mattresses/ overlays


AP mattresses/ overlays (air-filled sacs which inflate and deflate)

air-loss mattresses

Operating-table overlays

Sheepskins (synthetic/natural)

fluidised beds

air-loss beds – patients are supported on a series of air sacs through which warmed air passes

Air flotation beds

filled beds

297

distributing devices for the prevention of pressure ulcers?

patients are supported on a series of air sacs through which warmed air passes

298


Seating

Standard Chair

Tilt in space

Pressure relieving chairs

Cushionso foamo gelo fluido air/dry flotationo alternating pressure cushionso tilt

Wheelchair support surfaces

Other

Pillows

Postural support

Limb protectors: pads and cushions of different forms to protect bony prominences

As prev


No intervention


Proportion of participants developing new pressure ulcers (dichotomous outcome)(describe different categories ofulcer)

Important outcomes:




Time in hospital or



Standard Chair

Tilt in space

Pressure relieving chairs

Cushionsfoam-filled cushionsgel-filled cushionsfluid-filled cushionsair/dry flotation cushionsalternating pressure cushionstilt-in-space cushions

Wheelchair support surfaces

Pillows

Postural support


As prevention strategies

Each other

No intervention



Important outcomes:

t acceptability



Time in hospital or other healthcare setting (continuous data)

KCE Report 193S



KCE Report 193S


Healthpressure ulcer patients, therefore may have to be narratively summarisedo Shorto Manchester Short Assessment of Quality of Lifeo EQo WHOo Cardiff HRQoL toolo HUIo Pressure ulcer quality of lif


Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missingdata such as available case analysis or ITT (with the ap




Non-English language papers

The search strategy The databases to be


All years.



Population

Intervention

Comparison

Outcomes

Study designstudies will be meta

Unit of analysis



Health-related quality of life (continuous data) (although unlikely to be sensipressure ulcer patients, therefore may have to be narratively summarised

Short-form health survey (SF36)Manchester Short Assessment of Quality of LifeEQ-5DWHO-QOL BREFCardiff HRQoL toolHUIPressure ulcer quality of life (Gorecki)


Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missingdata such as available case analysis or ITT (with the appropriate assumptions)







All years.



Population – any population will be combined for meta-analysis except for different strata

Intervention – Different categories of device will not be combined for meta-analysis

Comparison – any comparison which fits the inclusion criteria will be meta-anal

Outcomes – single side effects will be meta-analysed separately from other side effects

Study design – randomised and quasi-randomised studies will be meta-analysed together. Blinded and unblindedstudies will be meta-analysed together. Crossover trials will be meta-analysed together with parallel trials

Unit of analysis – patients, clusters (hospital wards), individual pressure ulcers

299

related quality of life (continuous data) (although unlikely to be sensi tive enough to detect changes in

Cochrane reviews will be included if they match our inclusion criteria and have appropriate assumptions for missingpropriate assumptions)


nalysis except for different strata

analysis

analysed

analysed separately from other side effects


patients, clusters (hospital wards), individual pressure ulcers – for those where patients are the

300


unit of analysis and the patient has multiple ulcers it should be the first presscategories of ulcer)

Minimum duration of treatment = no minimum.



Use available case analysis for dealing with missing data if there is a 10% dor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.

MIDs: 0.75 to 1.25 for dichotomous variables and 0.5 x standard deviation for contin

Analysis Strata:




People with sensory i

Patients with a BMI >40

Subgroups:


Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separat

Different ulcer locations

Other terms Support surfaces, pressure relieving, pressure reducing, pressure preventing




unit of analysis and the patient has multiple ulcers it should be the first presscategories of ulcer)




Use available case analysis for dealing with missing data if there is a 10% dor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data..

MIDs: 0.75 to 1.25 for dichotomous variables and 0.5 x standard deviation for contin





Patients with a BMI >40

Subgroups:


Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separat

Different ulcer locations: sacral, heel and others.

Support surfaces, pressure relieving, pressure reducing, pressure preventing


KCE Report 193S

unit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different

Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take the



Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separat ely)


KCE Report 193S

7.2. search strategy



Search strategy Re-distributing devices

Date 27th Mar 2012


Search strategy 1 pressure ulcer/

2 decubit*.ti,ab.

3 (pressure adj (sore* or ulcer* or damage)).ti,ab




7 or/1-6



10 controlled clinical trial.pt.

11 randomi#ed.ab.

12 placebo.ab.

13 drug therapy.fs.

14 randomly.ab.

15 trial.ab.

16 groups.ab.

17 or/9-16


19 trial.ti.

20 or/9-12,14,18-19

21 letter/

22 editorial/



sore*).ti,ab.





ontrolled clinical trial.pt.


301

Results


8894

3865

6062

501

50

622

13487

10757

322734

83763

285035

134079

1512984

175416

246425

1148425

2901023

158570

102055

789946

752856

302491

302


23 news/



26 comment/

27 case report/


29 or/21-28


31 29 not 30





36 exp Rodentia/


38 or/31-37

39 Meta-Analysis/





44 (search strategy or search criteria or systematic search or study selection or data extraction).ab.


46 (medline or pubmed or cochrane or embindex or bids or cancerlit).ab.

47 cochrane.jw.

48 or/39-47

49 20 or 48





exp Animals, Laboratory/



Analysis as Topic/



st* or bibliograph* or hand search* or manual search* or relevant journals).ab.



(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citation

KCE Report 193S

Results


ase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citation

143966

302413

4185

493095

1558286

83156

3025178

674026

3010191

3594930

665788

5218

365269

2460341

1032770

7127677

32205

11873

42057

50096

19856

21391

19634

61940

7944

145126

893674

KCE Report 193S


50 49 not 38

51 8 and 50

52 exp beds/

53 (mattress* or cushion* or foam or transfoam or overlay* or pa

54 (pressure adj2 (device* or support* or constant)).ti,ab.

55 (static adj air).ti,ab.

56 (air adj (suspension or bag*)).ti,ab.

57 (pressure adj2 (relie* or reduc* or alleviat* or redistribut* or re

58 water suspension*.ti,ab.

59 (elevation adj2 device*).ti,ab.

60 (clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus orcairwave).ti,ab.

61 ((turn* or tilt*) adj2 (bed* or frame*)).ti,ab.

62 (kinetic adj (therapy or table*)).ti,ab.

63 net bed*.ti,ab.

64 (positioning or repositioning or re

65 or/52-64

66 (seat* or chair* or wheelchair* or pillow*).ti,ab.

67 wheelchairs/

68 65 or 66 or 67

69 51 and 68

70 limit 69 to yr="2010 -Curr

Notes


(mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel).ti,ab.

(pressure adj2 (device* or support* or constant)).ti,ab.

(air adj (suspension or bag*)).ti,ab.

(pressure adj2 (relie* or reduc* or alleviat* or redistribut* or re-distribut* or alternat*)).ti,ab.

water suspension*.ti,ab.

(elevation adj2 device*).ti,ab.

(clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or

((turn* or tilt*) adj2 (bed* or frame*)).ti,ab.

tic adj (therapy or table*)).ti,ab.

(positioning or repositioning or re-positioning).ti,ab.

(seat* or chair* or wheelchair* or pillow*).ti,ab.

Current"

303

Results


782841

995

3372

250061

6845

72

439

16888

280

10

448

454

77

9

33140

309311

36394

3172

344756

323

49

304



Date 27th Mar 2012

Database Embase-OVID

Search strategy 1 random*.ti,ab.

2 factorial*.ti,ab.








10 or/1-9


12 note.pt.

13 editorial.pt.



16 or/11-15

17 randomized controlled tria

18 16 not 17


20 nonhuman/



23 animal model/

24 exp Rodent/




((doubl$ or singl$) adj blind$).ti,ab.


ocedure/










KCE Report 193S

Results

711167

18452

60004

136181

549213

33346

107813

15595

318508

1177104

775094

511290

399508

1825147

134926

3380104

794389

3354078

1321445

3806953

1498332

408085

629106

2520889

1103508

KCE Report 193S


26 or/18-25


28 meta-analysis/




32 (search strategy or search criteria or systematic se


34 (medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citationindex or bids or cancerlit).ab.


36 cochrane.jw.

37 or/27-36

38 decubitus/

39 decubit*.ti,ab.



42 ((moist* or friction or shear) adj2 (sore* or ulcer*


44 or/38-43


46 (10 or 37) not 26

47 45 and 46

48 (mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel).ti,ab.






aly* or metanaly* or metaanaly*).ti,ab.






d) adj2 (data or trials or studies or results)).ab.


sore*).ti,ab.








305

Results

arch or study selection or data extraction).ab.


or damage or wound* or injur* or lesion*)).ti,ab.

8855378

48030

61737

54972

58719

24411

26081

24044

75039

31034

11048

222072

12420

4747

7047

655

759

53

16890

13015

1103384

1435

265218

7910

100

513

20059

306






57 (kinetic adj (therapy or tabl

58 net bed*.ti,ab.



61 exp bed/

62 exp wheelchair/

63 or/48-62

64 47 and 63

65 limit 64 to yr="2010 -Current"

Notes


,ab.




(kinetic adj (therapy or table*)).ti,ab.



Current"

KCE Report 193S

Results


370

13

525

525

100

9

38650

40750

7588

5032

378050

427

69

KCE Report 193S



Date 27th Mar 2012

Database CINAHL

Search strategy S26 S7 and S24 Limiters –

S25 S7 and S24

S24 S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20 or S21 or S22 or S23

S23 seat* or chair* or wheelchair* or pillow*

S22 positioning or repositioning or re

S21 net bed*

S20 kinetic and (therapy or table*)

S19 (turn* or tilt*) and (bed* or frame*)

S18 clinifloat or maxifloat or vaperm or therarest or sheepskincairwave

S17 elevation N2 device*

S16 water suspension*

S15 pressure and (relie* or reduc* or alleviat* or redistribut* or re

S14 air suspension or air bag*

S13 static air

S12 pressure and (device* or support* or constant)

S11 mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel

S10 (MH "Wheelchairs+")

S9 (MH "Pillows and Cushions")

S8 (MH "Beds and Mattresses+")

S7 S1 or S2 or S3 or S4





S2 decubit*


Notes


– Published Date from: 20101201-20121231; English Language; Exclude MEDLINE records

S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20 or S21 or S22 or S23

at* or chair* or wheelchair* or pillow*

positioning or repositioning or re-positioning

kinetic and (therapy or table*)

(turn* or tilt*) and (bed* or frame*)

clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or

elevation N2 device*

pressure and (relie* or reduc* or alleviat* or redistribut* or re-distribut* or alternat*)

air suspension or air bag*

pressure and (device* or support* or constant)

mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel

(MH "Wheelchairs+")

(MH "Pillows and Cushions")

(MH "Beds and Mattresses+")

or S5 or S6



sore*



307

Results

20121231; English Language; Exclude MEDLINE records


or hammock or foot waffle or silicore or pegasus or


133

3354

48691

12957

7537

4

370

1366

57

6

0

14412

131

12

8690

9244

2956

456

2576

9607

1368

69

155

8277

474

7513

308



Date 27th Mar 2012

Database Cochrane (- CDSR [3/2012]; DARE; Central [3/2012]; NHS EED; HTA)

Search strategy #1 MeSH descriptor Pressure Ulcer explode all trees



#4 (bedsore* or bed-sore*):ti,ab,



#7 (#1 OR #2 OR #3 OR #4 OR #5 OR #6)

#8 MeSH descriptor Beds explode all trees

#9 MeSH descriptor Wheelchairs explode all trees

#10 (mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel):ti,ab,kw

#11 (pressure NEAR/2 (device* or support* or constant)):ti,ab,kw

#12 (static NEAR/2 air):ti,ab,kw

#13 (air NEAR/2 (suspension or bag*)

#14 (pressure NEAR/2 (relie* or reduc* or alleviat* or redistribut* or re

#15 water suspension*:ti,ab,kw

#16 (elevation NEAR/2 device*):ti,ab,kw#17 (clinifloat or maxifloat or vaperm or therarest or sheepskincairwave):ti,ab,kw

#18 ((turn* or tilt*) NEAR/2 (bed* or frame*)):ti,ab,kw


#20 net bed*:ti,ab,kw

#21 (positioning or repositioning or re

#22 (seat* or chair* or wheelchair* or pillow*):ti,ab,kw#23 (#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR#22)

#24 (#7 AND #23)

#25 (#24), from 2010 to 2012

Notes


CDSR [3/2012]; DARE; Central [3/2012]; NHS EED; HTA)



sore*):ti,ab,kw

(incontinen* near/2 dermatitis):ti,ab,kw


(#1 OR #2 OR #3 OR #4 OR #5 OR #6)

MeSH descriptor Beds explode all trees

Wheelchairs explode all trees

(mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel):ti,ab,kw

(pressure NEAR/2 (device* or support* or constant)):ti,ab,kw

(static NEAR/2 air):ti,ab,kw

(air NEAR/2 (suspension or bag*)):ti,ab,kw

(pressure NEAR/2 (relie* or reduc* or alleviat* or redistribut* or re-distribut* or alternat*)):ti,ab,kw

water suspension*:ti,ab,kw

(elevation NEAR/2 device*):ti,ab,kw(clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or

((turn* or tilt*) NEAR/2 (bed* or frame*)):ti,ab,kw


(positioning or repositioning or re-positioning):ti,ab,kw

(seat* or chair* or wheelchair* or pillow*):ti,ab,kw(#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR

(#24), from 2010 to 2012

KCE Report 193S

Results


distribut* or alternat*)):ti,ab,kw

or hammock or foot waffle or silicore or pegasus or

(#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR

481

348

821

32

10

63

1161

243

127

7516

800

4

8

3643

118

5

53

47

47

289

8906

2653

22993

498

48

KCE Report 193S


Figure 1 – Flow chart of clinical article selection

Titles andabstracts identified,n =220

Full copiesretrieved andassessed foreligibility, n =58

Excluded, n

Publicationsincluded in review,n = 51

Excluded, n = 7


of clinical article selection

Excluded, n = 162

309

310


Reference Reason for exclusion

ALLEN1993 No clinical outcomes, only interface pressurerecorded

ANDREWS1989 Did not fulfil study design criteria

BALLARD1997 Data recorded were comfort data; no pressureulcer outcomes

BARHYTE1995 Did not fulfil study design criteria. No datapresented

BLISS1967 Did not fulfil study design criteria. Patients wererecruited to the trial on the basis of their riskscore

BLISS1995 Whilst 8 surfaces were evaluated in thisprospective trial, not all surfaces were in thetrial at the same time, therefore, the surfaceswere not truly compared with one anothercontemporaneously. Furthermore, it waspossible for patients to be reinto the study, which occurred frequently, with atotal of 457 mattress trials reported for only 238patients. The data were not prepatient only by mattress trial.

Duplicate citation of Bliss 1994 [conferenceabstract]

BRANIFF-MATTHEWS1997

Healing and prevention outcome data were notseparated.

BRIENZA2001 Study of pressure measurement

BUCHNER1995 Did not fulfil study desanti-decubitus management not reported anddecided by nurses. Number of pillows providedto third arm of the study was limited and not



cal outcomes, only interface pressure

Did not fulfil study design criteria

Data recorded were comfort data; no pressure

Did not fulfil study design criteria. No data

ot fulfil study design criteria. Patients wererecruited to the trial on the basis of their risk

Whilst 8 surfaces were evaluated in thisprospective trial, not all surfaces were in thetrial at the same time, therefore, the surfaces

not truly compared with one anothercontemporaneously. Furthermore, it waspossible for patients to be re-randomised backinto the study, which occurred frequently, with atotal of 457 mattress trials reported for only 238patients. The data were not presented bypatient only by mattress trial.

Duplicate citation of Bliss 1994 [conference

Healing and prevention outcome data were not

Study of pressure measurement

Did not fulfil study design criteria. Criteria fordecubitus management not reported and

decided by nurses. Number of pillows providedto third arm of the study was limited and not


given to all participants.

CADUE2008 More relevant to heel ulcer review.

CHALONER2000 Did not fulfil study design criteria,randomisation corrupted, authors reported thatrandomisation was compromised on the basisof bed availability

COLIN1996 No clinical outcomes recorded; onlymeasurements taken were for transcutaneousoxygen tension

CONINE1991 Did not fulfil study design criteria

DEBOISBLANC1993 Outcome incidence of pneumonia, no pressureulcer outcomes

DEFLOOR1997 Compared turning

DEFLOOR2000 Did not compare surfaces

DEFLOOR2004 Compared turning

DELLAVALLE2001 Outcome of interface pr

ECONOMIDES1995 Wound breakdown rather than pressure ulcers

EWING1964 More relevant to heel ulcer review.

FLAM1995 Outcome skin temperature and skin moisturelevel, no pressure ulcer outcomes

FLEISCHER1997 Did not fulfil study design criteria

GEELKERKEN1994 Did not fulfil study design criteria. No datapresented.

GENTILELLO1988 More relevant to repositioning review

GILAGUDO2009 Outcome measure of interface pressure

KCE Report 193S


given to all participants.

More relevant to heel ulcer review.

not fulfil study design criteria,randomisation corrupted, authors reported thatrandomisation was compromised on the basisof bed availability

No clinical outcomes recorded; onlymeasurements taken were for transcutaneousoxygen tension


Outcome incidence of pneumonia, no pressureulcer outcomes

Compared turning

Did not compare surfaces

Compared turning

Outcome of interface pressure

Wound breakdown rather than pressure ulcers


Outcome skin temperature and skin moisturelevel, no pressure ulcer outcomes


Did not fulfil study design criteria. No datapresented.

More relevant to repositioning review

Outcome measure of interface pressure

KCE Report 193S


GILCREAST2005 More relevant to heel ulcer review.

GRAY2008 Not an RCT, but a clinical audit

GRINDLEY1996 Patients were crossed between interventiongroups at 3 days. Outcome used was theassessment of patient comfort.

GUNNINGBERG1998 Did not fulfil study design criteria. Study of riskcalculation rather than prevention

HAALBOOM1994 Did not fulfil study design criteria

HAMPTON1998 Did not fulfil study design criteria

HAWKINS1997 Did not fulfil study design criteria

HEYNEMAN2009 Meta-analysis of 2 previously published RCTs(Vanderwee 2005; Vanderwee 2007).Vanderwee 2005 alrreview. Vanderwee 2007 excluded as it is aturning trial

HOLZGREVE1993 Full paper unavailable. Insufficient informationto assess

HUANG2009 Evaluated dressings

INMAN1999 Comparisono of a bed rental versus a bedpurchase strategy,surfaces

JACKSICH1997 Did not fulfil study design criteria

JESURUM1996 Did not fulfil study design criteria

KOO1995 Did not fulfil study design criteria, study ofinterface pressure in healthy volunteers

MARCHAND1993 Did not fulfil study design criteria, was aretrospective chart audit




t a clinical audit

Patients were crossed between interventiongroups at 3 days. Outcome used was theassessment of patient comfort.

Did not fulfil study design criteria. Study of riskcalculation rather than prevention




analysis of 2 previously published RCTs(Vanderwee 2005; Vanderwee 2007).Vanderwee 2005 already included in thisreview. Vanderwee 2007 excluded as it is a

Full paper unavailable. Insufficient information

Comparisono of a bed rental versus a bedpurchase strategy, not a comparison of



Did not fulfil study design criteria, study ofinterface pressure in healthy volunteers

study design criteria, was aretrospective chart audit


MCMICHAEL2008 Outcome measure of interface pressure

NEANDER1996 Paper in germanRCT. There were no data on how the decisionto include patients in the control andintervention groups was made

OOKA1995 Did not fulfil study design criteria, conveniencesample used

PHILLIPS1999 N of 1 trial design, only one participant in thetrial

REGAN1995 This study reported an audit of pressure ulcerincidence after implementationcomprehensive pressure ulcer policy; it is not aprospective rCT

REYNOLDS1994 This study Did not fulfil study design criteria

ROSENTHAL1996 Did not fulfil study design criteria. Outcomemeasure of interface pressure

SCOTT1995 Insufficient informatidecision

SCOTT1999 No clinical outcomes, healthy volunteer study ofinterface pressures

SCOTT2000 Not an RCT of beds and mattresses

STONEBERG1986 Historical control group

SUAREZ1995 Controlled clinical trial which recorded onlypressure measurements

SUMMER1989 More relevant to repositioning review

TAKALA1994 Not an RCT, outcome measure of interfacepressure

311



Paper in german – translator state it was not anRCT. There were no data on how the decisionto include patients in the control and

ervention groups was made

Did not fulfil study design criteria, conveniencesample used

N of 1 trial design, only one participant in the

This study reported an audit of pressure ulcerincidence after implementation of acomprehensive pressure ulcer policy; it is not aprospective rCT

This study Did not fulfil study design criteria

Did not fulfil study design criteria. Outcomemeasure of interface pressure

Insufficient information available to make adecision

No clinical outcomes, healthy volunteer study ofinterface pressures

Not an RCT of beds and mattresses

Historical control group

Controlled clinical trial which recorded onlyressure measurements

More relevant to repositioning review

Not an RCT, outcome measure of interfacepressure

312


THOMAS1994 Did not fulfil study design criteria

TIMMONS2008 Did not fulfil study design criteria. Review of aproduct not a trial

TORRAIBOU2002 Evaluated dressings

TURNAGE-CARRIER2008


TYMEC1997 More relevant to heel ulcer review.

VANDERWEE2007 Compared turning

VANDERWEE2008 Literature review of previously conductedstudies

WELLS1984 Only recorded interface pressuremeasurements

WILD1991 Interface pressure measurements

ZERNIKE1997 Incidence of pressure ulcers not reported

ZERNIKE1994 Unable to assess due to information inresearch paper. Email address provided wasno longer valid and we were unable to findother contact details.




Did not fulfil study design criteria. Review of a



Literature review of previously conducted

Only recorded interface pressure

Interface pressure measurements

Incidence of pressure ulcers not reported

Unable to assess due to information inresearch paper. Email address provided was

d and we were unable to find



A Cochrane review by McInnes et al (2011search and was adapted for this review. We quality assureCochrane review and as it was of very high quality and matched themajority of our protocol we used the information within it to populate ourreview for the summary of studies, forest plots and for the qualityassessment of studies. Changes odifferences in the protocol or to adapt for the purposes of GRADE.


We removed 7 of the 53 studies that were included in the Cochranereview. Four studies

74 , 75 , 76 , 77were removed from this review as they

included only heel ulcers and will be covered in the heel ulcer preventionreview (see Error! Reference source not found.(Economides, 1995)

78was excluded as it looked at wound breakdown

rather than incidence of pressure ulcers. Two other studies (Gentilello,1988

70and Summer, 1989

71) were excluded from this review as they were

deemed more relevant to the repositioning review.

Five additional studies79 , 80 , 81 , 82 ,

were not included in the review, and have been extracted (see Appendix7).

Fifty-one studies in total were included in this review.

This review identified studies in different settings: operating, 114

, intensive care units82 , 102 , 106 , 118

101 , 109 , 113 , 117 , 119, accident and emergency ward

facilities89 , 90 , 91 , 93 , 108

, nursing homeshospital wards

80 , 84 , 86 , 87 , 98 , 104 , 114

study setting16 , 88 , 100 , 101 , 103 , 105 , 112

Various types of redistributing devices are used, and the Cochranereview

73categorised them as:

Low-tech (non-powered) constant low pressure sup

High-tech support surfaces;

Other support surfaces (e.g. operating table overlay, turningbeds/frames, wheelchair cushions and limb protectors).

KCE Report 193S

by McInnes et al (2011)73

was identified from theor this review. We quality assured the McInnes

Cochrane review and as it was of very high quality and matched themajority of our protocol we used the information within it to populate ourreview for the summary of studies, forest plots and for the qualityassessment of studies. Changes or additions were made based ondifferences in the protocol or to adapt for the purposes of GRADE.

We removed 7 of the 53 studies that were included in the Cochranewere removed from this review as they

d only heel ulcers and will be covered in the heel ulcer preventionReference source not found.). One other study

was excluded as it looked at wound breakdownrather than incidence of pressure ulcers. Two other studies (Gentilello,

) were excluded from this review as they weredeemed more relevant to the repositioning review.

, 83were identified in our search, which

were not included in the review, and have been extracted (see Appendix

one studies in total were included in this review.16 , 78 -126

This review identified studies in different settings: operating theatre85 , 95 , 111

118 , 120 , 122 , 125, orthopaedic ward

92 , 94 , 97 ,

, accident and emergency ward99

, extended care, nursing homes

79 , 83 , 96 , 107 , 110,different types of

114 , 126. Several studies did not specify the

112 , 116 , 121 , 123 , 124, 127.

Various types of redistributing devices are used, and the Cochrane

powered) constant low pressure support surfaces;

ther support surfaces (e.g. operating table overlay, turningr cushions and limb protectors).

KCE Report 193S

The high-tech support surfaces included:

Alternating-pressure mattresses/overlays: patient liesthat inflate and deflate sequentially to relieve pressure at differentanatomical sites for short periods; these may incorporate a pressuresensor

Air-fluidised beds: warmed air circulates through fine ceramic beadscovered by a permeable sheet; allowing support over a larger contactarea (CLP)

Low-air-loss beds: patients are supported on a series of air sacsthrough which warmed air passes (CLP)

The other support surfaces included:

Turning beds/frames: these work by aiding manual repositpatient, or by motor driven turning and tilting.

Operating table overlays: mode of action as above.

Wheelchair cushions: either conforming cushions that reduce contactpressures by increasing surface area in contact, or mechanicalcushions e.g. alternating pressure.

Limb protectors: pads and cushions of different forms to protect bonyprominences.

As part of our protocol we were required to look at grades 2 pressureulcers and above as well as all grades of ulcer. This deviates from theMcInnes Cochrane review however they do state that studies that comparethe incidence of pressure ulcers of grade 2 or greater are more likely to bereliable. They included studies regardless of whether grade 1 ulcers weredescribed separately. Grading systems are variable but from the studieswhich reported grades 2 and above separately used the EPUAP orNPUAP classification system (see table of grading systems below). Forthose studies that did not use the EPUAP/NPUAP and reported grades ofulcer separately the distinction was usually a break in the skin or blister.

The McInnes Cochrane also found that methods for measuring secondaryoutcomes such as comfort, durability, reliability and acceptability were notwell developed. Where data were presented theyCharacteristics of included studies table, but did not incorporate into theiranalysis. As these were critical outcomes for this review,these outcomes in the GRADE evidence tables.


pressure mattresses/overlays: patient lies on air-filled sacsthat inflate and deflate sequentially to relieve pressure at differentanatomical sites for short periods; these may incorporate a pressure

fluidised beds: warmed air circulates through fine ceramic beadse sheet; allowing support over a larger contact

loss beds: patients are supported on a series of air sacs

Turning beds/frames: these work by aiding manual reposit ioning of the

Operating table overlays: mode of action as above.

Wheelchair cushions: either conforming cushions that reduce contactpressures by increasing surface area in contact, or mechanical

Limb protectors: pads and cushions of different forms to protect bony

As part of our protocol we were required to look at grades 2 pressureulcers and above as well as all grades of ulcer. This deviates from the

es Cochrane review however they do state that studies that comparethe incidence of pressure ulcers of grade 2 or greater are more likely to bereliable. They included studies regardless of whether grade 1 ulcers were

are variable but from the studieswhich reported grades 2 and above separately used the EPUAP orNPUAP classification system (see table of grading systems below). Forthose studies that did not use the EPUAP/NPUAP and reported grades of

he distinction was usually a break in the skin or blister.

Cochrane also found that methods for measuring secondaryoutcomes such as comfort, durability, reliability and acceptability were notwell developed. Where data were presented they did give details in the

table, but did not incorporate into theirAs these were critical outcomes for this review, we have included

The McInnes Cochrane did Meta-anathan one trial for an outcome which compared similar devices. The resultswere pooled using a fixed effect model, but if heterogeneity (Iabove and the p value was less than 0.10) was found they used a randomeffects model. They state that theyconstant for different lengths of followwere followed-up for different lengths of time.

No studies were found for standard or pressurespace wheelchairs, postural support or limb protectors.

7.3.3. Glossary of terms

Table 6 –Glossary of terms (NPUAP 2007)

Term Definition

Physical concepts related to support surfaces

Static Not active or moving; stationary. However withregards to support surfaces the description hasnow changed to mean ‘non

Dynamic Relating to energy or to objects in motion.However with regards to support surfaces thedescription has now ch‘powered’.

Friction (frictionalforce)

The resistance to motion in a parallel directionrelative to the common boundary of twosurfaces

Coefficient of friction A measurement of the amount of frictionexisting between two surfaces

Envelopment The ability of a support surface to conform, soto fit or mold around irregularities in the body

Fatigue The reduced capacity of a surface or itscomponents to perform as specified. Thischange may be the result of intended orunintended use and/or prchemical, thermal, or physical forces

313

analyse studies where there was morethan one trial for an outcome which compared similar devices. The resultswere pooled using a fixed effect model, but if heterogeneity (I

2= 50% or

above and the p value was less than 0.10) was found they used a random-ects model. They state that they assumed that the risk ratio remained

constant for different lengths of follow-up and so were pooled if participantsup for different lengths of time.

No studies were found for standard or pressure-relieving chairs, tilt-in-space wheelchairs, postural support or limb protectors.

Glossary of terms (NPUAP 2007)128

Definition

Physical concepts related to support surfaces

Not active or moving; stationary. However withregards to support surfaces the description hasnow changed to mean ‘non-powered’

Relating to energy or to objects in motion.However with regards to support surfaces thedescription has now changed to mean‘powered’.

The resistance to motion in a parallel directionrelative to the common boundary of twosurfaces

A measurement of the amount of frictionexisting between two surfaces

The ability of a support surface to conform, soto fit or mold around irregularities in the body

The reduced capacity of a surface or itscomponents to perform as specified. Thischange may be the result of intended orunintended use and/or prolonged exposure tochemical, thermal, or physical forces

314

Force A push-pull vector with magnitude (quantity)and direction (pressure, shear) that is capableof maintaining or altering the position of a body

Immersion Depth of penetration (sinking) intosurface

Life expectancy The defined period of time during which aproduct is able to effectively fulfil its designatedpurpose

Mechanical load Force distribution acting on a surface

Pressure The force per unit area exerted perpendicular tothe plane of interest

Pressureredistribution

The ability of a support surface to distribute loadover the contact areas of the human body. Thisterm replaces prior terminology of pressurereduction and pressure relief surfaces

Pressure reduction This term is no longer used to describe classesof support surfaces. The term is pressureredistribution; see above

Pressure relief This term is no longer used to describe classesof support surfaces. The term is pressureredistribution; see above

Shear (shear stress) The force per unit area exerted parallel to theplane of interest

Shear strain Distortion or deformation of tissue as a result ofshear stress

Components of support surfaces

Air A low density fluid with minimal resistance toflow

Cell/bladder A means of encapsulating a support medium

Viscoelastic foam A type of porous polymer material that conformsin proportion to the applied weight. The air


pull vector with magnitude (quantity)and direction (pressure, shear) that is capableof maintaining or altering the position of a body

Depth of penetration (sinking) into a support

The defined period of time during which aproduct is able to effectively fulfil its designated

Force distribution acting on a surface

The force per unit area exerted perpendicular to

The ability of a support surface to distribute loadover the contact areas of the human body. Thisterm replaces prior terminology of pressurereduction and pressure relief surfaces

is no longer used to describe classesof support surfaces. The term is pressureredistribution; see above

This term is no longer used to describe classesof support surfaces. The term is pressureredistribution; see above

The force per unit area exerted parallel to the

Distortion or deformation of tissue as a result of

A low density fluid with minimal resistance to

eans of encapsulating a support medium

A type of porous polymer material that conformsin proportion to the applied weight. The air

exists and enters the foam cells slowly whichallows the material to respond slower than astandard elast

Elastic foam A type of porous polymer material that conformsin proportion to the applied weight. Air entersand exits the foam cells more rapidly, due togreater density (non memory)

Closed cell foam A non-permeable structure in whbarrier between cells, preventing gases orliquids from passing through the foam

Open cell foam A permeable structure in which there is nobarrier between cells and gases or liquids canpass through the foam

Gel A semisolid system consistisolid aggrtegates, colloidal dispersions orpolymers which may exhibit elastic properties(can range from a hard gel to a soft gel)

Pad A cushioncomfort, protection or positioning

Viscous fluid A fluid with a relatively high resistance to flow ofthe fluid

Elastomer Any material that can be repeatedly stretched toat least twice its original length; upon releasethe stretch will return to approximately itsoriginal length

Solid A substance that dounder stress. Under ordinary conditions retainsits size and shape

Water A moderate desnity fluid with moderateresistance to flow

Features of support surfaces

Air fluidised A feature of a support surface that providespressure r

KCE Report 193S

exists and enters the foam cells slowly whichallows the material to respond slower than astandard elastic foam (memory foam)

A type of porous polymer material that conformsin proportion to the applied weight. Air entersand exits the foam cells more rapidly, due togreater density (non memory)

permeable structure in which there is abarrier between cells, preventing gases orliquids from passing through the foam

A permeable structure in which there is nobarrier between cells and gases or liquids canpass through the foam

A semisolid system consisting of a network ofsolid aggrtegates, colloidal dispersions orpolymers which may exhibit elastic properties(can range from a hard gel to a soft gel)

A cushion-like mass of soft material used forcomfort, protection or positioning

luid with a relatively high resistance to flow of

Any material that can be repeatedly stretched toat least twice its original length; upon releasethe stretch will return to approximately itsoriginal length

A substance that does not flow perceptiblyunder stress. Under ordinary conditions retainsits size and shape

A moderate desnity fluid with moderateresistance to flow

A feature of a support surface that providespressure redistribution via a fluid-like mediaum

KCE Report 193S

created by forcing air through beads ascharacterised by immersion and envelopment

Alternating pressure A feature of a support surface that providespressure redistribution via cyclic changes inloading and unloadingfrequency, duration , amplitude, and rate ofchange parameters

Lateral rotation A feature of a support surface that providesrotation about a longitudinal axis ascharacterised by degree of patient turn, durationand frequency

Low air loss A feature of a support surface that provides aflow of air to assist in managing the heat andhumidity (microclimate) of the skin

Zone A segment with a single pressure redistributioncapability

Multi-zoned surface A surface in which different segmdifferent pressure redistribution capabilities

Categories of support surfaces


created by forcing air through beads ascharacterised by immersion and envelopment

A feature of a support surface that providespressure redistribution via cyclic changes inloading and unloading as characterised byfrequency, duration , amplitude, and rate of

A feature of a support surface that providesrotation about a longitudinal axis ascharacterised by degree of patient turn, duration

A feature of a support surface that provides aflow of air to assist in managing the heat andhumidity (microclimate) of the skin

A segment with a single pressure redistribution

A surface in which different segments can havedifferent pressure redistribution capabilities

Reactive supportsurface

A powered and nonwith the capability to change its load distributionproperties only in response to applied load

Active supportsurface

A powered support surface, with the capabilityto change its load distribution properties, with orwithout applied load

Integrated bed system A bed frame and support surface that arecombined into a single unit whereby the surfaceis unable to function separately

Non-powered Any support surface not requiring or usingexternal sources of energy for operation(Enedrgy = D/C or A/C)

Powered Any support surface requiring or using externalsources of energy to operate (Energy = D/C orA/C)

Overlay An additional support surface designed to beplaced directly on top of an existing surface

Mattress A support surface designed to be placeddirectly on the existing bed frame

315

A powered and non-powered support surfacewith the capability to change its load distributionproperties only in response to applied load

A powered support surface, with the capabilityto change its load distribution properties, with orwithout applied load

A bed frame and support surface that arecombined into a single unit whereby the surfaceis unable to function separately

Any support surface not requiring or usingexternal sources of energy for operation(Enedrgy = D/C or A/C)

Any support surface requiring or using externalsources of energy to operate (Energy = D/C or

An additional support surface designed to beplaced directly on top of an existing surface

A support surface designed to be placeddirectly on the existing bed frame

316

7.3.4. Summary of included studies



Anderson1983

129Standard hospital mattress vsalternating air mattress vs waterfilled mattress (air mattress forcamping filled with water)

Aronovitch1999

85Alternating pressure system intraand postoperatively(MICROPULSE) vs conventionalmanagement (gel pad (ACTIONPAD) or standard pad in operatingroom and a replacement mattress(PRESSURE GUARD II)postoperatively)

Bennett 199884

Low air loss hydrotherapy(Permeable fast drying filter sheetover low-air-loss cushions(circulating air) (clensicair) vsstandard care (standard bed orfoam, air, alternating-pressuremattresses, skin care notstandardised)

Brienza 201079

Skin protection cushion (SPC) vssegmented foam cushion (SFC)

The skin protection cushion was acommercially available cushion withan incontinence cover. Cushionswere selected from three whichwere designed to improve tissuetolerance by reducing peakpressures near bony prominences,accommodating orthopaedicdeformities through immersion,


luded in the review

Population Outcomes

alternating air mattress vs water-filled mattress (air mattress for

Patients in acute setting at high riskof pressure ulcer development(Anderson scale) and withoutpressure ulcers

Incidence of pressure ulcers (allgrades)

system intra

) vs conventional(ACTIONin operating

room and a replacement mattress

Patients undergoing electivesurgery under general anaesthetic

Occurrence of pressure ulcer within7 days of surgery (all grades ofulcer)

le fast drying filter sheet

standard care (standard bed orpressure

Acute and long-term care patientsincontinent of urine and/or faceswith pressure ulcers grade 2 orbelow

Number of patients who developedpressure ulcers grade 2of patients with non-blanchableerythema (grade 1)

protection cushion (SPC) vssegmented foam cushion (SFC)

The skin protection cushion was acommercially available cushion withan incontinence cover. Cushions

selected from three whichwere designed to improve tissuetolerance by reducing peakpressures near bony prominences,accommodating orthopaedicdeformities through immersion,

Elderly, nursing home populationwho used wheelchairs as primarymeans of seating and mobility andwere at-risk for developing pressureulcers.

Incidence of pressure ulcers(different areas of the body) (allgrades of ulcer)

KCE Report 193S

Study length

Incidence of pressure ulcers (all 10-day follow-up

Occurrence of pressure ulcer within7 days of surgery (all grades of

7-day follow-up

Number of patients who developedpressure ulcers grade 2-4; number

blanchable

60-day follow-up

Incidence of pressure ulcers(different areas of the body) (all

6 months

KCE Report 193S


enveloping small irregularities atthe seating interface withoutcausing height pressure gradients,and dissipating heat and moisture.Solid seat inserts were provided.The segmented foam cushion wasa cross-cut, 7.6cm thick,segmented foam cushion with fittedincontinence cover and solid seatinsert.

Cavicchioli2007

86High-tech (HILL-ROM, DUO 2)mattress on alternating low-pressure setting vs high-tech (HILLROM DUO 2), mattress oncontinuous low-pressure setting

Cobb 199787

Low air loss bed (KINAIR) vs staticair mattress overlay (EHOBWAFFLE)

Collier 199688

Comparison of 8 foam mattresses:new standard hospital mattress vspressure-redistributing foammattresses (CLINIFLOAT,OMNIFOAM, SOFTFORM, STM5,THERAREST, TRANSFOAM,VAPOURLUX)

Conine 199089

Alternating-pressure overlay vssilicore overlay over standardhospital mattress (spring or foam)

All patients received usual careincluding 2-3 hourly turning; dailybed baths; weekly bath/shower; useof heel, ankle and other protectors

Conine 199390

Slab cushion bevelled at base toprevent seat sling vs contoured


Population Outcomes

enveloping small irregularities atthe seating interface without

ressure gradients,and dissipating heat and moisture.Solid seat inserts were provided.The segmented foam cushion was

cut, 7.6cm thick,segmented foam cushion with fittedincontinence cover and solid seat

)

HILL-

pressure setting

Acute and long-term careparticipants deemed at risk ofpressure ulceration (Braden score<17 activity or mobility sub-scales <3)

Number of participants withincidence of pressure ulcer (grade1 and 2)

vs static Hospital and ICU patientsconsidered high risk on Bradenscore

Number of participants withincidence pressure ulcer (grade 1and 2)

esses:new standard hospital mattress vs

OMNIFOAM, SOFTFORM, STM5,

Patients on a general medical ward,no further details

Incidence of pressure ulcers (allgrades of ulcers)

pressure overlay vs

hospital mattress (spring or foam)

All patients received usual care3 hourly turning; daily

bed baths; weekly bath/shower; used other protectors

Patients with chronic neurologicaldiseases

Incidence of pressure ulcers(including grade 1)

Slab cushion bevelled at base toprevent seat sling vs contoured

Extended care patients at high risk Incidence of pressure ulcers (all

317

Study length

Number of participants withincidence of pressure ulcer (grade

2-week follow-up

Number of participants withincidence pressure ulcer (grade 1

40-day follow-up

Incidence of pressure ulcers (all Not clear butassessed weekly

Incidence of pressure ulcers 3-month follow-up

Incidence of pressure ulcers (all 3-month follow-up

318


foam cushion with a posterior cutout in the area of ischialtuberosities and an anterior ischialbar

Conine 199491

Gel cushion with foam base (JAY)vs foam cushion

Cooper 199892

Dry flotation mattress (ROHO)dry flotation mattress (SOFFLEX)

Daechsel 198593

Alternating-pressure mattress vssilicore overlay

Demarre 201280

Alternating low pressure airmattress with multi-stage inflationand deflation of the air cells(CLINACTIV, HILL-ROM) vsstandard Alternating low pressureair mattress with single stage, steepinflation and deflation of air cells(HILL-ROM).

Exton-Smith1982

94Alternating-pressure mattress withtwo layers of air cells (PEGASUSAIRWAVE SYSTEM) vs alternatingpressure large cell ripple mattres

Feuchtinger2006

95Operating table with waterfilledwarming mattress and a 4-cmthermoactive viscoelastic foamoverlay vs standard OR tableconfiguration (OR table withwaterfilled warming mattress)

Gebhardt 1996125

Alternating-pressure air mattresses(shallow small cell overlays,


Population Outcomes

foam cushion with a posterior cut

tuberosities and an anterior ischial

of pressure ulcers grades of ulcers)

with foam base (JAY) Elderly patients in an extended carehospital deemed at high risk ofpressure ulcers

Pressure ulcer incidence (all gradesof ulcers)

vsotation mattress (SOFFLEX)

Mixed emergency orthopaedictrauma ward patients with waterlowrisk scores of >/=15

Incidence of grade 2 ulcers andabove

pressure mattress vs Neurological conditions in a long-term care hospital at high risk

Incidence of grade 1 and abovepressure ulcers

stage inflation

Alternating low pressureair mattress with single stage, steepinflation and deflation of air cells

Hospitalised patients. The wardswere neurology, rehabilitation,cardiology, dermatology,pneumology oncology and chroniccare or a combination of differenttypes of medical conditions

Incidence of all grades of ulcer andgrade 2 ulcer or greater; withdrawaldue to discomfort; time to developnew pressure ulcers

pressure mattress withtwo layers of air cells (PEGASUS

alternating-large cell ripple mattress

Geriatric patients, with fracturedneck of femur and long-staypatients without pressure ulcers ofgrade 2 or greater, Norton score<14.

Incidence of grade 2 ulcer orgreater

Operating table with waterfilled

thermoactive viscoelastic foam

Patients scheduled for cardiacsurgery with extracorporalcirculation, not required to be freeof pressure ulcers

Number of participants withincidence of pressure ulcers (allgrades and grades 2 and above)

pressure air mattresses Patients in ICU with a Norton score<13 with no pressure ulcers

Support provided; incidence ofpressure ulcers (all grades and

KCE Report 193S

Study length

idence (all grades 3-month follow-up

Incidence of grade 2 ulcers and 7-day follow-up

and above 3-month follow-up

Incidence of all grades of ulcer andgrade 2 ulcer or greater; withdrawaldue to discomfort; time to develop

14 days

Incidence of grade 2 ulcer or 2-week follow-up

Number of participants withincidence of pressure ulcers (allgrades and grades 2 and above)

5-day follow-up

provided; incidence ofpressure ulcers (all grades and

unclear

KCE Report 193S


medium depth large cell overlays,deep mattresses and deeppulsating low air loss bed) vsconstant low-pressure supports(fibre overlays, foammattresses/overlays, static airoverlays, gel overlay, water overlay,bead overlay, low air lossmattresses, static air overlay, lowair-loss beds and air-fluidised beadbeds)

Geyer 200196

Pressure-reducing wheelchaircushions (a commercial cushion,chosen by nurse based on patient,from a group of cushions designedspecifically to improve tissuetolerance in sitting by providingmore surface area and/or reducingpeak pressure near the ischialtuberosities, sacrum and coccygealareas. A fitted incontinence coverwas also included vs standardinch convoluted foam (EGGRATEcushion

Goldstone1982

97Bead bed system (BEAUFORT)(includes bead-filled mattress onA&E trolley; bead-filled operatingtable overlay; bead-filled sacralcushion for operating table; beadfilled boots to protect heels onoperating table

Gray 199498

Pressure-redistributing foammattress (SOFTFOAM) vs standard130mm NHS foam mattress


Population Outcomes

medium depth large cell overlays,

pressure supports

overlays, gel overlay, water overlay,

mattresses, static air overlay, low-fluidised bead

grades 2 and above); cost

commercial cushion,chosen by nurse based on patient,from a group of cushions designed

tolerance in sitting by providingmore surface area and/or reducingpeak pressure near the ischialtuberosities, sacrum and coccygeal

as. A fitted incontinence covervs standard 3-

EGGRATE)

Elderly patients in nursing homes;wheelchair users with Braden score</=18

Number of participants withincidence of pressure ulcer (allgrades)

(BEAUFORT)filled mattress on

filled operatingfilled sacral

cushion for operating table; bead-

Over 60 years with femur fracture Pressure ulcer incidenceof pressure ulcers)

vs standardPatients with orthopaedic trauma,vascular and medical oncologyunits without breaks in the skin

Incidence of pressure ulcers (grade2 or greater ulcer)

319

Study length

grades 2 and above); cost

Number of participants withincidence of pressure ulcer (all

12-month follow-up

ncidence (all grades Follow-up not clear

Incidence of pressure ulcers (grade 10-day follow-up

320


Gray 1998127

Pressure-redistributing foammattress (TRANSFOAM) vspressure-redistributing foammattress (TRANSFOAMWAVE)

Grisell 200881

A neoprene air filled bladder (dryflotation) device (ROHO) vs adisposable polyurethane foamprone head positioner (OSI) vs aprone view protective helmetsystem with a disposablepolyurethane foam head positioner)

Gunningberg2000

9910cm visco-elastic foam mattress(TEMPUR-PEDIC) on arrival inA&E, and visco-elastic foamoverlay on standard ward mattressvs standard A&E trolley mattress(5cm) and ward mattress (10cmfoam)

Hampton 1997100

Alternating-pressure mattress(CAIRWAVE SYSTEM) vsalternating pressure mattress(AIRWAVE SYSTEM)

Hofman 1994101

Cubed foam mattress(COMFORTEX DECUBE) vsstandard hospital foam mattress(standard polypropylene SG40)

Inman 1993102

Low-air-loss air-suspension beds(KINAIR) vs standard Intensivecare unit bed (patients rotatedevery 2 hours)

Jolley 2004103

Australian medical sheepskinmattress overlay (leather-backedwith a dense uniform 25 mm wool


Population Outcomes

mattress (TRANSFOAMWAVE)

General hospital patients admittedfor bed-rest or surgery with intactskin, no terminal illness

Incidence of pressure ulcegrades of ulcers)

A neoprene air filled bladder (dryflotation) device (ROHO) vs adisposable polyurethane foamprone head positioner (OSI) vs a

helmetsystem with a disposablepolyurethane foam head positioner)

Elective surgery patients – thoracic,lumbar or thora-columbar spinalsurgery that required pronepositioning

Incidence of all pressure ulcers andof grade 2 and above pressureulcers

elastic foam mattresson arrival in

overlay on standard ward mattressy mattress

(10cm

Patients admitted with a suspectedhip fracture via an A&E department;over 65 years; did not havepressure ulcers

Grade 2 to 4 incidence; meancomfort rating

Little detail, average age 77 years;number of patients at high-veryhigh risk

Incidence of pressure ulcers (grade2 and above)

ess(standard polypropylene SG40)

Patients with a femoral-neckfracture and risk score >8 (Dutchconsensus scale)

Incidence of ulcers (grade 2 orgreater)

bedsIntensive

bed (patients rotated

Patients >17 years with APACHE IIscore >15

Incidence of pressure ulcers (ulcersper patient and patients with ulcers)(grade 2 or greater)

backedwith a dense uniform 25 mm wool

Low to moderate risk of developinga pressure ulcer; aged >18 years.

Number of participants withincidence of pressure ulcer (allgrades of pressure ulcers)

KCE Report 193S

Study length


Incidence of all pressure ulcers andof grade 2 and above pressure

No details

Grade 2 to 4 incidence; mean Follow-up untildischarge or 14 dayspostoperatively

Incidence of pressure ulcers (grade 20 days maximumfollow-up

Incidence of ulcers (grade 2 or 2-week followup

Incidence of pressure ulcers (ulcersper patient and patients with ulcers)

Average 17 daysfollow-up

Number of participants withncidence of pressure ulcer (allgrades of pressure ulcers)

Unclear follow-upperiod; average 7days.

KCE Report 193S


pile vs usual care determined bystaff (repositioning and any otherpressure-redistributing device orprevention strategy with/withoutlow-tech constant pressurerelieving devices

Kemp 1993104

Convoluted foam overlay (either 3inch overlay with density of 1.42lbper cubic foot (acute settings) or a4 inch overlay with unknowndensity (long-term settings)) vssolid foam overlay (4 inches solidsculptured overlay with density to1.33lb per cubic foot)

Keogh 2001105

Profiling bed with a pressurereducing foam mattress/cushion vsflat-based bed with a pressurerelieving/redistributingmattress/cushion

Laurent 1998106

Standard mattress in ICU; standardmattress postoperatively vsalternating pressure mattress(NIMBUS)in ICU; standard mattresspostoperatively vs standardmattress in ICU; Constant lowpressure mattress (TEMPUR)postoperatively vs alternatingpressure mattress (NIMBUS) inICU; constant low pressuremattress (TEMPUR)postoperatively

Lazzara 1991107

Air-filled (SOFCARE) overlay vs gelmattress


Population Outcomes

ermined bystaff (repositioning and any other

orprevention strategy with/without

(either 3y with density of 1.42lb

per cubic foot (acute settings) or a

vs(4 inches solid

sculptured overlay with density to

>65 years, inpatients with BradenScore of </=16 from generalmedicine, acute geriatric medicineand long term care. Free frompressure ulcers.

Incidence of pressure ulcers (allgrades of pressure ulcers)

reducing foam mattress/cushion vsbased bed with a pressure

Patients from 2 surgical and 2medical wards; >18 years; waterlowscore of 15-25; tissue damage nogreater than grade 1

Incidence of pressure ulcers (allgrades of ulcers); healing ofexisting grade 1 ulcers

Standard mattress in ICU; standard

in ICU; standard mattress

in

Adults over 15 years of age,admitted for major cardiovascularsurgery

Incidence of ulcers of grade 2above

) overlay vs gel Nursing home residents at risk ofpressure ulcers (Norton score >15)

Incidence of pressure ulcer (allgrades and grade 2 or greaterulcers)

321

Study length

Incidence of pressure ulcers (allgrades of pressure ulcers)

1-month follow-up

Incidence of pressure ulcers (allgrades of ulcers); healing ofexisting grade 1 ulcers

5-10 days follow-up

Incidence of ulcers of grade 2 or unclear

essure ulcer (allgrades and grade 2 or greater

6-month follow-up

322


Lim 1988108

Foam slab cushion (2.5cm mediumdensity foam glued to 5cm firmchipped foam) vs contoured foamcushion (same foam as above; cutinto a customised shape to relievepressure on ischial tuberosities).

Malbrain 201082

Reactive dry floatation mattressoverlay (ROHO) vs the activealternating pressure mattress(NIMBUS 3)

McGowan 2000109

Standard hospital mattress, sheetand an Australian MedicalSheepskin overlay; sheepskin heeland elbow protectors as required vsstandard hospital mattress, sheetwith or without other low techconstant pressure devices asrequired.

Mistiaen 2009

Mistiaen 2010110

Australian medical sheepskin vsusual care

Cointerventions: usual interventionfor prevention of pressure ulcers instudy settings

Nixon 1998111

Dry visco-elastic polymer pad onoperating table vs standardoperating theatre table mattressplus Gamgee heel support


Population Outcomes

Foam slab cushion (2.5cm mediumdensity foam glued to 5cm firmchipped foam) vs contoured foamcushion (same foam as above; cut

shape to relievepressure on ischial tuberosities).

Residents of an extended carefacility; aged >/=60; free ofpressure ulcers but at high risk ofdeveloping one (Norton score</=14); using a wheelchair for >/=3hours/day; without progressivedisease or confined to bed

Incidence of all ulcers (grade 1 andabove)

dry floatation mattress ICU patients at high risk of pressureulcers (Norton score </= 8) andrequiring mechanical ventilation forat least 5 days with intact skin orwith PUs on admission

Incidence of pressure ulcers (allgrades of ulcers and grade 2 andabove)

Standard hospital mattress, sheet

Sheepskin overlay; sheepskin heelectors as required vs

standard hospital mattress, sheet

Orthopaedic patients aged >/60years; low or moderate risk (Bradenscale)

Incidence of ulcers (grade 1 andabove)

Australian medical sheepskin vs

Cointerventions: usual interventionfor prevention of pressure ulcers in

Patients from aged care facility(predominantly rehabilitationdepartment) and rehabilitationcentre. Grade 1 pressure ulcersincluded in sample


elastic polymer pad on

operating theatre table mattress

Patients >/=55 years; admitted forelective major general,gynaecological or vascular surgeryin supine or lithotomy position andfree of preoperative pressuredamage greater than grade 1


KCE Report 193S

Study length

Incidence of all ulcers (grade 1 and 5-month follow-up

Incidence of pressure ulcers (allof ulcers and grade 2 and

No details but meanstudy durationreported for patientswas 15 (s.d 14) inthe NIMBUS groupand 12.2 (s.d 5.5) inthe ROHO group

Incidence of ulcers (grade 1 and Discharge fromhospital, transfer to arehabilitation ward.



KCE Report 193S


Nixon 2006130

Alternating-pressure overlay(alternating cell height minimum8.5cm, max 12.25 cm) vsalternating-pressure mattress(alternating cell height min19.6cms, max 29.4cms)

Price 1999113

Low-pressure inflatable mattress(REPOSE SYSTEM) and cushionin polyurethane material) vsdynamic flotation Nimbus II plusalternating-pressuyre cushion for achair (ALPHA TRANSCELL): allother care standard best practice,including regular repositioning

Russell 2000114

Multi-cell pulsating dynamicmattress system (MICROPULSESYSTEM)in the operating roompostoperatively vs Conventionalcare (gel pad (ACTION PAD) inoperating room, standardmattress(HILL_ROM CENTRA with6 inch foam overlay or HILL-ROMCENTRA with 4 inch foam overlay)postoperatively)

Russell 2003131

Visco-polymer energy absorbingfoam mattress (CONFOR-MED 3inch layer viscoelastic foam and a 3inch layer of standard polyurethanefoam)/cushion combination vsstandard mattress/cushioncombination (KING’S FUND,LINKNURSE, SOFTFOAM,TRANSFOAM, KING’S FUNDMATTRESS with a SPENCO or


Population Outcomes

(alternating cell height minimumAcute or elective hospital patientsaged >/=55 years with limitedBraden activity and mobility score(1 or 2)

Incidence of pressure ulcer (grade2 and above)

pressure inflatable mattressand cushion

pluspressuyre cushion for a

: allandard best practice,

Patients with fractured neck offemur and Medley score of >25(very high risk) aged over 60 years

Incidence of pressure ulcers2 and above)

mattress system (MICROPULSEoperating roomand

postoperatively vs Conventionalin

(HILL_ROM CENTRA withROM

CENTRA with 4 inch foam overlay)

Patients >/= 18 years; undergoingscheduled cardiothoracic surgeryunder GA; surgery of at least 4hours duration; free of pressureulcers

Incidence and severity of pressureulcers

polymer energy absorbingMED 3

inch layer viscoelastic foam and a 3inch layer of standard polyurethane

MATTRESS with a SPENCO or

Elderly acute, orthopaedic andrehabilitation wards; >65 years;Waterlow score of 15-20

Development of non-blanchingerythema

323

Study length

Incidence of pressure ulcer (grade 30-day follow-up anda further 30-dayfollow-up

Incidence of pressure ulcers (grade 14-day follow-up

Incidence and severity of pressure 7-day follow-up

blanching Median 8-14(experimental) and 9-17 (control)

324


PROPAD mattress overlay)

Sanada 2003116

Double-layer cell overlay(TRICELL) – (two layers consistingof 24 narrow cylinder air cells, 10cm) vs single-layer air cell overlay(AIR DOCTOR single layerconsisting of 20 round air cells, 7.5cm) vs standard hospital mattress(PARACARE 8.5cm polyester)

Santy 1994117

Pressure-redistributing mattresses(CLINIFLOAT, OMNIFOAM,THERAREST, TRANSFOAM,VAPERM) vs NHS contract surface– standard foam (REYLON 150mm)

Schultz 1999132

Experimental mattress overlay inoperating room made of foam witha 25% indentation load deflection of30lb and density of 1.3 cubic feet vsusual care (padding as required,including gel pads, foammattresses, ring cushions (donutsetc…)

Sideranko1992

118Alternating air mattress (LAPIDUSAIRFLOAT SYSTEM 1.5 inch thick)vs static air mattress (GAY MARSOFCARE 4-inch thich) vs Watermattress (LOTUSs PXM 3666,4inch thick)

Stapleton 1986119

Large cell ripple bed pad (TALLEY)vs polyether foam pad 2 feet x 2feet x 3 inch thickness vs silicorebed pad (SPENCO)

Takala 1996133

Constant low pressure mattress


Population Outcomes

(two layers consisting, 10

layer air cell overlay

, 7.5hospital mattress

Acute care unit patients; Bradenscore of </=16; bed bound; free ofpressure ulcers

Incidence pressure ulcer (all gradesof ulcer and grade 2 and above)

redistributing mattresses

VAPERM) vs NHS contract surfacestandard foam (REYLON 150mm)

Patients aged >55 years with hipfracture, with or without pressureulcers

Incidence of pressure ulcers (allgrades).

mental mattress overlay inoperating room made of foam witha 25% indentation load deflection of30lb and density of 1.3 cubic feet vsusual care (padding as required,

mattresses, ring cushions (donuts

Patients admitted for surgery;aged>18 years; admitted with intactskin


LAPIDUS1.5 inch thick)GAY MAR

inch thich) vs WaterLOTUSs PXM 3666,4

Adult, surgical IC U patients;without existing skin breakdown


bed pad (TALLEY)vs polyether foam pad 2 feet x 2

silicore

Female elderly patients withfractured neck of femur; withoutexisting pressure ulcers; Nortonscore 14 or less

Incidence of pressure ulcers (allgrades and grade 2 or greater)

onstant low pressure mattress Non-trauma patients admitted to Incidence of pressure ulcers (all

KCE Report 193S

Study length

Incidence pressure ulcer (all gradesof ulcer and grade 2 and above)

Follow-up durationnot reported



Incidence of pressure ulcers (all Mean 9.4 daysfollow-up

Incidence of pressure ulcers (allgrades and grade 2 or greater)

Duration of follow-upunclear


KCE Report 193S


(CARITAL OPTIMA) (21 double airbags on a base) vs standardhospital foam mattress (10cm thickfoam density 35kg/m3)

Taylor 1999121

Alternating-pressure mattress withpressure-redistributing cushion(PEGASUS TRINOVA) vsalternative alternating-pressuresystem (unnamed) with pressureredistributing cushion

Theaker2005

122Alternating pressure mattress (KCITHERAPULSE) vs alternatingpressure mattress(HILL-ROMDUO)

Vanderwee2005

126Alternating pressure air mattress(aLPHA-X-CELL) vs visco-elasticfoam mattress (TEMPUR)

Vyhlidal1997

123Foam mattress overlay (IRIS 3000,4-inch thick 1.8lb density withdimpled surface) vs foam mattressreplacement (MAXIFLOAT)

Whitney1984

124Alternating-pressure mattress (1343-inch diameter air cells, 3 minutecycle) vs convoluted foam pad(EGGCRATE)

Patients in both groups were turnedevery 2 hours

Van Leen2011

83Combination of a standard 15cmcold foam mattress with a static airoverlay vs a standard 15cm coldfoam mattress


Population Outcomes

21 double air

hospital foam mattress (10cm thick

ICU grades of ulcers)

s withredistributing cushion

pressuresystem (unnamed) with pressure-

Hospital inpatients aged 16 or over;intact skin, requiring a pressure-relieving support


Alternating pressure mattress (KCITHERAPULSE) vs alternating

High risk patients in ICU Number of participants withincidence of pressure ulcers (allgrades of ulcers)

Alternating pressure air mattresselastic

Surgical, internal medicine orgeriatric hospital patients; at risk ofdeveloping pressure ulcer (Bradenscore <17)


Foam mattress overlay (IRIS 3000,

dimpled surface) vs foam mattress

Patients newly admitted to a skillednursing facility; free of pressureulcers but at risk (Braden score <18years)


pressure mattress (134inch diameter air cells, 3 minute

cycle) vs convoluted foam pad

in both groups were turned

Patients on medical –surgical units;relatively little skin breakdown;aged 19-91 years

Changes in skin conditionsgrades)

ombination of a standard 15cmcold foam mattress with a static air

a standard 15cm cold

Nursing home residents Incidence of pressure ulcers (grade2 and above)

325

Study length

e ulcers (all Discharge fromhospital or death

Number of participants withincidence of pressure ulcers (all

2 weeks follow-upafter discharge fromICU

Incidence of pressure ulcers (all unclear

Incidence of pressure ulcers (all 10-21 day follow-up

Changes in skin conditions (all 8-day follow-up

idence of pressure ulcers (grade 6 months follow-up

326

Table 8 – Classification systems used in the studies included in the review

ClassificationSystem

Grade 1

EPUAP/NPUAP Non-blanchableredness of intact skinIntact skin with non-blanchable erythema ofa localized area usuallyover a bonyprominence.Discoloration of theskin, warmth, edema,hardness or pain mayalso be present. Darklypigmented skin may nothave visible blanching.

Exton-Smith Persistent erythema

Stirling grade Discoloration of intactskin (light fingerpressure applied to thesite does not alter thediscoloration.

Torrance Redness to the skin –blanching occurs

Lowthian scale Discolorations of intactskin, including non-blanchable erythema,blue/purple and blackdiscoloration


Classification systems used in the studies included in the review

Grade 2 Grade 3 Grade 4

pigmented skin may not

Partial thickness skinloss or blister Partialthickness loss of dermispresenting as a shallowopen ulcer with a redpink wound bed,without slough. Mayalso present as anintact or open/rupturedserum-filled or sero-sanginous filled blister.

Full thickness skinloss (fat visible) Fullthickness tissue loss.Subcutaneous fat maybe visible but bone,tendon or muscle arenot exposed. Someslough may be present.May includeundermining andtunneling.

Full thickness tissueloss (muscle/bonevisible) Full thicknesstissue loss withexposed bone, tendonor muscle. Slough oreschar may be present.Often includeundermining andtunneling.

Localised blister Superficial sore Deep sore

Partial-thickness skinloss or damageinvolving epidermisand/or dermis

Full-thickness skin lossinvolving damage ornecrosis ofsubcutaneous tissuebut not extending tounderlying bone.

Full-thickness skin losswith extensivedestruction and tissuenecrosis extending tounderlying bone,tendon or joint capsule

2a redness to the skin –non-blanching occurs;2b superficial damageto the epidermis

Ulceration progressedthrough the dermis

ulceration extended intothe subcutaneous fat;

Partial thickness skinloss or damageinvolving the dermisand/or epidermis

Full thickness skin lossinvolving damager ornecrosis ofsubcutaneous tissue,but not through theunderlying fascia andnot extending to

Full thickness skin losswith extensivedestruction and tissuenecrosis extending tounderlying bone,tendon or joint capsule.

KCE Report 193S

Grade 5

Full thickness tissueloss (muscle/bone

Full thicknesstissue loss with

d bone, tendonor muscle. Slough oreschar may be present.Often includeundermining and

N/A

Extensive gangrenoussore.

thickness skin losswith extensivedestruction and tissuenecrosis extending tounderlying bone,tendon or joint capsule

N/A

ulceration extended intothe subcutaneous fat;

necrosis penetrating thedeep fascia andextending to muscle

ss skin losswith extensivedestruction and tissuenecrosis extending tounderlying bone,tendon or joint capsule.

N/A

KCE Report 193S

ClassificationSystem

Grade 1

Agency for Healthcare policy andresearch

Non-blanchingerythema or erythemanot resolving within 30minutes of pressurerelief. Epidermisremains intact.Reversible withintervention.

Dutch consensus Persistent erythema ofthe skin

Shea Involve the superficialbreakdown of theepidermis. Non-blanchable erythemawith edema, warmth,induration ordiscoloration. Reddiscoloration in lighterskin, blue or purple indarker skin


Grade 2 Grade 3 Grade 4

underlying bone,tendon or joint capsule

Partial thickness loss ofskin layers involvingepidermis and possiblypenetrating into but notthrough dermis. Maypresent as blisteringwith erythema and/orinduration; wound basemoist and pink; painful;free of necrotic tissue.

Full thickness tissueloss extending throughdermis to involvesubcutaneous tissue.Presents as shallowcrater unless coveredby eschar. May includenecrotic tissue,undermining, sinus tractformation, exudate,and/or infection. Woundbase is usually notpainful. If woundinvolves necrotic tissue,staging cannot beconfirmed, thereforeclassified as stage 4

Deep tissue destructionextending throughsubcutaneous tissue tofascia and may involvemuscle layers, jointand/or bone. Presentsas a deep crater. Mayinclude necrotic tissue,undermining, sinus tractformation, exudate,and/or infection. Woundbase is usually notpainful.

Blister formation Superficial (sub)cutaneous necrosis

Deep subcutaneousnecrosis.

Partial thicknessepidermal or skin lossthat extends throughthe epidermis andupper dermis

Full thickness deficitwith extension into thesubcutaneous tissue.Does not extend intothe fascia

Wound extends to themuscle, tendon, band or joint structures.Complications includeosteomyelitis,dislocations, orfractures. Also assessfor undermining.

327

Grade 5

eep tissue destructionextending throughsubcutaneous tissue to

ay involvemuscle layers, jointand/or bone. Presentsas a deep crater. Mayinclude necrotic tissue,undermining, sinus tractformation, exudate,and/or infection. Woundbase is usually not

N/A

eep subcutaneous N/A

Wound extends to themuscle, tendon, boneand or joint structures.Complications includeosteomyelitis,dislocations, orfractures. Also assessfor undermining.

N/A

328

7.4.1. Clinical evidence GRADE-tables

7.4.1.1. “Low-tech” constant low-pressure (CLP) supports

The Cochrane review compared standard foamSheepskin, static air-filled supports; water-filled supports; contoured or textured foam supports; geland alternative foam mattresses or overlays were considered to be lowstandard foam mattress is, and it can change over time, within countries, and even witin the Characteristics of included studies table, which we have put in our summary table. They have assumed that standard matwithin countries than between countries, and undertook subgroup analysis by country, although they did not pre

Standard foam hospital mattress compared with other “low

Table 10 – Clinical evidence profile: Constant lowQuality assessment


Inconsistency

Pressure ulcer incidence – Cubed foam mattress (COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40)

1Hofman(1994)

randomisedtrials

veryserious1



Pressure ulcer incidence – Softform mattress(COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40)

1Gray (1994) randomisedtrials

veryserious4


no seriousind


pressure (CLP) supports

The Cochrane review compared standard foam hospital mattresses with other low specification (low-tech), constant lowfilled supports; contoured or textured foam supports; gel-filled supports; bead

and alternative foam mattresses or overlays were considered to be low-tech CLP. However they point out that there is not an international definition of what astandard foam mattress is, and it can change over time, within countries, and even wit hin hospitals. If a description of the standard was given it was includedin the Characteristics of included studies table, which we have put in our summary table. They have assumed that standard mat

ween countries, and undertook subgroup analysis by country, although they did not pre -specify this.

Standard foam hospital mattress compared with other “low-tech” CLP

Clinical evidence profile: Constant low-pressure supports (CLP) vs standard foam mattresses (SFM) for pressure ulcer preventionQuality assessment No of patients


Constantlow-

pressuresupports

(CLP)

Standardfoam

mattresses(SFM)

Relative(95% CI)

(COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40) – grades 2



13/19(68.4%)

RR 0.34 (0.14to 0.85)

68.4%

(COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40) – grades 2-4 (no details of grading system)



none 6/90(6.7%)

27/80(33.8%)

RR 0.2 (0.09to 0.45)

33.8%

KCE Report 193S

tech), constant low-pressure (CLP) supports.filled supports; bead-filled supports; fibre-filled supports,

tech CLP. However they point out that there is not an international definition of what ahin hospitals. If a description of the standard was given it was included

in the Characteristics of included studies table, which we have put in our summary table. They have assumed that standard mat tresses are likely to vary lessspecify this.

am mattresses (SFM) for pressure ulcer preventionEffect Quality Importance

Relative(95% CI)

Absolute

grades 2-4 (Dutch consensus10)

RR 0.34 (0.14to 0.85)

452 fewerper 1000(from 103

fewer to 588fewer)

VERYLOW

Critical


fewer to 588fewer)

4 (no details of grading system)

RR 0.2 (0.09to 0.45)


fewer to 307fewer)

LOW

Critical


fewer to 308fewer)

KCE Report 193S

Quality assessment


Inconsistency

Pressure ulcer incidence – Cubed foam mattress (COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40)

1Hofman(1994)

randomisedtrials

veryserious1



Pressure ulcer incidence – Bead-filled mattress (BEAUFORT) vs standard hospital mattress

1Goldstone(1982)

randomisedtrials

Veryserious3



Pressure ulcer incidence – Water-filled mattress vs standard hospital mattress

1Andersen(1982)

randomisedtrials

veryserious5



Pressure ulcer incidence – Alternative foam pressure-reducing mattresses (CLINIFLOAT, OMNIFOAM, SOFTFORM, STM5, THERAREST, TRANSFOAM, VAPOURLUX) vs standard hospitalmattress – all grades (RCN and NPUAP grading system)12

2 (Collier,1996; Santy,

randomisedtrials

veryserious6






Constantlow-

pressuresupports

(CLP)

Standardfoam

mattresses(SFM)

Relative(95% CI)

(COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40) – all grades of ulcers (Dutch consensus



14/19(73.7%)

RR 0.48 (0.24to 0.96)

73.7%

(BEAUFORT) vs standard hospital mattress – all grades



21/43(48.8%)

RR 0.32 (0.14to 0.76)

48.8%

vs standard hospital mattress – all grades11



21/161(13%)

RR 0.35 (0.15to 0.79)

13%

reducing mattresses (CLINIFLOAT, OMNIFOAM, SOFTFORM, STM5, THERAREST, TRANSFOAM, VAPOURLUX) vs standard hospital



none 42/571(7.4%)

17/73(23.3%)

not pooled asCollier (1996)

329


Relative(95% CI)

Absolute

all grades of ulcers (Dutch consensus10)

RR 0.48 (0.24to 0.96)


fewer to 560fewer)

VERYLOW

Critical


fewer to 560fewer)

RR 0.32 (0.14to 0.76)


fewer to 420fewer)

VERYLOW

Critical


fewer to 420fewer)

RR 0.35 (0.15to 0.79)


fewer to 111fewer)

VERYLOW

Critical


fewer to 110fewer)

reducing mattresses (CLINIFLOAT, OMNIFOAM, SOFTFORM, STM5, THERAREST, TRANSFOAM, VAPOURLUX) vs standard hospital

not pooled asCollier (1996)

149 fewerper 1000

Critical

330

Quality assessment


Inconsistency

1994)

Pressure ulcer incidence – Hi-spec foam mattress/cushionSOFTFOAM, TRANSFOAM, KING’S FUND MATTRESS with a SPENCO or PROPAD mattress overlay

1 (Russell2003)

randomisedtrials

veryserious7



Comfort scores – very uncomfortable – pressure-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress

1 (Gray1994)

randomisedtrials

veryserious4



Comfort scores – uncomfortable – pressure-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress

1 (Gray1994)

randomisedtrials

veryserious4



Comfort scores – adequate – pressure-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress

1 (Gray1994)

randomisedtrials

veryserious4






Constantlow-

pressuresupports

(CLP)

Standardfoam

mattresses(SFM)

Relative(95% CI)

had 0 eventsbut 0.36 (0.22

to 0.59) forSanty (1994)13.3%

spec foam mattress/cushion – visco-polymer energy absorbing foam mattress (CONFORM-ED) vs standard mattress/cushion (KING’s FUND, LINKNURSE,OFTFOAM, TRANSFOAM, KING’S FUND MATTRESS with a SPENCO or PROPAD mattress overlay – all grades (Torrance scale)13



66/604(10.9%)

RR 0.78 (0.55to 1.11)

10.9%

reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress



none 0/90(0%)

0/80(0%)

not pooled



very seriousimprecision9

none 0/90(0%)

2/80(2.5%)

OR 0.12 (0.01to 1.91)

2.5%




none 6/90(6.7%)

44/80(55%)

RR 0.12 (0.05to 0.27)

KCE Report 193S


Relative(95% CI)

Absolute

had 0 eventsbut 0.36 (0.22

to 0.59) forSanty (1994)

(from 95fewer to 182

fewer)

LOW


fewer to 104fewer)

ED) vs standard mattress/cushion (KING’s FUND, LINKNURSE,

RR 0.78 (0.55to 1.11)


fewer to 12more)

VERYLOW

Critical


fewer to 12more)

not pooled not pooled LOW

Critical

OR 0.12 (0.01to 1.91)


fewer to 22more)

VERYLOW

Critical


fewer to 22more)

RR 0.12 (0.05to 0.27)

484 fewerper 1000

Critical

KCE Report 193S

Quality assessment


Inconsistency

Comfort scores – comfortable – pressure-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress

1 (Gray1994)

randomisedtrials

veryserious4



Comfort scores – very comfortable – pressure-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress

1 (Gray1994)

randomisedtrials

veryserious4



Comfort – Hi-spec foam mattress/cushion – visco-polymer energy absorbing foam mattress (CONFORMTRANSFOAM, KING’S FUND MATTRESS with a SPENCO or PROPAD mattress overlay

1 (Russell2003)

randomisedtrials

veryserious7



Length of stay in hospital (days) – Cubed foam mattress (COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40)

1 Hofman1994)

randomisedtrials

veryserious1



1 Unclear sequence generation and allocation concealment. No blinding. Unclear if incomplete outcome data was addressed.grades 2-4 ulcer outcome. Hofman (1994).2 Confidence interval crossed one MID point.




Constantlow-

pressuresupports

(CLP)

Standardfoam

mattresses(SFM)

Relative(95% CI)

55%




none 62/90(68.9%)

26/80(32.5%)

RR 2.12 (1.5to 2.99)

32.5%




0/80(0%)

OR 7.45 (2.2to 2

polymer energy absorbing foam mattress (CONFORM-ED) vs standard mattress/cushion (KING’s FUND, LINKNURSE, SOFTFOAM,TRANSFOAM, KING’S FUND MATTRESS with a SPENCO or PROPAD mattress overlay – all grades (Torrance scale)13



none 2.33 +/-0.98

N=323

2.46 +/-1.01N=383

(COMFORTEX DECUBE) vs standard hospital mattress (standard polypropylene SG40)


no serious very serious8 Median 21days

(range 5-64)

Median 23days (range

4-120)

Unclear sequence generation and allocation concealment. No blinding. Unclear if incomplete outcome data was addressed. Higher drop

331


Relative(95% CI)

Absolute

(from 402fewer to 523

fewer)

LOW


fewer to 523fewer)

RR 2.12 (1.5to 2.99)


more to 647more)

LOW

Critical


more to 647more)

OR 7.45 (2.2to 25.24)

- LOW

Critical

ED) vs standard mattress/cushion (KING’s FUND, LINKNURSE, SOFTFOAM,

- MD 0.13lower (0.28

lower to0.02 higher)

LOW

Critical

- Seefootnote8

VERYLOW

Important

Higher drop-out than event rate in CLP arm for

332

3 Inadequate sequence generation. Unclear allocation concealment and blinding. Incomplete outcome data was not addressed. Gol4 Unclear sequence generation, allocation concealment, blinding, addressing of incomplete o5 Unclear sequence generation, allocation concealment, blinding and addressing of incomplete outcome data. Andersen (1982).6 Unclear sequence generation, allocation concealment and addressing of iUnclear sequence generation, blinding and addressing of incomplete outcome data.7 Unclear allocation concealment. No blinding. Russell (2003).8 Data given as median and range so unable to analyse data in Revman.9 Confidence interval crossed both MID points.10 Dutch consensus grading system (1985): 0= normal skin; 1cutaneous necrosis.11 Bullae, black necrosis and skin defects were evidence of pressure ulcers.12 Collier (1996) used RCN grading and Santy (1994) used NPUAP 1989.13 Torrance scale, where blanching erythema represents a Torrrance grade I ulcer and non14 Limited number of events.


3 Inadequate sequence generation. Unclear allocation concealment and blinding. Incomplete outcome data was not addressed. Gol4 Unclear sequence generation, allocation concealment, blinding, addressing of incomplete outcome data and if groups similar at baseline (Gray 1994).5 Unclear sequence generation, allocation concealment, blinding and addressing of incomplete outcome data. Andersen (1982).6 Unclear sequence generation, allocation concealment and addressing of incomplete outcome data. No blinding. Unclear if groups were similar at baseline. Collier (1996).Unclear sequence generation, blinding and addressing of incomplete outcome data. Differential drop-out with higher drop-out in standard hospital mattress group7 Unclear allocation concealment. No blinding. Russell (2003).

ble to analyse data in Revman.

10 Dutch consensus grading system (1985): 0= normal skin; 1= persistent erythema of the skin; 2= blister formation; 3= superficial (sub

11 Bullae, black necrosis and skin defects were evidence of pressure ulcers.994) used NPUAP 1989.

13 Torrance scale, where blanching erythema represents a Torrrance grade I ulcer and non-blanching erythema represents a Torrance grade II ulcer.

KCE Report 193S

3 Inadequate sequence generation. Unclear allocation concealment and blinding. Incomplete outcome data was not addressed. Gol dstone (1982).utcome data and if groups similar at baseline (Gray 1994).

ncomplete outcome data. No blinding. Unclear if groups were similar at baseline. Collier (1996).out in standard hospital mattress group.Santy (1994).

= persistent erythema of the skin; 2= blister formation; 3= superficial (sub-cutaneous necrosis); 4= deep sub-

blanching erythema represents a Torrance grade II ulcer.

KCE Report 193S

Table 10 – Clinical evidence profile: Alternative foamQuality assessment


Inconsistency

Pressure ulcer incidence Various alternatives (pooled) – all grades of ulcer

5 (Collier 1996;Gray 1994;Hofman 1994;Russell 2003;Santy 1994)

randomisedtrials

veryserious1

very serious2

Pressure ulcer incidence (pooled) grades 2+ ulcer6 – pressure-

2 (Gray 1994;Hofman 1994)

randomisedtrials

veryserious1

No serious

1 Unclear sequence generation for three studies (Collier 1996, Gray 19941994, Hofman 2003 and Santy, 1994) No blinding in three studies (Collier 1996, Hofman 1994, Russell 2003) and unclear blindinUnclear if incomplete outcome data addressed in four studies (Collier 1996, Gray 1994,and Gray 1994) Different timing of outcome assessment in two studies (Collier 1996 and Gray 1994)mattrtress group (Santy 1994). Highr drop-out than event rate for incidence of pressure ulcers all grades and 2 and above (Hofman 1994)2 I2 = 77%, p=0.0043 I2 =84%, p=0.0024 Confidence interval crossed one MID point.5 Collier (1996) used RCN grading system, Gray (1994) had no details of grading system, Hofman (1994) used Dutch consensus, Russell (2003) used the Torrance scSanty (1994) used NPUAP 1989 grading system.


Clinical evidence profile: Alternative foam mattress vs standard foam mattress for pressure ulcer preventionQuality assessment No of patients


Alternativefoam

mattress

Standardfoam

mattress

Relative(95% CI)

all grades of ulcer5


Serious4 none 102/1240(8.2%)

124/776(16%)

RR 0.43(0.24 to0.76)

26.6%

-reducing foam mattress (SOFTFOAM) vs standard 130mm NHS foam mattress



none 10/107(9.3%)

40/99(40.4%)

RR 0.24(0.13 t0.45)

51.1%

1 Unclear sequence generation for three studies (Collier 1996, Gray 1994, Hofman 2003 and Santy 2004) Unclear allocation concealment in four studies (Collier 1996, Gray1994, Hofman 2003 and Santy, 1994) No blinding in three studies (Collier 1996, Hofman 1994, Russell 2003) and unclear blindin g in two studies (Gray 1994 and Santy 1994)Unclear if incomplete outcome data addressed in four studies (Collier 1996, Gray 1994, Hofman 1994 and Santy 1994) Unclear if similar at baseline in two studies (Collier 1996and Gray 1994) Different timing of outcome assessment in two studies (Collier 1996 and Gray 1994) Higher differential drop-out with higher rate in the standard hospital

out than event rate for incidence of pressure ulcers all grades and 2 and above (Hofman 1994)

ystem, Gray (1994) had no details of grading system, Hofman (1994) used Dutch consensus, Russell (2003) used the Torrance sc

333

mattress vs standard foam mattress for pressure ulcer preventionEffect Quality Importance

Relative(95% CI)

Absolute

RR 0.43(0.24 to0.76)


fewer)

VERYLOW

Critical


fewer

RR 0.24(0.13 to0.45)

307 fewer per1000 (from

222 fewer to352 fewer)

LOW

Critical



) Unclear allocation concealment in four studies (Collier 1996, Grayg in two studies (Gray 1994 and Santy 1994)

Hofman 1994 and Santy 1994) Unclear if similar at baseline in two studies (Collier 1996out with higher rate in the standard hospital

out than event rate for incidence of pressure ulcers all grades and 2 and above (Hofman 1994)

ystem, Gray (1994) had no details of grading system, Hofman (1994) used Dutch consensus, Russell (2003) used the Torrance sc ale,

334

Comparisons between alternative foam mattresses

Table 11 – Comparisons between alternative foam supports for pressure ulcer prevention

Quality assessment

No ofstudies

Design Risk ofbias


Pressure ulcer incidence – – Pressure-redistributing mattresses (CLINIFLOAT, OMNIFOAM, THERAREST, TRANSFOAM, VAPERM) vs standard nhs foam mattress (REYLON 150mm)grades (NPUAP)8

1Santy(1994)

randomisedtrials

veryserious1



Pressure ulcer incidence – Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000)

1(Vyhlidal(1997)

randomisedtrials

serious2 no seriousinconsistency


Pressure ulcer incidence – Solid foam overlay vs convoluted foam

1Kemp(1994)

randomisedtrials

Veryserious7




Comparisons between alternative foam mattresses

alternative foam supports for pressure ulcer prevention



Comparisonsbetween

alternative foamsupports

Control Relative(95% CI)

redistributing mattresses (CLINIFLOAT, OMNIFOAM, THERAREST, TRANSFOAM, VAPERM) vs standard nhs foam mattress (REYLON 150mm)



none 42/441(9.5%)

17/64(26.6%)

RR 0.36(0.22 to

0.59)

26.6%

Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000) – all ulcer grades9



12/20(60%)

RR 0.42(0.18 to

0.96)

60%

vs convoluted foam overlay – all ulcer grades (NPUAP)10



21/45(46.7%)

RR 0.66(0.37 to

1.16)

46.7%

KCE Report 193S


Relative(95% CI)

Absolute

redistributing mattresses (CLINIFLOAT, OMNIFOAM, THERAREST, TRANSFOAM, VAPERM) vs standard nhs foam mattress (REYLON 150mm) – all ulcer

RR 0.36(0.22 to

0.59)



LOW

Critical



RR 0.42(0.18 to

0.96)


fewer)

VERYLOW

Critical


fewer)

RR 0.66(0.37 to

1.16)


294 fewer to75 more)

LOW

Critical


294 fewer to

KCE Report 193S

Quality assessment

No ofstudies

Design Risk ofbias


Pressure ulcer incidence – pressure-reducing foam mattress (TRANSFOAM) vs pressure=reducing foam mattress (TRANSFOAMWAVE)

1Gray(1998)

randomisedtrials

Veryserious4



Pressure ulcer incidence – Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000)

1(Vyhlidal(1997)

randomisedtrials



Time to pressure ulcer development (days) – Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000)

1(Vyhlidal(1997)

randomisedtrials



1 Unclear sequence generation, allocation concealment, blinding and addressing of incomplete outcome data. Santy (1994).2 Unclear allocation concealment and blinding. Baseline differences. Vyhlidal (1997).3 Confidence interval crossed one MID.4 Unclear sequence generation and addressing of incomplete outcome data. Baseline data were provided for the treatment arm on5 Confidence interval crosse both MIDs and limited number of events.6 Not enough data to analyse in Revman.7 Unclear allocation concealment, blinding and baseline differences and did not address incomplete outcome data. Kemp (1993).




Comparisonsbetween

alternative foamsupports


reducing foam mattress (TRANSFOAM) vs pressure=reducing foam mattress (TRANSFOAMWAVE) – all ulcer grades


very serious5 none 1/50(2%)

1/50(2%)

RR 1 (to 15.55)

2%

Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000) – – – grades 2+ ulcer9



8/20(40%)

RR 0.38(0.12 to

1.21)

40%

Foam mattress replacement (MAXIFLOAT) vs foam mattress overlay (IRIS 3000) –


no serious very serious6 9.2 days 6.5days

P=0.3288

allocation concealment, blinding and addressing of incomplete outcome data. Santy (1994).2 Unclear allocation concealment and blinding. Baseline differences. Vyhlidal (1997).

4 Unclear sequence generation and addressing of incomplete outcome data. Baseline data were provided for the treatment arm on ly. Gray (1998).5 Confidence interval crosse both MIDs and limited number of events.

7 Unclear allocation concealment, blinding and baseline differences and did not address incomplete outcome data. Kemp (1993).

335


Relative(95% CI)

Absolute

75 more)

all ulcer grades11

RR 1 (0.06to 15.55)


more)

VERYLOW

Critical


more)

RR 0.38(0.12 to

1.21)



VERYLOW

Critical



P=0.3288 - VERYLOW

Important

ly. Gray (1998).

336

8 NPUAP 1989 grading system.9 Unclear grading system name, stage 0= no redness or breakdown; stage 1= erythema only,break in skin such as blisters, or abrasions; stage 3= break in skin exposing subcutaneous tissue; stage 4= break in skin extexposing muscle or bone.10 NPUAP1989.11 no details of grading system.

Comparisons between “low-tech” constant low

Table 12 – Comparisons between CLP supports for pressure ulcer prevention

Quality assessment


Inconsistency

Pressure ulcer incidence – Constant low pressure mattress (CARITAL OPTIMA) v

1Takala(1996)

randomisedtrials

Veryserious1


Pressure ulcer incidence – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1Cooper(1998)

randomisedtrials


Pressure ulcer incidence – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1Cooper(1998)

randomisedtrials



9 Unclear grading system name, stage 0= no redness or breakdown; stage 1= erythema only, redness does not disappear for 24 hours after pressure is relieved; stage 2=break in skin such as blisters, or abrasions; stage 3= break in skin exposing subcutaneous tissue; stage 4= break in skin ext ending through tissue and subcutaneous layers,

tech” constant low-pressure supports

Comparisons between CLP supports for pressure ulcer prevention



Comparisonsbetween CLP

supports


Constant low pressure mattress (CARITAL OPTIMA) vs standard foam mattress (10cm thick foam density 35kg/m3)– – all grades of ulcers (Shea)



7/19(36.8%)

RR 0.06(0 to0.99)

36.8%

dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO) – all grades of ulcers (Stirling grade)17


veryserious4

none 3/41(7.3%)

5/43(11.6%)

RR 0.63(0.16 to

2.47)

11.6%

dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO) grades 2+ ulcers (Stirling grade)17


veryserious4

none 1/41(2.4%)

0/43(0%)

RR 3.14(0.13 to75.02)0%

KCE Report 193S

redness does not disappear for 24 hours after pressure is relieved; stage 2=ending through tissue and subcutaneous layers,


Relative(95% CI)

Absolute

all grades of ulcers (Shea)17

RR 0.06(0 to0.99)


fewer to 368fewer)

VERY LOW

Critical


fewer to 368fewer)

RR 0.63(0.16 to

2.47)

43 fewer per1000 (from98 fewer to171 more)

VERY LOW

Critical


RR 3.14(0.13 to75.02)

- VERY LOW

Critical

-

KCE Report 193S

Quality assessment


Inconsistency

Pressure ulcer incidence – Gel mattress vs air-filled overlay (SOFCARE)

1Lazzara(1991)

randomisedtrials

veryserious5


Pressure ulcer incidence – Gel mattress vs air-filled overlay (SOFCARE)

1Lazzara(1991)

randomisedtrials

veryserious5


Pressure ulcer incidence – Static air mattress (GAY MAR SOFCARE)

1Sideranko(1992)

randomisedtrials

veryserious6


Pressure ulcer incidence – Foam overlay vs Silicore overlay (SPENCO)

1Stapleton(1986)

randomisedtrials

veryserious7






supports


(SOFCARE) – all grades of ulcers (NPUAP)17


veryserious4

none 8/33(24.2%)

10/33(30.3%)

RR 0.8(0.36 to

1.77)

15.2%

SOFCARE) – grades 2+ ulcers (NPUAP)17


veryserious4

none 4/33(12.1%)

5/33(15.2%)

RR 0.8(0.24 to

2.72)

15.2%

(GAY MAR SOFCARE) vs water mattress (LOTUS PXM 3666) – all grades of ulcers (grading system not reported)


veryserious4

none 1/20(5%)

2/17(11.8%)

RR 0.43(0.04 to

4.29)

11.8%

(SPENCO) – grades 2 and above17


veryserious4

none 14/34(41.2%)

12/34(35.3%)

RR 1.17(0.64 to

2.14)

29.4%

337


Relative(95% CI)

Absolute

RR 0.8(0.36 to

1.77)



VERY LOW

Critical


RR 0.8(0.24 to

2.72)

23 fewer per1000 (from2 fewer to35 fewer)

VERY LOW

Critical



all grades of ulcers (grading system not reported)17

RR 0.43(0.04 to

4.29)



VERY LOW

Critical



RR 1.17(0.64 to

2.14)

60 more per1000 (from


VERY LOW

Critical


127 fewer to

338

Quality assessment


Inconsistency

Pressure ulcer incidence – Australian medical sheepskin vs no sheepskin (all grades of ulcer)

3 (Jolley2004;McGowan2000;Mistiaen2009)

randomisedtrials

Veryserious8

serious9

Pressure ulcer incidence – Australian medical sheepskin vs no sheepskin (grade 2 + ulcers )

3 (Jolley2004;McGowan2000;Mistiaen2009)

randomisedtrials

Veryserious8


Pressure ulcer incidence – static air overlay (and cold foam mattress) vs cold foam mattress

1 Van Leen(2011)

randomisedtrials

veryserious12


Comfort – Australian medical sheepskin vs no sheepskin

1 Jolley(2004)

randomisedtrials

Veryserious8


Withdrawal due to discomfort

1 McGowan randomised serious8 no serious





supports


heepskin vs no sheepskin (all grades of ulcer)17


no serious none 59/644(9.2%)

120/637(18.8%)

RR 0.48(0.31 to

0.74)

16.6%

heepskin vs no sheepskin (grade 2 + ulcers )17



33/637(5.2%)

RR 0.56(0.32 to

0.97)

3.5%

static air overlay (and cold foam mattress) vs cold foam mattress– grade 2+ ulcers17



7/36(19.4%)

RR 0.(0.06 to

1.22)

19.4%


no serious very serious13 - - -

Withdrawal due to discomfort – Australian medical sheepskin vs no sheepskin

no serious no serious very serious14 - - -

KCE Report 193S


Relative(95% CI)

Absolute

402 more)

RR 0.48(0.31 to

0.74)


VERY LOW

Critical


RR 0.56(0.32 to

0.97)


VERY LOW

Critical


RR 0.27(0.06 to

1.22)


fewer to 43more)

VERY LOW

Critical


fewer to 43more)

Seefootnote13

VERY LOW

Critical

See Critical

KCE Report 193S

Quality assessment


Inconsistency

(2000) trials inconsistency

Patient acceptability – very uncomfortable – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1 Cooper(1998)

randomisedtrials


Patient acceptability – uncomfortable – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1 Cooper(1998)

randomisedtrials


Patient acceptability – adequate – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1 Cooper(1998)

randomisedtrials


Patient acceptability – comfortable – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)

1 Cooper(1998)

randomisedtrials


Patient acceptability – very comfortable – dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)





supports


indirectness

dry flotation mattress (SOFFLEX) vs dry flotation mattress (ROHO)


no serious none 0/41(0%)

0/43(0%)

Notpooled as

eventrate iszero



Serious2 none 0/41(0%)

5/43(11.6%)

OR 0.13(0.02 to

0.77)

11.6%



veryserious4

none 4/41(9.8%)

4/43(9.3%)

RR 1.05(0.28 t

3.92)

9.3%



veryserious4

none 24/41(58.5%)

24/43(55.8%)

RR 1.05(0.72 to

1.52)

55.8%


339


Relative(95% CI)

Absolute

footnote14 VERY LOW

Notpooled as

eventrate iszero

Not pooledas event

rate is zero

MODERATE

Critical

OR 0.13(0.02 to

0.77)


LOW

Critical


RR 1.05(0.28 to

3.92)


VERY LOW

Critical


RR 1.05(0.72 to

1.52)



VERY LOW

Critical



340

Quality assessment


Inconsistency

1 Cooper(1998)

randomisedtrials


Time to onset of first ulcer – Australian medical sheepskin vs no sheepskin

1 Jolley(2004)

randomisedtrials


Time to onset of first ulcer – Australian medical sheepskin vs no sheepskin

1 Mistiaen(2010E)

randomisedtrials


1 Unclear sequence generation but may have been block randomised and some outcome assessors may have been blinded but uncleadrop out than event rate for incidence of pressure ulcers.2 Confidence interval crossed one MID.3 Unclear blinding. Higher drop out than event rate for incidence of all grades of pressure ulcers and grades 2 and above pressure ulcers4 Confidence interval crossed both MIDs.5 Unclear allocation concealment, blinding and addressing incomplete outcome data. Lazzara (1991).6 Unclear sequence generation, allocation concealment, blinding, addressing incomplete outcome data, similarity at baseline.7 Unclear sequence generation, allocation concealment and blinding. Stapleton (1986).8 Unclear sequence generation (Jolley 2004), unclear allocation concealment (McGowan 2000) and no blinding (Jolley 2004, McGoUnclear addressing of incomplete outcome data (Mistainen 2009, 2010) and no addressing (Jolley 2004). Unclear if baseline difevent rate for incidence of all grades of pressure ulcers and grades 2 and above p9 I2 = 52%, p=0.12.10 Confidence interval crossed one MID.11 Ethical issues of not using repositioning. Limited details of sequence generation and allocation concealment. No details of bthan event rate for incidence of pressure ulcers .Van leen (2011)12 Comfort data not given for both groups. 10 patients in the sheepskin group complained about its comfort (too hot, 6; sensitivrequested its removal.13Study did not give details of comfort in both groups. Six patients in the experimental group withdrew before completion of dirritation, was too hot or uncomfortable.





supports



veryserious4

none 13/41(31.7%)

10/43(23.3%)

RR 1.36(0.67 to

2.76)

23.3%

Australian medical sheepskin vs no sheepskin


no serious Serious15 - - HR 0.39(95% CI0.22 to0.69)

Australian medical sheepskin vs no sheepskin


no serious very serious16 12 days 9 days -

1 Unclear sequence generation but may have been block randomised and some outcome assessors may have been blinded but uncleafor incidence of pressure ulcers. Takala (1996).

Higher drop out than event rate for incidence of all grades of pressure ulcers and grades 2 and above pressure ulcers

5 Unclear allocation concealment, blinding and addressing incomplete outcome data. Lazzara (1991).6 Unclear sequence generation, allocation concealment, blinding, addressing incomplete outcome data, similarity at baseline. Sideranko (1992).7 Unclear sequence generation, allocation concealment and blinding. Stapleton (1986).8 Unclear sequence generation (Jolley 2004), unclear allocation concealment (McGowan 2000) and no blinding (Jolley 2004, McGo wan 2000 and Mistainen 2009, 201Unclear addressing of incomplete outcome data (Mistainen 2009, 2010) and no addressing (Jolley 2004). Unclear if baseline dif ferences (Jolley 2004).event rate for incidence of all grades of pressure ulcers and grades 2 and above pressure ulcers(Jolley 2004, Mistiaen 2009, 2010)

Ethical issues of not using repositioning. Limited details of sequence generation and allocation concealment. No details of b linding of outcomVan leen (2011)

Comfort data not given for both groups. 10 patients in the sheepskin group complained about its comfort (too hot, 6; sensitiv e to the wool surface, 2; unco

Study did not give details of comfort in both groups. Six patients in the experimental group withdrew before completion of d ata collection because the sheepskin caused an

KCE Report 193S


Relative(95% CI)

Absolute

RR 1.36(0.67 to

2.76)


VERY LOW

Critical


HR 0.39(95% CI0.22 to0.69)

P<0.001 LOW

Important

- VERY LOW

Important

1 Unclear sequence generation but may have been block randomised and some outcome assessors may have been blinded but unclea , no allocation concealementr. Higher

Higher drop out than event rate for incidence of all grades of pressure ulcers and grades 2 and above pressure ulcers Cooper (1998).

anko (1992).

wan 2000 and Mistainen 2009, 2010).ferences (Jolley 2004). Higher drop out than

linding of outcome assessors. ). Higher drop out

e to the wool surface, 2; uncomfortable, 2) and

ata collection because the sheepskin caused an

KCE Report 193S

14 No data given for each arm but HR presented. Kaplan15 Not enough data to analyse in Revman.16 Takala (1996) used Shea 1975 grading system; Cooper (1998) used the Stirling grading system; Lazthe grading system; Stapleton (1986) adapted the grading system from Kenedi et al (1976) bed sore biomechanics study, where cbreak in skin (no crater) and category C= a break in skin (with crater) and category D= blackened tissue; Jolley (2004) and McGowan (2000) usedCare and Policy Research grading system; Mistiaen (2009, 2010) and Van Leen 2011 used the EPUAP grading

7.4.1.2. “High-tech” pressure supports

This section outlines three main groups of supports; alternating pressure (AP) supports, low

Alternating-pressure compared with constant low

Table 13 – Alternating-pressure vs standard foam mattress for pressure ulcer prevention

Quality assessment


Inconsistency

Pressure ulcer incidence – alternating air mattress/overlay vs standard foam mattress

2 (Andersen1982 Sanada2003)

randomisedtrials

veryserious1


Pressure ulcer incidence – alternating air mattress vs standard foam mattress

1 Sanada(2003)

randomisedtrials

veryserious1



No data given for each arm but HR presented. Kaplan-Meier survival curves used (p<0.001, log-rank test).

Takala (1996) used Shea 1975 grading system; Cooper (1998) used the Stirling grading system; Lazzara (1991) used NPUAP 1989 system; Sideranko (1992)did not reportthe grading system; Stapleton (1986) adapted the grading system from Kenedi et al (1976) bed sore biomechanics study, where c ategory A= superficial/blister, category B = a

crater) and category C= a break in skin (with crater) and category D= blackened tissue; Jolley (2004) and McGowan (2000) usedCare and Policy Research grading system; Mistiaen (2009, 2010) and Van Leen 2011 used the EPUAP grading system.

This section outlines three main groups of supports; alternating pressure (AP) supports, low-air loss beds and air-fluidised low beds.

pressure compared with constant low pressure

ssure vs standard foam mattress for pressure ulcer prevention



Alternating-pressure

Standardfoam

mattress

Relative(95% CI)

alternating air mattress/overlay vs standard foam mattress – all grades of ulcer3



none 13/221(5.9%)

31/188(16.5%)

RR 0.31(0.17 to

0.58)

25%

alternating air mattress vs standard foam mattress – grade 2+ ulcers3



6/27(22.2%)

RR 0.41(0.14 to

1.22)

22.2%

341

zara (1991) used NPUAP 1989 system; Sideranko (1992)did not reportategory A= superficial/blister, category B = a

crater) and category C= a break in skin (with crater) and category D= blackened tissue; Jolley (2004) and McGowan (2000) used the US Agency for Health

fluidised low beds.


Relative(95% CI)

Absolute

RR 0.31(0.17 to

0.58)


fewer)

LOW

Critical



RR 0.41(0.14 to

1.22)



VERYLOW

Critical



342

1 Unclear sequence generation, allocation concealment, blinding and addressing incomplete outcome data (Andersen 1982). Unclear blinding and no addressing ofoutcome data . Higher drop out than event rate for incidence of all grades of pressure ulcers and grades 2 and above pr2 Confidence interval crossed one MID point.3 Andersen 1982 used the classification of Bullae, black necrosis, and skin defects as evidence of pressure sores. Sanada (2

Alternating-pressure compared with constant low

Table 14 – Alternating-pressure (AP) vs constant low

Quality assessment

No of studies Design Risk of bias Inconsistency

Pressure ulcer incidence – Alternating pressure (all studies meta

11 (Conine 1990;Daechsel 1985;Stapleton 1986;Whitney 1984;;Gebhardt 1996;Andersen 1982; Price1999; Sideranko1992; Vanderwee,2005; Malbrain, 2010,Cavicchioli 2007)

randomisedtrials

veryserious1,2,3,4,5


Pressure ulcer incidence – Alternating pressure (various) vs Constant low pressuredevices– all grades of ulcer12

1 Gebhardt (1996) randomisedtrials

veryserious1


Pressure ulcer incidence – Alternating pressure vs Silicore or foam overlay

4 (Conine 1990;Daechsel 1985;Stapleton 1986;

randomisedtrials

veryserious2



llocation concealment, blinding and addressing incomplete outcome data (Andersen 1982). Unclear blinding and no addressing ofHigher drop out than event rate for incidence of all grades of pressure ulcers and grades 2 and above pr essure ulcers

3 Andersen 1982 used the classification of Bullae, black necrosis, and skin defects as evidence of pressure sores. Sanada (2 003) used NPUAP 1989 grading system.

re compared with constant low pressure

pressure (AP) vs constant low-pressure for pressure ulcer prevention




(AP)

Constantlow-

pressure

(all studies meta-analysed all had various types of alternating pressureand various types of constant low



serious6 none 125/785(15.9%)

170/837(20.3%)

23.1%

Constant low pressure (various) – one study which included patients with various types of alternating pres




none 15/115(13%)

39/115(33.9%)

33.9%

vs Silicore or foam overlay11 – all grades of ulcer and all types of patients12




81/186(43.5%)

KCE Report 193S

llocation concealment, blinding and addressing incomplete outcome data (Andersen 1982). Unclear blinding and no addressing of incompleteessure ulcers (Sanada 2003).

003) used NPUAP 1989 grading system.


Relative(95%CI)

Absolute

constant low-pressure – all grades of ulcer12

RR 0.85(0.65 to1.11)

30 fewerper 1000(from 71fewer to22 more)

VERYLOW

Critical


one study which included patients with various types of alternating pressure and constant low pressure

RR 0.38(0.22 to0.66)


265 fewer)

LOW

Critical


264 fewer)

RR 0.91(0.72 to1.16)


VERY

Critical

KCE Report 193S

Quality assessment


Whitney 1984)

Pressure ulcer incidence – Alternating pressure vs water or static air mattress

3 (Andersen 1982;Price 1999;;Sideranko 1992)

randomisedtrials

veryserious3

no seriousinconsisten

Pressure ulcer incidence – Alternating pressure setting on mattress (DUO 2)

1Cavicchioli (2007) randomisedtrials


Pressure ulcer incidence – Alternating pressure air mattress (ALPHA

1Vanderwee (2005) randomisedtrials

veryserious5






(AP)

Constantlow-

pressure

31.6%

vs water or static air mattress – all grades of ulcer12



veryserious7

none 13/226(5.8%)

12/232(5.2%)

5%

Alternating pressure setting on mattress (DUO 2) vs continuous low pressure mattress setting on mattress (DUO 2) – all grades of ulcer



veryserious8

none 2/69(2.9%)

1/71(1.4%)

1.4%

Alternating pressure air mattress (ALPHA-X-CELL) vs visco-elastic foam mattress (TEMPUR) – all grades of ulcer 12



veryserious7

none 34/222(15.3%)

35/225(15.6%)

15.6%

343


Relative(95%CI)

Absolute

fewer to70 more)

LOW


RR 1.31(0.51 to3.35)


122 more)

VERYLOW

Critical


117 more)


RR 2.06(0.19 to22.18)


298 more)

VERYLOW

Critical


297 more)

RR 0.98(0.64 to1.52)


VERYLOW

Critical

3 fewer

344

Quality assessment


Pressure ulcer incidence – Alternating pressure mattress (NIMBUS 3) vs dry flotation

1Malbrain (2010) randomisedtrials

veryserious9


Pressure ulcer incidence – Alternating pressure mattress vs Silicore

2 (Stapleton 1986;Whitney 1984)

randomisedtrials

veryserious2


Pressure ulcer incidence – Alternating pressure mattress vs Silicore

2 (Conine 1990;Daechsel 1985)

randomisedtrials

veryserious2


Pressure ulcer incidence – grade 2+ ulcers12

6 (Cavicchioli 2007; randomised very no serious





(AP)

Constantlow-

pressure

(NIMBUS 3) vs dry flotation mattress overlay (ROHO) – all grades of ulcer12



veryserious7

none 2/8(25%)

2/8(25%)

0%

vs Silicore – patients not singularly with chronic neurological conditions – all grades of ulcer



veryserious7

none 16/57(28.1%)

32/94(34%)

30.7%

vs Silicore overlay – patients with chronic neurological conditions – all grades of ulcer 12




49/92(53.3%)

42.1%

no serious no serious serious6 none 45/394 70/432

KCE Report 193S


Relative(95%CI)

Absolute

per 1000(from 56fewer to81 more)

RR 1(0.18 to5.46)


1000more)

VERYLOW

Critical

-


RR 0.89(0.54 to1.47)


160 more)

VERYLOW

Critical


144 more)

RR 0.92(0.7 to1.22)


117 more)

VERYLOW

Critical


RR 0.80 34 fewer Critical

KCE Report 193S

Quality assessment


Stapleton 1986;Gebhardt, 1996;Price, 1999;Vanderwee,2005;Malbrain 2010)

trials serious2 inconsistency

Drop out due to discomfort – Alternating pressure overlay vs Silicore

1Conine (1990) randomisedtrials

veryserious2


Comfort rating at 14 days dynamic flotation mattress (NIMBUS 2)

1Price (1999) randomisedtrials

veryserious3


Length of stay in hospital – Alternating pressure setting on mattress (DUO 2) vs continuous low pressure setting on mattress (DUO 2)

1Caviccioli (2007) randomisedtrials

veryserious5


1 No adequate sequence generation, allocation concealment and unclear blinding.2 Unclear sequence generation (Conine 1990, Daeschel 1985, Stapleton 1Unclear blinding (Daeschel 1985, Stapleton 1986, Whitney 1984). Unclear addressing of incomplete outcome data1985, Whitney 1984).3 Unclear sequence generation (Anderson 1982, Sideranko 1992). Unclear allocation concealment (Anderson 1982, Price 1999, Sid1982, Sideranko 1992) and no blinding (Price 1999). Unclearrate for incidence of all grades of pressure ulcers and comfort rating at 14 days. (Price 1999)4 Unclear sequence generation, allocation concealment. Differences between groups at baseline. Cavicchioli (2007).





(AP)

Constantlow-

pressure

inconsistency indirectness (11.4%) (16.9%)

14%

vs Silicore overlay



veryserious7

none 19/93(20.4%)

17/94(18.1%)

0%

(NIMBUS 2) and alternating pressure overlay cushion vs low pressure inflatable mattress (REPOSE SYSTEM)



veryserious7

none 26 24

Alternating pressure setting on mattress (DUO 2) vs continuous low pressure setting on mattress (DUO 2)



no serious very serious10 - -

1 No adequate sequence generation, allocation concealment and unclear blinding. Higher drop out than event rate for incidence of pressure ulcers (Gebhardt 1996)2 Unclear sequence generation (Conine 1990, Daeschel 1985, Stapleton 1986, Whitney 1984). Unclear allocation concealment (Conine 1990, Daeschel 1985, Stapleton 1986).Unclear blinding (Daeschel 1985, Stapleton 1986, Whitney 1984). Unclear addressing of incomplete outcome data (Daeschel, 1985)

3 Unclear sequence generation (Anderson 1982, Sideranko 1992). Unclear allocation concealment (Anderson 1982, Price 1999, Sid1982, Sideranko 1992) and no blinding (Price 1999). Unclear addressing of incomplete outcome data (Anderson 1982, Price 1999, Sideranko 1992).rate for incidence of all grades of pressure ulcers and comfort rating at 14 days. (Price 1999)

ent. Differences between groups at baseline. Cavicchioli (2007).

345


Relative(95%CI)

Absolute

(0.58 to1.11)

per 1000(from 71fewer to19 more)

VERYLOW


RR 1.13(0.63 to2.03)


186 more)

VERYLOW

Critical

-

(REPOSE SYSTEM) and cushion

- MD 7lower(19.01

lower to5.01

higher)

VERYLOW

Critical

- Seefootnote10

VERYLOW

Important

Higher drop out than event rate for incidence of pressure ulcers (Gebhardt 1996)986, Whitney 1984). Unclear allocation concealment (Conine 1990, Daeschel 1985, Stapleton 1986).

(Daeschel, 1985). Unclear baseline differences (Daeschel

3 Unclear sequence generation (Anderson 1982, Sideranko 1992). Unclear allocation concealment (Anderson 1982, Price 1999, Sid eranko 1992). Unclear blinding (Andersonaddressing of incomplete outcome data (Anderson 1982, Price 1999, Sideranko 1992). Higher drop out than event

346

5 Unclear blinding and addressing of incomplete outcome data. Vanderwee (2005).6 Confidence interval crossed one MID.7 Confidence interval crossed both MIDs.8 Confidence interval crossed both MIDs and limited number of events.9 Baseline difference; allocation concealment unclear; single blinding.10 There was no data presented, but the authors state that there was no difference in length of stay related to pressure ulcethe intervention or control mattresses.11 Conine (1990) and Daeschel (1985) included patients with chronic neurological conditions, which we identified as a group tincluded these studies together and in a subgroup test, no subgroup differences were found so the results are presented togetneurological conditions are also presented separately.12 Conine (1990) and Daechsel (1985) used Exton-Smith scale; Stapleton (1986) adapted the grading system from Kenedi et al (1976) bed sore biomechanics study, wherecategory A= superficial/blister, category B = a break in skin (no crater) and category C= a break in skin (withsystem where stage 0 = no redness or skin breakdown; stage 1= skin redness, fades in 15 minutes or less; stage II inflammatioless than one hour; stage III= inflammation of the skin fading time exceeds one hour; stage IV= skin break with redness of surrounding skin, rednhour; Gebhardt (1996) used a grading system by Bliss (1966) grade 1= persistent erythema; grade 2= epidedermis ; grade 4=cavity to subcutaneous tissue or deeper; Andersen (1982) used bullae, black necrosis and skin defects as eviHofman 1994 scale where 0=normal skin, 1= persistent erythema of the skin; 2= blister formation; 3= superficial subcutaneous necrosis; 4= deeSideranko (1992) did not report grading system; Vanderwee (2005) did not report grading system but grade 1EPUAP and Cavicchioli (2007) used EPUAP 2007.


5 Unclear blinding and addressing of incomplete outcome data. Vanderwee (2005).

d both MIDs and limited number of events.aseline difference; allocation concealment unclear; single blinding.(Malbrain, 2010)

10 There was no data presented, but the authors state that there was no difference in length of stay related to pressure ulce r development among high

11 Conine (1990) and Daeschel (1985) included patients with chronic neurological conditions, which we identified as a group t o be stratified. However the Cochrane reviincluded these studies together and in a subgroup test, no subgroup differences were found so the results are presented toget her. The results of those with and without chronic

Smith scale; Stapleton (1986) adapted the grading system from Kenedi et al (1976) bed sore biomechanics study, wherecategory A= superficial/blister, category B = a break in skin (no crater) and category C= a break in skin (with crater) and category D= blackened tissue; Whitney (1984) used asystem where stage 0 = no redness or skin breakdown; stage 1= skin redness, fades in 15 minutes or less; stage II inflammatio n of the skin, fading time exceeds 15 minutes,

stage III= inflammation of the skin fading time exceeds one hour; stage IV= skin break with redness of surrounding skin, rednhour; Gebhardt (1996) used a grading system by Bliss (1966) grade 1= persistent erythema; grade 2= epide rmal loss; grade 3= bluedermis ; grade 4=cavity to subcutaneous tissue or deeper; Andersen (1982) used bullae, black necrosis and skin defects as evi dence of pressure sores; Price (1999) used the

e where 0=normal skin, 1= persistent erythema of the skin; 2= blister formation; 3= superficial subcutaneous necrosis; 4= deeSideranko (1992) did not report grading system; Vanderwee (2005) did not report grading system but grade 1 was non-blanchable erythema or NBE; Malbrain (2010) used

KCE Report 193S

r development among high-risk patients placed on

o be stratified. However the Cochrane reviewher. The results of those with and without chronic

Smith scale; Stapleton (1986) adapted the grading system from Kenedi et al (1976) bed sore biomechanics study, wherecrater) and category D= blackened tissue; Whitney (1984) used a

n of the skin, fading time exceeds 15 minutes,stage III= inflammation of the skin fading time exceeds one hour; stage IV= skin break with redness of surrounding skin, redn ess fades longer than one

rmal loss; grade 3= blue-black discoloration or cavity extending todence of pressure sores; Price (1999) used the

e where 0=normal skin, 1= persistent erythema of the skin; 2= blister formation; 3= superficial subcutaneous necrosis; 4= dee p subcutaneous necrosis;blanchable erythema or NBE; Malbrain (2010) used

KCE Report 193S

Table 15 – Alternating pressure and Constant Low pressure in Intensive Care Unit/post Intensive Care Unitprevention

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Pressure ulcer incidence – Standard mattress in ICU/Standard foam mattress post

1Laurent(1998)

randomisedtrials

veryserious1


no seriousindirectn

Pressure ulcer incidence – Standard mattress in ICU/Standard foam mattress post

1Laurent(1998)

randomisedtrials

veryserious1



Pressure ulcer incidence – Alternating pressure (NIMBUS) ICU/SFM post

1Laurent(1998)

randomisedtrials

veryserious1


no seriousin

Pressure ulcer incidence – Standard ICU/Standard foam mattress post

1Laurent(1998)

randomisedtrials

veryserious1




Alternating pressure and Constant Low pressure in Intensive Care Unit/post Intensive Care Unit



AP and CLP inICU/post ICU

(factorialdesign)


Standard mattress in ICU/Standard foam mattress post -ICU vs alternating pressure mattress (NIMBUS) in ICU/Standard foam mattress



10/80(12.5%)

RR 0.4(0.13 to1.22)

12.5%

Standard mattress in ICU/Standard foam mattress post -ICU vs standard ICU/constant low pressure mattress (TEMPUR) post


veryserious3

none 14/80(17.5%)

11/75(14.7%)

RR 1.19(0.58 to2.46)

14.7%

Alternating pressure (NIMBUS) ICU/SFM post-ICU vs standard ICU/constant low pressure mattress (TEMPUR) post- ICU


veryserious3

none 10/80(12.5%)

11/75(14.7%)

RR 0.85(0.38 to1.89)

14.7%

Standard ICU/Standard foam mattress post-ICU vs Alternating pressure mattress (NIMBUS) ICU/Constant low pressure mattress (TEMPUR)CLP post


veryserious3

none 14/80(17.5%)

10/77(13%)

RR 1.35(0.64 to2.85)

347

(factorial design) for pressure ulcer


Relative(95% CI)

Absolute

ICU vs alternating pressure mattress (NIMBUS) in ICU/Standard foam mattress post-ICU

RR 0.4(0.13 to1.22)


fewer to 28more)

VERYLOW

Critical


fewer to 28more)

ICU vs standard ICU/constant low pressure mattress (TEMPUR) post -ICU

RR 1.19(0.58 to2.46)

28 more per1000 (from 62fewer to 214

more)

VERYLOW

Critical


more)

ICU

RR 0.85(0.38 to1.89)


more)

VERYLOW

Critical


more)

ICU vs Alternating pressure mattress (NIMBUS) ICU/Constant low pressure mattress (TEMPUR)CLP post -ICU

RR 1.35(0.64 to2.85)


more)

VERYLOW

Critical

348

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Pressure ulcer incidence – – Alternating pressure mattress (NIMBUS) ICU/SFM post

1Laurent(1998)

randomisedtrials

veryserious1



Pressure ulcer incidence – Standard ICU/constant low pressure mattress (TEMPUR) post

1Laurent(1998)

randomisedtrials

veryserious1



1 Unclear sequence generation, allocation concealment and no blinding. Laurent (1998).2 Confidence interval crossed one MID.3 Confidence interval crossed both MIDs.




AP and CLP inICU/post ICU

(factorialdesign)


13%

Alternating pressure mattress (NIMBUS) ICU/SFM post-ICU vs Alternating pressure mattress (NIMBUS) ICU/constant low pressure mattress (TEMPUR) post


veryserious3

none 10/80(12.5%)

10/77(13%)

RR 0.96(0.42 to2.18)

13%

Standard ICU/constant low pressure mattress (TEMPUR) post -ICU vs alternating pressure mattress (NIMBUS) ICU/constant low pressure mattress (TEMP


veryserious3

none 11/75(14.7%)

10/77(13%)

RR 1.13(0.51 to

2.5)

13%

1 Unclear sequence generation, allocation concealment and no blinding. Laurent (1998).

KCE Report 193S


Relative(95% CI)

Absolute


more)

ICU vs Alternating pressure mattress (NIMBUS) ICU/constant low pressure mattress (TEMPUR) post -ICU

RR 0.96(0.42 to2.18)


more)

VERYLOW

Critical


more)

ICU vs alternating pressure mattress (NIMBUS) ICU/constant low pressure mattress (TEMP UR) post-ICU

RR 1.13(0.51 to

2.5)


more)

VERYLOW

Critical


more)

KCE Report 193S

Comparisons between different alternating-pressure devices

Table 16 – Comparisons between alternating-pressure devices for pressure ulcer prevention

Quality assessment

No ofstudies

Design Risk ofbias


Pressure ulcer incidence – Alternating-pressure mattress (TRINOVA)

Taylor(1999)

randomisedtrials

veryserious4



Pressure ulcer incidence – alternating low pressure air mattress withsingle-stage inflation and deflation of air cells – ulcers of all grades

1Demarre(2012)

randomisedtrials

Veryserious8



Pressure ulcer incidence – alternating-pressure mattress with two layers of air cells (PEGASUS AIRWAVE SYSTEM)

1Exton-Smith(1982)

randomisedtrials

veryserious1



Pressure ulcer incidence – alternating-pressure mattress (PEGASUS AIRWAVE SYSTEM) vs alternating


pressure devices

pressure devices for pressure ulcer prevention



Comparisonsbetween

alternating-pressure devices


pressure mattress (TRINOVA)vs control – ulcers of all grades12


veryserious5

none 0/22(0%)

2/22(9.1%)

RR 0.2(0.01 to

9.1%

alternating low pressure air mattress with multi-stage inflation and deflation of air cells vs standard (CLINACTIV, HILLROM) alternating low pulcers of all grades12



56/312(17.9%)

RR 1.27(0

18%

pressure mattress with two layers of air cells (PEGASUS AIRWAVE SYSTEM) vs alternating-pressure large c



12/31(38.7%)

RR 0.42(0.17 to

38.7%

pressure mattress (PEGASUS AIRWAVE SYSTEM) vs alternating-pressure mattress (PEGASUS CAREWAVE SYSTEM)

349


Relative(95% CI)

Absolute

RR 0.2(0.01 to

3.94)


VERYLOW

Critical


standard (CLINACTIV, HILLROM) alternating low pressure air mattress with

RR 1.27(0.93 to

1.74)


VERYLOW

Critical


cell ripple mattress – grade 2+ ulcers12

RR 0.42(0.17 to

1.04)


fewer to 15more)

VERYLOW

Critical


fewer to 15more)

pressure mattress (PEGASUS CAREWAVE SYSTEM) – grade 2+ ulcers12

350

Quality assessment

No ofstudies

Design Risk ofbias


1 Hampton(1997)

randomisedtrials

veryserious3



Pressure ulcer incidence – alternating-pressure mattress (TRINOVA)Trinova vs control

1 Taylor(1999)

randomisedtrials

veryserious4



Pressure ulcer incidence – Alternating pressure overlay vs Alternating pres

1Nixon(2006)

randomisedtrials

Veryserious6



Pressure ulcer incidence – – alternating pressure bed (THERAPULSE) vs alternating pressure mattress (HILL

1Theaker(2005)

randomisedtrials

veryserious7



Pressure ulcer incidence – alternating low pressure air mattress withsingle-stage inflation and deflation of air cells – grade 2+ ulcers

1Demarre(2012)

randomisedtrials

Veryserious8






Comparisonsbetween





none 0/36(0%)

0/39(0%)

0%

pressure mattress (TRINOVA)Trinova vs control – grade 2+ ulcers12


veryserious5

none 0/22(0%)

2/22(9.1%)

RR 0.2(0.01 to

9.1%

Alternating pressure overlay vs Alternating pressure mattress – grade 2+ ulcers12


serious2 none 106/989(10.7%)

101/982(10.3%)

RR 1.04(0.81 to

10.3%

alternating pressure bed (THERAPULSE) vs alternating pressure mattress (HILL -ROM DUO) – grade 2+ ulcers12


veryserious5

none 3/30(10%)

6/32(18.8%)

RR 0.53(0.15 to

18.8%

alternating low pressure air mattress with multi-stage inflation and deflation of air cells vs standard (CLINACTIV, HILLROM) alternating lowgrade 2+ ulcers12


veryserious5

none 17/298(5.7%)

18/312(5.8%)

RR 0.99(0.5

KCE Report 193S


Relative(95% CI)

Absolute

- - LOW

Critical

-

RR 0.2(0.01 to

3.94)


VERYLOW

Critical


RR 1.04(0.81 to

1.35)


VERYLOW

Critical


RR 0.53(0.15 to

1.94)



VERYLOW

Critical




RR 0.99(0.52 to


Critical

KCE Report 193S

Quality assessment

No ofstudies

Design Risk ofbias


Withdrawal due to discomfort- alternating low pressure air mattress withsingle-stage inflation and deflation of air cells

1Demarre(2012)

randomisedtrials

Veryserious8



Comfort alternating-pressure mattress (TRINOVA)Trinova vs control

1Taylor(1996)

randomisedtrials

veryserious3



Length of stay in hospital (days) – who did develop a pressure sore

1 Theaker(2005)

randomisedtrials

veryserious3


no seriousindirectne

Length of stay in hospital (days) – who did not develop a pressure sore

1 Theaker(2005)

randomisedtrials

veryserious3


no sindirectness

Time to develop new pressure ulcer (days) – alternating low pressure air mattress withmattress with single-stage inflation and deflation of air cells

1Demarre(2012)

randomisedtrials



1 Inadequate sequence generation. Unclear allocation concealment, blinding and addressing of incomplete outcome data (Exton2 Confidence interval crossed one MID.3 Unclear sequence generation, allocation concealment,4 Unclear sequence generation, blinding, addressing incomplete outcome data. Selective reporting (Taylor 1999).5 Confidence interval crossed both MIDs.6 No blinding . High drop out in both groups. (Nixon 2006).




Comparisonsbetween



0%

alternating low pressure air mattress with multi-stage inflation and deflation of air cells vs standard (CLINACTIV, HILLROM) alternating low pressure air mattress with


veryserious5

none 11/298(3.7%)

17/312(5.4%)

RR 0.68(0.32 to

0%

pressure mattress (TRINOVA)Trinova vs control



very serious9 N=18 -

who did develop a pressure sore – alternating pressure bed (THERAPULSE) vs alternating pressure mattress (DUP)



very serious10 26 (range 23-37.3)

24 (range13-59)

who did not develop a pressure sore – alternating pressure bed (THERAPULSE) vs alternating pressure mattress (DUP)



very serious10 18 (range 5-127) 20 (range5-49)

alternating low pressure air mattress with multi-stage inflation and deflation of air cells vs standard (CLINACTIV, HILLROM) alternating low pressure air



very serious10 5.0 (IQR 3.0-8.5) 8.0 days(IQR 3.0-

8.5)

P=0.182

1 Inadequate sequence generation. Unclear allocation concealment, blinding and addressing of incomplete outcome data (Exton -Smith 1982).

3 Unclear sequence generation, allocation concealment, blinding, addressing incomplete outcome data, baseline differences (Hampton 1997).4 Unclear sequence generation, blinding, addressing incomplete outcome data. Selective reporting (Taylor 1999).

(Nixon 2006).

351


Relative(95% CI)

Absolute

1.88) 28 fewer to51 more)

VERYLOW

-


RR 0.68(0.32 to

1.42)


VERYLOW

Critical

-

- - VERYLOW

Critical

alternating pressure bed (THERAPULSE) vs alternating pressure mattress (DUP)

- - VERYLOW

Important

alternating pressure bed (THERAPULSE) vs alternating pressure mattress (DUP)

- - VERYLOW

Important

standard (CLINACTIV, HILLROM) alternating low pressure air

P=0.18211 - VERYLOW

Important

Smith 1982).

blinding, addressing incomplete outcome data, baseline differences (Hampton 1997).

352

7 Unclear sequence generation and addressing incomplete outcome data (Theaker 2005).8 No blinding of outcome assessors. High drop-out in both groups.9 Only comfort data for the intervention studied. 18/22 patients completed the comfort questionnaire, 11/18 (61.1%) described the mattress as being comfortable. Most 10/18(55.5%) found the mattress to be acceptable; overall opinion was that the mattress was unacceptable 5/18.10 Not enough data to analyse in Revman.11 Mann-Whitney U-test=113, p=0.182.12 Taylor (1999) no grading system reported but both sores were superficial one was nonused EPUAP 1999 grading system; Exton-Smith (1982) unclear grading system but included grades 3 and 4 which were superficial or deep sores; Hampton (1997) did notreport the grading system; Nixon (2006) used EPUAP 2004 and NPUAP 1999; Theaker (2005) used the Lowthian scale.

Low-air-loss (LAL) beds

Three studies evaluated the use of low-air-loss beds. Such devices provide a flow of air that assists in controlling the microclimate of the patient’s skin(NPUAP 2007).

Table 17 – Low Air Loss vs standard bed for pressure ulcer prevention

Quality assessment


Inconsistency

Pressure ulcer incidence – – low-air-loss bed (KINAIR) vs static air mattress overlay (EHOB WAFFLE)

1 Cobb (1997) randomisedtrials

veryserious1


Pressure ulcer incidence – low-air-loss bed (KINAIR/CLENSICAIR) vs static air mattress overlay (EHOB WAFFLE) or standard ICU bed or stpressure mattresses) – grade 2+ ulcers5

3 Bennett(1998) Cobb(1997) Inman(1993)

randomisedtrials

veryserious1


Pressure incidence – – low-air-loss bed (KINAIR) vs static air mattress overlay (EHOB WAFFLE) or standard ICU bed


7 Unclear sequence generation and addressing incomplete outcome data (Theaker 2005).out in both groups.(Demarre 2012)

d. 18/22 patients completed the comfort questionnaire, 11/18 (61.1%) described the mattress as being comfortable. Most 10/18(55.5%) found the mattress to be acceptable; overall opinion was that the mattress was unacceptable 5/18.

12 Taylor (1999) no grading system reported but both sores were superficial one was non-blanching erythema and one was a superficial break in the skin. Demarre (2012)Smith (1982) unclear grading system but included grades 3 and 4 which were superficial or deep sores; Hampton (1997) did not

report the grading system; Nixon (2006) used EPUAP 2004 and NPUAP 1999; Theaker (2005) used the Lowthian scale.

loss beds. Such devices provide a flow of air that assists in controlling the microclimate of the patient’s skin

Low Air Loss vs standard bed for pressure ulcer prevention

uality assessment No of patients


Low AirLoss

Standardbed

Relative(95% CI)

s bed (KINAIR) vs static air mattress overlay (EHOB WAFFLE) – all grades of ulcers5

no serious no seriousindirectness


12/61(19.7%)

RR 0.49(0.2 to1.23)

19.7%

loss bed (KINAIR/CLENSICAIR) vs static air mattress overlay (EHOB WAFFLE) or standard ICU bed or st andard care (standard bed or foam, air, alternating



none 20/153(13.1%)

41/166(24.7%)

-

19.7%

loss bed (KINAIR) vs static air mattress overlay (EHOB WAFFLE) or standard ICU bed – grade 2+ ulcers5

KCE Report 193S

d. 18/22 patients completed the comfort questionnaire, 11/18 (61.1%) described the mattress as being comfortable. Most 10/18

blanching erythema and one was a superficial break in the skin. Demarre (2012)Smith (1982) unclear grading system but included grades 3 and 4 which were superficial or deep sores; Hampton (1997) did not

loss beds. Such devices provide a flow of air that assists in controlling the microclimate of the patient’s skin


Relative(95% CI)

Absolute

RR 0.49(0.2 to1.23)


fewer to 45more)

VERYLOW

Critical


fewer to 45more)

andard care (standard bed or foam, air, alternating-

- 247 fewer per1000 (from 247

fewer to 247fewer)

LOW

Critical


fewer to 197fewer)

KCE Report 193S

Quality assessment


Inconsistency

2 Cobb (1997)Inman (1993)

randomisedtrials

veryserious1


Comfort – low air loss hydrotherapy (CLENSICAIR) vs standard care (standard bed or foam, air , alternating

1 Bennett(1998)

randomisedtrials

veryserious1


Patient acceptability – low air loss hydrotherapy (CLENSICAIR) vs standard care (standard bed or foam, air , alternating

1 Bennett(1998)

randomisedtrials

veryserious1


1 Unclear sequence generation (Cobb 1997, Inman 1993) and allocation concealment (Bennett 1998, Inman 1993). Unclear blindingNo addressing of incomplete outcome data (Inman 1993). Differences at baseline (Cobb 1997).2 Confidence interval crossed one MID point.3 Data on comfort only from intervention group and only 10/42 patients completed the questionnaire. 5/10 thought it wa4 It should be noted that there were more dropouts overall from the treatment than the control group 24/48 (35%) vs 2/58 (3%)hydrotherapy exited the study on the first day because either a patient or family member complained about the bed. This was due to being wet, cold or uncomspecialty bed. Two subjects were removed by the research investigators or nurses as a result of hypothermia within5 Bennett (1998) used NPUAP 1989; Cobb (1997) used NPUAP 1989 and Shea 1975; Inman (1993) used Shea 1975.


uality assessment No of patients


Low AirLoss

Standardbed

Relative(95% CI)



none 9/111(8.1%)

36/110(32.7%)

RR 0.25(0.13 to0.49)

34.5%

low air loss hydrotherapy (CLENSICAIR) vs standard care (standard bed or foam, air , alternating -pressure mattresses)



very low3 N=10 - -

low air loss hydrotherapy (CLENSICAIR) vs standard care (standard bed or foam, air , alternating -pressure mattresses)



very low4 - - -

1 Unclear sequence generation (Cobb 1997, Inman 1993) and allocation concealment (Bennett 1998, Inman 1993). Unclear blindingdressing of incomplete outcome data (Inman 1993). Differences at baseline (Cobb 1997).

3 Data on comfort only from intervention group and only 10/42 patients completed the questionnaire. 5/10 thought it wa s comfortable, 4/10 thought it was uncomfortable.4 It should be noted that there were more dropouts overall from the treatment than the control group 24/48 (35%) vs 2/58 (3%) (p=0.0001). Six subjects receiving low airloss

n the first day because either a patient or family member complained about the bed. This was due to being wet, cold or uncomspecialty bed. Two subjects were removed by the research investigators or nurses as a result of hypothermia within the first 24 hours of enrolment.5 Bennett (1998) used NPUAP 1989; Cobb (1997) used NPUAP 1989 and Shea 1975; Inman (1993) used Shea 1975.

353


Relative(95% CI)

Absolute

RR 0.25(0.13 to0.49)


fewer to 285fewer)

LOW

Critical


fewer to 300fewer)

- See footnote4VERYLOW

Critical

- See footnote5VERYLOW

Critical

1 Unclear sequence generation (Cobb 1997, Inman 1993) and allocation concealment (Bennett 1998, Inman 1993). Unclear blinding (Cobb 1997, Inman 1993, Bennett 1998).

s comfortable, 4/10 thought it was uncomfortable.(p=0.0001). Six subjects receiving low airloss

n the first day because either a patient or family member complained about the bed. This was due to being wet, cold or uncom fortable on thethe first 24 hours of enrolment.

354

7.4.1.3. Other support surfaces – operating room

Operating room mattresses (indentation load deflection) versus usual c

Table 18 – Indentation load deflection (IDL) (25%) operating room foam mattress (density 1.3 cubic feet, IDL 30lb) vs operating room usu(padding as required, including gel pads, foam mattresses, ring cushions (donuts) etc) for pressure ulcer pr

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers – all grades of pressure ulcer

1Schultz(1999)

randomisedtrials



Incidence of pressure ulcers – grade 2 + pressure ulcers

1Schultz(1999)

randomisedtrials



Patient acceptability – postoperative skin changes

1Schultz(1999)

randomisedtrials



1 No allocation concealment.2 Confidence interval crossed one MID point.3 Confidence interval crossed both MID points.4 No details given for number of patients in each arm for postoperative skin changes5 Patients on the experimental mattress (IDL) were significantly more likely to have skin changes than those on the usual cargiven.


perating room

ad deflection) versus usual care

Indentation load deflection (IDL) (25%) operating room foam mattress (density 1.3 cubic feet, IDL 30lb) vs operating room usu(padding as required, including gel pads, foam mattresses, ring cushions (donuts) etc) for pressure ulcer pr evention



ILDoperating

roommattress

Usualcare

Relative(95% CI)



34/207(16.4%)

RR 1.63(1.11 to

2.38)

16.4%


veryserious3

none 6/206(2.9%)

3/207(1.4%)

RR 2.01(0.51 to

7.93)

1.5%


no serious very serious4 - - P=0.0111

4 No details given for number of patients in each arm for postoperative skin changes .5 Patients on the experimental mattress (IDL) were significantly more likely to have skin changes than those on the usual car e operating room table, no further details were

KCE Report 193S

Indentation load deflection (IDL) (25%) operating room foam mattress (density 1.3 cubic feet, IDL 30lb) vs operating room usu al careevention


Relative(95% CI)

Absolute

RR 1.63(1.11 to

2.38)

103 more per1000 (from 18more to 227

more)

LOW

Critical

103 more per1000 (from 18more to 226

more)

RR 2.01(0.51 to

7.93)


more)

VERYLOW

Critical


more)

P=0.0111 See footnote5VERYLOW

Critical

e operating room table, no further details were

KCE Report 193S

Operating table overlay versus no overlay

Table 19 – Operating table overlay vs no overlay for pressure ulcer prevention

Quality assessment


Inconsistency

Pressure ulcer incidence – Viscoelastic polymer pad vs no overlay

1Nixon (1998) randomisedtrials


Pressure ulcer incidence – Viscoelastic foam overlay vs no overlay

1Feuchtinger(2006)

randomisedtrials

veryserious3


Pressure ulcer incidence – Viscoelastic foam overlay vs no overlay

1Feuchtinger(2006)

randomisedtrials

veryserious3


1 Difference at baseline. Standard mattress group had a longer length of operation, longer pregroup (Nixon 1998).2 Confidence interval crossed one MID.3 Unclear sequence generation, allocation concealment and addressing of incomplete outcome data (Feuchtinger 2006).4 Confidence interval crossed both MIDs.


ay vs no overlay for pressure ulcer prevention



Operatingtable

overlay

Nooverlay

Relative(95%

Viscoelastic polymer pad vs no overlay6



43/211(20.4%)

RR 0.53(0.33 to

0.85)

20.4%

Viscoelastic foam overlay vs no overlay6


veryserious4

none 13/85(15.3%)

9/90(10%)

RR 1.53(0.69 to

3.39)

10%

overlay – grade 2+ ulcers6


veryserious5

none 2/85 (2.4%) 1/90(1.1%)

RR 2.12(0.2 to22.93)

1.1%

1 Difference at baseline. Standard mattress group had a longer length of operation, longer pre -operative stay and more time in hypotensive state than the dry polymer pad

3 Unclear sequence generation, allocation concealment and addressing of incomplete outcome data (Feuchtinger 2006).

355


Relative(95% CI)

Absolute

RR 0.53(0.33 to

0.85)


fewer)

LOW

Critical


fewer)

RR 1.53(0.69 to

3.39)


more)

VERYLOW

Critical


more)

RR 2.12(0.2 to22.93)


more)

VERYLOW

Critical


more)

operative stay and more time in hypotensive state than the dry polymer pad

356

5 Confidence interval crossed both MID points and limite6 Nixon (1998) used the Torrance 1983 grading system; Feuchtinger (2006)used EPUAP 2005 grading system.

Face pillows in operating room

Table20 – Disposable polyurethane foam prone head positioner (OSI) vs neoprene air filled bladder (

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Incidence of pressure ulcers – all grades of ulcer4

1Grisell(2008)

randomisedtrials

Veryserious1


Incidence of pressure ulcers – grades 2+ ulcers4

1Grisell(2008)

randomisedtrials

Veryserious1


1 Grisell (2008): No details of baseline data. No blinding.grading system.


5 Confidence interval crossed both MID points and limited number of events.6 Nixon (1998) used the Torrance 1983 grading system; Feuchtinger (2006)used EPUAP 2005 grading system.

Disposable polyurethane foam prone head positioner (OSI) vs neoprene air filled bladder ( dry flotation) device (ROHO



OSI facepillow

ROHOface

pillow(95% CI)



0/22(0%)

Peto OR 12.55(3.11to 50.57)

0%


Veryserious2,3

none 2/22(9.1%)

0/22 (0%) Peto OR 7.75

0%

1 Grisell (2008): No details of baseline data. No blinding.higher drop-out than event rate. 2 Limited number of events. 3 Confidence interval crossed both MID points. 4 NPUAP

KCE Report 193S

dry flotation) device (ROHO)


Relative(95% CI)

Absolute

Peto OR 12.55(3.11to 50.57)

- VERYLOW

Critical

-

Peto OR 7.75(0.47 to128.03)

- VERYLOW

Critical

-

2 Limited number of events. 3 Confidence interval crossed both MID points. 4 NPUAP

KCE Report 193S

Table 21 – Disposable polyurethane foam prone headprone head positioner (DUPACO)

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Incidence of pressure ulcers – all grades of ulcer4

1Grisell(2008)

randomisedtrials

Veryserious1


Incidence of pressure ulcers – grades 2+ ulcer4

1Grisell(2008)

randomisedtrials

Veryserious1


1 Grisell (2008): No details of baseline data. No blinding.2 Limited number of events. 3 Confidence interval crossed both MID points. 4 NPUAP grading system.

Table 22 – Neoprene air filled bladder (dry flotation) device (ROHO)vs prone view protective helmet system with a disposable polyurethane foprone head positioner (DUPACO)

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Incidence of pressure ulcers – all grades of ulcers2

1Grisell(2008)

randomisedtrials

Veryserious1


Incidence of pressure ulcers – grades 2+ ulcers2

1Grisell(2008)

randomisedtrials

Veryserious1


1 Grisell (2008): No details of baseline data. No blinding.2 NPUAP grading system.


Disposable polyurethane foam prone head positioner (OSI)vs prone view protective helmet system with a disposable polyurethane foam



OSI facepillow

Dupacoface pillow



0/22(0%)

Peto OR 12.55(3.11 to 50.57)

0%


Veryserious2,3

none 2/22(9.1%)

0/22 (0%) Peto OR 7.75(0.47 to1280%

of baseline data. No blinding.higher drop-out than event rate.3 Confidence interval crossed both MID points. 4 NPUAP grading system.

rene air filled bladder (dry flotation) device (ROHO)vs prone view protective helmet system with a disposable polyurethane fo



ROHOface pillow

Dupacoface pillow



none 0/22(0%)

0/22(0%)

0%



none 0/22(0%)

0/22(0%)

0%

1 Grisell (2008): No details of baseline data. No blinding.Higher drop-out than event rate.

357

prone view protective helmet system with a disposable polyurethane foam


Relative(95% CI)

Absolute

Peto OR 12.55(3.11 to 50.57)

- VERYLOW

Critical

-

Peto OR 7.75(0.47 to128.03)

- VERYLOW

Critical

-

rene air filled bladder (dry flotation) device (ROHO)vs prone view protective helmet system with a disposable polyurethane fo am


face pillowRelative(95% CI)

Absolute

notpooled

notpooled

LOW

Critical

notpooled

notpooled

notpooled

LOW

Critical

notpooled

358

Other mattresses intra- and post-operatively

Table 23 – Multi-cell pulsating dynamic mattress system (MICROPULSE) vs standard mattress

Quality assessment



Pressure ulcer incidence – all grades of ulcer4

2Aronovitch(1999)Russell(2000)

randomisedtrials

veryserious1



Pressure ulcer incidence – grade 2+ ulcers4

1Aronovitch(1999);

randomisedtrials

veryserious1



Length of stay in hospital

1Aronovitch(1999)

randomisedtrials

veryserious1



1 Unclear sequence generation (quasi-randomised), allocation concealment and blindinggroup were at higher risk at baseline (Knoll score) Unclear2 Data given only for those who developed ulcers – 6/8 who developed ulcersthose developing ulcers was 14 days, which was6.7 days longer than the hospital’s average of 7.3 days for this Diagnosis Relarepresents and increase in length of stay of 92%.


cell pulsating dynamic mattress system (MICROPULSE) vs standard mattress for surgical patients for pressure ulcer prevention



MicropulseSystem for

surgicalpatients

Usual care (gel padin operating roomand a replacement

mattresspostoperatively)



none 3/188(1.6%)

14/180(7.8%)

7.9%


Serious3 none 0/90 (0%) 6/80 (7.5%)

7.5%



very serious2 - -

randomised), allocation concealment and blinding and higher drop out than event rate(Aronovitch 1999).Unclear sequence generation method , no blinding and higher drop out than event rate

6/8 who developed ulcers had a length of stay longer than average for the specific diagnosis. Average length of stay forthose developing ulcers was 14 days, which was6.7 days longer than the hospital’s average of 7.3 days for this Diagnosis Rela ted Group. The authors state that

KCE Report 193S

for surgical patients for pressure ulcer prevention


Relative(95% CI)

Absolute

RR 0.21(0.06 to

0.7)


fewer to 73fewer)

LOW

Critical


fewer to 74fewer)

RR 0.07(0 to1.2)


fewer to 15more)

VERYLOW

Critical


fewer to 15more)

- Seefootnote2

VERYLOW

Important

(Aronovitch 1999). The conventional managementand higher drop out than event rate (Russell 2000).

had a length of stay longer than average for the specific diagnosis. Average length of stay forted Group. The authors state that this

KCE Report 193S

3 Confidence interval crossed one MID point.4 Aronovitch (1999)used NPUAP and WOCN and Russell (2000) used NPUAP 1997.

Table 24 – Visco-elastic foam (TEMPUR-PEDIC)

Quality assessment


Inconsistency

Proportion with incidence of pressure ulcers – grade 2+ ulcers

1Gunningberg(2000)

randomisedtrials


Proportion with incidence of pressure ulcers – all grades of ulcers

1Gunningberg(2000)

randomisedtrials


1 No details of allocation concealment.2 Confidence interval crossed both MID points.3 EPUAP 1999 grading system.


4 Aronovitch (1999)used NPUAP and WOCN and Russell (2000) used NPUAP 1997.

PEDIC) A&E overlay and ward mattress vs standard A&E overlay and ward mattress



Accident andemergencyoverlay and

ward mattress


grade 2+ ulcers3



none 4/48(8.3%)

8/53(15.1%)

RR 0(0.18 to

15.1%

all grades of ulcers3



none 12/48 (25%) 17/53(32.1%)

RR 0.78(0.42 to

32.1%

359

s standard A&E overlay and ward mattress


Relative(95% CI)

Absolute

RR 0.55(0.18 to

1.72)



VERYLOW

Critical



RR 0.78(0.42 to

1.46)


186 fewer to148 more

VERYLOW

Critical



360

7.4.1.4. Profiling beds

Table 25 – Profiling bed with a pressure-reducing foam mattress vs flat

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Proportion with incidence of pressure ulcers – all grades of ulcer

1 Keogh(2001)

randomisedtrials

veryserious1


1 Unclear blinding, unclear addressing of incomplete outcome data and2 EPUAP 1991 grading system.

7.4.1.5. Seat cushions: comparison between different cushions

Table 26 – Seat cushions for pressure ulcer prevention

Quality assessment


Inconsistency

Pressure ulcer incidence – Slab foam cushion v Bespoke contoured foam

2 Conine(1993) Lim(1988)

randomisedtrials

veryserious1


Pressure ulcer incidence – Gel Cushion with foam base (JAY)

1Conine(1994)

randomisedtrials



reducing foam mattress vs flat-based bed with a pressure-reducing mattress



Profilingbed

Flat-basedbed




none 0/35 (0%) 0/35 (0%)

0%

1 Unclear blinding, unclear addressing of incomplete outcome data and higher drop out than event rate.

omparison between different cushions

Seat cushions for pressure ulcer prevention



Seatcushions


v Bespoke contoured foam cushion 6



none 104/151(68.9%)

102/149(68.5%)

RR 1.01(0.86 to1.17)

68.8%

with foam base (JAY) v Foam cushion 6



30/73(41.1%)

RR 0.61(0.37 to 1)

41.1%

KCE Report 193S

reducing mattress


basedRelative

(95%CI)

Absolute

0/35 (0%) - - LOW

Critical

-


Relative(95% CI)

Absolute

RR 1.01(0.86 to1.17)


116 more)

LOW

Critical


117 more)

RR 0.61(0.37 to 1)

160 fewer per1000 (from 259fewer to 0 more)

LOW

Critical


KCE Report 193S

Quality assessment


Inconsistency

Pressure ulcer incidence – Pressure reducing cushion (not specified

1Geyer(2001)

randomisedtrials


pressure ulcer incidence – skin protection cushion vs segmented foam cushion

1Brienza(2010)

randomisedtrials

Veryserious5


pressure ulcer incidence – skin protection cushion vs segmented foam cushion

1Brienza(2010)

randomisedtrials

Veryserious5


Patient acceptability – withdrawal due to discomfort – Gel Cushion with foam base (JAY)

1Conine(1994)

randomisedtrials


1 Unclear sequence generation, allocation concealment and blinding (Conine 1993, Lim 1988).2 Unclear sequence generation and allocation concealment (Conine 1994).3 Confidence interval crossed one MID.4 Confidence interval crossed both MIDs.5 Baseline differences. The study could not control for other support surfaces.6 Conine (1993) and (1994) used Exton Smith 1982; Lim (1988) used NPUAP 1989; Geyer (2001) used NPUAP 1992; Brienza (2010) us




Seatcushions


(not specified – chosen by nurse based on patient)v Standard 3inch convoluted foam cushion (EGGRATE)


very serious4 none 6/15(40%)

10/17(58.8%)

RR 0.68(0.33 to1.42)

58.8%

skin protection cushion vs segmented foam cushion – sitting related ischial tuberosities6



8/119(6.7%)

RR 0.13(0.02 to1.04)

0%

mented foam cushion – combined ischial tuberosities and sacral/coccyx6



21/119(17.6%)

RR 0.60(0.31 to1.17)

0%

Gel Cushion with foam base (JAY) vs Foam cushion


very serious4 none 1/83(1.2%)

6/80(7.5%)

RR 0.16(0.02 to1.30)

0%

1 Unclear sequence generation, allocation concealment and blinding (Conine 1993, Lim 1988).2 Unclear sequence generation and allocation concealment (Conine 1994).

Baseline differences. The study could not control for other support surfaces.6 Conine (1993) and (1994) used Exton Smith 1982; Lim (1988) used NPUAP 1989; Geyer (2001) used NPUAP 1992; Brienza (2010) us

361


Relative(95% CI)

Absolute

fewer to 0 more)

3inch convoluted foam cushion (EGGRATE)6

RR 0.68(0.33 to1.42)


fewer to 247more)

VERYLOW

Critical


fewer to 247more)

RR 0.13(0.02 to1.04)


fewer to 3 more)

VERYLOW

Critical

-

RR 0.60(0.31 to1.17)


fewer to 30more)

VERYLOW

Critical

-

RR 0.16(0.02 to1.30)


fewer to 22more)

VERYLOW

Critical

-

6 Conine (1993) and (1994) used Exton Smith 1982; Lim (1988) used NPUAP 1989; Geyer (2001) used NPUAP 1992; Brienza (2010) us ed NPUAP 2001.

362

7.4.2. Forest plots

Constant low-pressure supports (CLP) vs standard foam mattresses (SFM)

Figure 2 – Pressure ulcer incidence – grades 2+ ulcers

Study or Subgroup

1.1.1 Cubed foam mattress

Hofman 1994Subtotal (95% CI)

Total events



1.1.3 Softform mattress

Gray 1994Subtotal (95% CI)

Total events



Events

4

4

6

6

Total

1717

9090

Events

13

13

27

27

Total

1919

8080

Weight

100.0%100.0%

100.0%100.0%

CLP SFM


upports (CLP) vs standard foam mattresses (SFM)

grades 2+ ulcers

Weight

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

0.34 [0.14, 0.85]0.34 [0.14, 0.85]

0.20 [0.09, 0.45]0.20 [0.09, 0.45]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours CLP Favours SFM

KCE Report 193S

KCE Report 193S

Figure 3 – Pressure ulcer incidence – all grades of ulcer

Study or Subgroup

1.3.1 Cubed foam mattress

Hofman 1994Subtotal (95% CI)

Total events



1.3.2 Bead-filled mattress

Goldstone 1982Subtotal (95% CI)

Total events



1.3.3 Water-filled mattress

Andersen 1982Subtotal (95% CI)

Total events



1.3.4 Alternative foam

Collier 1996

Santy 1994Subtotal (95% CI)

Total events


Test for overall effect: Z = 4.03 (P < 0.0001)

1.3.5 Softform mattress

Gray 1994Subtotal (95% CI)

Total events



1.3.6 Hi-spec foam mattress/cushion

Russell 2003Subtotal (95% CI)

Total events



Events

6

6

5

5

7

7

0

42

42

6

6

48

48

Total

1717

3232

155155

130

441571

9090

562562

Events

14

14

21

21

21

21

0

17

17

27

27

66

66

Total

1919

4343

161161

9

6473

8080

604604

Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Random, 95% CI

0.48 [0.24, 0.96]0.48 [0.24, 0.96]

0.32 [0.14, 0.76]0.32 [0.14, 0.76]

0.35 [0.15, 0.79]0.35 [0.15, 0.79]

0.36 [0.22, 0.59]0.36 [0.22, 0.59]

0.20 [0.09, 0.45]0.20 [0.09, 0.45]

0.78 [0.55, 1.11]0.78 [0.55, 1.11]

Favours CLP SFM Risk Ratio


all grades of ulcer

M-H, Random, 95% CI

0.48 [0.24, 0.96]0.48 [0.24, 0.96]

0.32 [0.14, 0.76]0.32 [0.14, 0.76]

0.35 [0.15, 0.79]0.35 [0.15, 0.79]

Not estimable

0.36 [0.22, 0.59]0.36 [0.22, 0.59]

0.20 [0.09, 0.45]0.20 [0.09, 0.45]

0.78 [0.55, 1.11]0.78 [0.55, 1.11]


M-H, Random, 95% CI

0.1 0.2 0.5 1 2 5 10Favours CLP Favours SFM

363

364

Figure 4 – Patient acceptability – very uncomfortable

Figure 5 – Patient acceptability – uncomfortable

Figure 6 – Patient acceptability – adequate

Study or Subgroup

Gray 1994

Total (95% CI)

Total events


Test for overall effect: Not applicable

Events

0

0

Total

90

90

Events

0

0

Total

80

80

Softform mattress Std Foam

Study or Subgroup

Gray 1994

Total (95% CI)

Total events



Events

0

0

Total

90

90

Events

2

2

Total

80

80

100.0%


Study or Subgroup

Gray 1994

Total (95% CI)

Total events



Events

6

6

Total

90

90

Events

44

44

Total

80

80

Weight

100.0%

100.0%

Softform Std Foam


very uncomfortable

uncomfortable

Weight Peto, Fixed, 95% CI

Not estimable

Not estimable

Peto Odds Ratio Peto Odds Ratio

Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours softform Favours foam

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

0.12 [0.01, 1.91]

0.12 [0.01, 1.91]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours softform Favours foam

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.12 [0.05, 0.27]

0.12 [0.05, 0.27]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours foam Favours softform

KCE Report 193S

KCE Report 193S

Figure 7 – Patient acceptability – comfortable

Figure 8 – Patient acceptability – very comfortable

Figure 9 – Patient acceptability – comfort

Study or Subgroup

Gray 1994

Total (95% CI)

Total events



Events

62

62

Total

90

90

Events

26

26

Total

80

80


Study or Subgroup

Gray 1994

Total (95% CI)

Total events



Events

11

11

Total

90

90

Events

0

0

Total

80

80

100.0%


Study or Subgroup

Russell 2003

Total (95% CI)



Mean

2.33

SD

0.98

Total

323

323

Mean

2.46

SD

1.01

Total

383

383

CLP SFM


very comfortable

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

2.12 [1.50, 2.99]

2.12 [1.50, 2.99]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

Peto, Fixed, 95% CI

7.45 [2.20, 25.24]

7.45 [2.20, 25.24]


Peto, Fixed, 95% CI


Total Weight

100.0%

100.0%

IV, Fixed, 95% CI

-0.13 [-0.28, 0.02]

-0.13 [-0.28, 0.02]

Mean Difference Mean Difference

IV, Fixed, 95% CI

-100 -50 0 50 100Favours CLP Favours SFM

365

366

7.4.3. Alternative foam mattress vs standard

Figure 10 – Pressure ulcer incidence – all grades of ulcer (studies pooled)

Figure 11 – Pressure ulcer incidence – grades 2+ ulcers (studies pooled)

Study or Subgroup

2.1.1 Various alternatives (pooled)

Collier 1996

Gray 1994

Hofman 1994

Russell 2003

Santy 1994Subtotal (95% CI)

Total events

Heterogeneity: Tau² = 0.25; Chi² = 12.50, df = 3 (P = 0.006); I² = 76%


Total (95% CI)

Total events




Events

0

6

6

48

42

102

102

Total

130

90

17

562

4411240

1240

Events

0

27

14

66

17

124

124

Total

9

80

19

604

64776

776

Weight

19.9%

22.7%

30.3%

27.2%100.0%

100.0%

Alternative Foam Std Foam

Study or Subgroup

Gray 1994

Hofman 1994

Total (95% CI)

Total events



Events

6

4

10

Total

90

17

107

Events

27

13

40

Total

80

19

99

Weight

70.0%

30.0%

100.0%

Experimental Std Foam


Alternative foam mattress vs standard foam mattress

all grades of ulcer (studies pooled)

grades 2+ ulcers (studies pooled)



Weight

19.9%

22.7%

30.3%

27.2%100.0%

100.0%

M-H, Random, 95% CI

Not estimable

0.20 [0.09, 0.45]

0.48 [0.24, 0.96]

0.78 [0.55, 1.11]

0.36 [0.22, 0.59]0.43 [0.24, 0.76]

0.43 [0.24, 0.76]


M-H, Random, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Alternative Favours SFM

Weight

70.0%

30.0%

100.0%

M-H, Fixed, 95% CI

0.20 [0.09, 0.45]

0.34 [0.14, 0.85]

0.24 [0.13, 0.45]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours experimental Favours control

KCE Report 193S

KCE Report 193S

Comparisons between alternative foam supports



Study or Subgroup

3.1.1 Alternative foam vs standard foam

Santy 1994

3.1.2 Maxifloat foam mattress vs Iris foam overlay

Vyhlidal 1997

3.1.3 Solid foam vs convoluted foam

Kemp 1993

3.1.4 Transfoam mattress vs Transfoamwave mattress

Gray 1998

Events

42

5

12

1

Total

441

20

39

50

Events

17

12

21

1

Total

64

20

45

50

Foam 1 Foam 2

Study or Subgroup

Vyhlidal 1997

Total (95% CI)

Total events



Events

3

3

Total

20

20

Events

8

8

Total

Maxifloat foam overlay Iris foam overlay


Comparisons between alternative foam supports

all grades of ulcer

grades 2+ ulcers

3.1.4 Transfoam mattress vs Transfoamwave mattress

M-H, Fixed, 95% CI

0.36 [0.22, 0.59]

0.42 [0.18, 0.96]

0.66 [0.37, 1.16]

1.00 [0.06, 15.55]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Foam 1 Favours Foam 2

Total

20

20

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.38 [0.12, 1.21]

0.38 [0.12, 1.21]

Iris foam overlay Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours maxifloat Favours iris

367

368

Comparisons between CLP supports


Study or Subgroup

4.1.1 Optima vs SFM

Takala 1996Subtotal (95% CI)

Total events



4.1.2 Sofflex vs ROHO

Cooper 1998Subtotal (95% CI)

Total events



4.1.3 Gel mattress vs air-filled overlay

Lazzara 1991Subtotal (95% CI)

Total events



4.1.4 Static air mattress vs water mattress

Sideranko 1992Subtotal (95% CI)

Total events



4.1.5 Foam overlay vs Silicore overlay

Stapleton 1986Subtotal (95% CI)

Total events



4.1.6 Sheepskin vs no sheepskin (Including all pressure ulcers regardless of Grade)

Jolley 2004 (1)

McGowan 2000 (2)

Mistiaen 2009 (3)Subtotal (95% CI)

Total events



4.1.8 Static air overlay (and cold foam mattress) vs cold foam mattress

Van Leen, 2011Subtotal (95% CI)

Total events



Events

0

0

3

3

8

8

1

1

14

14

21

14

24

59

2

2

Total

2121

4141

3333

2020

3434

218

155

271644

3838

Events

7

7

5

5

10

10

2

2

12

12

37

43

40

120

7

7

Total

1919

4343

3333

1717

3434

223

142

272637

3636

Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

33.9%

30.5%

35.5%100.0%

100.0%100.0%

M-H, Random, 95% CI

CLP1 CLP2

(1) This study evaluates all patients with pressure ulcers regardless of grade




all grades of ulcer

4.1.6 Sheepskin vs no sheepskin (Including all pressure ulcers regardless of Grade)

4.1.8 Static air overlay (and cold foam mattress) vs cold foam mattress

M-H, Random, 95% CI

0.06 [0.00, 0.99]0.06 [0.00, 0.99]

0.63 [0.16, 2.47]0.63 [0.16, 2.47]

0.80 [0.36, 1.77]0.80 [0.36, 1.77]

0.42 [0.04, 4.29]0.43 [0.04, 4.29]

1.17 [0.64, 2.14]1.17 [0.64, 2.14]

0.58 [0.35, 0.96]

0.30 [0.17, 0.52]

0.60 [0.37, 0.97]0.48 [0.31, 0.74]

0.27 [0.06, 1.22]0.27 [0.06, 1.22]

Risk Ratio




Risk Ratio

M-H, Random, 95% CI

0.1 0.2 0.5 1 2 5 10Favours CLP1 Favours CLP2

KCE Report 193S

KCE Report 193S

Figure 14 – Pressure ulcer incidence – grade 2+ ulcers

Study or Subgroup

4.2.1 Sofflex vs ROHO

Cooper 1998Subtotal (95% CI)

Total events



4.2.2 Gel mattress vs air-filled overlay

Lazzara 1991Subtotal (95% CI)

Total events



4.2.3 Sheepskin vs no sheepskin (grade 2 + pressure ulcers only)

Jolley 2004

McGowan 2000

Mistiaen 2009Subtotal (95% CI)

Total events



4.2.4 static air overlay (and cold foam mattress) vs cold foam mattress

Van Leen, 2011Subtotal (95% CI)

Total events



4.2.5 Foam overlay vs silicore

Stapleton 1986Subtotal (95% CI)

Total events




Events

1

1

4

4

12

0

6

18

2

2

14

14

Total

4141

3333

218

155

271644

3838

3434

Events

0

0

5

5

20

5

8

33

7

7

12

12

Total

4343

3333

223

142

272637

3636

3434

Weight

100.0%100.0%

100.0%100.0%

59.0%

17.1%

23.8%100.0%

100.0%100.0%

100.0%100.0%

Group 1 Group 2


grade 2+ ulcers

4.2.3 Sheepskin vs no sheepskin (grade 2 + pressure ulcers only)

4.2.4 static air overlay (and cold foam mattress) vs cold foam mattress


Weight

100.0%100.0%

100.0%100.0%

59.0%

17.1%

23.8%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

3.14 [0.13, 75.02]3.14 [0.13, 75.02]

0.80 [0.24, 2.72]0.80 [0.24, 2.72]

0.61 [0.31, 1.22]

0.08 [0.00, 1.49]

0.75 [0.26, 2.14]0.56 [0.32, 0.97]

0.27 [0.06, 1.22]0.27 [0.06, 1.22]

1.17 [0.64, 2.14]1.17 [0.64, 2.14]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Group 1 Favours Group 2

369

370

Figure 15 – Patient acceptability – very uncomfortable

Figure 16 – Patient acceptability – uncomfortable

Figure 17 – Patient acceptability – adequate

Study or Subgroup

Cooper 1998

Total (95% CI)

Total events



Events

0

0

Total

41

41

Events

0

0

Total

43

43

Weight

SOFFLEX ROHO

Study or Subgroup

Cooper 1998

Total (95% CI)

Total events



Events

0

0

Total

41

41

Events

5

5

Total

43

43

Weight

100.0%

100.0%

SOFFLEX ROHO

Study or Subgroup

Cooper 1998

Total (95% CI)

Total events



Events

4

4

Total

41

41

Events

4

4

Total

43

43

Weight

100.0%

100.0%

SOFFLEX ROHO


very uncomfortable

uncomfortable

Weight M-H, Fixed, 95% CI

Not estimable

Not estimable

Odds Ratio Odds Ratio

M-H, Fixed, 95% CI


Weight

100.0%

100.0%

Peto, Fixed, 95% CI

0.13 [0.02, 0.77]

0.13 [0.02, 0.77]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours SOFFLEX Favours ROHO

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.05 [0.28, 3.92]

1.05 [0.28, 3.92]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours ROHO Favours SOFFLEX

KCE Report 193S

KCE Report 193S

Figure 18 – Patient acceptability – comfortable

Figure 19 – Patient acceptability – very comfortable

Alternating-pressure vs standard foam mattress


Study or Subgroup

Cooper 1998

Total (95% CI)

Total events



Events

24

24

Total

41

41

Events

24

24

Total

43

43

Weight

100.0%

100.0%

SOFFLEX ROHO

Study or Subgroup

Cooper 1998

Total (95% CI)

Total events



Events

13

13

Total

41

41

Events

10

10

Total

43

43

Weight

100.0%

100.0%

SOFFLEX ROHO

Study or Subgroup

Andersen 1982

Sanada 2003

Total (95% CI)

Total events



Events

7

6

13

Total

166

55

221

Events

21

10

31

Total

161

27

188

Alternating Pressure SFM


comfortable

very comfortable

vs standard foam mattress

all grades of ulcer

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.05 [0.72, 1.52]

1.05 [0.72, 1.52]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.36 [0.67, 2.76]

1.36 [0.67, 2.76]


M-H, Fixed, 95% CI


Total

161

27

188

Weight

61.4%

38.6%

100.0%

M-H, Fixed, 95% CI

0.32 [0.14, 0.74]

0.29 [0.12, 0.73]

0.31 [0.17, 0.58]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours AP Favours SFM

371

372


Study or Subgroup

Sanada 2003

Total (95% CI)

Total events



Events

5

5

Total

55

55

Events

6

6

Total

27

27

Alternating Pressure SFM


grades 2+ ulcers

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.41 [0.14, 1.22]

0.41 [0.14, 1.22]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours AP Favours SFM

KCE Report 193S

KCE Report 193S

7.4.4. Alternating-pressure vs constant low-pressure

Figure 22 – Pressure ulcer incidence – all grades of

Study or Subgroup

6.1.1 AP (various) vs CLP (various)

Gebhardt 1996Subtotal (95% CI)

Total events



6.1.2 AP vs Silicore or foam overlay

Conine 1990

Daechsel 1985

Stapleton 1986

Whitney 1984Subtotal (95% CI)

Total events



6.1.3 AP vs water or static air mattress

Andersen 1982

Price 1999

Sideranko 1992Subtotal (95% CI)

Total events



6.1.4 AP vs continuous low pressure mattress

Cavicchioli 2007Subtotal (95% CI)

Total events



6.1.5 AP vs visco-elastic foam mattress

Vanderwee 2005Subtotal (95% CI)

Total events



6.1.6 AP (NIMBUS 3) vs ROHO dry flotation

Malbrain, 2010Subtotal (95% CI)

Total events



Total (95% CI)

Total events




Events

15

15

39

4

11

5

59

7

1

5

13

2

2

34

34

2

2

125

Total

115115

72

16

32

25145

166

40

20226

6969

222222

88

785

Events

39

39

45

4

26

6

81

7

2

3

12

1

1

35

35

2

2

170

Total

115115

76

16

68

26186

155

40

37232

7171

225225

88

837

Weight

15.2%15.2%

27.3%

4.4%

14.2%

5.5%51.4%

5.8%

1.2%

3.7%10.7%

1.2%1.2%

19.3%19.3%

2.3%2.3%

100.0%

M-H, Random, 95% CI

0.38 [0.22, 0.66]0.38 [0.22, 0.66]

0.91 [0.69, 1.21]

1.00 [0.30, 3.32]

0.90 [0.51, 1.58]

0.87 [0.30, 2.48]0.91 [0.72, 1.16]

0.93 [0.34, 2.60]

0.50 [0.05, 5.30]

3.08 [0.82, 11.59]1.31 [0.51, 3.35]

2.06 [0.19, 22.18]2.06 [0.19, 22.18]

0.98 [0.64, 1.52]0.98 [0.64, 1.52]

1.00 [0.18, 5.46]1.00 [0.18, 5.46]

0.85 [0.65, 1.11]

AP CLP Risk Ratio


pressure

all grades of ulcer and condition

M-H, Random, 95% CI

0.38 [0.22, 0.66]0.38 [0.22, 0.66]

0.91 [0.69, 1.21]

1.00 [0.30, 3.32]

0.90 [0.51, 1.58]

0.87 [0.30, 2.48]0.91 [0.72, 1.16]

0.93 [0.34, 2.60]

0.50 [0.05, 5.30]

3.08 [0.82, 11.59]1.31 [0.51, 3.35]

2.06 [0.19, 22.18]2.06 [0.19, 22.18]

0.98 [0.64, 1.52]0.98 [0.64, 1.52]

1.00 [0.18, 5.46]1.00 [0.18, 5.46]

0.85 [0.65, 1.11]

Risk Ratio

M-H, Random, 95% CI

0.1 0.2 0.5 1 2 5 10Favours AP Favours CLP

373

374

Figure 23 – Pressure ulcer incidence – with and without neurological conditions


Study or Subgroup

6.2.1 AP vs Silicore or foam overlay - not neurological condition

Stapleton 1986

Whitney 1984Subtotal (95% CI)

Total events



6.2.2 AP vs Silicore or foam overlay - neurological condition

Conine 1990

Daechsel 1985Subtotal (95% CI)

Total events



Total (95% CI)

Total events




Events

11

5

16

39

4

43

59

Total

32

2557

72

1688

145

Events

26

6

32

45

4

49

81

Total

68

2694

76

1692

186

Weight

23.7%

32.0%

62.3%

68.0%

100.0%

AP Silicore or foam

Study or Subgroup

Cavicchioli 2007

Gebhardt 1996

Malbrain, 2010

Price 1999

Stapleton 1986

Vanderwee 2005

Total (95% CI)

Total events



Events

1

0

0

1

11

34

47

Total

69

23

8

40

32

222

394

Events

1

8

1

2

26

35

73

Total

71

20

8

40

68

225

432

Weight

1.5%

14.0%

2.3%

3.1%

25.6%

53.5%

100.0%

AP CLP


with and without neurological conditions

grade 2+ ulcers

6.2.1 AP vs Silicore or foam overlay - not neurological condition

6.2.2 AP vs Silicore or foam overlay - neurological condition


Weight

23.7%

8.4%32.0%

62.3%

5.7%68.0%

100.0%

M-H, Fixed, 95% CI

0.90 [0.51, 1.58]

0.87 [0.30, 2.48]0.89 [0.54, 1.47]

0.91 [0.69, 1.21]

1.00 [0.30, 3.32]0.92 [0.70, 1.22]

0.91 [0.71, 1.17]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours AP Favours silicore or foam

Weight

1.5%

14.0%

2.3%

3.1%

25.6%

53.5%

100.0%

M-H, Fixed, 95% CI

1.03 [0.07, 16.13]

0.05 [0.00, 0.84]

0.33 [0.02, 7.14]

0.50 [0.05, 5.30]

0.90 [0.51, 1.58]

0.98 [0.64, 1.52]

0.80 [0.58, 1.11]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours AP Favours CLP

KCE Report 193S

KCE Report 193S

Figure 25 – Drop-out due to discomfort

Figure 26 – Comfort rating at 14 days

Study or Subgroup

6.4.1 AP vs Silicore

Conine 1990Subtotal (95% CI)

Total events



Total (95% CI)

Total events




Events

19

19

19

Total

9393

93

Events

17

17

17

Total

9494

94

Weight

100.0%100.0%

100.0%

AP CLP

Study or Subgroup

6.5.1 dynamic flotation mattress and alternating-pressure cushion vs low-pressure inflatable mattress and cushion

Price 1999Subtotal (95% CI)



Total (95% CI)




Mean

60

SD

25

Total

2626

26

Mean

67

SD

18

Total

2424

24

Weight

100.0%100.0%

100.0%

IV, Fixed, 95% CI

-7.00 [-19.01, 5.01]-7.00 [-19.01, 5.01]

-7.00 [-19.01, 5.01]

AP CLP Mean Difference


Weight

100.0%100.0%

100.0%

M-H, Fixed, 95% CI

1.13 [0.63, 2.03]1.13 [0.63, 2.03]

1.13 [0.63, 2.03]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours AP Favours CLP

6.5.1 dynamic flotation mattress and alternating-pressure cushion vs low-pressure inflatable mattress and cushion

IV, Fixed, 95% CI

-7.00 [-19.01, 5.01]-7.00 [-19.01, 5.01]

-7.00 [-19.01, 5.01]


IV, Fixed, 95% CI

-100 -50 0 50 100Favours AP Favours CLP

375

376

Alternating-pressure and constant low-pressure in ICU/post


Study or Subgroup

8.1.1 Standard ICU/SFM post-ICU vs Nimbus AP ICU/SFM post-ICU

Laurent 1998

8.1.2 Standard ICU/SFM post-ICU vs standard ICU/Tempur CLP post-ICU

Laurent 1998

8.1.3 Nimbus AP ICU/SFM post-ICU vs standard ICU/Tempur CLP post-ICU

Laurent 1998

8.1.4 Standard ICU/SFM post-ICU vs Nimbus AP ICU/Tempur CLP post-ICU

Laurent 1998

8.1.5 Nimbus AP ICU/SFM post-ICU vs Nimbus ICU/Tempur post-ICU

Laurent 1998

8.1.6 Standard ICU/Tempur post-ICU vs Nimbus ICU/Tempur post-ICU

Laurent 1998

Events

4

14

10

14

10

11

Total

80

80

80

80

80

75

Events

10

11

11

10

10

10

Total

80

75

75

77

77

77

M-H, Fixed, 95% CI

Comparison 1 Comparison 2


pressure in ICU/post-ICU

all grades of ulcer

8.1.1 Standard ICU/SFM post-ICU vs Nimbus AP ICU/SFM post-ICU

8.1.2 Standard ICU/SFM post-ICU vs standard ICU/Tempur CLP post-ICU

8.1.3 Nimbus AP ICU/SFM post-ICU vs standard ICU/Tempur CLP post-ICU

8.1.4 Standard ICU/SFM post-ICU vs Nimbus AP ICU/Tempur CLP post-ICU

8.1.5 Nimbus AP ICU/SFM post-ICU vs Nimbus ICU/Tempur post-ICU

8.1.6 Standard ICU/Tempur post-ICU vs Nimbus ICU/Tempur post-ICU

M-H, Fixed, 95% CI

0.40 [0.13, 1.22]

1.19 [0.58, 2.46]

0.85 [0.38, 1.89]

1.35 [0.64, 2.85]

0.96 [0.42, 2.18]

1.13 [0.51, 2.50]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Comparison 1 Favours Comparison 2

KCE Report 193S

KCE Report 193S

Comparisons between alternating-pressure devices


Study or Subgroup

8.1.1 Airwave vs Large Cell Ripple

Exton-Smith 1982Subtotal (95% CI)

Total events



8.1.2 Airwave vs Pegasus Carewave

Hampton 1997Subtotal (95% CI)

Total events



8.1.3 Trinova vs control

Taylor 1999Subtotal (95% CI)

Total events



8.1.4 TheraPulse vs Duo

Theaker 2005Subtotal (95% CI)

Total events



8.1.5 AP overlay vs AP mattress

Nixon 2006Subtotal (95% CI)

Total events



8.1.6 Multi-stage inflation vs single-stage inflation

Demarre, 2012Subtotal (95% CI)

Total events




Events

5

5

0

0

0

0

3

3

106

106

68

68

Total

3131

3636

2222

3030

989989

298298

Events

12

12

0

0

2

2

6

6

101

101

56

56

Total

3131

3939

2222

3232

982982

312312

Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

0.42 [0.17, 1.04]0.42 [0.17, 1.04]

Not estimable

0.20 [0.01, 3.94]0.20 [0.01, 3.94]

0.53 [0.15, 1.94]0.53 [0.15, 1.94]

1.04 [0.81, 1.35]1.04 [0.81, 1.35]

1.27 [0.93, 1.74]1.27 [0.93, 1.74]

AP device Control Risk Ratio


pressure devices

all grades of ulcer

M-H, Fixed, 95% CI

0.42 [0.17, 1.04]0.42 [0.17, 1.04]

Not estimableNot estimable

0.20 [0.01, 3.94]0.20 [0.01, 3.94]

0.53 [0.15, 1.94]0.53 [0.15, 1.94]

1.04 [0.81, 1.35]1.04 [0.81, 1.35]

1.27 [0.93, 1.74]1.27 [0.93, 1.74]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours AP Favours Control

377

378


Study or Subgroup

8.2.1 Airwave vs Large Cell Ripple

Exton-Smith 1982Subtotal (95% CI)

Total events



8.2.2 Airwave vs Pegasus Carewave

Hampton 1997Subtotal (95% CI)

Total events



8.2.4 TheraPulse vs Duo

Theaker 2005Subtotal (95% CI)

Total events



8.2.5 AP overlay vs AP mattress


Total events





Total events



Events

5

5

0

0

3

3

106

106

17

17

Total

3131

3636

3030

989989

298298

Events

12

12

0

0

6

6

101

101

18

18

Total

3131

3939

3232

982982

312312

Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

AP device Control


grade 2+ ulcers

Weight

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

0.42 [0.17, 1.04]0.42 [0.17, 1.04]

Not estimableNot estimable

0.53 [0.15, 1.94]0.53 [0.15, 1.94]

1.04 [0.81, 1.35]1.04 [0.81, 1.35]

0.99 [0.52, 1.88]0.99 [0.52, 1.88]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours AP device Favours control

KCE Report 193S

KCE Report 193S

Figure 30 – Withdrawal due to discomfort

Low-air-loss vs standard bed



Study or Subgroup



Total events



Total (95% CI)

Total events




Events

11

11

11

Total

298298

298

Events

17

17

17

Total

312312

312

Weight

100.0%100.0%

100.0%

multi-stage single-stage

Study or Subgroup

Cobb 1997

Total (95% CI)

Total events



Events

6

6

Total

62

62

Events

12

12

Total

61

61

Weight

100.0%

100.0%

Low air loss Standard

Study or Subgroup

Bennett 1998

Cobb 1997

Inman 1993

Events

8

6

6

Total

42

62

49

Events

4

12

25

Total

56

61

49

Low Air Loss Standard ICU bed


all grades of ulcer

grade 2+ ulcers

Weight

100.0%100.0%

100.0%

M-H, Fixed, 95% CI

0.68 [0.32, 1.42]0.68 [0.32, 1.42]

0.68 [0.32, 1.42]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours multi-stage Favours single-stage

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.49 [0.20, 1.23]

0.49 [0.20, 1.23]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Low air loss Favours standard

M-H, Fixed, 95% CI

2.67 [0.86, 8.27]

0.49 [0.20, 1.23]

0.24 [0.11, 0.53]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Low Air Loss Favours Std ICU bed

379

380

Figure 33 – Pressure ulcer incidence – grade 2+ ulcers (pooled)

Operating table overlay vs no overlay


Study or Subgroup

Cobb 1997

Inman 1993

Total (95% CI)

Total events



Events

3

6

9

Total

62

49

111

Events

11

25

36

Total

61

49

110

Weight

30.7%

69.3%

100.0%

Low air loss Standard bed

Study or Subgroup

13.1.1 Viscoelastic polymer pad vs no overlay


Total events



13.1.2 Viscoelastic foam overlay vs no overlay

Feuchtinger 2006Subtotal (95% CI)

Total events



Events

22

22

13

13

Total

205205

8585

Events

43

43

9

9

Total

211211

9090

Weight

100.0%100.0%

100.0%100.0%

Overlay No Overlay


grade 2+ ulcers (pooled)

all grades of ulcer

Weight

30.7%

69.3%

100.0%

M-H, Fixed, 95% CI

0.27 [0.08, 0.91]

0.24 [0.11, 0.53]

0.25 [0.13, 0.49]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours low air loss Favours standard

Weight

100.0%100.0%

100.0%100.0%

M-H, Random, 95% CI

0.53 [0.33, 0.85]0.53 [0.33, 0.85]

1.53 [0.69, 3.39]1.53 [0.69, 3.39]


M-H, Random, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Overlay Favours No Overlay

KCE Report 193S

KCE Report 193S


Indentation load deflection operating room foam



Study or Subgroup

Feuchtinger 2006

Total (95% CI)

Total events



Events

2

2

Total

85

85

Events

1

1

Total

90

90

Weight

100.0%

100.0%

Overlay No Overlay

Study or Subgroup

Schultz 1999

Total (95% CI)

Total events



Events

55

55

Total

206

206

Events

34

34

Total

207

207

Weight

100.0%

100.0%

IDL Usual care

Study or Subgroup

Schultz 1999

Total (95% CI)

Total events



Events

6

6

Total

206

206

Events

3

3

Total

207

207

Weight

100.0%

100.0%

IDL Usual care


grades 2+ ulcers

foam mattress vs operating room usual care

all grades of ulcer

grade 2+ ulcers

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

2.12 [0.20, 22.93]

2.12 [0.20, 22.93]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours overlay Favours no overlay

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

1.63 [1.11, 2.38]

1.63 [1.11, 2.38]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours IDL Favours usual care

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

2.01 [0.51, 7.93]

2.01 [0.51, 7.93]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours IDL Favours usual care

381

382

Micropulse system for surgical patients


Figure 39 – Pressure ulcer incidence – grade 2+ ulcer

Visco-elastic A&E overlay and ward mattress vs standard A&E overlay and ward mattress


Study or Subgroup

Aronovitch 1999

Russell 2000

Total (95% CI)

Total events



Events

1

2

3

Total

90

98

188

Events

7

7

14

Total

80

100

180

Weight

100.0%

Micropulse System Std Care

Study or Subgroup

Aronovitch 1999

Total (95% CI)

Total events



Events

0

0

Total

90

90

Events

6

6

Total

80

80

Weight

100.0%

100.0%

Micropulse System Std Care

Study or Subgroup

Gunningberg 2000

Total (95% CI)

Total events



Events

4

4

Total

48

48

Events

8

8

Total

53

53

Weight

100.0%

100.0%

Visco-elastic foam Standard


all grades of ulcer

grade 2+ ulcer

elastic A&E overlay and ward mattress vs standard A&E overlay and ward mattress

grade 2+ ulcers

Weight

51.7%

48.3%

100.0%

M-H, Fixed, 95% CI

0.13 [0.02, 1.01]

0.29 [0.06, 1.37]

0.21 [0.06, 0.70]


M-H, Fixed, 95% CI

0.1 0.2 0.5 1 2 5 10Favours Micropulse Favours Standard

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.07 [0.00, 1.20]

0.07 [0.00, 1.20]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.55 [0.18, 1.72]

0.55 [0.18, 1.72]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours visco-elastic Favours foam

KCE Report 193S

KCE Report 193S


Profiling bed vs flat-based bed


Study or Subgroup

Gunningberg 2000

Total (95% CI)

Total events



Events

12

12

Total

48

48

Events

17

17

Total

53

53

Weight

100.0%

100.0%

Visco-elastic foam Standard

Study or Subgroup

Keogh 2001

Total (95% CI)

Total events



Events

0

0

Total

35

35

Events

0

0

Total

35

35

Weight

Profiling bed Foam mattress


all grades of ulcer

all grades of ulcer

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.78 [0.42, 1.46]

0.78 [0.42, 1.46]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours visco-elastic Favours standard


Not estimable

Not estimable


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours foam Favours profiling bed

383

384

Seat cushions


Study or Subgroup

17.1.1 Slab foam v Bespoke contoured foam

Conine 1993

Lim 1988Subtotal (95% CI)

Total events



17.1.2 Jay Gel Cushion v Foam

Conine 1994Subtotal (95% CI)

Total events



17.1.3 Pressure reducing cushion v Standard foam cushion

Geyer 2001Subtotal (95% CI)

Total events



17.1.4 Skin protection cushion vs segmented foam cushion - sitting-acquired pressure ulcer - ischial tuberosities ulcers

Brienza, 2010Subtotal (95% CI)

Total events



17.1.5 Skin protection cushion vs segmented foam cushion - combined ischial tuberosities and sacral/coccyx pressure ulcers

Brienza, 2010Subtotal (95% CI)

Total events




Events

85

19

104

17

17

6

6

1

1

12

12

Total

125

26151

6868

1515

113113

113113

Events

84

18

102

30

30

10

10

8

8

21

21

Total

123

26149

7373

1717

119119

119119

Weight

82.5%

17.5%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

100.0%100.0%

M-H, Fixed, 95% CI

1.00 [0.84, 1.18]

1.06 [0.75, 1.49]1.01 [0.86, 1.17]

0.61 [0.37, 1.00]0.61 [0.37, 1.00]

0.68 [0.33, 1.42]0.68 [0.33, 1.42]

0.13 [0.02, 1.04]0.13 [0.02, 1.04]

0.60 [0.31, 1.17]0.60 [0.31, 1.17]

Cushion 1 Cushion 2 Risk Ratio


all grades of ulcer

17.1.4 Skin protection cushion vs segmented foam cushion - sitting-acquired pressure ulcer - ischial tuberosities ulcers

17.1.5 Skin protection cushion vs segmented foam cushion - combined ischial tuberosities and sacral/coccyx pressure ulcers

M-H, Fixed, 95% CI

Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Cushion 1 Favours Cushion 2

KCE Report 193S

KCE Report 193S

Figure 44 – Withdrawal due to discomfort

Face pillows

Figure 45 – Incidence of pressure ulcers – all grades of ulcer

Figure 46 – Incidence of pressure ulcers – grade 2+ ulcers

Study or Subgroup

Conine 1994

Total (95% CI)

Total events



Events

1

1

Total

83

83

Events

6

6

Total

80

80

Jay gel cushion Foam cushion

Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

10

10

Total

22

22

Events

0

0

Total

22

22

OSI face pillow ROHO face pillow

Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

2

2

Total

22

22

Events

0

0

Total

22

22

OSI face pillow ROHO face pillow


all grades of ulcer

grade 2+ ulcers

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.16 [0.02, 1.30]

0.16 [0.02, 1.30]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Jay gel Favours foam

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

12.55 [3.11, 50.57]

12.55 [3.11, 50.57]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100OSI positioner pillow ROHO pillow

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

7.75 [0.47, 128.03]

7.75 [0.47, 128.03]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours OSI Favours ROHO

385

386




Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

10

10

Total

22

22

Events

0

0

Total

22

22

Weight

100.0%

100.0%

OSI face pillow Dupaco face pillow

Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

2

2

Total

22

22

Events

0

0

Total

22

22

OSI face pillow Dupaco face pillow

Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

0

0

Total

22

22

Events

0

0

Total

22

22

ROHO face pillow Dupaco face pillow


all grades of ulcer

grade 2+ ulcers

all grades of ulcer

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

12.55 [3.11, 50.57]

12.55 [3.11, 50.57]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours OSI face pillow Favours Dupaco face pillo

Total

22

22

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

7.75 [0.47, 128.03]

7.75 [0.47, 128.03]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours OSI Favours Dupaco

Total

22

22


Not estimable

Not estimable


Peto, Fixed, 95% CI


KCE Report 193S

KCE Report 193S



Table 26 – MCINNES2011


Author and year:McInnes 2011

Title: Supportsurfaces for pressureulcers prevention(Review)

Journal: CochraneDatabase ofSystematic Reviews2011, Issue 4.

N of studies: 53

Inclusion criteria:

Population: peoplereceiving health carewho were thought at riskof developing pressureulcers, in any settings.Patients could haveexisting pressure ulcersbut only the incidence ofnew pressure ulcerswas looked at.

Studies: RCTs andquasi-randomised trialscomparing supportsurfaces and measuredthe incidence of newpressure ulcers.

Exclusion criteria: see

Study or Subgroup

Grisell, 2008

Total (95% CI)

Total events



Events

0

0

Total

22

22

Events

0

0

Total

22

22

ROHO face pillow Dupaco face pillow


grade 2+ ulcers


Comparison

Outcomemeasures

Quality assessment

who were thought at risk

existing pressure ulcersbut only the incidence of

randomised trials

surfaces and measured

see

Low-tech CLP supportsurfaces:

Standard foammattresses

Alternative foammattresses/overlays (egconvoluted foam, cubedfoam)

Gel-filledmattresses/overlays

Fibre-filledmattresses/overlays

Air-filledmattresses/overlays

Water-filledmattresses/overlays

Bead-filledmattresses/overlays

Sheepskins

High-tech support

Primaryoutcomes:incidence ofpressure ulcers

Grades of newpressure ulcers

Does the review address anappropriate question relevantto the guideline reviewquestion? yes

Does the review collect thetype of studies you considerrelevant to the guidelinereview question? yes

Was the literature searchsufficiently rigorous toidentify all relevant studies?yes

Was study quality assessedreported? yes

Was an adequate descriptionof the methodology used andincluded, and the methodsused are appropriate to thequestion? yes

Secondaryoutcomes: costof the devices;patient comfort;durability/longevity of the devices;acceptability ofthe devices forhealthcare staff;quality of life

Total

22


Not estimable

Not estimable


Peto, Fixed, 95% CI


387

Quality assessment Comments

Does the review address ane question relevant

to the guideline reviewquestion? yes


Was the literature searchsufficiently rigorous toidentify all relevant studies?

y quality assessedreported? yes


Quality grade:very low risk ofbias

388


above.

Population:

Studies: only reportingsubjective measures ofoutcome; only reportedproxy measures such asinterface pressure.

Details of studiesincluded: 27 studiesincluded participantswithout pre-existingpressure ulcers; 8included patients withgrade 1 or abovepressure ulcers; 4 didnot specify the gradingof the pre-existing ulcersand one included peoplewith grade 4 pressureulcers only. 12 studiesthe baseline skin statuswas unclear.

Five studies evaluateddifferent operating tablesurfaces; 9 evaluateddifferent surfaces inintensive care units; 8confined evaluation toorthopaedic patients;one involved both A&Eand ward setting; fivewere in extended care



Comparison

Outcomemeasures

Quality assessment

proxy measures such as

existing ulcersand one included people

the baseline skin status

ferent operating table

surfaces:

AP mattresses/overlays

Air-fluidised beds

Low-air-loss beds

Other support surfaces

Turning beds/frames

Operating table overlays

Wheelchair cushions

Limb protectors

KCE Report 193S


KCE Report 193S


facilities; 3 were innursing homes, 7involved two or moredifferent hospital wards;15 did not specify thestudy setting.

11 trials evaluatedcushions, 4 evaluatedsheepskins, 4 looked atturning beds/tables; 16examined overlays and2 looked at mattress; 3evaluated foamsurfaces, 2 evaluatedwaffle surfaces. Manystudies had multipleinterventions.

Many studies had asmall sample size andonly 20 reported a priorisample size calculation.



Comparison

Outcomemeasures

Quality assessment

different hospital wards;

sheepskins, 4 looked at

only 20 reported a priorisample size calculation.

389


390

Table 27 – BRIENZA2010


Author and year:Brienza 2010

Title: A randomizedclinical trial onpreventing pressureulcers withwheelchair seatcushions

Journal: J Am GeriatrSoc (2010)December; 58 (12),2308-2314.

Sequence generation:1:1 randomisationscheme prepared bya research teammember who wasindependent to thosewho had contact withparticipants.Randomised blocksof varying lengthused.

Allocationconcealment:adequate, see above.

Blinding: not possibledue to thedifferences inconfiguration andweight of the

Patient group: Elderly,nursing home populationwho used wheelchairsas primary means ofseating and mobility andwere at-risk fordeveloping pressureulcers.

All patients

Randomised N: 232(222 receivedintervention)

Completed N: 190

Drop-outs: 42

Age: 86.7 (s.d 7.6 years)

Ethnicity: 92.2% white.

Gender: 84.9% female.

Group 1 (SPC)

Randomised N: 113

Completed N: 86

Dropouts: 27 (6 did notreceive intervention, 5voluntarily withdrew, 16other)

Age:86.8 (s.d 7.4)

Gender (f):91 (80.5%)



Comparison

Outcomemeasures

Effect sizes

nursing home population

seating and mobility and

Age: 86.7 (s.d 7.6 years)

Gender: 84.9% female.

27 (6 did not

voluntarily withdrew, 16

Group 1: skin protectioncushion (SPC)

Group 2: segmented foamcushion (SFC)

Treatment started withseating assessment byoccupational therapist trainedin seating and mobility.

SPC group had acommercially availablecushion with an incontinencecover. Selected from agroup of three designed toimprove tissue tolerance byreducing peak pressuresnear bony prominences,accommodating orthopaedicdeformities throughimmersion, enveloping smallirregularities at the seatinginterface without causinghigh pressure gradients, anddissipating heat andmoisture. Solid seat insertswere provided. Multiple SPCgroup cushions were neededto allow for cushion selectionbased upon specific clinicalconditions. Clinical judgment

Outcome 1:Incidence of asitting-acquiredpressure ulcer –ischialtuberosities ulcers

Group 1 (SPC)(0.9%)

Group 2 (SFC)(6.7%)

P<0.04

Stage 1 ulcers (n(n=7), and unstageable (n=1)

Outcome 2:Incidence ofcombined ischialtuberosities andsacral/coccyxpressure ulcers:

Group 1(10.6%)

Group 2(17.6%)

33 participants had 38 IT andsacral /coccyx pressureulcers. Stage 1 (n=6), stage2 (n=29), stage 3 (n=2),unstageable (n=1).

P: NS

KCE Report 193S

Comments

(SPC): 1/113

(SFC): 8/119

Stage 1 ulcers (n=1), stage 2(n=7), and unstageable (n=1)


Limitations:baselinedifferences. Thestudy could notcontrol for othersupport surfaces.

Additionaloutcomes: N/A

Notes: a pilotstudy wasconducted prior tothe clinical trial toassist indevelopingmethods and todetermineappropriatesample size.

The authors statethat the RCTcould havelowered the risklevel as thewheelchair fit andfunction wasmonitored and

(SPC): 12/113

(SFC): 21/119

33 participants had 38 IT andsacral /coccyx pressure

Stage 1 (n=6), stage2 (n=29), stage 3 (n=2),unstageable (n=1).

KCE Report 193S


cushions, outcomeassessors weremasked.

Addressing incompleteoutcome data:missing data wasdue to voluntarywithdrawal death orother – examplesgiven. ITT analysisused. Missing datacovered with flowdiagram.

Statistical analysis:Rate of pressureulcers ITT analysis.Kaplan-Meier used toestimate thecumulative incidenceof pressure ulcers,with the log-rankstatistic used toassess differencesby treatment group.

Baseline differences:no statisticallysignificantdifferences exceptambulation. Slightlyfewer males in theSFC group (10.9%)than the SPC group(19.5%).

ethnicity (white):103(91.2%)

BMI:24.6 (s.d 4.4)

Total Bradenscore:15.4 (s.d 1.4)

Incontinent:97 (90.7%)

Ambulation: 0 feet: 67(62.6%); </= 10 feet: 14(13.1%), >10 feet: 26(24.3%)

Could not walkunassisted: 62.6%

Could walk 3 meters orless:13.1%

Could walk 3 meters ormore: 24.3%

Group 2 (SFC)

Randomised N: 119

Completed N: 94

Dropouts: 25 (4 did notreceive intervention, 6voluntary withdrawn, 14other, 1 discharged).

Age:86.6 (s.d 7.8)

Gender (f):106 (89.1)

ethnicity (white):111(93.3%)

BMI:25.0 (s.d 5.2)



Comparison

Outcomemeasures

Effect sizes


(62.6%); </= 10 feet: 14

ld walk 3 meters or

Could walk 3 meters or

25 (4 did not

voluntary withdrawn, 14

and expertise of the teamwas used to select aparticular SPC cushionbased on its compatibilitywith the subject’s clinicalneeds and preferences.

SFC group received a 7.6cmthick, segmented foamcushion fitted with anincontinence cover, and solidseat insert. This cushionwas chosen as the controlbecause it is representativeof a large number ofcushions currently used innursing homes.

Both groups: interfacepressure measurement datawas used to monitor theeffects of adjustments madeto the wheelchair. Eachparticipant received a new,properly fitted wheelchair.Two models were used. Onechair (Guardian Escort wasused and floor to seat heightis fixed at 51 cm,adjustments are possible, butnot easily accomplished.Subjects needing analternate seat-to-floor height

391

Comments

adjustedregularly.Pressure mappingused to assist inselection of skinprotection wheelchair cushions.

392


Study power/samplesize: powercalculation done 90%power required asample size of 234.

Setting: 12 nursinghomes (profit andnonprofit) in theGreater PittsburghArea. 180 licensedbeds.

Length of study: 6months.

Assessment of PUs:Sitting-acquiredpressure ulcer wasthose occurringprimarily over theischial tuberositieswhile sacral ulcersprimarily result fromexcessive loading inbed.

Weekly skin and riskassessments(Braden Score) wereperformed by aresearch nursemasked to thetreatmentassignment.Assessmentscontinued until first

Total Bradenscore:15.5 (s.d 1.5)


Ambulation: 0 feet: 86(76.1%), </=10 feet: 5(4.4%); > 10 feet 22(19.5%)

Could not walkunassisted: 76.1%

Could walk 3 meters orless: 4.4%

Could walk more than3 meters: 19.5%

Inclusion criteria: LTCresident 65 years of ageor older; Braden scoreof </=18 ( at risk fordeveloping pressureulcers; combinedBraden Activity andMobility Subscale score</=5; absence of ischialarea pressure ulcers;tolerance for dailywheelchair sitting time>/=6 hours; and abilityto accommodate seatingand positioning needswith the wheelchairselected for use in thisstudy.



Comparison

Outcomemeasures

Effect sizes

97 (85.8%)

Could walk 3 meters or

ore than

LTCresident 65 years of age

Mobility Subscale score</=5; absence of ischial

to accommodate seating

were given a Breezy Ultra 4wheelchair. The differencebetween groups for differentwheelchair was non-significant.

Wheelchairs and cushionswere checked weekly be theseating specialist andrepaired or adjusted asneeded.

KCE Report 193S

Comments

KCE Report 193S


incidence of apressure ulcer,discharge from thefacility, voluntarywithdrawal from thestudy, death, or thestudy end date 6months from theinitiation of theseating intervention.


Exclusion criteria:Body weight exceeding113kg (exceedswheelchair weightcapacity); hip widthexceeding 51cm(exceeds wheelchairwidth capacity);wheelchair seatingrequirements for headsupport, seat depth>46cm, oraccommodation ofsevere orthopaedicdeformities of the pelvis,lower extremities orback that exceed thecapability of the studywheelchairs; and currentuse of any cushioningmaterial(s) other thanthe SFC or equivalent,or a lower qualitycushion.



Comparison

Outcomemeasures

Effect sizes

deformities of the pelvis,

lchairs; and current

393

Comments

394

Table 28 – DEMARRE2012


Author and year:Demarre 2012

Title: Multi-stageversus single-stageinflation and deflationcycle for alternatinglow pressure airmattresses to preventpressure ulcers inhospitalised patients:a randomised-controlled clinical trial

Journal: InternationalJournal of NursingStudies, 47 (2012),416-426.

Type of study: multi-centre RCT

Sequence generation:randomised on 1:1ratio by simplerandomisation. Thesequence was basedon computer-generated list ofrandom numbers.

Allocation concealment:Nurses contactedresearcher andreceived a number fortype of allocated

Patient group:hospitalised patients.The wards wereneurology (n=6),rehabilitation (n=3),cardiology (n=2),dermatology (n=1),pneumology (n=1),oncology (n=1) andchronic care (n=1) or acombination of differenttypes of medicalconditions (n=2).

All patients

Randomised N: 610

Completed N: 307

Drop-outs: 303

Group 1

Randomised N: 298

Completed N: 152

Dropouts: 146

(PU category II-IV(n=17), losses tofollow-up because of:technical problems(n=3), discomfort(n=11), reason not


Intervention

Comparison

Outcomemeasures

Effect sizes

chronic care (n=1) or acombination of different

Group 1: ALPAM with multi-stage inflation and deflationof the air cells. The inflationcurve of the air cell wasidentical to the deflationcurve of t air cell. The headzone contained 3 air cellswith a continuous lowpressure, the heel zonecontained 7 cells with acontinuous ultra low pressureand the back and sacrumzone contained 10alternating low pressurecells. A sensor at the sacralzone measured the appliedpressure of the body on themattress. The deviceconsisted of a mattress and acontrol unit. Cycle times forinflation and deflation werebetween 10 and 12 minutes.The air cell width was 10cm.

Group 2: standard ALPAM.An ALPAM with a standardsingle-stage, steep inflationand deflation of the air cells.All air cells were alternating,the cycle time was 10minutes and the air cell widthwas 10cm. An externalmanual control unit was used

Outcome 1:Cumulativeincidence ofpressure ulcergrade II-IV (%developing a newpressure ulcer):

Group 1:17/298 (

Group 2: 18/312 (

P=0.97

Outcome 2: Non-blanchableerythema(pressure Grade1)

Group 1: 51/298 (17.1%)

Group 2: 38/312 (12.2%)

P=0.08

Outcome 3:excludingpressure ulcers(Grade II-IV)occurring in thefirst 3 days afteradmission in thestudy (whichcould have beencaused by tissuedamage prior tostart of study)

Group 1: (3.4%)

Group 2: (4.2%)

P=0.61

Binary logistic regressionanalysis: OR 1.17 (95% CI0.553-2.455), x2 = 0.16,df=1, p=0.687)

Outcome 4: Ti todevelop apressure ulcer(median time)

Group 1: 5.0 days (IQR 3.08.5)

Group 2: 8.0 days (IQR 3.08.5)

Mann-Whitney Up=0.182.

KCE Report 193S

Comments

17/298 (5.7%)

18/312 (5.8%)

Funding:Financiallysponsored byGhent Universityas part of a PhDstudy. Authorsstate that themattresses andcushions wereprovided by Hill-Rom but they didnot influence thestudy.

Limitations: Noblinding ofoutcomeassessors. Highdrop-out in bothgroups. Bothgroups had somepatients withpatients who hadgrade I ulcersalready (15.4%).

Additionaloutcomes:Incidence ofgrade II, grade III,Grade IV,

51/298 (17.1%)

38/312 (12.2%)

(3.4%)

(4.2%)

Binary logistic regressionanalysis: OR 1.17 (95% CI

2.455), x2 = 0.16,df=1, p=0.687)

5.0 days (IQR 3.0-

8.0 days (IQR 3.0-

Whitney U-test = 113,

KCE Report 193S


mattress (first oncomputer generatedlist).

Blinding: blinding notpossible due todifferences in externalcontrol unit of themattresses studied.No information wasgiven to the nursesregarding thedifferences inmattresses. Outcomeassessors notblinded.

Addressing incompleteoutcome data: flowchart with detailedreasons for drop-outgiven. High drop-out(in both groups). ITTanalysis used.

Statistical analysis: datapresented in %s andmeans if normallydistributed data andmedians of notnormally distributed.T-tests used innormally distributedcontinuous data.Mann-Whitney u-testsfor non-normallydistributed

defined (n=3), transferto another ward (n=15),discharge to home(n=40), death (n=15),discharge to anotherinstitution (n=42))

Group 2

Randomised N: 312

Completed N: 155

Dropouts: 157

(PU category II-IV(n=18), losses tofollow-up because of:technical problems(n=3), discomfort(n=16), reason notdefined (n=5), transferto another ward (n=22),discharge to home(n=41), death (n=14),discharge to anotherinstitution (n=37),withdrawal of consent(n=1))

Inclusion criteria: atrisk for pressure ulcerdevelopment accordingto the Braden scale.

Exclusion criteria:having a pressure ulcer


Intervention

Comparison

Outcomemeasures

Effect sizes

defined (n=3), transferto another ward (n=15),

defined (n=5), transferto another ward (n=22),

development according

having a pressure ulcer

to adjust the mattress to thepatient's weight.

Both mattresses werecovered with an identicalmattress cover. No standardrepositioning protocol wasused in bed. An identicalseating protocol was used inboth groups. All patientswere seated on a static aircushion. The control unitwas disconnected duringtransport of the patient,resulting in an inflatedmattress for 2 hours withoutalternating air cells.

Outcome 5:acceptability ofthe devices –number whowithdrew due todiscomfort

Group 1: 11/298

Group 2: 17/312

395

Comments

11/298

17/312

incontinence-associateddermatitis.Incidence forvarious areas –pelvic area(sacral; hip); heelarea (heel, ankle);other. Probabilityto remainpressure free.

396


continuous data. Chi-square and Fisher'sexact tests forcategorical variables.

Baseline differences: nosignificant differences

Study power/samplesize: powered for 600patients (300 in eachgroup).

Setting: 25 wards from5 Belgian hospitals.

Length of study: 14days follow-up

Assessment of PUs:pressure ulcersclassified by EPUAPclassification system.Skin assessment dailyby nurses.Transparent plasticdisc method used toobserve non-blanchable erythema(Grade 1).


Grade II-IV onadmission; theexpected admissiontime in the hospital was< 3 days; aged < 18years; there was a 'donot resuscitate code'specifying ending alltherapeuticinterventions; weightwas less than 30kg ormore than 160kg(mattressspecification);

Informed consent couldnot be obtained frompatient or his/her legalrepresentative.


Intervention

Comparison

Outcomemeasures

Effect sizes

time in the hospital was

Informed consent could

patient or his/her legal

KCE Report 193S

Comments

KCE Report 193S

Table 29 – VANLEEN2011


Author and year: VanLeen (2011)

Title: Pressure relief,cold foam or staticair? A single center,prospective,controlledrandomized clinicaltrial in a Dutchnursing home

Journal: Journal ofTissue Viability(2011), 20,30-34.

Type of study: singlecentre RCT.

Sequence generation:numbered envelopes

Allocationconcealment:numbered envelopes


Addressing incompleteoutcome data: ITTanalysis used. Statethat those who dieddid not developpressure ulcers.

Statistical analysis:using SPSS 15.0. Nofurther details.

Patient group: nursinghome residents

All patients

Randomised N: 83

Completed N: 74

Drop-outs: 5 diedduring study in group 1and 4 died during studyin group 2, none of thepatients who dieddeveloped a pressureulcer during theirparticipation.

Group 1

Randomised N: 42

Completed N: 38

Dropouts: 4 (died)

Age (mean, s.d): 81.1(8.37)

Gender (females): 33

Norton 5-8 at start ofstudy: 26 (61.9%)


Diagnoses

Dementia: 31 (73.8%)

CVA: 8 (19%)



Comparison

Outcomemeasures

Effect sizes

nursing

and 4 died during study

Group 1: combination of astandard 15cm cold foammattress with a static airoverlay

Group 2: a standard 15cmcold foam mattress

All patients: when out of bed,sitting on a static air pillowfollowing the institutionalPUPP. At night, nobodyreceived repositioningconforming to this PUprotocol.

No repositioning was allowedbefore development of agrade 2 pressure ulcer.

Outcome 1:development ofgrade 2, 3 and 4pressure ulcers(EPUAPclassification) atthe heel or in thesacral/hip region.

Incidence ofpressure ulcers:

Group 1: 2/42 ITT (4.8%)

Group 2: 7/41 ITT (17.1%)

P=0.088 (Fisher’s exact test)

(95% CI 1.3% to 25.9%)

Outcome 2:Incidence ofGrade 2 ulcers:

Group 1: 0/42

Group 2: 1/41

Outcome 3:Incidence ofGrade 3 ulcers:

Group 1: 1/42

Group 2: 5/41

Outcome 4:Incidence ofGrade 4 ulcers

Group 1: 0/42

Group 2: 0/41

Outcome 5:

397

Comments

2/42 ITT (4.8%)

7/41 ITT (17.1%)

P=0.088 (Fisher’s exact test)

(95% CI 1.3% to 25.9%)

Funding: nofunding.

Limitations:Ethical issues ofnot usingrepositioning.Limited details ofsequencegeneration andallocationconcealment. Nodetails of blindingof outcomeassessors. Smallstudy.

Additionaloutcomes:incidence ofpressure ulcers ingroups at Nortonscale risk 5-8 and9-12, for Grade2,3 and 4 ulcers

The authorsprotocol iscontrary tonationalguidelines forpressure ulcer

0/42

1/41

1/42

5/41

0/42

0/41

398


Baseline differences:there were morepatients in theintervention groupwith a very lowNorton score (morepressure ulcer pronepatients).

Study power/samplesize: power of 80%required 38 patientsin each group

Setting: Nursinghome, DeNaaldhorst, theNetherlands.

Length of study:patients werefollowed for a periodof 6 months.

Assessment of PUs:not reported.

Risk of pressureulcers assessed byNorton scale.

Multiple ulcers: notreported.

Rheumatoid arthritis: 1(2.4%)

Encephalopathy: 0

m. Parkinson: 1 (2.4%)

Diabetes: 0

Arthrosis: 0

Hip fracture: 1 (2.4%)

COPD: 0

Group 2

Randomised N: 41

Completed N: 36

Dropouts: 5 (died)

Age (mean, s.d): 83.1(7.86)

Gender (females): 34(82.9%)



Diagnoses:

Dementia: 31 (75.6%)

CVA: 4 (9.8%)

Rheumatoid arthritis: 0

Encephalopathy: 1(2.4%)

m. Parkinson: 1 (2.4%)

Diabetes: 1 (2.4%)



Comparison

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

preventionregardingrepositioning for 2reasons:interference insleep and thehigher workloadfor nursing staffand theaccompanyinghigher costs.

KCE Report 193S


Arthrosis: 1 (2.4%)

Hip fracture: 1 (2.4%)

COPD: 1 (2.4%)

Inclusion criteria: age>65, Norton scorebetween 5-12; informedconsent of patients orrepresentatives in caseof mental disorders.

Exclusion criteria: apressure ulcer in theprevious 6 months

Table 30 – GRISELL2008


Author and year: Grisell(2008)

Title: Face tissuePressure in PronePositioning: acomparison of threeface pillows while in theprone position forspinal surgery.

Journal: SPINE, 33 (26),2938-2941.

Type of studyprospectiverandomised trial.

Patient group:elective surgerypatients – thoracic,lumbar or thora-columbar spinalsurgery that requiredprone positioning

All patients

Randomised N: 66

Completed N: 66

Drop-outs: 0



Comparison

Outcomemeasures

Effect sizes

2; informed

Intervention

Comparison

Outcomemeasures

Effect sizes

surgery that required

3 different types of facepillows that are used forprone positioning in theoperating room:

Group 1: a neoprene airfilled bladder (dry flotation)device by ROHO

Group 2: the OSI(orthopaedic systemc inc)(disposable polyurethanefoam prone head positioner)

Group 3: the Prone ViewProtective Helmet system (a

Outcome 1:incidence ofpressure ulcers

Group 1: 0/22

Group 2: 10/22

Group 3: 0/22

Outcome 2:incidence of stage1 pressure ulcers

Group 1: 0/22

Group 2: 8/22

Group 3: 0/22

Outcome 3:incidence of stage2 pressure ulcers

Group 1: 0/22

Group 2: 8/22

Group 3: 0/22

399

Comments

Comments

0/22

10/22

0/22

Funding: Notreported.

Limitations:Aimed at tissueinterfacepressures ratherthan incidence ofpressure ulcers.No details ofallocationconcealment orblinding ofoutcome

0/22

8/22

0/22

0/22

8/22

0/22

400


Sequence generation:randomisation listmentioned and wasconsulted forassignment ofpositioner before startof surgery.Randomisation list wasgenerated usingwebsitewww.randomization.com – which usesrandomly permutatedblocks to assign eachsubject to a pillow.

Allocation concealment:no details

Blinding: the patient wasunaware of theirassigned positionertype at all times. Nodetails of otherblinding.

Addressing incompleteoutcome data: allpatients completed thestudy.

Statistical analysis:Nonparamatericstatistical methodsused because of smallsample sizes. Mann-Whitney U was used toanalyse measures of

Group 1

Randomised N: 22

Completed N: 22

Dropouts: 0

Group 2

Randomised N: 22

Completed N: 22

Dropouts: 0

Group 2

Randomised N: 22

Completed N: 22

Dropouts: 0

Inclusion criteria:aged 18 to 65 years(inclusive);presenting to theoperating room forelective thoracic,lumbar, or thora-columbar spinalsurgery that requiredprone positioningwere included.

Exclusion criteria:patients with anyfacial skin ailment orlesion (rash, abrasion


Intervention

Comparison

Outcomemeasures

Effect sizes

surgery that required

facial skin ailment orlesion (rash, abrasion

disposable polyurethanefoam head positioner)

All patients: positioned proneon a Jackson table usingstandard positioning. A lowprofile pressure sensor waspositioned between thesubject’s forehead and thepillow and between thesubject’s chin and the pillow.

Procedures lasted from 1 to12 hours.

KCE Report 193S

Comments

assessors. Smallsample size. Nodetails ofpopulationcharacteristicsand baselinedifferences.

Did not stratify byage, gender,surgery type,surgery locationor surgery length(other than therequirement thatsurgery last atleast 1 hour)

Additionaloutcomes: tissueinterface pressure

Studies mainaims wereregarding tissuepressures.

No statistics wereused to evaluatethe lengths ofprocedures butthe authors statethat the averagetime for the

KCE Report 193S


central tendency andvariability of the tissuepressures measured.The Friedman analysiswas used to evaluateand assess thedifferences across timeat each of the timevariables measured.

Baseline differences: nodetails

Study power/sample size:80% power required 20patients in each group.

Setting: surgery

Length of study: nodetails except range ofsurgery times.

Assessment of PUs:Authors say anypressure ulcers seenwere staged accordingto the NPUAP stagingsystem.

Multiple ulcers: therewere multiple ulcers butgave details of numberof patients.

infection, redness,inflammation,bruising); history ofincreased intraocularpressure orglaucoma; patientspresented foremergent spinalsurgery; patients forsurgery that includedany cervical level;patients whose majorlanguage was notEnglish.


Intervention

Comparison

Outcomemeasures

Effect sizes

increased intraocular

surgery that included

patients whose major

401

Comments

procedures oneach of thepositioners wassimilar.

402

Table 31 – MISTIAEN2010E


Author and year: Mistiaen(2010)

Title: The effectiveness ofthe Australian MedicalSheepskin for theprevention of pressureulcers in somatic nursinghome patients: Aprospective multicenterrandomized-controlledtrial (ISRCTN17553857)

Journal: Wound Rep Reg(2010), 18, 572-579.

Type of study: multicenterprospective RCT

Sequence generation:Randomisation schemecreated in SPSS byassigning theintervention to a randomsample of around 50% ina list of 1,500 numbersand assigning the controlgroup to the restRandomisation was doneon admission day or atleast within 48 hoursafter admission.

Allocation concealment:Adequate. The sequencegeneration was then

Patient group:nursing homepatients

All patients

Randomised N:588

Completed N: 543

Drop-outs: 45

Group 1

Randomised N:295

Completed N: 271

Dropouts: 24

Gender (female %):71%

Age mean (range):78 (26-97)

Barthel score mean:9.9

Patients with risk onpressure ulcer %(Braden score</=20): 70

Patients with risk onpressure ulcer %(Braden score


Intervention

Comparison

Outcomemeasures

Effect sizes

Gender (female %):

Barthel score mean:

Patients with risk on


Group 1: All usual care andthe application of theAustralian MedicalSheepskin (AMS) (hi-temp,urine resistant, size XXL) asan overlay on top of thestandard mattress in the areaof the buttocks. An extraAMS at the bottom of the bedand in the (wheel) chair wasalso permitted. Theapplication of the AMSstarted no later than 48 hoursafter admission. The AMSwas then applied during thefirst 30 days after admissionor until a patient died or wasdischarged, whichever camefirst.

All other usual pressure ulcerpreventive interventions suchas mobilisation andrepositioning could be addedas co-interventions as far aswere usual care in thenursing homes. All othernursing care could becontinued as usual (includingincontinence materials)

Group 2: Control groupreceived usual care only,

Outcome 1:incidence ofsacral pressureulcers in the first30 days afteradmission

Group 1: 24/271 (8.9%) ACA

Group 2: 40/272 (14.7%)ACA

Two-sided x2

Outcome 2:incidence ofpressure ulcerson other areas

Group 1: 16.4%

Group 2: 15.1%

X2, p=0.69

Outcome 4:comfort of thesheepskin asexperienced bythe patients (self-developed seven-timequestionnaire witha five-point ratinganswer structure)– softness,itching, smell,warmth, tickling,comfort, if wouldrecommend toother patients;additionalcomments

(209 filled out questionnaire)

Too warm: one third

Recommend AMS to otherpatients: 52%, no judgement26%, would not recommend22%.

Compliance to AMS:

Group 1: 1/3 of patients inthe sheepskin groupdiscontinued the use of theMAS, mostly within the firstweek and mainly becausethey found it too warm. Thesheepskin was almost neverapplied under the heels or inthe chair.

In the controthe observable days was

KCE Report 193S

Comments

24/271 (8.9%) ACA

40/272 (14.7%)

2, p=0.035

Funding: grantfrom the EfficacyResearchProgram, round2007, of theNetherlandsOrganisation forHealth Researchand Development.

Limitations: noblinding . Unclearaddressing ofincompleteoutcome data.

Additionaloutcomes: onsetday of pressureulcers; usual carecomponents byintervention group(table given). Nosignificantdifferences inusual carecomponent.

16.4%

15.1%

(209 filled out questionnaire)

Too warm: one third

Recommend AMS to otherpatients: 52%, no judgement26%, would not recommend

liance to AMS:

Group 1: 1/3 of patients inthe sheepskin groupdiscontinued the use of theMAS, mostly within the firstweek and mainly becausethey found it too warm. Thesheepskin was almost neverapplied under the heels or in

In the control group, 1.7% ofthe observable days was

KCE Report 193S


blinded on a paper listnumbered 1 through1,500 by a secretary notfurther involved in theproject. The admittingnurse called the principalinvestigator who thendisclosed the allocationfrom the blinded list tothe nurse, who thendisclosed to patient.

Blinding: Not possible toblind if someone is in theexperimental group ornot, only the patientallocation itself wasblinded to all partiesinvolved. Checking wasdone to see thatallocated interventionwas correctly applied.

Addressing incompleteoutcome data: ITTanalysis was used. Mainreason for not obtainingoutcome data wasprimarily nursesforgetting to send theforms or discarded byaccident when a patientdied or was dischargedhome or transferred toanother institution or lostin the mail.

</=18): 47

BMI mean: 24.6

Group 2

Randomised N:293

Completed N: 272

Dropouts: 21

Gender (female %):67%

Age mean (range):78 (27-98)

Barthel score mean:9.4



BMI mean: 25.6

Inclusion criteria:patients newlyadmitted for aprimarily somaticreason, adult (aged


Intervention

Comparison

Outcomemeasures

Effect sizes

Gender (female %):

Barthel score mean:



reason, adult (aged

including all the pressure-reducing interventions andother preventive actions,normally taken in theparticipating nursing homes.The application, in any form,of the AMS was forbidden inthis group during the first 30days after admission.

In both groups: theregular/usual mattresseswere applied (differed frominstitution to institution andeven ward to ward). Woundcare specialists were allowedto start with a specialpressure-reducing mattressfor a patient during theobservation period when theyconsidered this necessaryand was required to be notedon the daily observationform.

spent with an AMS, thisoccurred in the beginning ofthe study period, because itwas then not entirely clear tothe nurses when they wereallowed to give an AMS tothe patients.

Mean onset dayof pressureulcers (daysafter admission):

Group 1: 12

Group 2: 9

Outcome 5: easeof use of thesheepskin asexperienced bythe carepersonnel(measured bygroup interviewswith ward nurseson threeoccasions)

Nurses did not encounterdifficulties in using AMS indaily practice, but it did makeit slightly more difficult tochange bed linen in bedridden patients. Also thedirty sheepskins neededseparate linen bags causedsome inconvenience.

Outcome 6:quality of life(visual analogscale 0=worsthealth status ever100=the best thatcould beimagined) mean

Group 1: 62.1

Group 2: 61.3

Student’s t test p=0.71

Mean quality of life forpatients without sacralpressure ulcers:

403

Comments

spent with an AMS, thisoccurred in the beginning ofthe study period, because itwas then not entirely clear tothe nurses when they wereallowed to give an AMS tothe patients.

Nurses did not encounterdifficulties in using AMS indaily practice, but it did makeit slightly more difficult tochange bed linen in bed-ridden patients. Also thedirty sheepskins neededseparate linen bags causedsome inconvenience.

62.1

61.3

Student’s t test p=0.71

Mean quality of life forpatients without sacralpressure ulcers:

404


Characteristics of lost tofollow-up patients vsanalysed patients weregiven (no statisticallysignificant differences)

Statistical analysis:primary outcome(incidence) wasconducted withmultilevel binary logisticregression analysis.

Baseline differences: Nodifference for gender,age, Braden score,Barthel score and BMI ormedical diagnosis orprior surgery in monthbefore admission. nosignificant differencesbetween nursing homesin the proportion ofpatients that wererandomised to theintervention or controlgroup.

Study power/sample size:80% power 750 (2x375)required.

Setting: 8 nursing homes(23 nursing wards), theNetherlands.

Length of study:observations continued

18 years and older),expected stay >1week

Exclusion criteria:pressure ulcers onthe sacrum atadmission, havingdarkly pigmentedskin (because ofdifficulty indiagnosing grade 1pressure ulcer), andknown allergy towool; admitted for aprimarily psycho-geriatric reason.


Intervention

Comparison

Outcomemeasures

Effect sizes

18 years and older),

:

nosing grade 1pressure ulcer), and

wool; admitted for a

Group 1: 63

Group 2: 53

Student’s t test, p=0.003

KCE Report 193S

Comments

Student’s t test, p=0.003

KCE Report 193S


until day 30 afteradmission

Assessment of PUs: dailyskin observations, usedEPUAP grading system.Used photographicseries of the variouspressure ulcer grades aswell as transparent disksthat nurses pressedagainst erythema byhand to see whether thearea blanched underpressure. If uncertainthey called a specialisednurse. All cases ofpressure ulcers werereported to a wound carespecialist who checkedthe observation, gavecare instructions andmonitored the progressof the ulcer.

Risk assessment: Bradenscale.



Intervention

Comparison

Outcomemeasures

Effect sizes

405

Comments

406

Table 32 – MALBRAIN2010


Author and year:Malbrain 2010

Title: A pilotrandomised controlledtrial comparingreactive air and activealternating pressuremattresses in theprevention andtreatment of pressureulcers among medicalICU patients

Journal: Journal ofTissue Viability (2010),19, 7-15

Sequence generation:envelopes shuffled.

Allocation concealment:envelopes wereidentical, shuffled andplaced in a box but nomention of opaque.

Blinding: singleblinded

Addressing incompleteoutcome data:adequate

Statistical analysis: T-test and Fisher’s exacttest.

Baseline differences:

Patient group: patients inICU with high pressureulcer risk (Nortonscore</=8 requiringmechanical ventilation forat least 5 days, with eitherintact skin or pressureulcers

All patients

Randomised N: 16

Completed N: 15

Drop-outs: one deathknow that developed asacral persistenterythema (category 1)immediately prior todeath.

Group 1

Randomised N: 8

Completed N: 8

Dropouts: 0

Age (years): 71.5 (s.d11.8)

Sex F/M: 3/5

BMI (kg/m2): 22.1 (s.d2.7)

Pre-albumin (mg/dl):



Comparison

Outcomemeasures

Effect sizes

patients inICU with high pressure

mechanical ventilation forat least 5 days, with eitherintact skin or pressure

butknow that developed a

erythema (category 1)

71.5 (s.d

22.1 (s.d

Group 1: ROHO dryfloatation mattress overlay

Group 2: the NIMBUS 3active alternating pressuremattress

Both groups were givenstandard treatmentaccording to Belgianconsensus protocol.Repositioning every 2hours from semi-Fowler tothe right/left lateral 30degrees position. Two-way stretch sheet and alow friction slide sheetused for repositioning.Pillow between calves andinterface, which isstandard protocol inBelgium. Additionalnutritional support. All hadindwelling urinarycatheters. Skin wasinspected daily anddocumented.

Outcome 1:incidence ofpressure ulcers(all grades)

Group 1: 2/8 (25%)

Group 2: 2/8 (25%)

KCE Report 193S

Comments

2/8 (25%)

2/8 (25%)

Funding: nodetails

Limitations: verysmall samplesize; unclearallocationconcealment.Single blinded.Baselinedifferences.

Additionaloutcomes:healing of ulcers.

Notes:

KCE Report 193S


statistically significantdifference in age andper-albumin.

Study power/samplesize: power calculationnot given but verysmall sample size.

Setting: ICU, Belgium

Length of study: notreported but averagegiven for both groupsas 15 (s.d 14) in theNIMBUS group and12.2 (s.d 5.5) in theROHO group

Assessment of PUs:PUSH tool

Multiple ulcers: all wererecorded.

20.3 (s.d 12.4)

Norton score: 7 (s.d 0)

APACHE II score: 20.4(s.d 7.5)

SOFA score: 11.4 (s.d3.2)

CRP day 1 (mg/dl): 10.1(s.d 14.1)

% Semi-Fowler position:58.1 (s.d 7.5)

% lateral decubitus: 41(s.d 17.2)

Group 2

Randomised N: 8

Completed N: 7

Dropouts: 1 died

Age (years): 56.9 (s.d16.3)

Sex F/M: 5/3

BMI (kg/m2): 24.2 (s.d6.5)

Pre-albumin (mg/dl): 6.7(s.d 3.6)

Norton score: 7.4 (s.d1.1)

APACHE II score: 22.8(s.d 4.6)

SOFA score: 11.8 (s.d2.7)



Comparison

Outcomemeasures

Effect sizes

7 (s.d 0)

20.4

11.4 (s.d

10.1

Fowler position:

41

56.9 (s.d

24.2 (s.d

6.7

7.4 (s.d

22.8

11.8 (s.d

407

Comments

408


CRP day 1 (mg/dl): 10.3(s.d 8.2)

% Semi-Fowler position:54.9 (s.d 11.8)

% lateral decubitus: 37.1(s.d 11.2)

Inclusion criteria:patients in ICU with highpressure ulcer risk(Norton score</=8requiring mechanicalventilation for at least 5days, with either intactskin or pressure ulcers

Exclusion criteria: ifconsent refused or if attime admitted not at leastone of the mattressesavailable.



Comparison

Outcomemeasures

Effect sizes

10.3

Fowler position:

37.1

patients in ICU with high

ventilation for at least 5days, with either intactskin or pressure ulcers

used or if attime admitted not at least

KCE Report 193S

Comments

KCE Report 193S

8. HEEL ULCER PREVENTIO


Table 1 – Protocol review

Protocol Heel prevention (devices)

Review question What are the most clinically effective pressure re


Intervention Heel-specific devices:

Air-filled booties

Foam foot pro

Gel foot protectors

Pillows and other aids

Splints or other medical devices

Sheepskins for heels (synthetic and natural)

Pressure Relief Ankle Foot Orthosis



No intervention

Outcomes Critical outcom

Proportion of participants developing new pressure ulcers (dichotomous outcome)(describe different categories ofulcer)

Important outcomes:



Time to develop new pressure

Time in hospital or NHS care (continuous data)

Healthpressure ulcer patients, therefore may have to be narratively summar


HEEL ULCER PREVENTION (DEVICES)

Heel prevention (devices)

What are the most clinically effective pressure re-distributing devices for the prevention of heel pressure ulcers?


specific devices:

filled booties

Foam foot protectors

Gel foot protectors

Pillows and other aids

Splints or other medical devices

Sheepskins for heels (synthetic and natural)

Pressure Relief Ankle Foot Orthosis


Each other

No intervention



Important outcomes:






409

distributing devices for the prevention of heel pressure ulcers?


related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes in

410


o Short


o EQ

o WHO


o HUI



Cochrane reviewsdata such as available case analysis or ITT (with the appropriate assumptions)


Exclusion Studies



The search strategy The databases to be searched are:


All years.

Studies will be restricted t


Population

Interventi

Comparison

Outcomes

Study dstudies will be meta

Unit of analysisunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different





EQ-5D

WHO-QOL BREF

Cardiff HRQoL tool

HUI










All years.




Intervention – Different categories of device will not be combined for meta-analysis

Comparison – any comparison which fits the inclusion criteria will be meta-analysed

Outcomes – single side effects will be meta-analysed separately from other side effects


Unit of analysis – patients, clusters (hospital wards), individual pressure ulcersunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different

KCE Report 193S

will be included if they match our inclusion criteria and have appropriate assumptions for missingdata such as available case analysis or ITT (with the appropriate assumptions)


nalysis except for different strata

analysis

analysed



tal wards), individual pressure ulcers – for those where patients are theunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different

KCE Report 193S


categories of ulcer)

Minimum duration of treatment = no mini



Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if canauthor’s data.


Analysis Strata:

The following groups will be considered separately as strata if da




Bariatric patients (BMI >40)

Subgroups:

The following groups will be considered separately

Different categories of pressure ulcer (from category 2 upwards where outcomes are reported separately)

Adjunctive therapies

Other terms Support surfaces, pressure relieving, pressure reducing,




categories of ulcer)




Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data..






Bariatric patients (BMI >40)

Subgroups:



Adjunctive therapies

Support surfaces, pressure relieving, pressure reducing, pressure preventing


411

Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsnot work out the available case analysis will take the


as subgroups if data are present and there is inconsistency:



412




Search strategy Heel ulcer prevention

Date 27th Mar 2012


Search strategy(part I –Protectiondevices)

1 letter/

2 editorial/

3 news/



6 comment/

7 case report/


9 or/1-8

10 randomized controlled trial/ or random*.

11 9 not 10





16 exp Rodentia/


18 or/11-17

19 pressure ulcer/

20 decubit*.ti,ab.

21 (pressure adj (sore* or










sore*).ti,ab.

KCE Report 193S

Results

761962

307832

150655

303553

4270

502495

1566420

83724

3056534

681405

3041380

3612470

669993

5302

368581

2474809

1037887

7181325

8951

3879

6110

502

KCE Report 193S




25 or/19-24



28 wheelchairs/

29 (bed or beds).ti,ab.

30 (cutfoam or padding or sheepskin* or sheep

31 (alternat* adj2 pressure).ti,ab.

32 shoes/

33 exp orthotic devices/

34 (orthotic adj2 (device* o

35 (shoe* or boot* or footwear or foot

36 (orthos* or insole).ti,ab.

37 ((contact or walk*) adj2 cast*).ti,ab.

38 (aircast* or scotchcast*).ti,ab.

39 ((foot or feet or heel*) adj2 (pressure or protect* or device*))

40 ((foot or feet or heel* or leg*) adj2 trough*).ti,ab.

41 (heel* adj2 (lift* or splint* or float* or glove* or suspen* or elevat*)).ti,ab.

42 or/27-41

43 26 and 42

44 43 not 18



47 randomi#ed.ab.

48 placebo.ab.

49 drug therapy.fs.

50 randomly.ab.





eat* or chair* or wheelchair* or pillow*).ti,ab.

(cutfoam or padding or sheepskin* or sheep-skin* or gels).ti,ab.

(alternat* adj2 pressure).ti,ab.

(orthotic adj2 (device* or therap* or treat*)).ti,ab.

(shoe* or boot* or footwear or foot-wear).ti,ab.

(orthos* or insole).ti,ab.

((contact or walk*) adj2 cast*).ti,ab.

(aircast* or scotchcast*).ti,ab.

((foot or feet or heel*) adj2 (pressure or protect* or device*)).ti,ab.

((foot or feet or heel* or leg*) adj2 trough*).ti,ab.

(heel* adj2 (lift* or splint* or float* or glove* or suspen* or elevat*)).ti,ab.



413

Results


53

632

13575

10837

36666

3185

70275

35899

279

4352

8586

512

15002

13416

341

102

1022

5

160

178049

1575

1400

325596

83986

289033

135372

1524063

177256

414


51 trial.ab.

52 groups.ab.

53 or/45-52


55 trial.ti.

56 or/45-48,50,54-55

57 Meta-Analysis/


59 (meta analy* or metanaly* or metaan



62 (search strategy or search criteria or systematic search or study selection o



65 cochrane.jw.

66 or/57-65

67 56 or 66

68 44 and 67

Part II (supportsurfaces)

1 pressure ulcer/

2 decubit*.ti,ab.

3 (pressure adj (sore* or ul




7 or/1-6





Analysis as Topic/








sore*).ti,ab.




trolled trial.pt.

KCE Report 193S

Results

r data extraction).ab.


249935

1158597

2925297

159527

103737

797665

33150

12045

43213

51468

20401

21932

20026

63574

8354

148230

903308

207


8894

3865

6062

501

50

622

13487

10757

322734

KCE Report 193S



11 randomi#ed.ab.

12 placebo.ab.

13 drug therapy.fs.

14 randomly.ab.

15 trial.ab.

16 groups.ab.

17 or/9-16


19 trial.ti.

20 or/9-12,14,18-19

21 letter/

22 editorial/

23 news/



26 comment/

27 case report/


29 or/21-28


31 29 not 30



34 exp Animal Experime


36 exp Rodentia/











415

Results

83763

285035

134079

1512984

175416

246425

1148425

2901023

158570

102055

789946

752856

302491

143966

302413

4185

493095

1558286

83156

3025178

674026

3010191

3594930

665788

5218

365269

2460341

1032770

416


38 or/31-37

39 Meta-Analysis/



42 ((systematic* or evidence*) adj2 (review* or overview*)).t





47 cochrane.jw.

48 or/39-47

49 20 or 48

50 49 not 38

51 8 and 50

52 exp beds/

53 (mattress* or cushion* or foam or trans








61 ((turn* or tilt*) adj2 (bed* or fr


63 net bed*.ti,ab.



Analysis as Topic/






med or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citation











KCE Report 193S

Results


med or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citation


7127677

32205

11873

42057

50096

19856

21391

19634

61940

7944

145126

893674

782841

995

3372

250061

6845

72

439

16888

280

10

448

454

77

9

KCE Report 193S


65 or/52-64


67 wheelchairs/

68 65 or 66 or 67

69 51 and 68




Date 27th Mar 2012



1 decubitus/

2 decubit*.ti,ab.



5 ((moist* or friction or shear)


7 or/1-6



10 exp wheelchair/

11 (bed or beds).ti,ab.

12 (cutfoam or padding or sheepskin* or sheep

13 (alternat* adj2 pressure).ti,ab.

14 orthopedic shoe/

15 shoe/



Current"


sore*).ti,ab.





am or padding or sheepskin* or sheep-skin* or gels).ti,ab.

(alternat* adj2 pressure).ti,ab.

417

Results

33140

309311

36394

3172

344756

323

49

Results

adj2 (sore* or ulcer* or damage or wound* or injur* or lesion*)).ti,ab.

12517

4766

7117

659

767

56

17007

13126

41012

5086

89020

35692

299

193

5740

418


16 orthotics/

17 (orthotic adj2 (device* or therap* or treat*)).ti,ab.

18 (shoe* or boot* or footwear or foot

19 (orthos* or insole).ti,ab.

20 ((contact or walk*) adj2 cast*).ti,ab.

21 (aircast* or scotchcast*).ti,ab.

22 ((foot or feet or heel*) adj2 (pressure or protect* or device*)).ti,ab.

23 ((foot or feet or heel* or leg*) adj2 trough*).ti,ab.

24 (heel* adj2 (lift* or splint* or float* or glove* or suspen* or elevat*)).ti,ab.

25 or/9-24

26 8 and 25

27 random*.ti,ab.




31 (assign* or allocat* or volunteer* or placebo*).ti,a





36 or/27-35


38 note.pt.

39 editorial.pt.



42 or/37-41



(orthotic adj2 (device* or therap* or treat*)).ti,ab.

(shoe* or boot* or footwear or foot-wear).ti,ab.

s* or insole).ti,ab.

((contact or walk*) adj2 cast*).ti,ab.

(aircast* or scotchcast*).ti,ab.

((foot or feet or heel*) adj2 (pressure or protect* or device*)).ti,ab.

((foot or feet or heel* or leg*) adj2 trough*).ti,ab.

splint* or float* or glove* or suspen* or elevat*)).ti,ab.










domized controlled trial/ or random*.ti,ab.

KCE Report 193S

Results

2943

614

18106

16100

396

127

1219

5

178

203234

1719

717655

18594

60412

137024

553050

33588

108289

15735

320112

1186128

778574

514042

401605

1831335

135434

3393890

801083

KCE Report 193S


44 42 not 43


46 nonhuman/



49 animal model/

50 exp Rodent/


52 or/44-51


54 meta-analysis/




58 (search strategy or search cr




62 cochrane.jw.

63 or/53-62

64 36 or 63

65 26 and 64

66 65 not 52


1 random*.ti,ab.

2 factorial*.ti,ab.













((pool* or combined) adj2 (data or trials or studies or results)).ab.



419

Results

iteria or systematic search or study selection or data extraction).ab.


3367763

1323451

3824666

1504918

410580

633405

2532293

1107552

8891638

48857

62389

55834

59625

24583

26269

24389

75972

31350

11048

224468

1344623

300

290

711167

18452

60004

136181

420







10 or/1-9


12 note.pt.

13 editorial.pt.



16 or/11-15

17 randomized controlled trial/ or random*.

18 16 not 17


20 nonhuman/



23 animal model/

24 exp Rodent/


26 or/18-25


28 meta-analysis/

29 (meta analy* or metana



32 (search strategy or search criteria or systematic search or study




ble blind procedure/













KCE Report 193S

Results

selection or data extraction).ab.

549213

33346

107813

15595

318508

1177104

775094

511290

399508

1825147

134926

3380104

794389

3354078

1321445

3806953

1498332

408085

629106

2520889

1103508

8855378

48030

61737

54972

58719

24411

26081

KCE Report 193S




35 ((pool* or combined) adj2 (data

36 cochrane.jw.

37 or/27-36

38 decubitus/

39 decubit*.ti,ab.



42 ((moist* or friction or shear) adj2 (sore* or ulcer* or damage or w


44 or/38-43


46 (10 or 37) not 26

47 45 and 46

48 (mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel).ti,ab.










58 net bed*.ti,ab.







sore*).ti,ab.





adj2 (device* or support* or constant)).ti,ab.




tion adj2 device*).ti,ab.





421

Results


ound* or injur* or lesion*)).ti,ab.


24044

75039

31034

11048

222072

12420

4747

7047

655

759

53

16890

13015

1103384

1435

265218

7910

100

513

20059

370

13

525

525

100

9

422



61 exp bed/

62 exp wheelchair/

63 or/48-62

64 47 and 63




Date 27th Mar 2012

Database CINAHL


S25 S22 NOT S23 Limiters

S24 S22 NOT S23

S23 PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orPT cartoon or PT commentary or PT computeor PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTpictorial or PT poetry or PT proceedings or PT “quesmaterials or PT website

S22 S7 and S21

S21 S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20

S20 heel* AND (lift* OR splint* OR float* OR glove* OR suspen*

S19 (foot or feet or heel* or leg*) and trough*

S18 (foot OR feet OR heel*) AND (pressure OR protect* OR device*)

S17 contact N2 cast* OR walk* N2 cast*

S16 orthotic N2 treat* OR orthotic N2 therap* OR orthotic N2 device*

S15 alternat* N2 pressure

S14 bed or beds or cutfoam or padding or sheepskin* or sheep



Current"

S22 NOT S23 Limiters – English Language; Exclude MEDLINE records

PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orPT cartoon or PT commentary or PT computer program or PT editorial or PT games or PT glossary or PT historical materialor PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTpictorial or PT poetry or PT proceedings or PT “questions and answers” or PT response or PT software or PT teaching

S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20

heel* AND (lift* OR splint* OR float* OR glove* OR suspen* OR elevat*)

(foot or feet or heel* or leg*) and trough*

(foot OR feet OR heel*) AND (pressure OR protect* OR device*)

contact N2 cast* OR walk* N2 cast*

orthotic N2 treat* OR orthotic N2 therap* OR orthotic N2 device*

ure

bed or beds or cutfoam or padding or sheepskin* or sheep-skin* or gels or shoe* or boot* or footwear or foot

KCE Report 193S

Results

38650

40750

7588

5032

378050

427

69

Results

PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orr program or PT editorial or PT games or PT glossary or PT historical material

or PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTtions and answers” or PT response or PT software or PT teaching

skin* or gels or shoe* or boot* or footwear or foot-wear or

455

1485

1001377

2467

42142

178

22

3452

152

233

131

KCE Report 193S


orthos* or insole or aircast* or scotchcast*

S13 (MH "Orthopedic Footwear")

S12 (MH "Seating")

S11 (MH "Orthoses+")

S10 (MH "Shoes+")



S7 S1 or S2 or S3 or S4 or S5 or S6



S4 bedsore* OR bed-sor


S2 decubit*




S25 S7 and S24

S24 S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19 or S20 or S21 or S22 or S23


S22 positioning or repositioning or re

S21 net bed*

S20 kinetic and (therapy or table*)

S19 (turn* or tilt*) and (bed* or frame*)

S18 clinifloat or maxifloat or vaperm or therarest or sheepskin or hammockcairwave

S17 elevation N2 device*

S16 water suspension*

S15 pressure and (relie* or reduc* or alleviat* or redistribut* or re

S14 air suspension or air bag*


orthos* or insole or aircast* or scotchcast*

(MH "Orthopedic Footwear")

seat* or chair* or wheelchair* or pillow*




sore*



– Published Date from: 20101201-20121231; English Language; Exclude MEDLINE records


r* or wheelchair* or pillow*

positioning or repositioning or re-positioning

kinetic and (therapy or table*)

(turn* or tilt*) and (bed* or frame*)

clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or

elevation N2 device*

pressure and (relie* or reduc* or alleviat* or redistribut* or re-distribut* or alternat*)

air suspension or air bag*

423

Results

26061

94

633

5830

2320

13034

2982

9652

1377

70

155

8313

476

7543



or foot waffle or silicore or pegasus or

133

3354

48691

12957

7537

4

370

1366

57

6

0

14412

131

424


S13 static air

S12 pressure and (device* or support* or constant)

S11 mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel


S9 (MH "Pillows and Cushions")

S8 (MH "Beds and Mattresses+")

S7 S1 or S2 or S3 or S4 or S5 or S





S2 decubit*




Date 27th Mar 2012

Database Cochrane (- CDSR [3/2012]; DARE; Central [3/2012]; NHS EED; HTA)


#1 MeSH descriptor Pressure Ulcer explode all trees



#4 (bedsore* or bed-sore*)



#7 (#1 OR #2 OR #3 OR #4 OR #5 OR #6)

#8 (seat* or chair* or wheelchair* or pillow*):ti,ab,k



essure and (device* or support* or constant)

mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel

(MH "Wheelchairs+")

(MH "Pillows and Cushions")

(MH "Beds and Mattresses+")




sore*


sure Ulcer")

R [3/2012]; DARE; Central [3/2012]; NHS EED; HTA)



sore*):ti,ab,kw



(#1 OR #2 OR #3 OR #4 OR #5 OR #6)

(seat* or chair* or wheelchair* or pillow*):ti,ab,kw

MeSH descriptor Wheelchairs explode all trees

KCE Report 193S

Results


12

8690

9244

2956

456

2576

9607

1368

69

155

8277

474

7513

Results


487

349

829

33

10

63

1171

2687

128

KCE Report 193S


#10 MeSH descriptor Shoes explode all trees

#11 MeSH descriptor Orthotic Devices explode all trees

#12 (bed or beds or cutfoam or padding or sheepskin* or sheeporthos* or insole or aircast* or scotchcast*):ti,ab,kw

#13 (alternat* near/2 pressure):ti,ab,kw

#14 (orthotic near/2 (device* or therap* or treat*)):ti,ab,kw

#15 ((contact or walk*) near/2 cast*):ti,ab,kw

#16 ((foot or feet or heel*) near/2

#17 ((foot or feet or heel* or leg*) near/2 trough*):ti,ab,kw

#18 (heel* near/2 (lift* or splint* or float* or glove* or suspen* or elevat*)):ti,ab,kw

#19 (#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #

#20 (#7 AND #19)




#3 (pressure near/2 (sore* or ulcer* or damag




#7 (#1 OR #2 OR #3 OR #4 OR #5 OR #6)

#8 MeSH descriptor Beds explode all trees


#10 (mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel):ti,ab,kw

#11 (pressure NEAR/2 (device* or support* or constant)):ti,ab,kw

#12 (static NEAR/2 air):ti,ab,kw

#13 (air NEAR/2 (suspension or bag*)):ti,ab,kw

#14 (pressure NEAR/2 (relie* or reduc* or alleviat* or redistribut* or re

#15 water suspension*:ti,ab,kw

#16 (elevation NEAR/2 device*):ti,ab,kw


MeSH descriptor Shoes explode all trees

MeSH descriptor Orthotic Devices explode all trees

(bed or beds or cutfoam or padding or sheepskin* or sheep-skin* or gels or shoe* or boot* or footwearorthos* or insole or aircast* or scotchcast*):ti,ab,kw

(alternat* near/2 pressure):ti,ab,kw

(orthotic near/2 (device* or therap* or treat*)):ti,ab,kw

((contact or walk*) near/2 cast*):ti,ab,kw

((foot or feet or heel*) near/2 (pressure or protect* or device*)):ti,ab,kw

((foot or feet or heel* or leg*) near/2 trough*):ti,ab,kw

(heel* near/2 (lift* or splint* or float* or glove* or suspen* or elevat*)):ti,ab,kw

(#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18)



sore*):ti,ab,kw



(#1 OR #2 OR #3 OR #4 OR #5 OR #6)

tor Beds explode all trees

MeSH descriptor Wheelchairs explode all trees

(mattress* or cushion* or foam or transfoam or overlay* or pad or pads or gel):ti,ab,kw

(pressure NEAR/2 (device* or support* or constant)):ti,ab,kw

:ti,ab,kw

(air NEAR/2 (suspension or bag*)):ti,ab,kw

(pressure NEAR/2 (relie* or reduc* or alleviat* or redistribut* or re-distribut* or alternat*)):ti,ab,kw

water suspension*:ti,ab,kw

(elevation NEAR/2 device*):ti,ab,kw

425

Results

skin* or gels or shoe* or boot* or footwear or foot-wear or

234

713

12750

44

450

53

146

1

24

15727


distribut* or alternat*)):ti,ab,kw

481

348

821

32

10

63

1161

243

127

7516

800

4

8

3643

118

5

426


#17 (clinifloat or maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus orcairwave):ti,ab,kw



#20 net bed*:ti,ab,kw

#21 (positioning or repositioning or re

#22 (seat* or chair* or wheelchair* or pillow*):ti,ab,kw

#23 (#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR#22)

#24 (#7 AND #23)

#25 (#24), from 2010 to 2012


maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or



positioning or repositioning or re-positioning):ti,ab,kw

(seat* or chair* or wheelchair* or pillow*):ti,ab,kw


0 to 2012

KCE Report 193S

Results

maxifloat or vaperm or therarest or sheepskin or hammock or foot waffle or silicore or pegasus or


53

47

47

289

8906

2653

22993

498

48

KCE Report 193S


Figure 1 – Flow diagram of clinical article selection for what is the clinical and costprevention of heel ulcers? review



Excluded, n =


Excluded, n = 65


Flow diagram of clinical article selection for what is the clinical and cost -effectiveness of pressure

Excluded, n = 694

427

effectiveness of pressure-redistributing devices for the

428



ANON1993 Ordered for devices for prevention review

ARONOVITCH1998 Ordered for devices for prevention review

BALES2012 Literature review

BERTHE2007 Ordered for devices for prevention review

BHATNAGAR1997 Commentary

BRIENZA2010 Ordered for devices for prevention review

BROWN2000 Ordered for devices for prevention review

CHALONER2000 Ordered for devices for

CHENEWORTH1994 Literature review

DEFLOOR2000B Not our outcomes

DEKEYSER1994 Not our outcomes

DEMARRE2012 Ordered for devices for prevention review

DONNELLY2011 Ordered for devices for prevention review

EKSTEEN2006 Not an RCT

EVANS2009 Not an RCT/abstract not freely available

EVANS2009A Abstract

EWING1964 Cochrane excluded as it was ctoo small and suffering from risk of bias tothe extent that its results coulregarded as valid. Does not mentionpressure ulcers but ‘rheels and ankles’.



Ordered for devices for prevention review








Not an RCT/abstract not freely available

Cochrane excluded as it was consideredtoo small and suffering from risk of bias tothe extent that its results could not beregarded as valid. Does not mention

but ‘reddening of skin of’.

FAWCETT2004 Abstract

FERRELL1993 Ordered for devices for prevention review;economic study

FINNEGAN2008 Not our outcomes

GIL-AGUDO2009 Not our outcomes

GONZALEZ DELLAVALLE2001

Not our outcomes

GOOSSENS2008 Not our outcomes

GRAY2000 Ordered for devices for prevention review

GRINDLEY1996 Not our outcomes

GRISELL2008 Ordered for devices for prevention review

HAMPTON2010 Not an RCT

HEYNEMAN2009 Pooled 2 RCTs (which were included inreview)

HUANG2011 Not our outcomes

HUBER2008 Not our outcomes

ISMAIL2001 Ordered for devices for prevention review.But the paper states that ‘those whodeveloped pressure sore were not turnedat night’ unclear if just these patients or allpatients.

JAN2011 Not RCT

JESURUM1996 Ordered for devices for prevention review.

JOLLEY2010 Ordered for devices for prevention review

JUNKIN2009 Systematic review

KCE Report 193S

Abstract

Ordered for devices for prevention review;economic study

Not our outcomes

Not our outcomes

Not our outcomes

Not our outcomes


Not our outcomes


Not an RCT

Pooled 2 RCTs (which were included inreview)

Not our outcomes

Not our outcomes

Ordered for devices for prevention review.But the paper states that ‘those whodeveloped pressure sore were not turnedat night’ unclear if just these patients or allpatients.

Not RCT

Ordered for devices for prevention review.


Systematic review

KCE Report 193S

LOCKYERSTEVENS1993 Not an RCT

MACFARLANE2006 Not an RCT

MAKHSOUS2009 Ordered for devices for prevention review

MAYROVITZ2003 Not an RCT

MAYROVITZ2004 Not an RCT

MCINNES2012 Ordered for devices for prevention review

MILNE2011 Abstract not freely available

MISTIAEN2008 Cost-effectiveness study protocol

MISTIAEN2010A Ordered for devices for prevention review

MISTIAEN2010E Ordered for devices for prevention review

NICOSIA2007 Meta-analysis which included deviceswhich were not specific to the heel

NIXON2006B Erratum for study ordered for devices forprevention review

PINZURI1991 Not an RCT

RAFTER2011 Ordered for devices for prevention review

RUSSELL2000 Ordered for devices for prevention review

RUSSELL2000B Ordered for devices for prevention review




Abstract not freely available

effectiveness study protocol



analysis which included deviceswhich were not specific to the heel

Erratum for study ordered for devices for

devices for prevention review



RUSSELL2003 Ordered for devices for prevention review

RUSSELL2003A Ordered for devices for prevention review

SANTAMARIA2012 Abstract

SCOTT1999 Ordered for devices for prevention review

SILVERTHORN2011 Not pressure ulcers

SIMMS2011 Abstract

STERZI2003 Ordered for devices for prevention review

STONE2011 Abstract

TACCONE2009 Not an RCT

VANLEEN2011 Ordered fo

WILLIAMS1995 Commentary

VUOLO2010 Commentary

ZERNIKE1994 Not our outcomes

ZERNIKE1997 Not our outcomes

429



Abstract


Not pressure ulcers

Abstract


Abstract

Not an RCT


Commentary

Commentary

Not our outcomes

Not our outcomes

430

8.3. Clinical evidenceFive randomized controlled trials were included in the review.

8.3.1. Summary of included studies


Study Intervention/comparison

Cadue 200874

Foam body support and standardpressure prevention protocol(half-seated position, watermattress preventive massage 6times/day) versus standardpressure ulcer protocol (asabove)

Donnelly 2011134

Heel elevation (Heeliftsuspension boot) plus pressureredistributing support surfaceversus standard care pluspressure-redistributing surfacealone

Gilcreast 200576

Bunny boot (fleece) high cushionheel protector vs egg crate heellift positioner vs foot wafflecushion

Tymec 199777

Foot waffle vs hospital pillowunder both legs from below kneeto the Achilles tendon

Torra 2009135

Special polyurethane foamhydrocellular dressing for theprotection of the heel (AllevynHeel) vs protective bandage ofthe heel (Soffban and gauzebandage).


Five randomized controlled trials were included in the review.74 , 76 , 77 , 134 , 135

Summary of studies included in the review

Population Outcomes

Foam body support and standardpressure prevention protocol

mattress preventive massage 6

Patients in an intensive caresetting

Number of participantsdeveloping non-blanchingpressure ulcer or worse on theheel

suspension boot) plus pressure-redistributing support surface

redistributing surface

Post-hip fracture patients Incidence of heel ulcers (allcategories)

Bunny boot (fleece) high cushionheel protector vs egg crate heel

Military tertiary-care academicmedical centre patients ofmoderate or high risk of pressureulcer development, Braden score≤14

Pressure ulcer incidence

Foot waffle vs hospital pillowunder both legs from below knee

Patients in selected nursing unitsof large hospital; Braden score<1 (risk); intact skin on heels

Number of pressure ulcersdeveloped

hydrocellular dressing for theprotection of the heel (AllevynHeel) vs protective bandage of

Nursing home patients and homecare program patients fromprimary health care centres.

Number of participants withpressure ulcers

KCE Report 193S

Study length

blanchingpressure ulcer or worse on the

Maximum follow-up 3months

Incidence of heel ulcers (all 12 days

Pressure ulcer incidence Follow-up period unclear

Number of pressure ulcers unclear

Number of participants with 8 weeks

KCE Report 193S

8.3.2. Clinical evidence GRADE-tables

Table 7 – Bunny boot fleece cushion heel protector versus egg crate heel lift positioner for prevention of heel pressure ulcersward, cardiology patients

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency

Incidence of patients with heel ulcers

1Gilcreast(2005)

randomisedtrials

veryserious1


1 Inadequate allocation concealment; no blinding; limited details of baseline data; unclear how many patients were randomiseddrop-outs came from but there was 29% of patients who did not have follow2 Confidence interval crossed both MID points.


Bunny boot fleece cushion heel protector versus egg crate heel lift positioner for prevention of heel pressure ulcers



Bunnyboot

Eggcrate

Relative(95% CI)


veryserious2

none 3/77(3.9%)

4/87(4.6%)

RR 0.85(0.2 to3.67)

4.6%

1 Inadequate allocation concealment; no blinding; limited details of baseline data; unclear how many patients were randomised to each group and therefore which aouts came from but there was 29% of patients who did not have follow-up data.

431

Bunny boot fleece cushion heel protector versus egg crate heel lift positioner for prevention of heel pressure ulcers – ICU, med, surgical

Effect Quality Importanceof outcome

Relative(95% CI)

Absolute

RR 0.85 7 fewer per1000 (from 37fewer to 123

more)

VERYLOW

Criticaloutcome


more)

to each group and therefore which arms the

432

Table 8 – Bunny boot fleece cushion heel protector versus foot waffle air cushion for prevention of heel pressure ulcardiology patients

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency


1Gilcreast(2005)

randomisedtrials

veryserious1



Table 9 – Eggcrate heel lift positioner versus foot waffle air cushion for prevention of heel pressure ulcerspatients

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency


1Gilcreast(2005)

randomisedtrials

veryserious1




Bunny boot fleece cushion heel protector versus foot waffle air cushion for prevention of heel pressure ul



Bunnyboot

Footwaffle

Relative(95% CI)


veryserious2

none 3/77(3.9%)

5/76(6.6%)

RR 0.59(0.15 to

2.39)

6.6%

1 Inadequate allocation concealment; no blinding; limited details of baseline data; unclear how many patients were randomised to each group and therefore which aouts came from but there was 29% of patients who did not have follow-up data.

Eggcrate heel lift positioner versus foot waffle air cushion for prevention of heel pressure ulcers –



Eggcrate Footwaffle

Relative(95% CI)


veryserious2

none 4/87(4.6%)

5/76(6.6%)

RR 0.7(0.19 to

2.51)

6.6%

KCE Report 193S

Bunny boot fleece cushion heel protector versus foot waffle air cushion for prevention of heel pressure ul cers – ICU, med, surgical ward,


Relative(95% CI)

Absolute

RR 0.59(0.15 to


fewer to 91more)

VERYLOW

Criticaloutcome


fewer to 92more)


– ICU, med, surgical ward, cardiology


Relative(95% CI)

Absolute

RR 0.7(0.19 to


fewer to 99more)

VERYLOW

Criticaloutcome


more)

KCE Report 193S


Table 10 – Foot waffle heel elevation device versus heel elevation pillow for prevention of heel pressure ulcersunits at a hospital

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of patients with heel pressure ulcers

1 Tymec(1997)

randomisedtrials

veryserious1



Time to pressure ulcer

1 Tymec(1997)

randomisedtrials

veryserious1



1 unclear allocation concealment, blinding and reporting of incomplete outcome data.2 Confidence interval crossed both MID points.3 No standard deviations so could not analyse in Revman.


1 Inadequate allocation concealment; no blinding; limited details of baseline data; unclear how many patients were randomised to each group and therefore whichouts came from but there was 29% of patients who did not have follow-up data.

Foot waffle heel elevation device versus heel elevation pillow for prevention of heel pressure ulcers



Footwaffle

Pillow Relative(95% CI)


veryserious2

none 0/26(0%)

1/26(3.8%)

Peto OR0.14 (0 to

6.82)

3.9%


No serious Very serious3 10days

13days

-

1 unclear allocation concealment, blinding and reporting of incomplete outcome data.

3 No standard deviations so could not analyse in Revman.

433


Foot waffle heel elevation device versus heel elevation pillow for prevention of heel pressure ulcers – patients from selected nursing


Absolute


more)

VERYLOW

Criticaloutcome


more)

Log-rank testp=0.036

VERYLOW

Criticaloutcome

434

Table 11 – Eggcrate suspension boot heel elevation device plus pressureredistributing surface alone e.g. cut foam mattress, mattress overlays and alternating pressure mattresses) for prevention of– older patients with fractured hips

Quality assessment

No ofstudies

Design Risk ofbias



1Donnelly(2011)

randomisedtrials



Comfort

1Donnelly(2011)

randomisedtrials



1 No blinding of patients or health care practitioners. Underpowered.2 Comfort – Themed analysis of participants’ opinions –rated them as comfortable overall. Poor concordance reasons were the weight and bulk of the boot (36%), heat (particularly a3 Could not analyse in Revman as data not for both arms of the trial.


Eggcrate suspension boot heel elevation device plus pressure-redistributing support surface versus standard care (pressureredistributing surface alone e.g. cut foam mattress, mattress overlays and alternating pressure mattresses) for prevention of



Heelelevation

device

Standardcare

Relative(95% CI)



none 0/120(0%)

17/119(14.3%)

Peto OR0.12 (0.04

to 0.31)

14.3%



Very serious3 Seefootnote2

Seefootnote2

Seefootnote

1 No blinding of patients or health care practitioners. Underpowered.– 32% of subjects felt the boots interfered with sleep and 41% felt that they adversely affected movement in bed, 59%

rated them as comfortable overall. Poor concordance reasons were the weight and bulk of the boot (36%), heat (particularly a t night) (31%) and discomfort (24%).ta not for both arms of the trial.

KCE Report 193S

redistributing support surface versus standard care (pressure-redistributing surface alone e.g. cut foam mattress, mattress overlays and alternating pressure mattresses) for prevention of heel pressure ulcers


Relative(95% CI)

Absolute

Peto OR0.12 (0.04

to 0.31)


fewer to 136fewer)

MODERATE

Criticaloutcome


fewer to 136fewer)

footnote2See

footnote2

VERY LOW

Criticaloutcome

41% felt that they adversely affected movement in bed, 59%t night) (31%) and discomfort (24%).

KCE Report 193S

Table 12 – Foam support surface (Perpendicular foam blocks covered with jersey) plus usual care versus usual care (halfmattress preventive massage 6 times/day) for prevention of heel pressure u

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of patients with heel ulcers (follow-up 3 months)

1 Cadue(2008)

randomisedtrials



Mean time to pressure ulcer

1 Cadue(2008)

randomisedtrials


no serioindirectness

1 Unclear blinding. No a priori sample size calculation and small sample size.


Foam support surface (Perpendicular foam blocks covered with jersey) plus usual care versus usual care (halfmattress preventive massage 6 times/day) for prevention of heel pressure u lcers – ICU patients



Foambody

supportplus usual

care

Usualcare

Relative(95% CI)

up 3 months) – all grades



none 3/35(8.6%)

19/35(54.3%)

RR 0.16(0.05 to0.49)

54.3%



Very serious2 5.6 days 2.8days

-

1 Unclear blinding. No a priori sample size calculation and small sample size. 2 No standard deviations could not analyse data in Revman.

435

Foam support surface (Perpendicular foam blocks covered with jersey) plus usual care versus usual care (half -seated position, water


Absolute

RR 0.16 456 fewer per1000 (from


MODERATE

Criticaloutcome



P=0.01 VERY LOW

Criticaloutcome

2 No standard deviations could not analyse data in Revman.

436

Table 13 – Polyurethane hydrocellular foam dressing versus protective bandage4

Quality assessment

No ofstudies

Design Risk ofbias

Inconsistency Indir


1Torra(2009)

randomisedtrials

veryserious1


no sindirectness

1 Open study. Unclear how many in each group but relative risk reported. No details of allocation concealment and randomisatioutcome data. ; 2 Limited number of events. ; 3 Absolute values not available as number of pat4 The study names it a dressing but from the photos it looks to be a device.


Polyurethane hydrocellular foam dressing versus protective bandage4 – nursing home and home care program patients



Polyurethanehydrocellular

foam dressing

Protectivebandage

Relative(95% CI)


Serious2 none 3.3% 44% RR 13.42(95% CI

3.3

1 Open study. Unclear how many in each group but relative risk reported. No details of allocation concealment and randomisati on method3 Absolute values not available as number of patients in each group not given.

4 The study names it a dressing but from the photos it looks to be a device.

KCE Report 193S

nursing home and home care program patients


Relative(95% CI)

Absolute

RR 13.42(95% CI

3.3 to 54)

N/A3VERYLOW

Criticaloutcome

on method. Unclear addressing of incomplete

KCE Report 193S

8.3.3. Forest plots

8.3.3.1. Heel pressure-redistributing devices for the prevention of heel pressure ulcers

Figure 2 – Bunny boot vs. egg crate – incidence of heel pressure ulcers

Figure 3 – Bunny boot vs. foot waffle- incidence of heel pressure ulcers

Figure 4 – Egg crate vs. foot waffle- incidence of heel pressure ulcers

Study or Subgroup

Gilcreast, 2005

Total (95% CI)

Total events



Events

3

3

Total

77

77

Events

4

4

Total

87

87

Weight

100.0%

100.0%

Bunny boot Egg crate

Study or Subgroup

Gilcreast, 2005

Total (95% CI)

Total events



Events

3

3

Total

77

77

Events

5

5

Total

76

76

Weight

100.0%

100.0%

Bunny boot Foot waffle

Study or Subgroup

Gilcreast, 2005

Total (95% CI)

Total events



Events

4

4

Total

87

87

Events

5

5

Total

76

76

Weight

100.0%

100.0%

Eggcrate Foot waffle


redistributing devices for the prevention of heel pressure ulcers

incidence of heel pressure ulcers



Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.85 [0.20, 3.67]

0.85 [0.20, 3.67]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours bunny boot Favours egg crate

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.59 [0.15, 2.39]

0.59 [0.15, 2.39]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours bunny boot Favours foot waffle

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.70 [0.19, 2.51]

0.70 [0.19, 2.51]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours eggcrate Favours foot waffle

437

438

Figure 5 – Foot waffle vs. pillow- incidence of heel pressure ulcers

Figure 6 – Heel elevation device vs. standard care

Figure 7- Foam body support vs. usual care- incidence of heel pressure ulcers

Figure 8 – Protective bandage vs. polyurethane foam hydrocellular dressing

Study or Subgroup

Tymec, 1997

Total (95% CI)

Total events



Events

0

0

Total

26

26

Events

1

1

Total

26

26

Weight

100.0%

100.0%

Foot waffle Pillow

Study or Subgroup

Donnelly, 2011

Total (95% CI)

Total events



Events

0

0

Total

120

120

Events

17

17

Total

119

119

Weight

100.0%

100.0%

Heel elevation device Standard care

Study or Subgroup

Cadue, 2008

Total (95% CI)

Total events



Events

3

3

Total

35

35

Events

19

19

Total

35

35

Weight

100.0%

100.0%

Foam body support Usual care

Study or Subgroup

Torra, 2009

Total (95% CI)



log[Risk Ratio]

2.5967

SE

0.71

Weight

100.0%

100.0%

IV, Fixed, 95% CI

13.42 [3.34, 53.96]

13.42 [3.34, 53.96]



Heel elevation device vs. standard care- incidence of heel pressure ulcers

ncidence of heel pressure ulcers

Protective bandage vs. polyurethane foam hydrocellular dressing

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

0.14 [0.00, 6.82]

0.14 [0.00, 6.82]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours foot waffle Favours pillow

Weight

100.0%

100.0%

Peto, Fixed, 95% CI

0.12 [0.04, 0.31]

0.12 [0.04, 0.31]


Peto, Fixed, 95% CI

0.01 0.1 1 10 100Favours heel elevation Favours standard care

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.16 [0.05, 0.49]

0.16 [0.05, 0.49]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours foam body support Favours usual care

IV, Fixed, 95% CI

13.42 [3.34, 53.96]

13.42 [3.34, 53.96]


IV, Fixed, 95% CI

0.01 0.1 1 10 100Favours bandage Favours foam dressing

KCE Report 193S

KCE Report 193S


Table 14 – CADUE2008 [foreign language but in support surfaces for prevention Cochrane Review)


Author and year:Cadue (2008)

Title: Prevention ofheel pressure soreswith a foam body-support device. Arandomisedcontrolled trial in amedical intensivecare unit; 37 (1suppl. Part 1); 30-60.

Journal: PresseMedical 2008

Type of study: RCT

Sequencegeneration:‘randomisation tablewas used to allocate70 patients into 2groups’. The twogroups were formedrandomly byfollowing arandomisation table(yes)

Allocationconcealment:translated as sealedenvelope (yes)

Patient group: patientsin intensive care setting

All patients

Randomised N: 70

Completed N: 70

Group 1

Randomised N: 35

Completed N: 35

Dropouts: 0

Group 2

Randomised N: 35

Completed N: 35

Dropouts: 0

Inclusion criteria:patients in an intensivecare setting with aWaterlow Score >10, noexisting heel pressureulcers, >/=18 years orover.

Exclusion criteria: not


CADUE2008 [foreign language but in support surfaces for prevention Cochrane Review)

Intervention

Comparison

Outcomemeasures

Effect sizes

patients

Inclusion criteria:patients in an intensivecare setting with aWaterlow Score >10, no

reulcers, >/=18 years or

Exclusion criteria: not

Group 1: Foam bodysupport and standardpressure prevention protocol(half-seated position, watermattress preventativemassage 6 times/day)

Group 2: Standard pressureulcer protocol (see above)

Outcome 1:number ofparticipantsdeveloping non-blanchingpressure ulcer orworse on the heel

Group 1: 3/35 (8.6%)

Group 2: 19/35 (55.4%)

Outcome 2: meantime without anypressure ulcer

Group 1: 5.6 days

Group 2: 2.8 days

P=0.01

439

Comments

Group 1: 3/35 (8.6%)

Group 2: 19/35 (55.4%)

Funding: do notknow

Limitations:Unclear blinding.No a priorisample sizecalculation andsmall samplesize.

Additionaloutcomes: *

Notes: Abstract,with full paper notavailable inEnglish.Extraction takenfrom CochraneReview onsupport surfacesin the preventionof pressureulcers.

days

Group 2: 2.8 days

440


Blinding: translatedto: thephysiotherapist andnurse assessed thestage of the lesiondaily – but it is notclear if they wereblinded (unclear)

Addressingincomplete outcomedata: 70 patientswere included, 35 ineach group. Tablepresented theprinciple results andnotes that ‘n=35’which has beeninterpreted that datawere presented on35 patients in eachgroup. No mentionwas found of anywithdrawals (yes)

Analysis: do notknow

Statistical analysis:do not know

Baseline differences:translated as atinclusion there wasno significantdifference betweenthe two groups in the

stated


Intervention

Comparison

Outcomemeasures

Effect sizes

KCE Report 193S

Comments

KCE Report 193S


theoretical risk ofdeveloping pressureulcers or any of themain factors knownto contribute to theoccurrence ofbedsores.

Study power/samplesize: no a priorisample sizecalculation given

Setting: do not know

Length of study:maximum follow-up30 days

Categorisation ofPus:

Assessment of PUs:do not know



Intervention

Comparison

Outcomemeasures

Effect sizes

441

Comments

442

Table 15 – GILCREAST2005


Author and year:Gilcreast (2005)

Title: Researchcomparing three heelulcer-preventiondevices

Journal: Journal ofwound ostomy andcontinence nursing,32 (2), 112-120.

Type of study: RCT

Sequence generation:drawing of cards

Allocationconcealment:inadequate (non-numbered envelopes)

Blinding: no- 1 nursewas performing allresearch tasks andwas not blinded to thedevice to which theparticipant wasassigned.

Addressingincomplete outcomedata: gives details ofwhy patients were notfollowed up butunclear which groupthey were from.

Patient group: patientsmoderate or high riskof pressure ulcerdevelopment (69% ofparticipants were inICU)

All patients

Randomised N: 338(not clear howdistributed among the3 groups).

Completed N: 240

Dropouts: 29% – 53not included, as didnot wear the devicesfor at least 48 hours;45 not included asthey were non-compliant.

Group 1

Randomised N:unclear

Completed N: 77

Dropouts: unclear

Group 2

Randomised N:


Intervention

Comparison

Outcomemeasures

Effect sizes

atients Group 1: Bunny boot (fleece)high cushion heel protector

Group 2: Egg crate heel liftpositioner

Group3: foot waffle

The investigators attemptedto control for all extraneousvariables by monitoring allfactors relating to pressureulcer development.


Group 1: 3/77 (4%)

Group 2: 4/87 (5%)

Group 3: 5/76 (7%)

KCE Report 193S

Comments

Group 1: 3/77 (4%)

Group 2: 4/87 (5%)

Group 3: 5/76 (7%)

Funding:TriServiceNursing ResearchProgram

Limitations:Inadequateallocationconcealment; noblinding; limiteddetails of baselinedata; unclear howmany patientswere randomisedto each groupand thereforewhich arms thedrop-outs camefrom but therewere 29% ofpatients who didnot have follow-up data.

Additionaloutcomes: *

Notes: *

KCE Report 193S


Analysis: no ITTanalysis.

Statistical analysis:chi-square , analysisof variance andlogistic regressionanalysis

Baseline differences:limited baselineinformation presented(unclear). Baselineimbalance in sex.

Study power/samplesize: a prioricalculation of 80%power required 550participants totalsample of 338 patientswas obtained.

Setting: militarytertiary-care academicmedical centre.

Length of study:follow-up periodunclear

Categorisation of PUs:NPUAP

Assessment of PUs:skin assessed daily


unclear

Completed N: 87

Dropouts: unclear

Group 2

Randomised N:unclear

Completed N: 76

Dropouts: unclear

Inclusion criteria:patients with moderateor high risk ofpressure ulcerdevelopment (Bradenscore</= 14).

Exclusion criteria:

Patients with hipsurgery; patientsanticipated to beadmitted for <72hours; those with pre-existing heel pressureulcers.


Intervention

Comparison

Outcomemeasures

Effect sizes

patients with moderate

443

Comments

444

Table 16 – TYMEC1997


Author and year:Tymec 1997

Title: A comparisonof two pressure-relieving devices onthe prevention ofheel pressure ulcers

Journal: Advancesin wound care, 1997,10 (1), 39-44.

Type of study:factorial design RCT

Sequencegeneration: blockrandomisation listand the patient’sposition order wasdetermined by a cointoss

Allocationconcealment: notreported (unclear)

Blinding: notreported (unclear)

Addressingincomplete outcomedata: thenumber/group notreported. 8/52developed grade 1pressure ulcers and

Patient group: patientsfrom nursing units ofhospital with a lowBraden score (at risk)

All patients

Randomised N: 52

Completed N: 44

Dropouts: 8 developedgrade 1 pressure ulcersand were removed fromthe study.

f/m: 23/29

Age, mean (range):66.6 s.d 16.5 years(27-90 years)

Mean Braden score atadmission: 11.8

Respiratory conditions:21

Cancer: 6

Stroke: 5

Group 1


Completed N: notreported


Intervention

Comparison

Outcomemeasures

Effect sizes

ade 1 pressure ulcersand were removed from

Group 1: Foot waffle (FDAapproved, non-abrasive vinylboot with built in foot cradleand inflated air chamber

Group 2: Hospital pillowunder both legs from belowknee to the Achilles tendon.

In this hospital the standardpillow is a 20-ounce (+/-2ounces) polyfiber-filledpillow.

Outcome 1:number of heelpressure ulcersdeveloped

Group 1: 0/26

Group 2: 1/26

Logistic regressionpillow/foot waffle0.44 , p=0.001, OR 4.38

Outcome 2: timeuntil pressureulcer occurred(mean survivaltime)

Group 1: 10 days

Group 2: 13 days

Kaplan Meierdifference

Log-rank tests p=0.036

KCE Report 193S

Comments

Logistic regressionpillow/foot waffle -1.48, s.e0.44 , p=0.001, OR 4.38


Limitations:unclear allocationconcealment,blinding, reportingof incompleteoutcome data.

Additionaloutcomes: tissueinterfacepressures.

Notes: number ofother ulcers eg.Metatarsal, top offoot.

Group 1: 10 days

Group 2: 13 days

Kaplan Meier – significant

rank tests p=0.036

KCE Report 193S


were removed fromthe study, so itwould appear thatthe 52 participantswere followed-up.


Statistical analysis:logistic regression

Baseline differences:no details given forcharacteristics of thegroups

Study power/samplesize: powercalculation for 80%power required 52sample size.

Setting: selectednursing units of alarge hospital


Categorisation ofPUs: AHCPRguideline pressureulcer stages

Assessment of PUs:skin inspection



Group 2


Completed N: notreported


Inclusion criteria:Braden score of <<16(risk); intact skin onheels.

Exclusion criteria: notreported.


Intervention

Comparison

Outcomemeasures

Effect sizes

445

Comments

446

Table 17 – DONNELLY2011


Author and year:Donnelly 2011

Title: An RCT todetermine the effect ofa heel elevation devicein pressure ulcerprevention post-hipfracture

Journal: Journal ofwound care, 20 (7), 309-318

Type of study: RCT

Sequence generation:computer-generatedblock randomisationschedule (permutedblocks of 20)

Allocationconcealment:randomisation schedulewas held and managedby a senior researchnurse manager notdirectly involved in thestudy.

Blinding: authors statethat it was not possibleto blind either thepatient or theinvestigator as theintervention was very

Patient group: post-hip fracture patients.

All patients

Randomised N: 239

Completed N: 227

f/m: 184/55

age (mean, range):81 years (65-100)

Group 1

Randomised N: 120

Completed N: 111

Dropouts: 9(deterioratingmedical conditionn=6, lost-to follow-upn=1, adverse eventpossibly linked to theintervention n=1,patient withdrewconsent n=1).

Group 2

Randomised N: 119

Completed N: 116

Dropouts: 3 (lost tofollow up n=1,


Intervention

Comparison

Outcomemeasures

Effect sizes

up

possibly linked to the

Group 1: Heel elevation(Heelift Suspension Boot)plus pressure-redistributingsupport surface

Group 2: standard care pluspressure-redistributingsupport surface alone).

Mattress type determined byward nurses according toperceived need. Theirchoice was recorded andanalysed as a covariate.

Outcome 2:incidence of heelulcers (allcategories)

Group 1: 0/120

Group 2: 17/119

Outcome 3:comfort (themedanalysis)

Group 1: 32% of subjects feltthe boots interfered withsleep and 41% felt that theyadversely affecmovement in bed, 59% ratedthem as comfortable overall.Poor concordance reasonswere the weight and bulk ofthe boot (36%), heat(particularly at night) (31%)and discomfort (24%).

KCE Report 193S

Comments

Group 1: 0/120

Group 2: 17/119

Funding:researchsupported by aSpecial NursingResearchFellowshipfunded by theResearch andDevelopmentOffice for Healthand Social Carein NorthernIreland.

Limitations: Noblinding of patientor investigator;underpowered.

Additionaloutcomes:

Notes: *

Group 1: 32% of subjects feltthe boots interfered withsleep and 41% felt that theyadversely affectedmovement in bed, 59% ratedthem as comfortable overall.Poor concordance reasonswere the weight and bulk ofthe boot (36%), heat(particularly at night) (31%)and discomfort (24%).

KCE Report 193S


distinctive. Outcomeassessor was blinded.

Addressing incompleteoutcome data: yes, flowdiagram given

Analysis: ITT

Statistical analysis: Chi-squared test forassociation forproportion of patientsdeveloping one or morePU. Kaplan-Meier forgroup survival. CoxHazards RegressionalModel to analyse thepotential impact ofcovariates.

Baseline differences:no statisticallysignificant differencesat baseline.

Study power/samplesize: powered for 240patients per group togive 87.5% power,whereas had half thisamount.

Setting: fracture traumaunit of a major tertiaryreferral centre


deteriorating medicalcondition n=1,recruited incorrectlyn=1)

Inclusion criteria:aged 65 years orover on day offracture; suffered ahip fracture,including any bonyinjury to the femoralhead or femoralneck, in the previous48 hours

Exclusion criteria:did not give written,informed consent, orindicate willingnessto participate througha process ofinclusionary consent;existing heelpressure damage(NPUAP); and/orhistory of previouspressure ulceration;patients for whomthe investigator ormedical/nursingteam consideredunsuitable.


Intervention

Comparison

Outcomemeasures

Effect sizes

deteriorating medical

neck, in the previous

informed consent, or

participate through

inclusionary consent;

447

Comments

448


Categorisation of PUs:NPUAP scale.

Assessment of PUs:skin risk assessmenttool – modified Knollrisk assessment tool


Table 18 – TORRA2009


Author and year:Torra 2009

Title: Preventingpressure ulcers onthe heel: a Canadiancost study

Journal:DermatologyNursing 2009, 21 (5),268-272.

Type of study:multicentre RCT

Sequencegeneration: nodetails of method

Allocationconcealment: nodetails

Patient group: Nursinghome patients andhome care programpatients from primaryhealth care centres.

All patients

Randomised N: 130

Completed N: 111

Dropouts: 19 – 6 died,8 left study (fourbecause of settingchange and the otherfour following clinicaldecision), 4 abandonedthe study (died)

Group 1


Intervention

Comparison

Outcomemeasures

Effect sizes

Intervention

Comparison

Outcomemeasures

Effect sizes

Group 1: specialpolyurethane foamhydrocellular dressing for theprotection of the heel(Allevyn Heel) and normalmeasures of preventingpressure ulcers. Dressingswere fixed with a socket or anet bandage.

Group 2: protective bandageof the heel (Soffban andgauze bandage). Thebandage covered all theankle articulation. Normalmeasures for preventingpressure ulcers.


Group 1: 3.3%

Group 2: 44%

RR: 13.42 (95% CI 3.31 to54.3)

P<0.001

KCE Report 193S

Comments

Comments

%

44%

RR: 13.42 (95% CI 3.31 to


Limitations: openstudy. Unclearhow many ineach group butrelative riskreported. Nodetails ofallocationconcealment andrandomisationmethod. Unclearaddressing ofincompleteoutcome data.

Additional

KCE Report 193S


Blinding: open study

Addressingincomplete outcomedata: no details bygroup.

Analysis: no details

Statistical analysis:no details

Baseline differences:no statisticallysignificantdifferences

Study power/samplesize: no a prioripower calculationgiven but 130entered study

Setting: nursinghomes and threehome careprogrammes fromprimary care centres.

Length of study: 8weeks

Categorisation ofPUs: no details

Assessment of PUs:no details

Multiple ulcers: nodetails

Randomised N: unclear


Dropouts: unclear

Group 2



Dropouts: unclear

Inclusion criteria:patients at risk ofdeveloping pressureulcers according toBraden Scale; patientswho could give consentto participate in thestudy

Exclusion criteria:patients with existingpressure ulcers inheels; patients withdiabetes; patients usingspecial preventionsurfaces; patients usingdevices for relievinglocal pressure at heels


Intervention

Comparison

Outcomemeasures

Effect sizes


diabetes; patients using

surfaces; patients using

449

Comments

outcomes:

Notes: TheAllevyn heel issaid to be adressing butlooks to be also adevice for theheel.

Another studyTorra I Bou et al(2002) was theoriginal study butthis was a foreignlanguage paper.

450

9. NUTRITION AND HYDRATION


Table 1 – Review protocol

Protocol Nutrition/hydration

Review question What are the most clinically effective interventions with nutrition or hydration for the prevention of pressure ulcers forpeople with and without nutritional deficiency


With and without nutritional deficiencies

Intervention Nutritional interventions (supplementation or special diet)

Hydration

As prevent

Comparison Usual diet (participant’s usual diet or the standard hospital diet)

Other supplementation

Other special diet


Proportion of participants developing new pressure ulcers (dichotomous outcome)ulcer)

Important outcomes:

Patients acceptability


Time to develop new pressure ulcer (time to even

Time in hospital

Side effects (nausea, vomiting, diarrhoea) (dichotomous data)

Healthpressure ulcer patients,

o Short


ION

Nutrition/hydration

What are the most clinically effective interventions with nutrition or hydration for the prevention of pressure ulcers fore with and without nutritional deficiency?


With and without nutritional deficiencies

Nutritional interventions (supplementation or special diet)

Hydrational strategies

As preventive strategies

Usual diet (participant’s usual diet or the standard hospital diet)

Other supplementation

Other special diet



Important outcomes:

Patients acceptability of supplements – e.g. measured by compliance, tolerance, reports of unpalatability




Side effects (nausea, vomiting, diarrhoea) (dichotomous data)



KCE Report 193S

What are the most clinically effective interventions with nutrition or hydration for the prevention of pressure ulcers for

Proportion of participants developing new pressure ulcers (dichotomous outcome) (describe different categories of

measured by compliance, tolerance, reports of unpalatability

related quality of life (continuous data) (although unlikely to be sensitive enough to detect changes in

KCE Report 193S



o EQ

o WHO


o HUI




Cohort studies will be considered if




Search strategy The databases to be searched are:


All years.


The review strategy How will individual PICO characteristics be combined in a meta

Population

Interventionwill not be combined for meta

Outcomes

Study destudies will be meta

Unit of analysisunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe differentcategories of ulcer)


Nutrition/hydration


EQ-5D

WHO-QOL BREF

Cardiff HRQoL tool

HUI


High quality systematic reviews of RCTs and/or RCTs only








All years.


How will individual PICO characteristics be combined in a meta-analysis?:


ention – Different types of nutritional supplementation and hydration strategies and nutritional interventionswill not be combined for meta-analysis

Outcomes – single side effects eg nausea will be meta-analysed separately from other side effects


Unit of analysis – patients, clusters (hospital wards), individual pressure ulcersunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe differentcategories of ulcer)

451


analysis?:

for different strata

and hydration strategies and nutritional interventions



al wards), individual pressure ulcers – for those where patients are theunit of analysis and the patient has multiple ulcers it should be the first pressure ulcer occurring (describe different

452


Minimum duration of treatment = no minimimprovements.



Use available case analysis for dealing with missing data if there is a 10% differential or higher between theor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data.


Analysis Strata – where included studies are split up at outset as separate reviews (dissimilar groups andconfident that the intervention will work very differently in the two (or more) strata.recommendations on these.



With and without nutritional deficiency

Different nutritional supplements

Hydrational strategies and nutritional interventions

Subgroup analysison the basis of pre

The following groups will be considered separately as subgroups:


Other terms



Nutrition/hydration

Minimum duration of treatment = no minimum, but would expect at least a fortnight before they showimprovements.



Use available case analysis for dealing with missing data if there is a 10% differential or higher between theor if the missing data is higher than the event rate, if cannot work out the available case analysis will take theauthor’s data..


where included studies are split up at outset as separate reviews (dissimilar groups andconfident that the intervention will work very differently in the two (or more) strata.recommendations on these.

lowing groups will be considered separately as strata if data are present:


With and without nutritional deficiency

Different nutritional supplements

Hydrational strategies and nutritional interventions

roup analysis – combining all the studies together initially and then looking at any inconsistency between studiesbasis of pre-defined subgroups.

The following groups will be considered separately as subgroups:



KCE Report 193S

, but would expect at least a fortnight before they show

Use available case analysis for dealing with missing data if there is a 10% differential or higher between the groupsor if the missing data is higher than the event rate, if cannot work out the available case analysis will take the


where included studies are split up at outset as separate reviews (dissimilar groups and we need to beconfident that the intervention will work very differently in the two (or more) strata. The GDG will make separate

lowing groups will be considered separately as strata if data are present:

combining all the studies together initially and then looking at any inconsistency between studies


KCE Report 193S

9.2. search strategy


There were no limitations on sample size and only direct studies relating to pressure ulcers and nutrition or hydrationcomparisons or outcomes were considered. Only randomised controlled trials were included. Abstracts were not included unlesscontrolled trial full papers for the comparison. No studies wer



Search strategy

Date 27th Mar 2012


Search strategy(part I – nurition)

1 pressure ulcer/

2 decubit*.ti,ab.



5 or/1-4


7 exp diet/

8 exp food/

9 exp nutritional support/

10 enteral nutrition/

11 exp parenteral nutrition/

12 malnutrition/

13 exp diet therapy/

14 dh.fs.

15 (nutri* or food* or diet*).ti,ab.

16 or/7-15

17 6 and 16



20 randomi#ed.ab.


There were no limitations on sample size and only direct studies relating to pressure ulcers and nutrition or hydration were included. No indirect interventions,comparisons or outcomes were considered. Only randomised controlled trials were included. Abstracts were not included unlesscontrolled trial full papers for the comparison. No studies were found for hydrational interventions to prevent the occurrence of pressure ulcers.


sore*).ti,ab.


exp nutritional support/

exp parenteral nutrition/

(nutri* or food* or diet*).ti,ab.



453

were included. No indirect interventions,comparisons or outcomes were considered. Only randomised controlled trials were included. Abstracts were not included unless there were no randomised

e found for hydrational interventions to prevent the occurrence of pressure ulcers.

Results

9086

3915

6200

508

13124

10393

170157

944480

35531

14514

20532

4931

37786

34571

662638

1465966

753

322698

84030

284036

454

Search strategy

21 placebo.ab.

22 drug therapy.fs.

23 randomly.ab.

24 trial.ab.

25 groups.ab.

26 or/18-25


28 trial.ti.

29 or/18-21,23,27-28

30 letter/

31 editorial/

32 news/



35 comment/

36 case report/


38 or/30-37


40 38 not 39





45 exp Rodentia/


47 or/40-46

48 Meta-Analysis/






zed controlled trial/ or random*.ti,ab.





ysis as Topic/


KCE Report 193S

Results

134576

1518236

174415

246780

1145216

2903459

159472

102183

789656

750353

299086

142410

306887

4116

487891

1571028

82116

3034289

672095

3019416

3624822

675879

5199

371043

2493649

1040004

7176100

31869

12015

41158

KCE Report 193S

Search strategy



53 (search strategy or sear



56 cochrane.jw.

57 or/48-56

58 (29 or 57) not 47

59 17 and 58


Part II(hydration)

1 pressure ulcer/

2 decubit*.ti,ab.



5 or/1-4


7 fluid therapy/

8 dehydration/

9 drinking/

10 (hydrat* or rehydrat* or re

11 or/7-10

12 6 and 11

13 letter/

14 editorial/

15 news/



18 comment/

19 case report/







Current"


sore*).ti,ab.


(hydrat* or rehydrat* or re-hydrat* or dehydrat* or de-hydrat*).ti,ab.

455

Results

ch criteria or systematic search or study selection or data extraction).ab.


48805

19812

21689

19180

60492

8210

142473

780799

106

59

9086

3915

6200

508

13124

10393

12793

9572

11760

63383

87489

95

750353

299086

142410

306887

4116

487891

1571028

456

Search strategy


21 or/13-20


23 21 not 22





28 exp Rodentia/


30 or/23-29

31 12 not 30


Search strategy

Date 27th Mar 2012


Search strategy(part I –nutrition)

1 decubitus/

2 decubit*.ti,ab.



5 or/1-4


7 exp diet/

8 exp food/

9 exp diet therapy/

10 exp nutritional support/

11 exp artificial feeding/

12 exp food intake/









sore*).ti,ab.


exp nutritional support/

KCE Report 193S

Results

82116

3034289

672095

3019416

3624822

675879

5199

371043

2493649

1040004

7176100

86

Results

12024

4568

6772

630

15589

11928

153794

526257

186661

10892

49886

168353

KCE Report 193S

Search strategy

13 exp malnutrition/

14 (nutri* or food* or diet*).ti,ab.

15 or/7-14

16 6 and 15

17 random*.ti,ab.









26 or/17-25


28 note.pt.

29 editorial.pt.



32 or/27-31


34 32 not 33


36 nonhuman/



39 animal model/

40 exp Rodent/


42 or/34-41


(nutri* or food* or diet*).ti,ab.



* or volunteer* or placebo*).ti,ab.






comment*).ti.





457

Results

90561

734983

1319130

1068

665174

17410

57063

129012

518363

31195

101701

14442

292701

1106203

750039

457705

385981

1762297

131461

3234388

740298

3210903

1264585

3741600

1475898

361812

612474

2401842

1065594

8534950

458

Search strategy


44 meta-analysis/






50 (medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citatiindex or bids or cancerlit).ab.


52 cochrane.jw.

53 or/43-52

54 (26 or 53) not 42

55 16 and 54


Part II(hydration)

1 decubitus/

2 decubit*.ti,ab.



5 or/1-4


7 rehydration/

8 fluid therapy/

9 drinking/

10 (hydrat* or rehydrat* or re

11 or/7-10

12 6 and 11


14 note.pt.

15 editorial.pt.







(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citati


Current"


sore*).ti,ab.


(hydrat* or rehydrat* or re-hydrat* or dehydrat* or de-hydrat*).ti,ab.

KCE Report 193S

Results


(medline or pubmed or cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or science citati on

45174

57412

49825

53088

22849

24490

21961

68666

28922

10982

205807

1031869

151

105

12024

4568

6772

630

15589

11928

3444

12893

9832

67509

89258

118

750039

457705

385981

KCE Report 193S

Search strategy



18 or/13-17


20 18 not 19


22 nonhuman/



25 animal model/

26 exp Rodent/


28 or/20-27

29 12 not 28


Search strategy

Date 27th Mar 2012

Database CINAHL


S10 s8 not s9

S9 PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orPT cartoon or PT commentary or PT computer program or PT editorial oror PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTpictorial or PT poetry or PT proceedings or PT “questions and answers” or PT rematerials or PT website


S7 S5 and S6

S6 nutri* or food* or diet*






ized controlled trial/ or random*.ti,ab.




PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orPT cartoon or PT commentary or PT computer program or PT editorial or PT games or PT glossary or PT historical materialor PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTpictorial or PT poetry or PT proceedings or PT “questions and answers” or PT response or PT software or PT teaching

Published Date from: 20020101-20111231; English Language; Exclude MEDLINE records

nutri* or food* or diet*

S1 or S2 or S3 or S4

sore*

459

Results

1762297

131461

3234388

740298

3210903

1264585

3741600

1475898

361812

612474

2401842

1065594

8534950

98

Results

PT anecdote or PT audiovisual or PT bibliography or PT biography or PT book or PT book review or PT brief item orPT games or PT glossary or PT historical material

or PT interview or PT letter or PT listservs or PT masters thesis or PT obituary or PT pamphlet or PT pamphlet chapter or PTsponse or PT software or PT teaching


109

974559

164

786

138288

8354

152

460

Search strategy


S2 decubit*


Part II(hydration)


S10 S5 and S9

S9 S6 or S7 or S8

S8 hydrat* or rehydrat* or re

S7 (MH "Fluid Therapy")

S6 (MH "Dehydration")



S3 pressure n1 sore* OR pressure n1 ulcer* OR

S2 decubit*



Search strategy

Date 27th Mar 2012

Database Cochrane (- CDSR [3/2012]; DARE; Central [3/2012];






#5 (#1 OR #2 OR #3 OR #4)

#6 Any MeSH descriptor with qualifier: DH

#7 (nutri* or food* or diet*):kw,ti,ab

#8 (#6 OR #7)

#9 (#5 AND #8)

#10 (#9), from 2002 to 2011




English Language; Exclude MEDLINE records

hydrat* or rehydrat* or re-hydrat* or dehydrat* or de-hydrat*

(MH "Fluid Therapy")


sore*



CDSR [3/2012]; DARE; Central [3/2012]; NHS EED; HTA)



sore*):ti,ab,kw

(#1 OR #2 OR #3 OR #4)

SH descriptor with qualifier: DH

(nutri* or food* or diet*):kw,ti,ab

(#9), from 2002 to 2011

KCE Report 193S

Results

8090

466

7352

29

72

5691

4196

2106

1724

8354

152

8090

466

7352

Results

472

340

805

31

1076

4606

42630

42630

65

35

KCE Report 193S

Part II(hydration)

S5 and S9 Limiters – English Language; Exclude MEDLINE records

S5 and S9

S6 or S7 or S8

hydrat* or rehydrat* or re-hydrat* or dehydrat* or de

(MH "Fluid Therapy")

(MH "Dehydration")


bedsore* OR bed-sore*


decubit*


9.2.3. Flow diagram for article selection

Figure 1 – Flow diagram for article selection



Excluded, n


Excluded, n = 6


English Language; Exclude MEDLINE records

hydrat* or dehydrat* or de-hydrat*


Excluded, n = 325

461

29

72

5691

4196

2106

1724

8354

152

8090

466

7352

462

9.2.4. Excluded studies

Author/title REF ID Reason for exclusion

Larsson 1990/ Effect of dietarysupplement on nutritional statusand clinical outcome in 501Geriatric Patients – A RandomisedStudy

None of our outcomes exceptmortality

Ek 1987/ Prediction of pressuresore development

Only outcome given is incidence ofpressure sores, b9.9% in the experimental and 12%in the control group. There are nodetails of how many patients werein the experimental and controlgroups. Is linked to Larsson 1990but it has a different numberwithdrawn so don’t think figurescan be used as the denominatorfor this outcome.

Neander 2004/A specificnutritional supplement reducesthe incidence of pressure ulcersin elderly people

Abstract

Okuwa2009/ The prevalence andincidence of pressure ulcers inhome care setting in Japan

Abstract

Gallart 2010/ Prevention ofpressure sores in patients withpoor perfusion tissue: a pilotstudy comparing oil vs milkhyperoxygenated fatty acids

Abstract

Sampson 2009/ Enteral tubefeeding for older people withadvanced dementia (Review)

Cochrane review but did notinclude RCTs.



None of our outcomes except

Only outcome given is incidence ofpressure sores, but this is given as9.9% in the experimental and 12%in the control group. There are nodetails of how many patients werein the experimental and controlgroups. Is linked to Larsson 1990but it has a different numberwithdrawn so don’t think figures

be used as the denominatorfor this outcome.

ane review but did notinclude RCTs.

9.3. Clinical evidenceNo studies were found for hydrational interventions to prevent theoccurrence of pressure ulcers. A Cochrane Review by Langer (2003)including four RCTs about the effect of nutritional interventions to preventpressure ulcers was found. We updated the Cochrane review with fourother studies, Dennis et al. (2005)(2007)

139and Oloffson et al. (2007)

(1998)138

and Oloffson et al. (2007)but rather pressure ulcers were an event or compduring these trials.

The literature search and Cochrane reviewers identified five RCTscomparing participants who received nutritional supplementation inaddition to their standard diet (which was the hospital standard diet) tothose who received only the standard hospital diet.studies all included older people who were in hospital.(2003)

143and Hartgrink et al. (1998)

Delmi et al. (1990)142

included patients witBourdel-Marchasson et al. (2000)Dennis et al. (2005)

137included stroke patients. Hartgrink et al. (1998)

gave patients a supplement of energy and protein by nasogastric tubecompared to the standard hospital diet. Studies followfrom 2 weeks to 6 months. The supplements included variouscompositions of protein, carbohydrate, vitamins and minerals.

One study138

included long-term patients with type 2Researchers gave the patients a diseaseand modified fat) formula compared to the standard high carbohydrateformula. Patients were followed up for 3 months.

Another study139

gave patients suffering from lung injury a macronutrientdiet plus lipids and vitamins compared to a macronutrient diet alone. Thesepatients were followed up for 7 days.

One RCT140

with femoral neck fracture patients who weenriched meals compared to normal postoperative care and followed themup for 4 months.

We have meta-analysed the results in contrast to the original Cochranereview

136to lump the studies together aiming to gain a greater confidence

in the evidence and then report on heterogeneity of studies if this exists.

KCE Report 193S

No studies were found for hydrational interventions to prevent theoccurrence of pressure ulcers. A Cochrane Review by Langer (2003)

136

including four RCTs about the effect of nutritional interventions to prevente ulcers was found. We updated the Cochrane review with four

(2005)137

, Craig et al. (1998)138

, Theilla et al.and Oloffson et al. (2007)

140. Dennis et al. (2005)

137, Craig et al.

(2007)140

were not looking at pressure ulcers,but rather pressure ulcers were an event or complication that occurred

The literature search and Cochrane reviewers identified five RCTscomparing participants who received nutritional supplementation inaddition to their standard diet (which was the hospital standard diet) to

received only the standard hospital diet.137,141 , 142 , 143,144

Thesestudies all included older people who were in hospital. Houwing et al.

(1998)144

included patients with hip fracture,included patients with fractured neck of the femur,

Marchasson et al. (2000)141

included critically ill patients andincluded stroke patients. Hartgrink et al. (1998)

144

gave patients a supplement of energy and protein by nasogastric tubee standard hospital diet. Studies follow-up period ranged

from 2 weeks to 6 months. The supplements included variouscompositions of protein, carbohydrate, vitamins and minerals.

term patients with type 2 diabetes.Researchers gave the patients a disease-specific (reduced-carbohydrateand modified fat) formula compared to the standard high carbohydrateformula. Patients were followed up for 3 months.

gave patients suffering from lung injury a macronutrientdiet plus lipids and vitamins compared to a macronutrient diet alone. Thesepatients were followed up for 7 days.

with femoral neck fracture patients who were given protein-enriched meals compared to normal postoperative care and followed them

analysed the results in contrast to the original Cochraneto lump the studies together aiming to gain a greater confidence

n report on heterogeneity of studies if this exists.

KCE Report 193S

We meta-analysed studies together that looked at nutritional supplementsin addition to standard hospital diet (which mainly included energy andprotein) versus the standard hospital diet.

137,141 ,

another meta-analysis of these studies of nutritional supplements and alsoincluded a study (Oloffson et al., 2007)

140with a protein diet compared to

the standard hospital diet since all of the interventions had a highproportion of protein.

Some of the studies gave the results separately by grade of pressure ulcerthat occurred as well as all grades of ulcers that occurred. We have splitthe results (see appendix 9) to show data for all pressure ulcers and forthose with grade 2-4 ulcers (with details of the classification system ofgrading).


Table 6 – RCTs and outcomes included in the review

Study Study design Population

Houwing2003

143RCT

Double blind

Older people with hip fracture

Bourdel-Marchasson2000

141

RCT

Unblinded

Critically ill older people

Hartgrink1998

144RCT

Unblinded

Older people with hip fracture

Delmi 1990142

RCT

Unblinded

Older people with fracturedneck of the femur


analysed studies together that looked at nutritional supplementsin addition to standard hospital diet (which mainly included energy and

, 142 , 143,144We conducted

of nutritional supplements and alsowith a protein diet compared to

the standard hospital diet since all of the interventions had a high

Some of the studies gave the results separately by grade of pressure ulcert occurred as well as all grades of ulcers that occurred. We have split

) to show data for all pressure ulcers and for4 ulcers (with details of the classification system of

RCTs and outcomes included in the review

Population Interventions/comparison Outcomes

Older people with hip fracture Standard diet with additionaloral supplementation (highprotein enriched with argininezinc and antioxidants) versusstandard diet with a placebo.

Incidence of pressureulcers; time to first day ofpressure ulcer; mortality.

Critically ill older people Standard diet with additionaloral supplementation (protein,fat, carbohydrate and mineralsand vitamins) versus standarddiet.

Incidence of pressureulcers

Older people with hip fracture Standard diet with tubefeeding (energy, protein,Nutricia) versus standard diet


Older people with fracturedneck of the femur

Standard diet with additionaloral nutrition supplements(protein, carbohydrate, lipid,calcium, vitamin A, vitamin D,vitamins E, B1, B2, B6, B12,


463

Follow-up period(weeks)

Incidence of pressureulcers; time to first day ofpressure ulcer; mortality.

28 days

Incidence of pressure 15 days.

Incidence of pressure 2 weeks

Incidence of pressure Assessed at 14, 21and 28 days andfollowed up at 6months

464

Study Study design Population

Craig 1998138

RCT double-blinded pilot study

LTC residents with type 2diabetes

Theilla2007

139RCT unblinded Critically ill, mechanically

ventilated patients sufferingfrom acute lung injury

Olofsson2007

140RCT Femoral neck fracture

patients

Dennis2005

137Multicentre RCT Elderly stroke patients in

hospital


Population Interventions/comparison Outcomes

C, nicotinamide, folate,calcium pantothenate, biotin,and minerals) versusstandard diet

LTC residents with type 2 Disease-specific (reduced-carbohydrate, modified-fat)formula vs standard high-carbohydrate formula


Critically ill, mechanicallyventilated patients sufferingfrom acute lung injury

Macronutrient diet plus lipids(elcosapentanoic acid,gamma-linolenic acid, vitaminsA, C and E) vs macronutrientdiet read to feed (high fat, lowcarbohydrate, enteral formula)


Femoral neck fracture Protein-enriched meals vsnormal postoperative care

Incidence of pressureulcers; time in hospital

Elderly stroke patients in Normal hospital diet plus oralsupplements vs normalhospital diet

Incidence of pressureulcers; length of stay inhospital

KCE Report 193S

Follow-up period(weeks)

ce of pressure 3 months

Incidence of pressure 7 days

Incidence of pressureulcers; time in hospital

4 months follow-up

Incidence of pressureulcers; length of stay in

6 months follow-up

KCE Report 193S


Table 7 – Protein, fat, carbohydrate, minerals and vitamins supplement (twice daily 200kcal, protein 30%, fat 20%, carbohydrate 50%, zincvitamin C 15mg) and standard diet versus standard dietpopulation

Quality assessment


Inconsistency

Incidence of PU – critically ill older patients

1Bourdel-Marchasson2000

randomisedtrials

veryseriousa


Acceptability of supplements – compliance – critically ill older patients

1Bourdel-Marchasson2000

randomisedtrials

veryseriousa


a Unclear details of sequence generation, no blinding and high levels of missing data in both groups. Difference at baselinehigher risk of pressure ulcers (Norton score and were more dependent (Kuntzmann score), however the level of serum albumin was lower in thewhich indicates a higher risk of pressure ulcers. The authors thought it was not easy to propose amanner as this could have a deleterious effect on the energy intake in the control group because in elderly hospitalised patifood could limit voluntary energy intake. The study was randomised by hospital wards (19) which were stratified according to tpatients. The nurses in the wards were trained by the research nurse and thb The confidence interval crossed one MID point.c 60% in the supplement group were compliant at end of the 1st week and this was 99% at the end of thefollow-up.d The nutritional intervention group had energy intake of 1081 +/595 kcal and the standard hospital diet group had 957 +/38.3+/-23.8g respectively, p<0.001.


otein, fat, carbohydrate, minerals and vitamins supplement (twice daily 200kcal, protein 30%, fat 20%, carbohydrate 50%, zincvitamin C 15mg) and standard diet versus standard diet – patients not specified as malnourished but thought at higher risk



Nutritionalsupplement

plus standardhospital diet

Standardhospital

diet

Relative(95% CI)


Seriousimprecisionb

Noned 118/295(40%)

181/377(48%)

RR 0.83

48%

critically ill older patients


N/A Noned See footnotec N/A

a Unclear details of sequence generation, no blinding and high levels of missing data in both groups. Difference at baseline for risk of pressure ulcerspressure ulcers (Norton score and were more dependent (Kuntzmann score), however the level of serum albumin was lower in the

which indicates a higher risk of pressure ulcers. The authors thought it was not easy to propose a placebo oral supplement with similar taste and consistency in a doublemanner as this could have a deleterious effect on the energy intake in the control group because in elderly hospitalised pati ents, the volume rather than the energy content of

od could limit voluntary energy intake. The study was randomised by hospital wards (19) which were stratified according to t heir specialty and recruitment for critically ill olderpatients. The nurses in the wards were trained by the research nurse and the dietician to monitor patients. Multivariate analyses took into account the intra

c 60% in the supplement group were compliant at end of the 1st week and this was 99% at the end of the second week. 7% of the control group had the supplement during

d The nutritional intervention group had energy intake of 1081 +/595 kcal and the standard hospital diet group had 957 +/ - 530 kcal, p=0.006 and protein 45.9 +/

465

otein, fat, carbohydrate, minerals and vitamins supplement (twice daily 200kcal, protein 30%, fat 20%, carbohydrate 50%, zinc 1.8mg,patients not specified as malnourished but thought at higher risk as critically ill older


Relative(95% CI)

Absolute

RR 0.83(0.7 to0.99)


fewer)

VERYLOW

Critical


fewer)

N/A N/A N/A Critical

for risk of pressure ulcers – the control group had apressure ulcers (Norton score and were more dependent (Kuntzmann score), however the level of serum albumin was lower in the nutritional intervention group

placebo oral supplement with similar taste and consistency in a double-blindents, the volume rather than the energy content of

heir specialty and recruitment for critically ill oldere dietician to monitor patients. Multivariate analyses took into account the intra-ward correlation.

second week. 7% of the control group had the supplement during

530 kcal, p=0.006 and protein 45.9 +/-27.8grams and

466

Table 8 – High protein enriched with arginine zinc and antioxidants supplement (energy 125kcal, protein 10g, l125 mg, vitamin E 50mg x-TE, carotenoids 1g) and standard diet versus placeboassumed as population had hip fracture

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of all pressure ulcers – older patients with hip fracture

1Houwing2003

randomisedtrials

Seriousa no seriousinconsistency

no seriousindi

Incidence of stage II pressure ulcers – older patients with hip fracture

1Houwing2003

randomisedtrials



Acceptability of treatment – compliance – older patients with hip fracture

1Houwing2003

randomisedtrials



a No details of sequence generation or allocation concealment.b Confidence interval crossed both MID points.c Approximately 70% of patients consumed the supplement for a week or more. 75% of the patients consumed 75% or more of their daily dose.


High protein enriched with arginine zinc and antioxidants supplement (energy 125kcal, protein 10g, lTE, carotenoids 1g) and standard diet versus placebo and standard diet – patients not specified as malnourished but




plus standarddiet

Placeboplus

standarddiet

Relative(95% CI)

older patients with hip fracture


veryseriousb

none 27/51(52.9%)

30/52(57.7%)

RR 0.92(0.65 to

1.3)

57.7%



veryseriousb

none 9/51(17.6%)

14/52(26.9%)

RR 0.66(0.31 to

1.38)

26.9%



N/A N/ASee footnotec N/A

a No details of sequence generation or allocation concealment.

of patients consumed the supplement for a week or more. 75% of the patients consumed 75% or more of their daily dose.

KCE Report 193S

High protein enriched with arginine zinc and antioxidants supplement (energy 125kcal, protein 10g, l -arginine 1.5mg, zinc 5mg, vitamin cpatients not specified as malnourished but


Relative(95% CI)

Absolute

RR 0.92(0.65 to

1.3)



VERYLOW

Critical



RR 0.66(0.31 to

1.38)



VERYLOW

Critical

-

N/A N/A N/A

Critical

of patients consumed the supplement for a week or more. 75% of the patients consumed 75% or more of their daily dose.

KCE Report 193S

Table 9 – Protein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C, nicotinamide, folate, calcium pbiotin, and minerals supplement (250ml supplement energy 254kcal, protein 20.4g, carbohydrate 29.5g, lipid 5.8g, calcium 525mvitamin Ds 25 IU) and standard hospital diet versus standard hospital diet

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers (at 6 months) – older patients with fractured neck of the femur

1Delmi1990

randomised trials

veryseriousa



Acceptability of treatment – compliance – older patients with fractured neck of the femur

1Delmi1990

randomised trials

veryseriousa



Time in hospital – older patients with fractured neck of the femur

1Delmi1990

randomised trials

veryseriousa



a No details of sequence generation, allocation concealment or blinding. High dropb Confidence interval crossed both MID points.c No standard deviations given.d This is the number at 6 months follow-up.e The supplement was said to be well-tolerated and completely ingested and no sidef A dietary survey of 50 daily measurements of foot intake showed energy intaThe supplement increased the intake of energy by 23%, protein 62%, calcium 130%. The supplements did not reduce the voluntary


rotein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C, nicotinamide, folate, calcium pbiotin, and minerals supplement (250ml supplement energy 254kcal, protein 20.4g, carbohydrate 29.5g, lipid 5.8g, calcium 525mvitamin Ds 25 IU) and standard hospital diet versus standard hospital diet – most patients nutritionally deficient

No of patients


Otherconsiderations

NutritionalSupplement plusstandard hospitaldiet

Standardhospital diet

Relative(95% CI)

older patients with fractured neck of the femur


veryseriousb nonef 0/25d

(0%)2/27d

(7.4%)RR 0.22(0.01 to

4.28)



N/Anonef See footnotee N/A N/A



Seriousc

nonef Median 24 days(range 13-157)

N=27

Median 40days (range

10-259)

N=32

P=0.09

a No details of sequence generation, allocation concealment or blinding. High drop-out. Baseline difference for plasmas level, which was lower in non

tolerated and completely ingested and no side-effects were observed.f A dietary survey of 50 daily measurements of foot intake showed energy intake was only 1100kcal (SD 300) per day – protein 34g (11) per day, calcium 400mg (250) per day.The supplement increased the intake of energy by 23%, protein 62%, calcium 130%. The supplements did not reduce the voluntary

467

rotein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C, nicotinamide, folate, calcium p antothenate,biotin, and minerals supplement (250ml supplement energy 254kcal, protein 20.4g, carbohydrate 29.5g, lipid 5.8g, calcium 525m g, vitamin A 750 IU,

ionally deficient

Effect Quality

Importance

lative(95% CI)

Absolute

RR 0.22(0.01 to

4.28)


243 more)

VERYLOW

Critical

N/A Critical

09 -

VERYLOW

Important

out. Baseline difference for plasmas level, which was lower in non-supplemented patients.

protein 34g (11) per day, calcium 400mg (250) per day.oral intake.

468

Table 10 – Nutritional supplement (360mL at 6.27kJmL and 62.5gL in protein) plus standard hospital diet versus standard hospital dietwere undernourished

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers – older stroke patients

1Dennis2005

randomisedtrials

VerySeriousa



Acceptability of supplements – compliance – older stroke patients

1Dennis2005

randomisedtrials



Length of time in hospital – older stroke patients

1Dennis2005

randomisedtrials



a Aim not to look at pressure ulcers and there were no details of pressure ulcers at start of the trial. No blinding to treatrate. Trial was stopped before they reached their target as no funding was available to continue beyond 2004 and to ensure the trial was closed in an orderly manner.b Confidence interval crossed one MID pointc Crude compliance rate of 79 (4%) did not receive any supplement. 48 of those who were supposed tof 98%.


ional supplement (360mL at 6.27kJmL and 62.5gL in protein) plus standard hospital diet versus standard hospital diet



Nutritionalsupplement plusstandard hospital

diet

Standardhospital

diet

Relative(95% CI)


Seriousb none 15/2016(0.7%)

26/2007(1.3%)

RR 0.57(0.31 to1.08)

1.3%

older stroke patients


N/A N/A See footnote c N/A


Seriousb none 34.0 (48.0)

N=2016

32.0 (46.0)

N=2007

a Aim not to look at pressure ulcers and there were no details of pressure ulcers at start of the trial. No blinding to treat ment allocationet as no funding was available to continue beyond 2004 and to ensure the trial was closed in an orderly manner.

c Crude compliance rate of 79 (4%) did not receive any supplement. 48 of those who were supposed t o only receive the normal diet had some supplements, crude compliance

KCE Report 193S

ional supplement (360mL at 6.27kJmL and 62.5gL in protein) plus standard hospital diet versus standard hospital diet – majority


Relative(95% CI)

Absolute

RR 0.57(0.31 to1.08)


fewer to 1more)

VERYLOW

Critical


fewer to 1more)


- MD 2.00higher (0.91lower to 4.91

higher)

LOW

Critical

ment allocation. . Higher drop-out rate than the eventet as no funding was available to continue beyond 2004 and to ensure the trial was closed in an orderly manner.

o only receive the normal diet had some supplements, crude compliance

KCE Report 193S

Table 11 – Tube fed energy, protein (1 litre Nutrion Steriflo Energypatients not specified as malnourished but assumed as older population with hip fracture

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of grade 2-4 pressure ulcers (Stage 0=normal skin, 1=persistent erythema of the skin, stage 2=blister formation, stage 3=superficial (sub)cutaneous necros4=subcutaneous necrosis, according to the Dutch consensus meeting for the prevention of pressure sores)

1Hartgrink1998

randomisedtrials

veryseriousa



Incidence of all pressure ulcers – older patients with hip fracture

1Hartgrink1998

randomisedtrials

veryseriousa



a No details of sequence generation, allocation concealment and no blinding. High dropdone as it was thought unethical to discomfort the control group with a nasogastric tube.b The confidence interval crossed both MID points.c The confidence interval crossed one MID point.


Tube fed energy, protein (1 litre Nutrion Steriflo Energy-plus – energy 1500kcal/l, protein 60 g/l) and standard diet versus standard dietas malnourished but assumed as older population with hip fracture



Nutritionalsupplement plus

standardhospital diet

Standardhospital

diet

Relative(95% CI)

(Stage 0=normal skin, 1=persistent erythema of the skin, stage 2=blister formation, stage 3=superficial (sub)cutaneous necrosutaneous necrosis, according to the Dutch consensus meeting for the prevention of pressure sores) – older patients with hip fracture


veryseriousb

none 25/48(52.1%)

30/53(56.6%)

RR 0.92(0.64 to

1.32)

56.6%



seriousc none 30/48(62.5%)

37/53(69.8%)

RR 0.90(0.68 to

1.19)

0%

a No details of sequence generation, allocation concealment and no blinding. High drop -out in both groups. Very few remained tube fed at 2 weeks (16/70). Blinding was notdone as it was thought unethical to discomfort the control group with a nasogastric tube.

469

energy 1500kcal/l, protein 60 g/l) and standard diet versus standard diet –


Relative(95% CI)

Absolute

(Stage 0=normal skin, 1=persistent erythema of the skin, stage 2=blister formation, stage 3=superficial (sub)cutaneous necros is, stageolder patients with hip fracture

RR 0.92(0.64 to

1.32)



VERYLOW

Critical



RR 0.90(0.68 to

1.19)



VERYLOW

Critical

-

groups. Very few remained tube fed at 2 weeks (16/70). Blinding was not

470

Table 12 – Disease-specific (reduced-carbohydrate, modified(high-carbohydrate) formula (1060kcal, 44.4g protein, 151.7g carbohydrate, 35.9g fat)care patients

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers – older long-term care patients with type 2 diabetes

1Craig1998

randomisedtrials

veryseriousa


no serindirectness

Adverse events – older long-term care patients with type 2 diabetes

1Craig1998

randomisedtrials

veryseriousa



a study aim was not to look at pressure ulcers, it was only an event experib Confidence interval crossed both MID points.c No statistically significant differences for number of adverse events reported.d Disease-specific formula was 1000kca


carbohydrate, modified-fat) formula (1000kcal, 41.8g protein,93.7g carbohydrate, 55.7g fat) versus standardcarbohydrate) formula (1060kcal, 44.4g protein, 151.7g carbohydrate, 35.9g fat) – patients not specified as malnourished but older long

No of patients


Disease-specific(reduced-

carbohydrate,modified-fat)

formula

Standard (high-carbohydrate)

formula

Relative(95% CI)

term care patients with type 2 diabetes


veryseriousb

none 7/17(41.2%)

8/15(53.3%)

RR 0.77

53.3%

rm care patients with type 2 diabetes


N/A N/A See footnotec N/A

a study aim was not to look at pressure ulcers, it was only an event experienced during the study. No details of sequence generation or allocation concealment.

c No statistically significant differences for number of adverse events reported.

KCE Report 193S

fat) formula (1000kcal, 41.8g protein,93.7g carbohydrate, 55.7g fat) versus standardnot specified as malnourished but older long-term


Relative(95% CI)

Absolute

RR 0.77(0.37 to

1.62)


fewer to 331more)

VERYLOW

Critical


fewer to 330more)


enced during the study. No details of sequence generation or allocation concealment.

KCE Report 193S

Table 13 – Macronutrient diet plus lipids (elcosapentanoic acid, gamma(high fat, low carbohydrate, enteral) formulac –

Quality assessment

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers – Critically ill, mechanically ventilated patients suffering from acute lung injury

1Theilla2007

randomisedtrials

veryseriousa


no sindirectness

Incidence of grade 2-4 pressure ulcers – Critically ill, mechanically ventilated patients suffering from acute lung injury

1Theilla2007

randomisedtrials

veryseriousa



a no details of sequence generation, allocation concealment. No blinding. BMI was higher in the intervention group at baseline.b Confidence interval crossed both MID points.c Formulas contained: EPA+GLA – 62.5g/L protein, 105.5g/L carbohydrate, 93.7g/L lipids, 317IU/L vitamin E, 844mg/L vitamin C181mg/L L-carnitine; the control group – 62.6g/L protein; 105.7g/L carbohydrate; 92.1g/L lipids, 85IU/L vitamin E, 317mg/L vitamin C, 160mg/L taurine, 160mg/L LThe lipids in EPA+GLA had 31.8% canola oil, 25% MCT, 20% fish oil, 3.2% soy lecithin the control group had 55.8% canola oil, 20% MCT, 14% corn oil, 7% high oleicsafflower oil and 3.2% soy lecithin.d Nutritional intake at baseline for EPA+GLA was 1053+/378kcal/day (57%), and 1420+/-437kcal/day (71%) at 7 days.


Macronutrient diet plus lipids (elcosapentanoic acid, gamma-linolenic acid, vitamins A, C and E) vs macronutrient diet ready to feed– patients not specified as malnourished

No of patients


Macronutrient dietplus lipids,

gamma-linolenicacid, vitamins A,C

and E

Macronutrient dietready to feed, high

fat, lowcarbohydrate,enteral formula

Critically ill, mechanically ventilated patients suffering from acute lung injury


veryseriousb noned 8/46

(17.4%)10/49

(20.4%)

20.4%

Critically ill, mechanically ventilated patients suffering from acute lung injury


veryseriousb noned 4/49

(8.2%)6/49

(12.2%)

12.2%

ion concealment. No blinding. BMI was higher in the intervention group at baseline.

62.5g/L protein, 105.5g/L carbohydrate, 93.7g/L lipids, 317IU/L vitamin E, 844mg/L vitamin C , 5.0 B62.6g/L protein; 105.7g/L carbohydrate; 92.1g/L lipids, 85IU/L vitamin E, 317mg/L vitamin C, 160mg/L taurine, 160mg/L L

, 25% MCT, 20% fish oil, 3.2% soy lecithin the control group had 55.8% canola oil, 20% MCT, 14% corn oil, 7% high oleic

d Nutritional intake at baseline for EPA+GLA was 1053+/-351kcal/day (49%)and 1624+/-512 (69%) at day 7; the nutritional intake at baseline for the control diet was 1055+/437kcal/day (71%) at 7 days.

471

linolenic acid, vitamins A, C and E) vs macronutrient diet ready to feed


Relative(95% CI)

Absolute

RR 0.85(0.37 to1.97)


198 more)

VERYLOW

Critical


198 more)

RR 0.71(0.21 to2.36)


167 more)

VERYLOW

Critical

, 5.0 B-carotene (mg/L), 316g/L Taurine,62.6g/L protein; 105.7g/L carbohydrate; 92.1g/L lipids, 85IU/L vitamin E, 317mg/L vitamin C, 160mg/L taurine, 160mg/L L -carnitine.

, 25% MCT, 20% fish oil, 3.2% soy lecithin the control group had 55.8% canola oil, 20% MCT, 14% corn oil, 7% high oleic

day 7; the nutritional intake at baseline for the control diet was 1055+/-

472

Table 14 – Protein-enriched mealsd vs normal postoperative care

Quality assessmen

No ofstudies

Design Risk ofbias


Incidence of pressure ulcers – Older femoral neck fracture patients

1Oloffson2007

randomisedtrials

Veryseriousa



Time in hospital (Better indicated by lower values) – Older femoral neck fracture patients

1Oloffson2007

randomisedtrials


no seriouindirectness

a Randomised to different wards. No blinding. Higher dropb Confidence interval crossed one MID point.c Limited number of events.d The intervention group had a nutritional journal for the first four days established the patients’ nutrition deficiencies.30 calories per kilo body weight to supply the extra energy requirement for the first four postoperative days or longer if required. At lunch an appprotein-enriched meals and a dessert at dinner. If the patients were malnourished on admission the nursneeded even more energy/caloiries. If had problems in these areas they consulted a dietitian. The patients in the interventidrinks 2x200ml daily while hospitalised. Additional nutritional and protein drinks were served after every meal for patients who nealso optimised to facilitate the intake of nutrition eg no unnecessary noise. The contro


enriched mealsd vs normal postoperative care – large proportion were malnourished



Protein-enriched

meals

Normalpostoperative

care

Relative(95% CI)

emoral neck fracture patients



14/74(18.9%)

RR 0.45(0.19 to1.04)

18.9%

emoral neck fracture patients


Seriousimprecisionc

none 27.4(14.9)days

N=83

39.8 (41.9)days

N=74

Higher drop-out rate than the event rate.

d The intervention group had a nutritional journal for the first four days established the patients’ nutrition deficiencies. Protein-enriched meals were calculated at approximatelyy weight to supply the extra energy requirement for the first four postoperative days or longer if required. At lunch an app

enriched meals and a dessert at dinner. If the patients were malnourished on admission the nurs es found out when or why they lost their appetite to see if the patientsneeded even more energy/caloiries. If had problems in these areas they consulted a dietitian. The patients in the interventi on group also received two nutritional land protein

2x200ml daily while hospitalised. Additional nutritional and protein drinks were served after every meal for patients who ne eded extra calories. The environment wasalso optimised to facilitate the intake of nutrition eg no unnecessary noise. The contro l group had conventional postoperative care routines.

KCE Report 193S


Relative(95% CI)

Absolute

RR 0.45(0.19 to1.04)


153 fewer to 8more)

VERYLOW

Critical


153 fewer to 8more)

- MD 12.4lower (22.47

to 2.33 lower)

LOW

Important

enriched meals were calculated at approximatelyy weight to supply the extra energy requirement for the first four postoperative days or longer if required. At lunch an app etiser was served with the

es found out when or why they lost their appetite to see if the patientson group also received two nutritional land protein

eded extra calories. The environment wasl group had conventional postoperative care routines.

KCE Report 193S

Table 15 – Oral supplements plus standard hospital diet versus standard hospital diet

Quality assessment


Inconsistency


5 (Bourdel-Marchasson 2000;Delmi 1990; Dennis2005; Hartgrink1998; Houwing2003)

randomisedtrials

veryseriousa


a Unclear details of sequence generation and allocation concealment. Majority of studies had a lack of blinding. Some trialsb Confidence interval crossed one MID point.

The results were pooled for all studies that included an oral supplement compared to normal hospital diet, as the main constituentsprotein and energy.


Oral supplements plus standard hospital diet versus standard hospital diet – mixed population




Standardhospital

diet




269/2516(10.7%)

48%

a Unclear details of sequence generation and allocation concealment. Majority of studies had a lack of blinding. Some trials had high level of missing data in both groups.

sults were pooled for all studies that included an oral supplement compared to normal hospital diet, as the main constituents

473


Relative(95% CI)

Absolute

RR 0.82(0.71 to

0.95)


VERYLOW

Critical


had high level of missing data in both groups.

sults were pooled for all studies that included an oral supplement compared to normal hospital diet, as the main constituents of the supplement were

474

Table 16 – Nutritional supplementation (supplements/diet containing protein and ene– mixed population

Quality assessment




6 (Bourdel-Marchasson2000; Delmi1990; Dennis2005;Hartgrink1998;Houwing2003; Oloffson2007)

randomisedtrials

veryseriousa


nindirectness

a Unclear details of sequence generation and allocation concealment. Majority of studies had a lack of blinding. Some trials had high level of missing dab Confidence interval crossed one MID point.

The results were pooled for all studies that included nutritionalsupplement were protein and energy. This included a study of nutritional supplements which were given by tube feeding.


Nutritional supplementation (supplements/diet containing protein and energy) plus standard hospital diet versus standard hospital diet



Nutritionalsupplements/diet

Standardhospital

diet

Relative(95%CI)



283/2590(10.9%)

RR 0.8(0.69 to0.92)

33.5%

e generation and allocation concealment. Majority of studies had a lack of blinding. Some trials had high level of missing da

The results were pooled for all studies that included nutritional supplementation compared to a normal hospital diet, as the main constituents of thesupplement were protein and energy. This included a study of nutritional supplements which were given by tube feeding.

KCE Report 193S

rgy) plus standard hospital diet versus standard hospital diet


Relative(95%

Absolute

RR 0.8(0.69 to0.92)


fewer)

VERYLOW

Critical


104 fewer)

e generation and allocation concealment. Majority of studies had a lack of blinding. Some trials had high level of missing da ta in both groups.

supplementation compared to a normal hospital diet, as the main constituents of thesupplement were protein and energy. This included a study of nutritional supplements which were given by tube feeding.

KCE Report 193S

9.3.3. Appendix II: Forest plots

Figure 2 – Incidence of pressure ulcers – Bourdel MarchassonProtein, fat, carbohydrate, minerals and vitamins supplement and standard dietversus standard diet

Figure 3 – Incidence of all pressure ulcers –versus standard diet

Figure 4 – Incidence of stage II pressure ulcersversus standard diet

Study or Subgroup

Bourdel-M 2000

Total (95% CI)

Total events



Events

118

118

Total

295

295

Events

181

181

Total

377

377

Weight

100.0%

100.0%

Supplement Standard hospital

Study or Subgroup

Houwing, 2003

Total (95% CI)

Total events



Events

27

27

Total

51

51

Events

30

30

Total

52

52

Weight

100.0%

100.0%


Study or Subgroup

Houwing 2003

Total (95% CI)

Total events



Events

9

9

Total

51

51

Events

14

14

Total

52

52

Weight

100.0%

100.0%



Bourdel MarchassonProtein, fat, carbohydrate, minerals and vitamins supplement and standard diet

– Houwing High protein enriched with arginine zinc and antioxidants

Incidence of stage II pressure ulcers – HouwingHigh protein enriched with arginine zinc and antioxidants supplement and standard diet

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.83 [0.70, 0.99]

0.83 [0.70, 0.99]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours supplement Favours standard hospital

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.92 [0.65, 1.30]

0.92 [0.65, 1.30]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.66 [0.31, 1.38]

0.66 [0.31, 1.38]


M-H, Fixed, 95% CI


475

Bourdel MarchassonProtein, fat, carbohydrate, minerals and vitamins supplement and standard diet

Houwing High protein enriched with arginine zinc and antioxidants supplement and standard diet

HouwingHigh protein enriched with arginine zinc and antioxidants supplement and standard diet

476

Figure 5 – Incidence of pressure ulcers – Delmi Protein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C,nicotinamide, folate, calcium pantothenate, biotin, and minerals supplement and standard diet versus standard diet

Figure 6 – Incidence of pressure ulcers – Dennis Standard hospital diet plus nutritional supplements (360mL at 6.27kJ/mL and 62.5g/L in protein)vs standard hospital diet

Figure 7 – Length of time in hospital – Dennis Standard hospital diet plus nutritional supplements (360mL at 6.2standard hospital diet

Figure 8 – Incidence of grade 2-4 pressure ulcers

Study or Subgroup

Delmi, 1990

Total (95% CI)

Total events



Events

0

0

Total

25

25

Events

2

2

Total

27

27

Weight

100.0%

100.0%

Supplement Standard hospital diet

Study or Subgroup

Dennis, 2005

Total (95% CI)

Total events



Events

15

15

Total

2016

2016

Events

26

26

Total

2007

2007

Weight

100.0%

100.0%

Supplement Normal hospital

Study or Subgroup

Dennis, 2005

Total (95% CI)



Mean

34

SD

48

Total

2016

2016

Mean

32

SD

46

Total

2007

2007

Weight

100.0%

100.0%

Supplement Normal hospital

Study or Subgroup

Hartgrink, 1998

Total (95% CI)

Total events



Events

25

25

Total

48

48

Events

30

30

Total

53

53

Weight

100.0%

100.0%



Delmi Protein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C,nicotinamide, folate, calcium pantothenate, biotin, and minerals supplement and standard diet versus standard diet

Dennis Standard hospital diet plus nutritional supplements (360mL at 6.27kJ/mL and 62.5g/L in protein)

Dennis Standard hospital diet plus nutritional supplements (360mL at 6.2

4 pressure ulcers – Hartgrink Tube fed energy, protein versus standard diet

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.22 [0.01, 4.28]

0.22 [0.01, 4.28]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.57 [0.31, 1.08]

0.57 [0.31, 1.08]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours supplement Favours normal hospital

Weight

100.0%

100.0%

IV, Fixed, 95% CI

2.00 [-0.91, 4.91]

2.00 [-0.91, 4.91]


IV, Fixed, 95% CI

-100 -50 0 50 100Favours supplement Favours normal hospital

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.92 [0.64, 1.32]

0.92 [0.64, 1.32]


M-H, Fixed, 95% CI


KCE Report 193S

Delmi Protein, carbohydrate, lipid, calcium, vitamin A, vitamin D, vitamins E, B1, B2, B6, B12, C,nicotinamide, folate, calcium pantothenate, biotin, and minerals supplement and standard diet versus standard diet

Dennis Standard hospital diet plus nutritional supplements (360mL at 6.27kJ/mL and 62.5g/L in protein)

Dennis Standard hospital diet plus nutritional supplements (360mL at 6.27kJ/mL and 62.5g/L in protein) vs

KCE Report 193S

Figure 9 – Incidence of all pressure ulcers – Hartgrink Tube fed energy, protein vers

Figure 10 – Incidence of pressure ulcers – Craig Diseaseformula

Figure 11 – Incidence of all pressure ulcers – Theilla Macronutrient diet plus lipids,ready to feed, high fat, low carbohydrate, enteral formula

Figure 12 – Incidence of grade 2-4 pressure ulcersmacronutrient diet ready to feed, high fat, low carbohydrate, enteral formula

Study or Subgroup

Hartgrink, 1998

Total (95% CI)

Total events



Events

30

30

Total

48

48

Events

37

37

Total

53

53

Weight

100.0%

100.0%


Study or Subgroup

Craig, 1998

Total (95% CI)

Total events



Events

7

7

Total

17

17

Events

8

8

Total

15

15

Weight

100.0%

100.0%

Disease-specific Standard

Study or Subgroup

Theilla, 2007

Total (95% CI)

Total events



Events

8

8

Total

46

46

Events

10

10

Total

49

49

Weight

100.0%

100.0%

Lipids and macronutrients Macronutrients

Study or Subgroup

Theilla, 2007

Total (95% CI)

Total events



Events

4

4

Total

46

46

Events

6

6

Total

49

49

Weight

100.0%

100.0%

Lipids and macronutrients Macronutrients


Hartgrink Tube fed energy, protein versus standard diet

Craig Disease-specific (reduced-carbohydrate, modified-fat formula vs standard high

Theilla Macronutrient diet plus lipids, gamma-linolenic acid, vitamins A,C and E vs macronutrient dietready to feed, high fat, low carbohydrate, enteral formula

4 pressure ulcers – Theilla Macronutrient diet plus lipids, gamma-linolenic acid, vitamins A,C andmacronutrient diet ready to feed, high fat, low carbohydrate, enteral formula

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.90 [0.68, 1.19]

0.90 [0.68, 1.19]


M-H, Fixed, 95% CI


Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.77 [0.37, 1.62]

0.77 [0.37, 1.62]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours disease-specific Favours standard

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.85 [0.37, 1.97]

0.85 [0.37, 1.97]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours Lipids + macro Favours Macronutrients

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.71 [0.21, 2.36]

0.71 [0.21, 2.36]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours lipids & macro Favours macronutrients

477

fat formula vs standard high-carbohydrate

linolenic acid, vitamins A,C and E vs macronutrient diet

linolenic acid, vitamins A,C and E vs

478

Figure 13 – Incidence of pressure ulcers – Oloffson 2007

Figure 14 – Time in hospital – Oloffson Protein

Figure 15 – Incidence of pressure ulcers –supplement vs standard hospital diet

Study or Subgroup

Oloffson, 2007

Total (95% CI)

Total events



Events

7

7

Total

83

83

Events

14

14

Protein-enriched meals Normal postoperative care

Study or Subgroup

Oloffson, 2007

Total (95% CI)



Mean

27.4

SD

14.9

Total

83

83

Mean

39.8

SD

41.9

Protein-enriched meals Normal postoperative care

Study or Subgroup

Bourdel-M 2000

Delmi, 1990

Dennis, 2005

Hartgrink, 1998

Houwing, 2003

Total (95% CI)

Total events



Events

118

0

15

25

27

185

Total

295

25

2016

48

51

2435

Events

181

2

26

30

30

269

Total

377

27

2007

53

52

2516

Weight

64.7%

1.0%

10.6%

11.6%

12.1%

100.0%

Oral supplement Standard hospital


Oloffson 2007 Protein-enriched meals vs normal postoperative care

Protein-enriched meals vs normal postoperative care

Bourdel-Marchasson, Delmi, Dennis, Hartgrink, Houwing Standard hospital diet plus nutritional

Total

74

74

Weight

100.0%

100.0%

M-H, Fixed, 95% CI

0.45 [0.19, 1.04]

0.45 [0.19, 1.04]

Normal postoperative care Risk Ratio Risk Ratio

M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours protein-enriched Favours normal

Total

74

74

Weight

100.0%

100.0%

IV, Fixed, 95% CI

-12.40 [-22.47, -2.33]

-12.40 [-22.47, -2.33]

Normal postoperative care Mean Difference Mean Difference

IV, Fixed, 95% CI

-100 -50 0 50 100Favours protein-enriched Favours normal

Weight

64.7%

1.0%

10.6%

11.6%

12.1%

100.0%

M-H, Fixed, 95% CI

0.83 [0.70, 0.99]

0.22 [0.01, 4.28]

0.57 [0.31, 1.08]

0.92 [0.64, 1.32]

0.92 [0.65, 1.30]

0.82 [0.71, 0.95]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours oral supplement Favours standard hospital

KCE Report 193S

enriched meals vs normal postoperative care

Marchasson, Delmi, Dennis, Hartgrink, Houwing Standard hospital diet plus nutritional

KCE Report 193S

Figure 16 – Incidence of pressure ulcers –nutritional supplement vs standard hospital diet


Table 17 – LANGER2003


Author and year:Langer 2003

Title: Nutritionalinterventions forpreventing andtreating pressure

ulcers (Review)

Journal: CochraneDatabase ofSystematic Reviews2003, Issue 4.

N of studies: 4

Inclusion criteria:

Population: People ofany age and sex with orwithout existingpressure ulcers, in anycare setting, irrespectiveof primary diagnosis. Apressure

ulcer was defined as anarea of localiseddamage to the skin and

underlying tissuecaused by pressure,shear, friction and/or a

Study or Subgroup

Bourdel-M 2000

Delmi, 1990

Dennis, 2005

Hartgrink, 1998

Houwing, 2003

Oloffson, 2007

Total (95% CI)

Total events



Events

118

0

15

25

27

7

192

Total

295

25

2016

48

51

83

2518

Events

181

2

26

30

30

14

283

Total

377

27

2007

53

52

74

2590

Weight

61.0%

0.9%

10.0%

11.0%

11.4%

5.7%

100.0%

Supplement/diet standard hospital


– Bourdel-Marchasson, Delmi, Dennis, Hartgrink, Houwing, Oloffson Standard hospital diet plusnutritional supplement vs standard hospital diet


Comparison

Outcomemeasures

Quality assessment

any age and sex with or

care setting, irrespective

ulcer was defined as an

Clearly described nutritionalsupplementation (enteral orparenteral nutrition) orspecial diet. Comparisonsbetween supplementarynutrition plus standard dietversus standard diet aloneand between different typesof supplementary nutrition(e.g. enteral vs. parenteral)were eligible.

Primary outcome:

Incidence ofpressure ulcers


Does the review collect thetype of studies you considerrelevant to threview question? yes

Was the literature searchsufficiently rigorous toidentify all relevant studies?yes

Was study quality assessedreported? Yes but the studyquality was in a narrative andno traffic lights or tables ofquality were reported.

Weight

61.0%

0.9%

10.0%

11.0%

11.4%

5.7%

100.0%

M-H, Fixed, 95% CI

0.83 [0.70, 0.99]

0.22 [0.01, 4.28]

0.57 [0.31, 1.08]

0.92 [0.64, 1.32]

0.92 [0.65, 1.30]

0.45 [0.19, 1.04]

0.80 [0.69, 0.92]


M-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours supplement/diet Favours standard hospital

479

Marchasson, Delmi, Dennis, Hartgrink, Houwing, Oloffson Standard hospital diet plus




Was the literature searchsufficiently rigorous toidentify all relevant studies?

Was study quality assessedYes but the study

quality was in a narrative andno traffic lights or tables ofquality were reported.

Quality grade:very low risk ofbias

480


combination

of these for the purposeof this review.

Studies: Randomisedcontrolled trials (RCTs)of parallel or crossoverdesign evaluating theeffect of enteral and/orparenteral nutrition on

the prevention andtreatment of pressureulcers by measuring the

incidence of new ulcers,ulcer healing rates orchanges in pressure

ulcer severity.Controlled clinical trials(CCT) were onlyconsidered

eligible for inclusion inthe absence of RCTs.

Exclusion criteria: seeabove for inclusioncriteria



Comparison

Outcomemeasures

Quality assessment

of these for the purpose

measuring the

incidence of new ulcers,

see


KCE Report 193S


dequate descriptionof the methodology used andincluded, and the methodsused are appropriate to thequestion? yes

KCE Report 193S

Table 18 – CRAIG1998


Author and year: Craig1998

138

Title: Use of a reduced-carbohydrate, modified-fat enteral formula forimproving metaboliccontrol and clinicaloutcomes in long-termcare residents with type2 diabetes: results of apilot trial

Journal: Nutrition, 1998,14 (6), 529-534.

Study type: RCT double-blinded pilot trial

Sequence generation:says randomised but nodetails of sequencegeneration

Allocation concealment:no details of allocationconcealment.

Blinding: double-blindedbut no details of whowas blinded.

Addressing incompleteoutcome data: adequate

Type of analysis:Available Case Analysis


Patient group: LTCresidents with type 2diabetes

All patientsrandomised N= 34

Completed: 27

Drop-outs: 7

Group 1:

Randomised N: 18

Completed: 16 at 4weeks, 14 at 12 weeks

Dropouts: 3 died

Age mean (sd): 82 (3),range 52-94 years

Males: not reported

Group 2:

Randomised N: 16

Completed:14 at 4weeks and 13 at 12weeks

Dropouts: 2 died, 1removed due touncontrolled bloodglucose levels.



Comparison

Outcomemeasures

Effect sizes

residents with type 2

weeks, 14 at 12 weeks

82 (3),

2 died, 1

Group 1: disease-specific(reduced-carbohydrate,modified-fat) formula

(Energy 1000 kcal, 41.8 gprotein, 16.7% kcal –source sodium andcalcium caseinates, 93.7gcarbohydrate, 33.3% kcal– source maltodextrin, soypolysaccharide; fructose;fat 55.7 g, 50%kcal –source high-oleic saffloweroil, soy oil).

Group 2: standard high-carbohydrate formula

(Energy 1060kcal, 44.4gprotein, 16.7% kcal –source sodium andcalcium caseinates;carbohydrate 151.7g(includes soy fiber thatprovides 39 kcal and 14gof total dietary fiber per L)carbohydrate, 53.3% kcal– source maltodextrin, soypolysaccharide; fat 35.9g,30.0% kcal – source high-oleic safflower oil, canolaoil, MCT oil.

Outcome 1:Incidence of PU:

Group 1: 7/17 (41.2%)

Group 2: 8/15 (53.3%

Relative risk:

95% CI: 0.37 to 1.62

481

Comments

7/17 (41.2%)

8/15 (53.3%)

Relative risk: 0.77

0.37 to 1.62

Funding:supported byRoss ProductsDivision, Ohio

Limitations:study aim was notto look atpressure ulcers, itwas only an eventexperiencedduring the study.No details ofsequencegeneration orallocationconcealment.Small samplesize.

Additionaloutcomes:

482


ANOVA for continuousdata; secondaryoutcomes Pearson chi-square test, Cochran-Mantel-Haenszel meanrank scores statistic fortreatment groupdifferences.

Baseline differences: nosignificant differences.

Study power/sample size:no power calculationvery small sample size

Setting: 2 long-term carefacilities in USA.

Length of study: 3months

Categorisation of Pus:not reported

Assessment of PUs:clinical outcomescollected daily but nodetails of how.


Age mean (sd): 80 (2),range 52-100.

Males: not reported

Inclusion criteria: atleast 50 years of age;history of type 2diabetes mellitus or haddocumentedhyperglycemia asevidenced by either aplasma glucose randommeasurement of>200mg/dL or a fastingplasma glucose>140mg/dL on towoccasions; required totalenteral nutrition supportby tube; were able totolerate a volume offormula that maintainedbody weight; informedconsent provided.

Exclusion criteria: seeabove.



Comparison

Outcomemeasures

Effect sizes

80 (2),

atleast 50 years of age;

diabetes mellitus or had

evidenced by either aom

>200mg/dL or a fasting

occasions; required totalenteral nutrition supportby tube; were able to

formula that maintainedbody weight; informed

see

KCE Report 193S

Comments

KCE Report 193S

Table 19 – THEILLA2007


Author and year:Theilla 2007

Title: A diet enrichedin eicosapentanoicacid, gamma-linolenic acid andantioxidants in theprevention of newpressure ulcerformation in criticallyill patients with acutelung injury: arandomised,prospective,controlled study

Journal: ClinicalNutrition, 26, 752-757.

Study type: RCT

Sequence generation:no details


Blinding: Not blinded.

Addressing incompleteoutcome data: nofurther drop-outsexcept those whowere excluded as didnot meet inclusion

Patient group: criticallyill, mechanicallyventilated patientssuffering from acutelung injury (secondaryoutcome from a largerstudy on acute lunginjury)

All patients

Randomised N=100

Completed N: 95

Drop-outs: 5 excludeddue to diarrhoea or foodintolerance (gastricresidue larger than250mL.

Group 1

Randomised N:

Completed N: 46

Dropouts:

Age (mean +/-SD): 57.0(18.7)

Gender (Male): 29(63.0)

Diagnostic categoryfor ICU admission:

Medical: 28 (60.9%)



Comparison

Outcomemeasures

Effect sizes

critically

ea or food

SD): 57.0

Group 1: same macronutrientdiet as control group plus alipids (elcosapentanoic acid(EPPA), gamma-linolenicacid (GLA)), vitamins A,Cand E

Group 2: macronutrient diet:ready to feed, high fat, lowcarbohydrate, enteralformula.

Outcome 1:incidence of allpressure ulcers

Group 1:8/46 (17.4%)

Group 2: 10/49 (20.4%)

Relative risk:

95% CI: 0.37 to 1.97

Outcome 2:incidence ofgrade 2-4pressure ulcers

Group 1: 4/49 (8.2%)

Group 2: 6/49 (12.2%)

Relative risk:

95% CI: 0.21 to 2.36

483

Comments

8/46 (17.4%)

10/49 (20.4%)

Relative risk: 0.85

0.37 to 1.97

Funding: nodetails of funding

Limitations: nodetails ofsequencegeneration,allocationconcealment. Noblinding. BMI washigher in theintervention groupat baseline.

Additionaloutcomes:pressure ulcers atday 7 (all ulcersincluding those atstart of study)

4/49 (8.2%)

6/49 (12.2%)

Relative risk: 0.71

0.21 to 2.36

484


criteria as haddiarrhoea or foodintolerance

Analysis: not reported

Statistical analysis:ANOVA withrepeated measure fordifference betweendependent variables.Chi-square test forassociationsbetween no-dependent variables

Baseline differences:BMI was significantlyhigher in the studygroup

Study power/samplesize: no a priorisample sizecalculation given andsmall sample size.

Setting: ICU, Israel.

Study length: 7 days

Categorisation of PUs:NPUAP

Assessment of PUs:NPUAP grading,assessed daily byresearchers.

Surgical: 18 (39.1%)

Trauma: 0

No. with pressureulcers: 7/46

Grade 1: n=5

Grade 2: n=1

Grade 3: n=1

BMI (SD): 28.9(6.2)kg/m2

Group 2

Randomised N:

ITT N:49

Dropouts:

Age (mean+/-SD):62.3(17.2)

Gender (Male): 28(57.1%)

Diagnostic categoryfor ICU admission:

Medical: 34 (69.4%)

Surgical: 15 (30.6%)

Trauma: 0

No. with pressureulcers: 14/49 (p=NS)

Grade 1: n=6

Grade 2: n=7

Grade 3: n=1



Comparison

Outcomemeasures

Effect sizes

SD):62.3

KCE Report 193S

Comments

KCE Report 193S


BMI (SD): 26.5(5.4)kg/m2, p=0.05

Inclusion criteria:patients with acute lunginjury defined by aPaO2/FIO2 ratio below250.

Exclusion criteria:patients with headtrauma, cerebralbleeding, coagulationdisorders, receivingsteroids in a dose>0.25mg/kg/daymethylprednisolone ornon-steroidal anti-inflammatory agents,patients less than 18years and pregnantpatients. If diarrhoeaoccurred more thanthree times.



Comparison

Outcomemeasures

Effect sizes

patients with acute lung

485

Comments

486

Table 20 – OLOFSSON2007


Author and year:Olofsson 2007

Title: Malnutrition in hipfracture patients: anintervention study

Journal: Journal ofClinical Nursing, 16(11),2027-2038.

Sequence generation:randomised topostoperative care in ageriatric ward with aspecial interventionprogramme or toconventional care in theorthopaedic department

Allocation concealment:sealed, opaqueenvelopes stratifiedaccording to operationmethod. Nurse on dutyat the orthopaedic dept,not involved in thestudy, opened theenvelope.

Blinding: the staff on theintervention ward wasaware of the nature ofthe study, and the staffworking on the controlward was informed that

Patient group:femoral neck fracturepatients

All patients

Randomised N: 199

Completed N: 157

Drop-outs: 42

Group 1

Randomised N: 102

Completed N: 83

Dropouts: 19(18.6%)

Six patients diedduring hospitalisationand five patients hadmissing MNA

(a)(91

were assessed at 4months), 3 patientsdied after discharge,one patient declinedto continue and fourpatients had missingMNA

(a).

Group 2

Randomised N: 97


Intervention

Comparison

Outcomemeasures

Effect sizes

femoral neck fracture

during hospitalisationand five patients had

died after discharge,

had missing

Group 1: protein enrichedmeals (calculated atapproximately 30 calories perkilo body weight) servedduring the first fourpostoperative days andlonger if necessary. At lunchan appetizer was alwaysserved with the protein-enriched meals and adessert at dinner. When theregistered nurses suspectedmalnourishment onadmission they found outwhen or why they had losttheir appetite to discoverwhether the patients neededeven more energy/calories.If there were problems inthese areas, a dietician wasconsulted.

They also received twonutritional and protein drinks(2x200ml) daily during wholehospitalisation period.Additional nutritional andprotein drinks were servedafter every meal for patientswho needed extra calories. Ifpatients could not sleep orwere anxious at night anextra meal was offered


Group 1: 7/83

Group 2: 14/74

P=0.054

Those who did developpressure ulcers were almostexclusively suffering fromsevere malnutrition.

Outcome 2: timein hospital

Group 1: 27.4 (14.9)

Group 2: 39.8 (41.9)

P=0.019

KCE Report 193S

Comments

7/83

14/74

Those who did developpressure ulcers were almostexclusively suffering fromsevere malnutrition.

Funding: grantsfrom theBorgerskapet inUmea ResearchFoundation, theDementia Fund,the VardalFoundation, theJoint Committeeof the NorthernHealth Region ofSweden, the JCKempe MemorialFoundation, theFoundation of theMedical Faculty,University ofUmea, the CountyCouncils ofVasterbotten andthe SwedishResearch Councilgrant.

Limitations:randomised todifferent wards.No blinding.Small study nopower calculation.

Additional

27.4 (14.9)

39.8 (41.9)

KCE Report 193S


a new care programmewas being implementedand that it was beingevaluated in thegeriatric interventionward.

Addressing incompleteoutcome data: explainswhat happened to allmissing data.

Statistical analysis:Student’s t-test wasused to analysedifferences in MNA

(a)

scores on admissionand at the four-monthfollow-up betweengroups.

Analysis: Available CaseAnalysis

Statistical analysis:Student’s t-test toanalyse differences inMNA

(a)scores

Baseline differences:there was asignificantly higherscore for theintervention group forheart failure at baseline.There were fourpatients missing data inthe control group and

Completed N: 74

Dropouts: 23 (23%)

Seven patients diedduring hospitalisation,8 patients hadmissing MNA

(a)(82

were assessed at 4months). Six patientsdied after discharge,1 patient moved toanother city and onepatient had missingMNA

(a).

Inclusion criteria:femoral neck fracture,aged 70 years orolder, admittedconsecutively to theorthopaedic dept ofone hospital, fromMay 2000 toDecember 2002.

Exclusion criteria:severe rheumatoidarthritis, severed hiposteoarthritis, severerenal failure,metastatic fractureand patients whowere bedriddenbefore their injury.


Intervention

Comparison

Outcomemeasures

Effect sizes

during hospitalisation,

months). Six patientsdied after discharge,

and one

femoral neck fracture,

osteoarthritis, severe

during the night shift. Theenvironment around the mealwas adjusted to facilitategood nutrition, by making themeal times nice andcomfortable with nounnecessary noise, bustle orstress. Any aspect thatmight improve the patients’nutrition was considered egthey could choose their ownfood or ask what they wantedto eat. All physical problemsthat led to patients eatingless were dealt with egconstipation, pain or bad oralhygiene.

Group 2: postoperative carein the orthopaedicdepartment in accordancewith conventionalpostoperative care routines(described in table). Staffingratio 1.01 nurses or aids perbed. Patients who needed alonger rehabilitation periodwere transferred to a generalgeriatric rehab ward but notto the ward where theintervention programme hadbeen implemented (n=30).Staffing ratio was 1.07nurses or aids per bed.

487

Comments

outcomes:compliance – thenutritional andprotein drinkswere servedduring the wholehospitalisationperiod in theintervention groupbut we do notknow exactly howmuch wereconsumed.Should be notedwhen interpretingthe results.Complicationsduringhospitalisationwere given inrelation to theMNA

(a)scores at

baseline in eachgroup (delirium,nutritiondifficulties,constipation,pressure ulcers,urinary tractinfection.

Study was part ofa multifactorial

488


one in the interventiongroup at this time.

Study power/sample size:small, no powercalculation.

Setting: orthopaedicdepartment, UmeaUniversity HospitalSweden.

Length of study: fourmonth follow-up

Categorisation of PUs:not reported

Assessment of PUs: notspecifically mentionedas not main aim ofstudy.

Other assessments: themini mental stateexamination, organicbrain syndrome scaleand the geriatricdepression scale wereused. The MNA

(a)was

used to assess thepatients’ nutritionalstatus.

Multiple ulcers: notreported.

MNA – mini nutritional assessment scale


Intervention

Comparison

Outcomemeasures

Effect sizes

All patients: received samepreoperative treatment in theorthopaedic department andhad same mean waiting forsurgery (25.1 hours in thecontrol group and 24.6 hoursin the intervention group,p=0.852).

KCE Report 193S

Comments

multidisciplinaryintervention study.

KCE Report 193S

Table 21 – DENNIS2005


Author and year:Dennis 2005

Title: Routine oralnutritionalsupplementation forstroke patients inhospital (FOOD): amulticentrerandomisedcontrolled trial

Journal: Lancet,2005, 365, 755-763.

study type:

Multicentre RCT

Sequence generation:computer-generated

Allocationconcealment:international co-ordinating centre andcomputer-generatedminimisationalgorithm balancedtreatment withineach country

Blinding: no blindingof assessment andtreatment allocation.

Addressing incompleteoutcome data:

Patient group: elderlystroke patients in hospital

All patients randomisedN= 4023

Completed:

Drop-outs:

Group 1:

Randomised N: 2016

Completed: 1767

Dropouts: 4 lost tofollow-up, 3 vital statusonly, 241 died

Age mean (sd): 71 (12)

Males: 1071 (53%)

Nutritional status:

Undernourished: 156(8%)

Normal: 1550 (77%)

Overweight: 310 (15%)

Glasgow coma scaleverbal normal: 1644(82%)

Group 2:

Randomised N: 2007


ics Intervention

Comparison

Outcomemeasures

Effect sizes

in hospital

All patients randomised

71 (12)

310 (15%)

Group 1: normal hospitaldiet plus oral supplements(360mL at 6.27 kJ/mL and62.5g/L in protein everyday)

Most centres usedcommercially availablesupplements of suitableconsistency for patients withmild swallowingimpairments eg liquid,yoghurt, pudding.

The supplements wereprescribed on drug-administration charts toincrease compliance and toallow monitoring ofcompliance by the hospitalcoordinator so that therewas an increase in the totalprotein and energy intake ofelderly patients in hospital.

Group 2: normal hospitaldiet

Outcome 1:Incidence of PU:

Group 1: 15/2016 (0.7%)

Group 2: 26/2007 (1.3%)

Relative risk:

95% CI: 0.31 to 1.08

Outcome 2:length of stay inhospital – meandays (s.d)

Group 1: 34.0

Group 2: 32.00 (

489

Comments

15/2016 (0.7%)

26/2007 (1.3%)

Relative risk: 0.57

: 0.31 to 1.08

Funding: grantsfrom the HTAboard of NHSresearch anddevelopment inthe UK, theStrokeAssociation, theChief ScientistOffice of theScottishExecutive, andChest, Heart andStroke Scotland.The RoyalAustralasianCollege ofPhysicianssupported the trialin Hawkes Bay,New Zealand.

Limitations: aimnot to look atpressure ulcersand there were nodetails ofpressure ulcers atstart of the trial.Pressure ulcerswere classified asa complication.

34.0 (48.00)

32.00 (46.00)

490


adequate

Analysis: primaryanalyses ITT

Statistical analysis:Log-rank test

Baseline differences:no differences

Study power/samplesize: yes based ondichotomousoutcome – dead orpoor outcome(MRS

(a)3-5) at

follow-up. 87% power6000 participants.

Setting: multicentre,UK

Length of study: 6-months follow-up



How outcomesrecorded: postalquestionnaire orstructured telephoneinterview frompatient, carer orproxy.

Multiple ulcers: not

Completed: 1740

Dropouts: 7 lost tofollow-up, 5 vital statusonly, 253 died

Age mean (sd): 71 (13)

Males: 1078 (54%)

Nutritional status:

Undernourished: 158(8%)

Normal: 1542 (77%)

Overweight: 307 (15%)

Glasgow coma scaleverbal normal: 1606(80%)

Inclusion criteria:patients admitted with arecent stroke (first orrecurrent stroke no morethan 7 days beforeadmission) could beenrolled if they passedtheir swallow screen, theresponsible clinician wasuncertain whether to useoral nutritionalsupplements and thepatient (or a relative)consented to enrolment.Enrolled within 30 days ofadmission, or within 30


ics Intervention

Comparison

Outcomemeasures

Effect sizes

71 (13)

307 (15%)

patients admitted with a

recurrent stroke no more

their swallow screen, theresponsible clinician wasuncertain whether to use

consented to enrolment.Enrolled within 30 days of

KCE Report 193S

Comments

The authors statethat the dataneeds to beinterpreted withcaution becausethey could notmask theassessment totreatmentallocation and itwas not feasiblefor local sourcedata to be verifiedfor the occurrenceof these. Trialwas stoppedbefore theyreached theirtarget as nofunding wasavailable tocontinue beyond2004 and toensure the trialwas closed in anorderly manner.

Additionaloutcomes:primary outcomeswere death orpoor outcome andoverall survival.

KCE Report 193S


reported. days of a stroke occurringin hospital.

Exclusion criteria:subarachnoidhaemorrhage

MRS is the modified Rankin scale which is a scale for measuring the degree of disability or dcauses of neurological disability. Scoring: 0 No symptoms at all; 1 No significant disability despite symptoms; able to carrunable to carry out all previous activities, but able to look after own affairs without assistance; 3 Moderate disability; reModerately severe disability; unable to walk without assistancerequiring constant nursing care and attention; 6 Dead.

Table 22 – HOUWING2003


Author and year:Houwing et al 2003

Title:

A randomised,double-blindassessment of theeffect of nutritionalsupplementation onthe prevention ofpressure ulcers inhip-fracture patients,Clinical Nutrition,22(4),401-405

Study type:

Multicentre RCT

Patient group: hipfracture patients

All patientsrandomised N=103

Drop-outs: 0

Group 1:

Randomised N: 51

Dropouts: 0

Age (mean):81.5+/-0.9

Sex (female): 40/51

Risk score CBO:


ics Intervention

Comparison

Outcomemeasures

Effect sizes

days of a stroke occurring

MRS is the modified Rankin scale which is a scale for measuring the degree of disability or dependence in the daily activities of people who have suffered a stroke or othercauses of neurological disability. Scoring: 0 No symptoms at all; 1 No significant disability despite symptoms; able to carr y out all usual duties and activities; 2 Slight disunable to carry out all previous activities, but able to look after own affairs without assistance; 3 Moderate disability; re quiring some help, but able to walk without assistance; 4Moderately severe disability; unable to walk without assistance and unable to attend to own bodily needs without assistance; 5 Severe disability; bedridden, incontinent and


Comparison

Outcomemeasures

Effect sizes

Group 1: Standard diet withadditional supplement.Supplement was a high-protein nutritionalsupplement enriched witharginine, zinc andantioxidants (400ml).

Given immediatelypostoperatively for 4 weeksor until discharge

Group 2: Standard diet withplacebo: a non-caloric,water-based drink containingonly sweeteners, colorants

Outcome 1:incidence of allpressure ulcers

Group 1:27/51 (55.1%)

Group 2:30/52 (58.8%)

Relative risk:

95% CI:-0.16 to 0.23

P value: 0.420

Outcome 2:Incidence ofgrade 2 pressureulcers

Group 1: 9/51 (17.6%)

Group 2: 14/52 (26.9%)

Relative risk:

95% CI: 0.31 to 1.38

491

Comments

Aim of study wasnot to look atpressure ulcers.

ependence in the daily activities of people who have suffered a stroke or othery out all usual duties and activities; 2 Slight disability;

quiring some help, but able to walk without assistance; 4and unable to attend to own bodily needs without assistance; 5 Severe disability; bedridden, incontinent and

Comments

27/51 (55.1%)

30/52 (58.8%)

Relative risk:0.037

0.16 to 0.23

0.420

Funding:NumicoResearch BV,Wageningen, theNetherlands

Limitations:Unclear selectionbias – no detailsof sequencegeneration orallocationconcealment.

Additionaloutcomes: total

9/51 (17.6%)

14/52 (26.9%)

Relative risk: 0.66

0.31 to 1.38

492




Blinding: double-blinded. Look andtaste of bothsupplements werenot identical butsupplements weregiven in similar,blinded packages tomask the differences.

Addressing incompleteoutcome data: nodropouts

Analysis: ITT

Statistical analysis:Distribution ofvariables evaluatedvisually byKolmogorov-Smirnovtest. Differences incontinuous variablesdetermined byStudent’s t-test orMann-Whitney U-test.Difference inincidence rates byFisher’s exact test.Results adjusted forage or length of

11.1+/-0.3

Group 2:

Randomised N: 52

Dropouts: 0

Age (mean): 80.5+/-1.3

Sex (female): 44/52

Risk score CBO:11.2+/-0.2

Inclusion criteria: hipfracture, patient with apressure risk score over8 according to the CBO-risk assessment tool(four-point scoring toolincluding: mental status,neurology, mobility,nutritional status,nutritional intake,incontinence, age,temperature, medicationand diabetes).

Exclusion criteria:terminal care, metastatichip fracture, insulin-dependent diabetes,renal disease (creatinine>176mmol/l, hepaticdisease, morbid obesity



Comparison

Outcomemeasures

Effect sizes

1.3

pressure risk score over-

including: mental status,

temperature, medication

terminal care, metastatic

renal disease (creatinine

disease, morbid obesity

and flavourings (400ml)

KCE Report 193S

Comments

max wound size(cm

3), first day

pressure ulcer,number of dayswith pressureulcer.

Notes: 57%developed PUwithin first 2 daysof the study and76% by the fourthday

KCE Report 193S


surgery by ANOVA.

Baseline differences:no significantdifference in baselinevalues.

Study power/samplesize: underpowered

Setting: three centresin the Netherlands

Length of study: 28days or untildischarge

Categorisation of PUs:EPUAP classificationsystem

Assessment of PUs:PU assessed daily bynursing staff


(BMI>40), need fortherapeutic dietincompatible withsupplementation andpregnancy or lactation.



Comparison

Outcomemeasures

Effect sizes

on.

493

Comments

494

Table 23 – BOURDEL-MARCHASSON2000


Author and year:Bourdel-Marchasson(2000) Title: A multi-centre trial of theeffects of oralnutritionalsupplementation incritically ill olderinpatients

Study type: multi-centre cluster-randomised RCT

Sequence generation:19 wards stratified byspecialty and thewards randomisedinto 2 groups. Nodetails on seq. gen.

Allocationconcealment: nodetails butmulticentre stratified

Blinding: not blinded(authors state it isnot easy to proposeplacebo oralsupplements withsimilar taste andconsistency in adouble-blind manner.Also it could have adeleterious effect on

Patient group: Criticallyill older patients.

All patients

Randomised N= 672

Drop-outs: 173

Group 1

Randomised N: 295

Completed N: 107

Dropouts: 188

Age mean (s.d): 83.6(7.3)

Male (%): 96 (32.5)

Other baseline data:

Stroke: 23.6%

Falls and gaitdisturbance: 13.7%

Heart failure anddyspnea: 13.1%

Infectious diseases:13.7%

Digestive diseases:3.2%

Delirium: 5.6%

Dehydration: 2.9%

Lower limb fractures:



Comparison

Outcomemeasures

Effect sizes

Critically Group 1:

standard diet of 1800kcal/dayplus 2 oral supplements of200kcal each (30% protein,20% fat, 50% carbohydratein addition to minerals andvitamins such as zinc 1.8mgand vitamin C (15mg)

Group 2: standard diet of1800kcal/day

Outcome 1:pressure ulcer(cumulative)incidence at endof follow-up

Group 1: 118/295 (40%)

Group 2: 181/377 (48%)

Relative risk:

95% CI: 0.70 to 0.99

KCE Report 193S

Comments

118/295 (40%)

181/377 (48%)

Relative risk: 0.83

0.70 to 0.99

Funding: ProjetHospitalier deRechercheClinique,Ministère de laSanté et del’ActionHumanitaire,DirectionGénérale de laSanté et laDirection desHôpitaux.

Limitations: 25died inIntervention and22 in controlgroup. No detailsof sequencegeneration forclusterrandomisation.No blinding.There werebaselinedifferences butauthor didmultivariateanalysis toaccount for thesedifferences.

KCE Report 193S


the energy intake inthe control groupbecause in elderlyhospitalised patientsthe volume ratherthan the energycontent of food couldlimit voluntaryenergy intake).

Addressing incompleteoutcome data: forsubjects who died orwere dischargedwithout pressureulcers before the day15, the date of deathor discharge wereconsidered ascensoring the data.

Analysis: ITT

Statistical analysis:Chi-square test forcategorical variablesand Student’s t testfor numericalvariables afterapplying the Fishertest. Multiple hazardregression Coxmodel to adjustanalysis.Homogeneity testused and a

0.3%

Cancer: 1.1%

Neurologic diseases:2.4%

Painful arthritis: 2.1%

Deep VeinThrombosis: 2.9%

Miscellaneous medicaldiseases: 15.3%

Group 2

Randomised N: 377

Completed N: 244

Dropouts: 133

Age mean (s.d):83.0(7.1)

Male (%): 139 (36.9)

Other baseline data:

Stroke: 6.8% (P<0.001)

Falls and gaitdisturbance: 20.2%(p=0.02)

Heart failure anddyspnea: 7.2%(p=0.009)

Infectious diseases:11% (N.S)

Digestive disease:14.4% (p<0.001)



Comparison

Outcomemeasures

Effect sizes

Miscellaneous medical

Stroke: 6.8% (P<0.001)

495

Comments

There was a veryhigh drop-out63% inintervention groupand 35% incontrol group.

Additionaloutcomes:

496


multivariate Coxproportional hazardmodel.

Baseline differences:the nutritional groupincluded morepatients with stroke,heart failure, anddyspnea and fewerwith antecedent falls,delirium, lower limbfractures anddigestive disease.The nutritional grouphad a lower risk ofpressure ulcers,were less dependent(Kuntzman score)and a lower serumalbumin level(indicates a higherrisk for pressureulcers)

Study power/samplesize: a priori powercalculation notreported but largesample size.

Setting: inpatients ofhospital wards inBordeaux orinpatients at geriatricunits in Southwest

Delirium: 9.9%(p=0.001)

Dehydration: 2.7%(N.S)

Lower limb fractures:4.1% (p=0.004)

Cancer: 4.8% (N.S)

Neurologic diseases:2.4% (N.S)

Painful arthritis: 2.1%(N.S)

DVT: 0 (N.S)

Miscellaneous medicaldiseases: 14.4% (N.S)

Inclusion criteria: olderthan 65 years, in theacute phase of a criticalillness, unable to moveby themselves, andunable to eatindependently atadmission.

Exclusion criteria:pressure ulcers atadmission.



Comparison

Outcomemeasures

Effect sizes

Miscellaneous medical

older

acute phase of a critical

KCE Report 193S

Comments

KCE Report 193S


France belonging toGAGE, a group forthe evaluation andimprovement ofhealth care for theelderly.

Length of study: 15days follow-up

Categorisation of PUs:

Assessment of PUs:

Assessment: Nortonscale to assess riskof developingpressure ulcers;Kuntzman scaleassessed theactivities of dailyliving. Ulcers gradedby four gradesdefined by theAgency for HealthCare Policy andResearch.




Comparison

Outcomemeasures

Effect sizes

497

Comments

498

Table 24 – HARTGRINK1998


Author and year:Hartgrink 1998

Title: Pressure soresand tube feeding inpatients with afracture of the hip: arandomised clinicaltrial

Journal: ClinicalNutrition 1998, 17 (6),287-292.

Study type: singlecentre parallel RCT

Sequence generation:no details.

Allocationconcealment: nodetails.



Analysis: per protocol


Baseline differences:no differences

Study power/samplesize: no powercalculation given.

Length of study: 2

Patient group: hip fracturepatients

All patients

Randomised N=140

Evaluable at admission: 129(11 did not fulfil entrycriteria)

Drop-outs: 11 excluded atadmission (randomisationnot correctly performed).

Evaluable at 1 week: 116Evaluable at 2 weeks: 101

Group 1

Randomised N: 70

Evaluable at admission: 62

Evaluable at 1 week: 54

Evaluable at 2 weeks: 48

Dropouts:

Age (mean): 84.0 (7.1)

Sex M/F: 10/52

Time from entry to operation(min) mean (SD): 20.0 (16.3)

Operation time (min): 58.2(22.4)

Pressure-sore risk score(mean, SD): 9.0 (1.3)



Comparison

Outcomemeasures

Effect sizes

Patient group: hip fracture


11 excluded atadmission (randomisationnot correctly performed).

Evaluable at 1 week: 116Evaluable at 2 weeks: 101




Time from entry to operation(min) mean (SD): 20.0 (16.3)

Operation time (min): 58.2

sore risk score

All patients receivedstandard hospital diet. Incase they wererandomised to tubefeeding, a nasogastrictube was given duringsurgery or within 12 hoursafterwards. Actualfeeding started within 24hours.

Group 1: Standardhospital diet plus tubefeeding (1 litre NutrisonSteriflo Engergy-plus(1500kcal/l energy, 60gram/l protein, Nutricia,Netherlands)).Administered with afeeding pump through apolyurethane nasogastricfeeding tube. Tubefeeding was to be givenfor 2 weeks andadministered between21:00 and 05:00 tominimise interference withthe normal hospital diet.Nurses kept record offood offered and food leftover. Calculation ofenergy and protein intake

Outcome 1:pressure soreincidence(grade 2 ormore) [no.evaluable at 2weeks]

Group 1: 25/48 (44%)

Group 2: 30/53 (57%)

Relative risk:

95% CI: 0.64 to 1.32

Outcome 3:

Pressure soreincidence (allgrades) [no.available at 2weeks]

Group 1: 30/48 (62.5%)

Group 2: 37/53 (69.8%)

Relative risk:

95% CI: 0.68 to 1.19

Outcome 2:pressure soreincidence(grade 2 ormore) [no.available at 1week]

Group 1:20/54 (28%)

Group 2: 30/62 (48%)

Relative risk: 0.77

95% CI: 0.50 to 1.18

Outcome 4:pressure soreincidence (allgrades) [no.available at 1weeks]

Group 1: 35/54 (64.8%)

Group 2: 41/62 (66%)

Relative risk: 0.98

95% CI: 0.75 to 1.28

KCE Report 193S

Comments

25/48 (44%)

30/53 (57%)

Relative risk: 0.92

0.64 to 1.32

Funding: notstated.

Limitations: nodetails ofsequencegeneration,allocationconcealment andno blinding. Highdrop-out in bothgroups. Thosewho were stilltube fed at 1 and2 weeks were 25and 16 patientsrespectively.

Additionalmortality:evaluable at week1 and week 2.

Group 1: 30/48 (62.5%)

Group 2: 37/53 (69.8%)

Relative risk: 0.90

0.68 to 1.19

20/54 (28%)

30/62 (48%)

Relative risk: 0.77

95% CI: 0.50 to 1.18

35/54 (64.8%)

41/62 (66%)

Relative risk: 0.98

95% CI: 0.75 to 1.28

KCE Report 193S


weeks treatment.

Categorisation of PUs:(Stage 0=normalskin, 1=persistenterythema of the skin,stage 2=blisterformation, stage3=superficial(sub)cutaneousnecrosis, stage4=subcutaneousnecrosis, accordingto the Dutchconsensus meetingfor the prevention ofpressure sores)



Group 2

Randomised N: 70




Dropouts:

Age (mean): 83.3 (8.1)

Sex M/F: 6/6

Time from entry to operation(min) mean (SD):21.1 (12.3)

Operation time (min):

63.1 (23.4)

Pressure-sore risk score(mean, SD):9.2 (1.3)

Inclusion criteria: fracturedhip; pressure-sore risk scoreof 8 points or more (calculatedas sum of points scored on 10risk indices – mental status,neurology, mobility, nutritionalstatus, incontinence, age,temperature, medication anddiabetes).

Exclusion criteria: Patientswith pressure sores of grade 2or more at admission (Dutchconsensus).



Comparison

Outcomemeasures

Effect sizes



aluable at 2 weeks: 53

Time from entry to operation(min) mean (SD):21.1 (12.3)

sore risk score

fracturedrisk score

of 8 points or more (calculatedas sum of points scored on 10

mental status,neurology, mobility, nutritionalstatus, incontinence, age,temperature, medication and

Patientsgrade 2

or more at admission (Dutch

by diet and tube feedingdone daily by dietician.

Group 2: standardhospital diet.

499

Comments

500

Table 25 – DELMI1990


Author and year:Delmi 1990

Title: dietarysupplementation inelderly patients withfractured neck of thefemur

Journal: Lancet 1990,28, 335 (8696); 1013-1016.

Study type: RCT



Blinding: no details


Analysis: not reported

Statistical analysis:unpaired t tests or Utests, and X2 andFisher’s exact testsfor analysis ofclinical course.

Baseline differences:the 250HD plasmalevel was lower in

Patient group: elderlypatients with fractures ofthe proximal femur.

All patients

Randomised N=59

Completed N: 49

Drop-outs: 10 died(not included inanalysis)

Group 1

Randomised N: 27

Completed N: 21

Dropouts: 6 died (notincluded in analysis)

Age (mean SD andrange): 80.4 (8.5,61-93)

Female/Male: 24/3

Triceps skinfold (mm):Women 12.1 (4.6)

Men 5,7,10

Upper armcircumference (mm):

Women 251 (30)

Men* 255, 260, 260



Comparison

Outcomemeasures

Effect sizes

of

93)

Triceps skinfold (mm):

Group 1:

Daily oral nutritionsupplements, for mean 28days in addition to standardhospital diet.

Group 2: control group

250ml oral nutritionalsupplement provided254kcal, 20.4g protein, 29.5gcarbohydrate, 5.8g lipid,525mg calcium, 750 IUvitamin A, 25 IU vitamin D3,vitamins E, B1, B2, B6, B12,C, nicotinamide, folate,calcium pantothenate, biotin,and minerals.

Outcome 1:pressure ulcersat first hospital(orthopaedic)

Group 1:2/27 (7.4%)

Group 2:3/32 (9.38%)

Relative risk:

95% CI:

Outcome 2:pressure ulcersat 2

ndhospital

(recovery)

Group 1:0/9 (0%)

Group 2:3/15 (20%)

Relative risk:

95% CI:

Outcome 3:pressure ulcersat 6 months[figures used inCR]

Group 1: 0/25 (0%)

Group 2: 2/27 (7.4%)

Relative risk:

95% CI:

Outcome 4: totallength of stay inorthopaedicward andrecoveryhospital

Group 1: median 24 days(range 13-157)

Group 2: 40 (10

Relative risk:

P=0.09

KCE Report 193S

Comments

:2/27 (7.4%)

:3/32 (9.38%)

Relative risk:


Limitations:small sample. Nodetails ofsequencegeneration,allocationconcealment orblinding.Difference atbaseline forplasma level.

Notes: mostpatients hadnutritionaldeficiencies. Theauthors state thatelderly are oftenmalnourished andpatients withfractured proximalfemur seemespecially under-nourished.

Supplement waswell tolerated andcompletelyingested so no

0/9 (0%)

3/15 (20%)

Relative risk:

0/25 (0%)

2/27 (7.4%)

Relative risk:

median 24 days157)

40 (10-259)

Relative risk:

KCE Report 193S


non-supplementedpatients (median9.0nmol/l, range 2.3-61.5 vs 14.9, 4.2-87,p<0.05).

Study power/samplesize: no powercalculation.

Setting: orthopaedicunit of Universityhospital of Geneva

Length of study:

assessments madeon days 14, 21 and28 and at 6 months.




Group 2

Randomised N: 32

Completed N: 28

Dropouts: 4 died (notincluded in analysis)

Age (mean SD andrange): 82.9 (1.9, 66-96)

Female/Male: 29/3

Triceps skinfold (mm):Women 11.4 (5.7)

Men* 4,7, 13

Upper armcircumference (mm):

Women 261 (41)

Men* 230, 270, 290

*Data for 3 men ineach group

Inclusion criteria:patients over 60 yearsold admitted betweenMarch 1

stand May 15

th

1985 with a femoralneck fracture after anaccidental fall. Allpatients were well-oriented, able tounderstand the aim ofthe study, and willing to



Comparison

Outcomemeasures

Effect sizes

Dropouts: 4 died (not

Triceps skinfold (mm):

the study, and willing to

501

Comments

side-effectsobserved.

Outcomes alsoreported but notspecified here:severe anaemia,cardiac failure,infection and GIulcer. These weregiven for firsthospital(orthopaedic), 2

nd

hospital(recovery) and at6 months.

502


cooperate.

Exclusion criteria:Fractures from violentexternal trauma andpathological fracturesdue to tumours or non-osteoporoticosteopathies; dementia;renal, hepatic orendocrine disease,gastrectomy ormalabsorption, ortreatment withphenytoin, steroids,barbiturates, fluoride, orcalcitonin.



Comparison

Outcomemeasures

Effect sizes

osteopathies; dementia;

barbiturates, fluoride, or

KCE Report 193S

Comments

KCE Report 193S

10. GRADE SYSTEM

10.1. Down- or upgrading the evidence

Study designInitial level ofevidence

Randomised trials High

Observational studies Low

The quality evidence is downgraded based on the following elements:conventions agreed within the GDG and with our international partner for this study in case of inconsistency and imprecision are explained bel


Lower if

Risk of Bias

-1 Serious

-2 Very serious

Inconsistency

-1 Serious

-2 Very serious

Indirectness

-1 Serious

-2 Very serious

Imprecision

-1 Serious

-2 Very serious

Publication bias

-1 Serious

-2 Very serious

The quality evidence is downgraded based on the following elements: risk of bias, inconsistency, indirectness, imprecision, publication biasd within the GDG and with our international partner for this study in case of inconsistency and imprecision are explained bel

503

Higher if

Large effect

+1 Large

+2 Very large

Dose response

+1 Evidence of a gradient

All plausible residual confounding

+1 Would reduce a demonstratedeffect

+1 Would suggest a spurious effect ifno effect was observed

risk of bias, inconsistency, indirectness, imprecision, publication bias. The specificd within the GDG and with our international partner for this study in case of inconsistency and imprecision are explained bel ow.

504

10.1.1. Risk of bias

Limitations in the study design and implementation may bias the estimatesof the treatment effect. Major limitations in studies decrease the confidencein the estimate of the effect.

10.1.2. Inconsistency

Results were considered to be heterogeneous in casevary widely across studies or the confidence intervals show minimal or nooverlap or a Chi square p<0.1 or I- squared inconsistency statistic of>50%. When no plausible explanation can be found for this heterogeneity,the quality of evidence was downgraded by one or two levels, dependingon the extent of uncertainty to the results contributed by the iin the results. In addition to the I- square and Chi square values, thedecision for downgrading was also dependent on factors such as whetherthe intervention is associated with benefit in all other outcomes or whetherthe uncertainty about the magnitude of benefit (or harm) of the outcomeshowing heterogeneity would influence the overall judgment about netbenefit or harm (across all outcomes).

145

10.1.3. Indirectness

Indirectness refers to differences in studycomparator and outcomes between the available evidence and theprotocol.

10.1.4. Imprecision

Results are often imprecise when studies include relatively few patientsand few events and thus have wide confidence intervalsestimate of effect. This, in turn, may mean that we are uncertain if there isan important difference between interventions or not. If this is the case, theevidence may be considered to be of lower quality of the evidence lowerthan it otherwise would be because of resulting uncertainty in the results.


Limitations in the study design and implementation may bias the estimatesations in studies decrease the confidence

Results were considered to be heterogeneous in case the point estimatesvary widely across studies or the confidence intervals show minimal or no

squared inconsistency statistic of>50%. When no plausible explanation can be found for this heterogeneity,the quality of evidence was downgraded by one or two levels, dependingon the extent of uncertainty to the results contributed by the inconsistency

square and Chi square values, thedecision for downgrading was also dependent on factors such as whetherthe intervention is associated with benefit in all other outcomes or whether

he magnitude of benefit (or harm) of the outcomeshowing heterogeneity would influence the overall judgment about net

population, intervention,comparator and outcomes between the available evidence and the

Results are often imprecise when studies include relatively few patientsconfidence intervals around the

ate of effect. This, in turn, may mean that we are uncertain if there isan important difference between interventions or not. If this is the case, theevidence may be considered to be of lower quality of the evidence lower

se of resulting uncertainty in the results.

The thresholds of important benefits or harms, or the minimal importantdifference (MID) for an outcome are important considerations fordetermining whether there is a “clinically important” difference betweeninterventions and in assessing imprecision.MID is defined as “the smallest difference in score in the outcome ofinterest that informed patients or informed proxies perceive as important,either beneficial or harmful, and thatconsider a change in the management”.the MID is considered to be “clinically important”. For dichotomousoutcomes, the MID is considered in terelative risks.

The difference between two interventions, as observed in the studies, wascompared against the MID when considering whether the findings were of“clinical importance”; this is useful to guide decisions.effect size was small (less than the MID), this finding suggests that theremay not be enough difference to strongly recommend one interventionover the other based on that outcome.

The default thresholds suggested by GRADE were a relaof 25% (relative risk of 0.75 for negative outcomes) or a relative riskincrease of 25% (risk ratio 1.25 for positive outcomes) for binary outcomes.For this guideline, default threshold suggested by GRADEFor continuous outcomes, a standardised mean difference (SMD) of 0.5was considered the MID for most outcomes.

The CI for the pooled or best estimate of effect was considered in relationto the MID, as illustrated in Figure 1threshold, there was uncertainty in the effect estimate in supporting ourrecommendations (because the CI was consistent with two decisions) andthe effect estimate was rated as imprecise.

KCE Report 193S

The thresholds of important benefits or harms, or the minimal importantdifference (MID) for an outcome are important considerations fordetermining whether there is a “clinically important” difference betweennterventions and in assessing imprecision. For continuous outcomes, theMID is defined as “the smallest difference in score in the outcome ofinterest that informed patients or informed proxies perceive as important,either beneficial or harmful, and that would lead the patient or clinician to

der a change in the management”.145

An effect estimate larger thanthe MID is considered to be “clinically important”. For dichotomousoutcomes, the MID is considered in terms of changes in both absolute and

The difference between two interventions, as observed in the studies, wascompared against the MID when considering whether the findings were of“clinical importance”; this is useful to guide decisions. For example, if theeffect size was small (less than the MID), this finding suggests that theremay not be enough difference to strongly recommend one interventionover the other based on that outcome.

The default thresholds suggested by GRADE were a relative risk reductionof 25% (relative risk of 0.75 for negative outcomes) or a relative riskincrease of 25% (risk ratio 1.25 for positive outcomes) for binary outcomes.For this guideline, default threshold suggested by GRADE was adopted.

tcomes, a standardised mean difference (SMD) of 0.5was considered the MID for most outcomes.

The CI for the pooled or best estimate of effect was considered in relation1. Essentially, if the CI crossed the MID

ld, there was uncertainty in the effect estimate in supporting ourrecommendations (because the CI was consistent with two decisions) andthe effect estimate was rated as imprecise.

KCE Report 193S

Relation with Evidence statements used in this report

Figure 1 illustrates how the clinical importance of effect estimates

Figure 1 – Six examples of point estimates and confidence intervals for relative risks


used in this report

ates how the clinical importance of effect estimates and imprecision were considered in the evidence statements throughout this guideline.

Six examples of point estimates and confidence intervals for relative risks

505

were considered in the evidence statements throughout this guideline.

506

The evidence statements are linked with the GRADEplots included in the evidence plots. The Point estimatesdetermine if a result is clinical important. In figure(more scenario’s are possible) of relative risks. The dotted linfrom which moment a result can be considered as ‘clinical important’ (i.e. arelative risk <0.75 or a relative risk >1.25). In the figure below this is thecase in examples 1,2 and 3. This is of course only a ‘rule of thumb’ thatwas discussed with the clinical experts of the GDG and the external expertpanel on a case-by-case basis.

The ‘Confidence Intervals’ are used to specify the level ofimprecision of the point estimates. When point estimates are based onsmall studies, for instance, confidence intervals are wide, indicating a highlevel of imprecision.

In case of a high level of precision the evidence statements areformulated as follows: ‘x studies showed intervention is more clinicaleffective than control’ (situation 1) or ‘x studies showed theredifference in effect between intervention and control” (

In case of ‘serious imprecision, ‘potentially’ is used as terminology: Xstudies showed intervention is potentially more clinically effectivepreventing pressure ulcers compared to controlshowed there is potentially no clinical difference in effect betweenintervention and control (situation 4)

In case of ‘very serious imprecision’ the wording ‘(situations 3 and 6)

The above examples are not set in stone. The formulation of evidencestatements could be altered after discussions within the GDG or with theexternal experts.

Evidence statements will be used as input together with otherconsiderations (e.g. costs; user-friendliness of an intervention,…) toformulate recommendations.

10.1.5. Publication Bias

Publication bias is a systematic underestimate or an overestimate of theunderlying beneficial or harmful effect due to the selective publication ofstudies.


are linked with the GRADE-tables and ForestPoint estimates are used tofigure 1 we show 6 examples

(more scenario’s are possible) of relative risks. The dotted line indicatesfrom which moment a result can be considered as ‘clinical important’ (i.e. arelative risk <0.75 or a relative risk >1.25). In the figure below this is thecase in examples 1,2 and 3. This is of course only a ‘rule of thumb’ that

of the GDG and the external expert

are used to specify the level of precision orof the point estimates. When point estimates are based on

nstance, confidence intervals are wide, indicating a high

the evidence statements are‘x studies showed intervention is more clinical

r ‘x studies showed there is no clinicaldifference in effect between intervention and control” (situation 5)

, ‘potentially’ is used as terminology: Xmore clinically effective at

(situation 2); X studiesclinical difference in effect between

’ the wording ‘May be’ is used

The above examples are not set in stone. The formulation of evidencestatements could be altered after discussions within the GDG or with the

Evidence statements will be used as input together with otherriendliness of an intervention,…) to

Publication bias is a systematic underestimate or an overestimate of theunderlying beneficial or harmful effect due to the selective publication of

11. ASSESSMENT OFGUIDELINES

A scoping review was carried out to prepare the development of theguidelines for the prevention and treatment of pressure ulcers.search strategy was performed to identify clinical practice guidelines onpresvention and/or treatment of pressure ulcers.search of electronic databases were search using indextext words. Following databases were included for this search:(OVID), CINAHL (EBSCO-interface), EmCochrane Collaboration. Secondly, websites of guideline developers andwound care organisations were searched using freeMedical Directors Association (AMDA), Australian Wound ManagementAssociation, Canadian Medical Association (CMA),für Qualitätsentwicklung in der Pflege (DNQP), European WoundManagement Association, Guidelines International Network (GIN),Autorité de Santé (HAS), Institute for Clinical Systems Improvement (ICSI),Joanna Briggs Institute (JBI), Kwaliteitsinstituut voor de Gezondheidszorg(CBO), Landelijke Eerstelijns Samenwerkings Afspraken (LEVA’S),National Institute of Health and Clinical Excellence (NICE), NationalPressure Ulcer Advisory Panel and European PrePanel (NPUAP and EPUAP), Registered Nurses' Association of Ontario(RNAO), Scottish Intercollegiate GuidelinesGuideline Clearinghouse, Verpleegkundigen & Verzorgenden Nederland,Wound, Ostomy, and Continence Nurses Society (WOCNS),international, Wounds UK, and 1

ste

lists of all retrieved guidelines were searched to identify additionalguidelines.

Eighteen clinical practice guidelineselectronic databases and websites of guidelines developers andnational/international wound care organization

KCE Report 193S

OF EXISTING

A scoping review was carried out to prepare the development of theguidelines for the prevention and treatment of pressure ulcers. A three stypsearch strategy was performed to identify clinical practice guidelines on thepresvention and/or treatment of pressure ulcers. The first step involved a

electronic databases were search using index-terms and free-. Following databases were included for this search: Medline

interface), Embase, and the Library of the, websites of guideline developers and

wound care organisations were searched using free-text words: AmericanMedical Directors Association (AMDA), Australian Wound Management

an Medical Association (CMA), Deutsches Netzwerkfür Qualitätsentwicklung in der Pflege (DNQP), European Wound

Guidelines International Network (GIN), HauteInstitute for Clinical Systems Improvement (ICSI),

Kwaliteitsinstituut voor de Gezondheidszorg(CBO), Landelijke Eerstelijns Samenwerkings Afspraken (LEVA’S),National Institute of Health and Clinical Excellence (NICE), NationalPressure Ulcer Advisory Panel and European Pressure Ulcer AdvisoryPanel (NPUAP and EPUAP), Registered Nurses' Association of Ontario

Guidelines Network (SIGN), US NationalGuideline Clearinghouse, Verpleegkundigen & Verzorgenden Nederland,

e Nurses Society (WOCNS), Woundsste

lijn Amsterdam. Thirdly, the referencelists of all retrieved guidelines were searched to identify additional

clinical practice guidelines were identified trough the search ofelectronic databases and websites of guidelines developers and

wound care organizations.146, 147 , 148 , 149-161

KCE Report 193S

The retrieved guidelines were evaluated by threusing the Appraisal of Guidelines Research & Evaluation II (AGREE II).The AGREE II scores, particularly the scores of the domain ‘Rigour ofdevelopment’, was used to guide the research team in the decisionprocess whether to (1) include, (2) exclude or (3) adapt a guideline.of the retrieved guidelines were considered to be suitable to be used in anADAPTE-process. The most common reason fabsence of a systematic search for evidence and a lack of qualof included studies.

It was decided to develop the guidelines de novo. However, the guidelinesof NPUAP/EPUAP

161and NICE

147 , 148 , 149were considered as useful to

support the formulation of best-practices for our purposes as they bothmade use of a systematic and extensive consultation process to gatherexpert opinion.


by three independent reviewersEvaluation II (AGREE II).

The AGREE II scores, particularly the scores of the domain ‘Rigour ofdevelopment’, was used to guide the research team in the decision-making

(1) include, (2) exclude or (3) adapt a guideline. Noneof the retrieved guidelines were considered to be suitable to be used in an

The most common reason for exclusion was thea systematic search for evidence and a lack of quality appraisal

It was decided to develop the guidelines de novo. However, the guidelineswere considered as useful to

practices for our purposes as they bothmade use of a systematic and extensive consultation process to gather

507

508

12. RECOMMENDATIONS: COM

Item

Recommendation(s)

Prevention Pressure Ulcers

Pressure ulcerpreventionshould betailored toindividual needsand situationsand should bebased on theprinciples ofshared decisionmaking – BestPractice

Prevention should take into account severalfactors such as the patients’ medicalcondition, the overall plan of care and thepatients’ preferences. The circumstancesand needs of the individual should be re-assessed regularly

When a patient is assessed to be at risk ofpressure ulcer development:o the aims of the preventive actions to betaken should be explained fully andunambiguously;o the possible risks and benefits related withthe preventive actions to be taken should bediscussed openly;o it should be confirmed that the patientwas able to fully understand the informationbeing provided;o the patient should be encouraged to clarifywhat he/she feels to be important, and thehealthcare provider should evaluate whetherthis is consistent with the aims of thepreventive actions to be taken;o Register the planned and agreed/refusedactions.

Training andeducation ofprofessionalcaregiversshould be anintegral part ofany pressure

Training and education should be tailoredboth to the needs of individual caregiver andto the responsibilities of the group ofprofessionals.


RECOMMENDATIONS: COMMENTS EXPERT PANELGOR LoE

Comments

Min Max Mean

Prevention Pressure Ulcers

Prevention should take into account several Best practice R10: can we be more precise about"regularly"?R9: I suppose that 'principles of shareddecision making' will be explained intext? I find the second sentence rathervague. What is meant bycirucumstances and needs? Is this riskassessment?

4 5

o the possible risks and benefits related withbe

was able to fully understand the information

o the patient should be encouraged to clarify

er

Best practice

R9: concerning the first bullet: to thepatient? Second bullet: discussed withthe caregivers? Third bullet: this willoften not be possible.R3: Oui pour la prise de décisionpartagée mais le soignant doit êtrecapable d'argumenter et de convaincrele patient à accepter une préventionadaptée à ses besoins,

3 5

egiver andBest practice

R10: we should take care that allresponsibilities are met within thegroup. It is not necessary that eachindividual knows everything, but thesum of the knowledge of the persons inthe group should be sufficient

4 5

KCE Report 193S

Mean Median % 4 or 5 Decision

5 5 100%

5 5 91%

We propose a"softer"formulation: anindividual plan ofcare isadopted basedon assessmentdata, identifiedrisk factors andpatient goals andpreferences. Theplan isdeveloped ininteraction withthe patient,significant othersand thehealthcareprofessional. Theplanned andagreed/refusedactions aredocumented.

5 5 100%

KCE Report 193S

Item

Recommendation(s)

ulcer preventionpolicy in allhealthcaresettings – BestPractice

Following components should beconsidered as part of eacheducational/training programme:o Risk assessment;o Skin assessment;o Selection and use of pressureredistributing devices;o Patient repositioning;o Methods of documenting riskassessments and preventive activities;o The importance of an interdisciplinaryapproach;o The education of patient and their informalcaregivers.

Risk assessment–Recommendation

Use a structured approach to riskassessment to identify individuals at risk ofdeveloping pressure ulcers. This structuredapproach should include all of followingcomponents:o Clinical judgement informed by knowledgeof key risk factors;o The use of a risk assessment tool. Asclinical studies do not demonstrate thesuperiority of one risk assessment tool overanother, decisions about which riskassessment tool (Braden, Norton,Waterlow…) to be used should be based onthe intended patient population (adults,children, elderly,…) and the intended caresetting (ICU, general wards, paediatrics,home care…) and the experience andexpertise of the healthcare staff;o A comprehensive skin assessment toevaluate any alterations to intact skin.

Risk assessment– Best practice

Pressure ulcer risk assessment should beperformed at each first patient contact.Reassessment should be undertaken atregular time intervals and if there is anychange in the individual’s medical condition.The decision on time intervals should bebased on an individual basis.


GOR LoE

Comments

Min Max Mean

o The education of patient and their informal

Best practice

R10: I suppose that nothing is toldabout the treatment, because this willbe in another document?R8: Je propose d'ajouter l'évaluationnutritionnelle du patient – identificationdes risques liés à la mal nutrition

4 5

assessment to identify individuals at risk ofdeveloping pressure ulcers. This structured

o Clinical judgement informed by knowledge

superiority of one risk assessment tool over

Waterlow…) to be used should be based on

Strong VeryLow

R9: First bullet: Can 'key risk factors' beexplained?R8: Il serait utile de faire la différenceentre plaie d'escarre et plaie demacération. En l'absence de preuvessur l'utilisation d'une échelled'évaluation et en soulignant l'intérêt dujugement clinique des soignants, ilserait utile de souligner l'intérêt dudéveloppement du jugement cliniquedans la formation des soignants.R3: et une anamnèse complète dupatient

4 5

change in the individual’s medical condition.

Best practiceR10: is it feasible to give the maximaltime interval that can be allowed?R1:risico evaluatie moet gebeuren opvastgestelde tijdstippen en niet opindividuele basis. Groot risico dat hetover het hoofd wordt gezien of nietwordt gedaan!

2 5

509


5 5 100%

We propose toadd:etiology and riskfactorspredisposing topressure ulcer;Classificationpressure ulcers;Differntialdiagnosis withother type of skinlesions;nutritional

5 5 100%

Risk factors (seebelow:this will beadded)

4 5 82%

510

Item

Recommendation(s)

Clinical judgement should take into accountseveral key risk factors such as reducedmobility, immobility, pressure, shear,sensory impairment, acute deterioration ofgeneral health status, incontinence, reducedlevel of consciousness, advanced age,previous history of pressure damage,vascular disaese, undernutrition, poorhydration status (Non-limitative list).

Risk assessment should be documentedand made accessible to all members of themultidisciplinary team

Skin assessment–recommendation

A comprehensive skin assessment shouldbe part of a structured risk assessmentapproach.

Skin assessment– best practice

Skin assessment should be an integral partof daily routine care and the frequency ofinspection may need to be increased inresponse to the evolution of the individual’scondition (improvement or deterioration).

Inspect the skin regularly for signs ofredness. A patient is considered at riskwhen non-blanchable erythema is observed.

Skin assessment should also includeassessment for localized heat, edema, orinduration (hardness).

Observe the skin for pressure damagecaused by medical devices.


GOR LoE

Comments

Min Max Mean

Clinical judgement should take into account

general health status, incontinence, reduced

Best practice 5 5

Best practice 5 5

Strong Low

R8: OK mais la recommandationdevrait être plus explicite sur ce qu'estune évaluation complète de la peau

4 5

Skin assessment should be an integral part

individual’s

Best practice

R9: Increase only in case ofdeterioration, I suppose,R8: et de l'évaluation du niveau derisque

4 5

blanchable erythema is observed.

Best practice 5 5

Best practiceR10: in the case these parameters areobserved: also important to documentchanges (increased heat eg is moreimportant) and important to think aboutother possible diagnoses (infection,eczema, other skin diseases)

2 5

Best practice 5 5

KCE Report 193S


5 5 100%

5 5 100%

5 5 100% We suggest toadd: "ahead-to-toe skinassessment withspecial attentionto vulnerableareas, especiallybonyprominences"

5 5 100%we will change"increased"into adapted andadd "and level ofrisk"We will removedaily

5 5 100%

delete "a patientconsidered atrisk …

5 5 90%

We will add frailskin, NBEand combine itwith the previousbest-practice

5 5 100%

KCE Report 193S

Item

Recommendation(s)

In patients with darkly pigmented skin,consider a patient at risk when there are:o purplish/bluish localised areas of skin;o localised heat which, if tissue becomesdamaged, is replaced by coolness;o localised oedema;o localised induration.

Any skin changes should be documentedand made accessible to all members of themultidisciplinary team.

Skin massage –recommendation

Skin massage and rubbing, particularly overbony prominences, should be avoided toprevent pressure ulcers

Skin massage –best practice

The clinical effectiveness of various types ofskin products (e.g. creams, ointments) beingintended for other purposes (such as skinhydration, skin protection) was not studiedfor this guideline. However, applying suchproducts requires a gentle massagetechnique; rubbing of the skin should beavoided.


GOR LoE

Comments

Min Max Mean

Best practice

R10: analogous remarks as above

3 5

Best practice

R10: also need for a system to "alert"the other teammembers when a certainskin change is observed

4 5

Skin massage and rubbing, particularly over Strong VeryLow

4 5

The clinical effectiveness of various types ofskin products (e.g. creams, ointments) being

hydration, skin protection) was not studied

Best practice

R9: As this was not studied for thisguideline, I would omit it form therecommendations. In my opinion youcannot include something that was notsubject of the guideline.R3: Ok pour doit être évité et nondevrait être évité

2 5

511


5 5 90%

5 5 100%

We will changethis in:Any skinchanges shouldbe documented,made accessibleandcommunicated tothemultidisciplinaryteam.

5 5 100%

4 5 91%

The clinicaleffectiveness ofvarious types ofskin products(e.g. creams,ointments)being intendedfor otherpurposes (suchas skinhydration, skinprotection) wasnot studied forthis guideline.Applying suchproducts requiresa gentleapplicationtechnique;rubbing of theskin should beavoided.

512

Item

Recommendation(s)

Repositioning –recommendation

A repositioning protocol (includingspecifications about posture and frequency)should be established and documented foreach person at risk for pressure ulcerdevelopment

Individuals being at risk for pressure ulcerdevelopment should be repositioned. Thefrequency and method for repositioning andthe posture should be determined andadapted based on an individual assessmentand should take into account:o the patient’s medical condition;o the patient’s skin condition;o the patient’s level of activity and mobility;o the patient’s comfort;o the patient’s overall plan of care;o the characteristics of the support surface.

Repositioning – lying positiono Repositioning using the 30° tilted side-lying position is recommended if theindividual can tolerate it and her/his medicalcondition allows (back supported andsacrum free).

Repositioningtechnique – bestpractices

Repositioning should be undertaken(alternately, right side, back, left side) or theprone position if the individual can toleratethis and her/his medical condition allows.Avoid postures that increase pressure, suchas the 90-degree side-lying position, or thesemi- recumbent position.

Increase the contact surface between thepatient and the support surface toredistribute and reduce the pressuremaximally on the patient’s skin andunderlying tissue.

Avoid the skin being exposed to pressure orshearing forces

Avoid positioning the individual on a bonyprominence, especially if non -blanchableerythema is present


GOR LoE

Comments

Min Max Mean

frequency)Strong Very

Low5 5

oning and

adapted based on an individual assessment

o the patient’s level of activity and mobility;

o the characteristics of the support surface.

Strong VeryLow

R9: Is the 'overall plan of care' anaddition to the other mentioned points?I find it rather vague.R8: L'évaluation du risque doit aussiêtre un élément à prendre en compte

3 5

individual can tolerate it and her/his medical

Strong VeryLow

R10: shouldn't there be an "it" afterallows?

5 5

(alternately, right side, back, left side) or thetolerate

Avoid postures that increase pressure, such

Best practice

R10: idem

5 5

Best practice 5 5

Avoid the skin being exposed to pressure or Best practice R9: This recommendation is somewhatsuperfluous in my opinion.

3 5

Best practice 4 5

KCE Report 193S


5 5 100%

5 5 91%

We will add 'levelof risk'

5 5 100%

5 5 100%

5 5 100%

5 5 91%

5 5 100%

KCE Report 193S

Item

Recommendation(s)

Manual handling devices should be usedcorrectly (Lift – don’t drag – the individual)in order to minimise shear and frictiondamage. After patient manipulation, anyslings, hoists, sleeves or other parts of thehandling equipment being used should beremoved immediately if they can damage theskin (a slide sheet can be tolerated andhelps to prevent shear forces in combinationwith a good posture).

If sitting upright in bed is needed, a head-ofbed elevation and slouched position(increasing pressure and shear on thesacrum and coccyx.) should be avoided.

Avoid pressure of medical devices or othermaterials directly on the skin and underlyingtissue (e.g. tubes, drainage systems,syringes, caps….).

Repositioningschedule – bestpractices

Assess the individual’s skin condition andgeneral comfort on a regular basis. If theindividual is not responding to therepositioning regimen as beingexpected(e.g. if Category I pressure ulceroccurs), the frequency, method, and appliedpostures for repositioning should bereconsidered, documented and madeaccessible to all members of themultidisciplinary team.

Repositioningseating – bestpractices

Position the individual so as to maintainhis/her full range of normal activities. Makesure that everything he/she needs is inreach;


GOR LoE

Comments

Min Max Mean

removed immediately if they can damage the

helps to prevent shear forces in combination

Best practice 3 5

of- Best practice

R9: I don't understand why a head ofbed elevation should be avoided ifsitting upright in bed is needed.

4 5

pressure of medical devices or othermaterials directly on the skin and underlying

Best practiceR10: if the device or material is thickenough, there could also be someproblems if there is not a directpressure (eg with clothes or bed linenin between.

4 5

occurs), the frequency, method, and applied

Best practice

R9: e.g. If Cat I PU occurs,,, is in factnot a good example as sometimespatiënts are only considered at risk ifthey have Cat I PU)

4 5

his/her full range of normal activities. MakeBest practice 4 5

513


5 5 82%

5 5 100%We suggest tochange this asfollows "morethan 30° head-ofbedelevation shouldbe avoided"

5 5 100%

5 5 100%

if a not pre-existingCategory Ipressure ulceroccurs

5 5 100%

514

Item

Recommendation(s)

The time that individuals at risk for pressureulcer development are seated in a chairshould be limited. The time a patient isseated in a chair should be determined andcontinuously adapted based on anindividual assessment and should take intoaccount comfort/dignity, the overall plan ofcare, the medical condition, and thecharacteristics of the pressure relievingdevices used;

The posture of individuals who spendsubstantial periods of time in a chair orwheelchair should take into account:distribution of weight of the support surface(cushion); postural alignment; pressurepoints and support of feet.

Repositioningpatient eucation– best practices

Individuals (or informal carers assisting theindividual) who are willing and able shouldbe taught the principles of weightdistribution and how to achieve this.

Redistributingdevices –

recommendation

The use of pressure redistributing devices(low-tech constant low pressure surfaces orhigh-tech support surfaces) isrecommended for individuals at risk ofpressure ulcers development.

Mattresses without pressure redistributingor relieving characteristics should beavoided to prevent ulcers development in atrisk individuals.


GOR LoE

Comments

Min Max Mean

The time that individuals at risk for pressure

seated in a chair should be determined and

individual assessment and should take intoaccount comfort/dignity, the overall plan of

Best practice 3 5

distribution of weight of the support surface

Best practice

R9: Support of feet is in upright sittingposition? Mention here also theprinciple of floating heels when sittingbackwards? Can't be mentioned thatsitting upright results in higher pressurethan sitting in backwards position?R8: Des propositions derepositionnement du patient devraientêtre ajoutées

4 5

(or informal carers assisting the Best practice 3 5

tech constant low pressure surfaces orStrong Very

Low5 5

avoided to prevent ulcers development in at

Strong VeryLow

R10: "ulcer development" in stead ofulcersR8: Chez les patients à risques

2 5

KCE Report 193S


5 5 91%

5 5 100%

We will takeNPUAPformulation for4.2 (p35)posturalalignment (by theuse ofpositioningdevices)

5 5 82%

5 5 100%

5 5 90%

KCE Report 193S

Item

Recommendation(s)

Pressure redistributing overlays arerecommended on the operating table.Consider the use of a visco elastic polymersupport surface on the operating table.

Redistributingdevices – best

practices

The Choice of pressure-relieving devices:o As clinical studies do not demonstrate thesuperiority of one pressure redistributingdevice over another, decisions about whichpressure redistributing device to use shouldbe based on an overall assessment of theindividual including level of risk, comfortand general health state. Appropriateness ofeach device in different care settings, andother considerations (e.g. cleaning, type ofmattress cover, Cardiopulmonaryresuscitation -function, disinfection andcost) can contribute to guide the choice.

Verify the functionality of the pressureredistributing device on a regular basis.

Use a pressure redistributing seat cushionfor an individual at risk of pressure ulcerdevelopment when in a seated position:o No seat cushion showed to out-performanother, therefore no recommendation canbe made about which type of cushion to usefor pressure redistribution purposes.


GOR LoE

Comments

Min Max Mean

Consider the use of a visco elastic polymer

Strong Low 4 5

o As clinical studies do not demonstrate the

device over another, decisions about whichpressure redistributing device to use should

and general health state. Appropriateness of

Best practice

R10: cardiopulmonary (minor case)R9: Is a very vague recommendation…R8: Ok mais pour tout patient à risqueun matelas réducteur de pression doitêtre recommandé

4 5

Best practice

R8: Sur base de quels critères ?;DD: I would complete this by "Ensureoptimal use of the device" (by checkingthe manual)". In daily practice thisseems to be often a bottleneckR3: Préciser davantage : à chaqueprise de service, pause horaire..

4 5

be made about which type of cushion to use

Best practice

R9: Also very vague. ° No "pressureredistributing" seat cushion,,,

3 5

515


5 5 100%

5 5 100% We will removethis part totherecommendation:As clinicalstudies do notdemonstrate thesuperiority of onepressureredistributingdevice overanother,decisions aboutwhich pressureredistributingdevice to useshould be basedon an overallassessment ofthe individualincluding level ofrisk, comfort andgeneral healthstate. (line 35)

5 5 100%

5 5 82%

ok

516

Item

Recommendation(s)

In addition to the use of overlays on theoperating table, other general preventivemeasures should be undertaken duringsurgery:o Position the patient in such a way as toreduce the risk of pressure ulcerdevelopment, especially by avoiding shearforces.o Elevate the heels completely (offloadthem) in such a way to redistribute theweight of the leg along the calf withoutputting all the pressure on the Achillestendon. The knee should be in flexion andsupported.o Several devices to redistribute pressure(e.g. face pillows for patients in a proneposition on the operating table) are availablebut no devices have shown to out-performanother, therefore no recommendation canbe made about which type to use forpressure redistribution purposes.

Heel ulcerprevention –recommendation

The use of devices that ensure that heels arefree of the surface of the bed in combinationwith a mattress with pressure-relievingcharacteristics is recommended forindividuals at risk for pressure ulcersdevelopment. No device has been shown toout-perform another, therefore norecommendation can be made about whichtype to use for pressure redistributionpurposes.


GOR LoE

Comments

Min Max Mean

tion on the operating table) are available

Best practice

R9: First and second bullet: add 'ifpossible'R8: Ok mais cette bonne pratiquedevrait se retrouver plus haut (sous larecommandation relative à la tabled'opération). Néanmoins, la mise endécharge de l'appui des talons ne doitpas être recommandée uniquementpour la salle d'op mais chez tous lespatients à risque.

4 5

The use of devices that ensure that heels arefree of the surface of the bed in combination

development. No device has been shown to

recommendation can be made about which

Strong VeryLow

R8: OK voir ci-dessusR1: hielen moeten verplicht inzweefstand geplaatst worden bij risico

4 5

KCE Report 193S


5 5 100%

5 5 100%

KCE Report 193S

Item

Recommendation(s)

Heel ulcerprevention – bestpractices

Heel-protection devices should elevate theheel completely (offload them) in such a wayas to distribute the weight of the leg alongthe calf without putting pressure on theAchilles tendon. The knee should be inslight flexion and supported.

Inspect the skin of the heels regularly.

Nutrition –recommendation

As clinical studies do not demonstrate thesuperiority of one nutritional intervention onanother, a specific diet cannot berecommended to prevent the developmentof pressure ulcers.

Nutrition best-practices

Best practice includes monitoring thenutritional status of individuals as part of ageneral assessment procedure and as anongoing process throughout an individual’sepisode of care. Initially, this assessmentshould include documentation andmonitoring of the following factors:o current weight and height;o recent weight loss;o usual eating habits;o recent changes in eating habits and intake.


GOR LoE

Comments

Min Max Mean

heel completely (offload them) in such a wayBest practice

R9: Pay also attention at heel PUprevention while sitting!R8: Un schéma devrait permettre demieux comprendre la position proposée(idem pour les différentes positions dupatient couché et assis)

4 5

Best practice R10: can we be more precise about"regularly"? Minimum frequency?

4 5

n onStrong Very

low

R3: préciser davantage

3 5

rt of a

ongoing process throughout an individual’s

o recent changes in eating habits and intake.

Best practice

R9: I wonder if it needed to describethe elements of the nutritionalassessment?R8: Il est important de noter que ceséléments doivent être suivis enpluridisciplinarité

4 5

517


5 5 100%We will start thephrase with"for bedriddenindividuals orindividualsrestricted to achair in " ma'sdecubitus.beintegrerenchange puttingpressure inapplyingpressurevoorste Geert +schem'sdecubitus.be

5 5 100%

5 5 90%

5 5 100%

518

Item

Recommendation(s)

If nutritional risk is suspected, practitionersshould undertake more detailed screening.A formal nutritional risk assessment scalemay be preferred to help with this andnutritionally compromised individualsshould be referred to a dietician.


GOR LoE

Comments

Min Max Mean

If nutritional risk is suspected, practitioners Best practice R10: does this include blood levels ofoligo-elements? Suggestion to foreseea multidiscplinary approach for thenutritionally compromised individuals(including a dietician, medical doctor)R9: This recommendation implies thata nutritional ris assessment scale isonly used in suspected risk patients.However, a general assessment canalso be done with a risk assessmentscale…R3: biologie: préalbumine, protéine,CRP

3 5

KCE Report 193S


5 5 90%

should referredto a dietician=> should bediscussedmultidisciplinary

KCE Report 193S

REFERENCES


1. Saleh M, Anthony D, Parboteeah S. The impact of pressure ulcerrisk assessment on patient outcomes among hospitalised patients.Journal of Clinical Nursing. 2009;18(13):1923

2. Webster J, Coleman K, Mudge A, Marquart L, Gardner G,Stankiewicz M, et al. Pressure ulcers: effectiveness of riskassessment tools. A randomised controlled trial (the ULCER trial).BMJ Quality and Safety. 2011;20(4):297

3. Bergstrom N, Braden BJ, Laguzza A, Holman V. Thefor Predicting Pressure Sore Risk. Nursing Research.1987;36(4):205-10.

4. Norton D, McLaren R, ExtonGeriatric Nursing Problems in Hospital. Edinburgh: ChurchillLivingstone; 1962.

5. Waterlow J. Pressure sore:Times. 1985;81:49-55.

6. Waterlow J. The Waterlow Pressure Ulcer prevention manual(revised 2005); 2005.

7. Cubbin B, Jackson C. Trial of a pressure area risk calculator forintensive therapy patients. Intensive Care Nursing

8. Jackson C. The revised Jackson/Cubbin Pressure Area RiskCalculator. Intensive & Critical Care Nursing. 1999;15(3):169

9. Quigley SM, Curley MAQ. Skin integrity in the pediatric population:preventing and managing pressure ulcers.of Pediatric Nurses. 1996;1(1):7

10. Pancorbo-Hidalgo PL, GarciaAlvarez-Nieto C. Risk assessment scales for pressure ulcerprevention: a systematic review. Journal of advanced nursing.2006;54(1):94-110.

11. Bergquist S, Frantz R. Braden scale: Validity in communityolder adults receiving home health care.Research. 2001;14(1):36-43.

12. Fuentelsaz GC. Validaci¢n de la escala EMINA: un instrumento devaloraci¢n del riesgo de desarrollar £lceras por presi¢n enpacientes hospitalizados. Enerferia Clinica. 2001;11(3):97

519

Saleh M, Anthony D, Parboteeah S. The impact of pressure ulcerrisk assessment on patient outcomes among hospitalised patients.Journal of Clinical Nursing. 2009;18(13):1923-9.

Mudge A, Marquart L, Gardner G,Stankiewicz M, et al. Pressure ulcers: effectiveness of risk-assessment tools. A randomised controlled trial (the ULCER trial).BMJ Quality and Safety. 2011;20(4):297-306.

Bergstrom N, Braden BJ, Laguzza A, Holman V. The Braden Scalefor Predicting Pressure Sore Risk. Nursing Research.

Norton D, McLaren R, Exton-Smith AN. An Investigation ofGeriatric Nursing Problems in Hospital. Edinburgh: Churchill-

Waterlow J. Pressure sore: a risk assessment card. Nursing

Waterlow J. The Waterlow Pressure Ulcer prevention manual

Cubbin B, Jackson C. Trial of a pressure area risk calculator forintensive therapy patients. Intensive Care Nursing. 1991;7(1):40-4.

Jackson C. The revised Jackson/Cubbin Pressure Area RiskCalculator. Intensive & Critical Care Nursing. 1999;15(3):169-75.

Quigley SM, Curley MAQ. Skin integrity in the pediatric population:preventing and managing pressure ulcers. Journal of the Societyof Pediatric Nurses. 1996;1(1):7-18.

Hidalgo PL, Garcia-Fernandez FP, Lopez-Medina IM,Nieto C. Risk assessment scales for pressure ulcer

prevention: a systematic review. Journal of advanced nursing.

Bergquist S, Frantz R. Braden scale: Validity in community-basedolder adults receiving home health care. Applied Nursing

43.

Fuentelsaz GC. Validaci¢n de la escala EMINA: un instrumento desarrollar £lceras por presi¢n en

Enerferia Clinica. 2001;11(3):97-103.

520

13. Boyle M, Green M. Pressure sores in intensive care: defining theirincidence and associated factors and assessing the utility of twopressure sore risk assessment tools. Australian Critical Care.2001;14(1):24-30.

14. Hagisawa S, Barbenel J. The limits of pressure sore prevention.Journal of the Royal Society of Medicine. 1999;92(11):576

15. Gunningberg L, Lindholm C, Carlsson M, Sj”d‚n PO.Implementation of risk assessment and classification of pressureulcers as quality indicators for patients with hip fractures. Journalof Clinical Nursing. 1999;8(4):396-406.

16. Andersen KE, Jensen O, Kvorning SA, Bach E. Prevention ofpressure sores by identifying patients at risk. BMJ.1982;284(6326):1370-1.

17. Anthony D, Reynolds T, Russell L, Anthony D, Reynolds T, RussellL. A regression analysis of the Waterlow score in pressure ulcerrisk assessment. Clinical Rehabilitation. 2003;17(2):216

18. Barnes D, Payton RG. Clinical application of the Braden Scale inthe acute-care setting. Dermatology Nursing. 1993;5(5):386

19. Bergstrom N, Demuth PJ, Braden BJ. A clinical trial of the BradenScale for Predicting Pressure Sore Risk. The Nursing Clinics ofNorth America. 1987;22(2):417-29.

20. Bergstrom N, Braden B, Kemp M, Champagne M, Ruby E.Predicting pressure ulcer risk: a multisite study of the predictivevalidity of the Braden Scale. Nursing Research. 1998;47(5):261

21. Braden BJ, Bergstrom N. Predictive validity of the Braden Scale forpressure sore risk in a nursing home population. Research inNursing & Health. 1994;17(6):459-70.

22. Capobianco ML, McDonald DD. Factors affecting the predictivevalidity of the Braden Scale. Advances in Wound Care.1996;9(6):32-6.

23. Chan WS, Pang SM, Kwong EW. Assessing predictive validity ofthe modified Braden scale for prediction of pressure ulcer risk oforthopaedic patients in an acute care setting. Journal of ClinicalNursing. 2009;18(11):1565-73.


Boyle M, Green M. Pressure sores in intensive care: defining theirincidence and associated factors and assessing the utility of two

essment tools. Australian Critical Care.

Hagisawa S, Barbenel J. The limits of pressure sore prevention.Journal of the Royal Society of Medicine. 1999;92(11):576-8.

Gunningberg L, Lindholm C, Carlsson M, Sj”d‚n PO.of risk assessment and classification of pressure

ulcers as quality indicators for patients with hip fractures. Journal

Andersen KE, Jensen O, Kvorning SA, Bach E. Prevention ofients at risk. BMJ.

Anthony D, Reynolds T, Russell L, Anthony D, Reynolds T, RussellL. A regression analysis of the Waterlow score in pressure ulcerrisk assessment. Clinical Rehabilitation. 2003;17(2):216-23.

on RG. Clinical application of the Braden Scale incare setting. Dermatology Nursing. 1993;5(5):386-8.

Bergstrom N, Demuth PJ, Braden BJ. A clinical trial of the BradenScale for Predicting Pressure Sore Risk. The Nursing Clinics of

Bergstrom N, Braden B, Kemp M, Champagne M, Ruby E.Predicting pressure ulcer risk: a multisite study of the predictivevalidity of the Braden Scale. Nursing Research. 1998;47(5):261-9.

lidity of the Braden Scale forpressure sore risk in a nursing home population. Research in

Capobianco ML, McDonald DD. Factors affecting the predictivevalidity of the Braden Scale. Advances in Wound Care.

Chan WS, Pang SM, Kwong EW. Assessing predictive validity ofthe modified Braden scale for prediction of pressure ulcer risk oforthopaedic patients in an acute care setting. Journal of Clinical

24. Compton F, Hoffmann F, Hortig T, Strauss M, Frey J, Zidek W, etal. Pressure ulcer predictors in ICU patients: nursing skinassessment versus objective parameters.[Erratum appears in JWound Care. 2008 Nov;17(11):493]. Journal of Wound Care.2008;17(10):417-24.

25. Curley MAQ, Razmus IS, Roberts KE, Wypij D. Predictingpressure ulcer risk in pediatric patients: the Braden Q Scale.Nursing Research. 2003;52(1):22

26. de Souza DM, Santos VL, Iri HK, Sadasue Oguri MY. Predictivevalidity of the Braden Scale for Pressureresidents of long-term care facilities. Geriatric Nursing.2010;31(2):95-104.

27. Edwards M. The levels of reliability and validity of the Waterlowpressure sore risk calculator. Journal of Wound Care.1995;4(8):373-8.

28. Feuchtinger J, Halfens R, Dassen T. Pressure ulcer riskassessment immediately after cardiac surgerydifference? A comparison of three pressure ulcer risk assessmentinstruments within a cardiac surgery population. Nursing in CriticalCare. 2007;12(1):42-9.

29. Goodridge DM, Sloan JA, LeDoyen YM, McKenzie JA, Knight WE,Gayari M. Risk-assessment scores, prevention strategies, and theincidence of pressure ulcers among the elderly in four Canadianhealth-care facilities. The Canadian Journal of Nursing.1998;30(2):23-34.

30. Halfens RJG, Van Achterberg T, Bal RM. Validity and reliability ofthe Braden scale and the influence of other risk factors: a multicentre prospective study. International Journal of Nursing Studies.2000;37(4):313-9.

31. Hatanaka N, Yamamoto Y, Ichihara K, Mastuo S, Nakamura Y,Watanabe M, et al. A new predictive indicator for development ofpressure ulcers in bedridden patients based on common laboratorytests results. Journal of Clinical Pathology. 2008;61(4):514

32. Jalali R, Rezaie M. Predicting pressure ulcer risk: comparing thepredictive validity of 4 scales. Advances in skin & wound care.2005;18(2):92-7.

KCE Report 193S

mann F, Hortig T, Strauss M, Frey J, Zidek W, etal. Pressure ulcer predictors in ICU patients: nursing skinassessment versus objective parameters.[Erratum appears in JWound Care. 2008 Nov;17(11):493]. Journal of Wound Care.

y MAQ, Razmus IS, Roberts KE, Wypij D. Predictingpressure ulcer risk in pediatric patients: the Braden Q Scale.Nursing Research. 2003;52(1):22-33.

de Souza DM, Santos VL, Iri HK, Sadasue Oguri MY. Predictivevalidity of the Braden Scale for Pressure Ulcer Risk in elderly

term care facilities. Geriatric Nursing.

Edwards M. The levels of reliability and validity of the Waterlowpressure sore risk calculator. Journal of Wound Care.

J, Halfens R, Dassen T. Pressure ulcer riskassessment immediately after cardiac surgery--does it make adifference? A comparison of three pressure ulcer risk assessmentinstruments within a cardiac surgery population. Nursing in Critical

Goodridge DM, Sloan JA, LeDoyen YM, McKenzie JA, Knight WE,assessment scores, prevention strategies, and the

incidence of pressure ulcers among the elderly in four Canadiancare facilities. The Canadian Journal of Nursing.

Halfens RJG, Van Achterberg T, Bal RM. Validity and reliability ofthe Braden scale and the influence of other risk factors: a multi-centre prospective study. International Journal of Nursing Studies.

, Yamamoto Y, Ichihara K, Mastuo S, Nakamura Y,Watanabe M, et al. A new predictive indicator for development ofpressure ulcers in bedridden patients based on common laboratorytests results. Journal of Clinical Pathology. 2008;61(4):514-8.

Rezaie M. Predicting pressure ulcer risk: comparing thepredictive validity of 4 scales. Advances in skin & wound care.

KCE Report 193S

33. Kim E, Lee S, Lee E, Eom M. Comparison of the predictive validityamong pressure ulcer risk assessment scales for spatients. Australian Journal of Advanced Nursing. 2009;26(4):8794.

34. Kwong E, Pang S, Wong T, Ho J, Shao-ling X, Lipressure ulcer risk with the modified Braden, Braden, and Nortonscales in acute care hospitals in MainlandResearch. 2005;18(2):122-8.

35. Langemo DK, Olson B, Hunter S, Hanson D, Burd C, CathcartSilberberg T. Incidence and prediction of pressure ulcers in fivepatient care settings. Decubitus. 1991;4(3):25

36. Lewicki LJ, Mion LC, Secic M. Sensitivity and specificity of theBraden Scale in the cardiac surgical population. Journal of WoundOstomy & Continence Nursing. 2000;27(1):36

37. Lincoln R, Roberts R, Maddox A, Levine S, Patterson C. Use ofthe norton pressure sore risk assessment scoring system withelderly patients in acute care. Journal of Enterostomal Therapy.1986;13:132-8.

38. Lindgren M, Unosson M, Krantz AM, Ek AC. A risk assessmentscale for the prediction of pressure sore development: reliabilityand validity. Journal of advanced nursing. 2002;38(2):190

39. Lothian P. Identifying and protecting patients who may getpressure sores. Nursing Standard. 1989;18(4):26

40. Lyder CH, Yu C, Emerling J, Mangat R, Stevenson D, EmpleoFrazier O, et al. The Braden Scale for pressure ulcer risk:evaluating the predictive validity in Black and Latino/Hispanicelders. Applied Nursing Research. 1999;12(2):60

41. Ongoma C, Schmollgruber S, Langley GC. Predictive validity ofpressure risk assessment scales in a private sectorintensive care unit. Southern African Journal of Critical Care.2005;21(2):78-86.

42. Page KN, Barker AL, Kamar J. Development and validation of apressure ulcer risk assessment tool for acute hospital patients.Wound Repair & Regeneration. 2011;19(1):31


Kim E, Lee S, Lee E, Eom M. Comparison of the predictive validityamong pressure ulcer risk assessment scales for surgical ICUpatients. Australian Journal of Advanced Nursing. 2009;26(4):87-

ling X, Li-jun T. Predictingpressure ulcer risk with the modified Braden, Braden, and Nortonscales in acute care hospitals in Mainland China. Applied Nursing

Langemo DK, Olson B, Hunter S, Hanson D, Burd C, Cathcart-Silberberg T. Incidence and prediction of pressure ulcers in fivepatient care settings. Decubitus. 1991;4(3):25-36.

Secic M. Sensitivity and specificity of theBraden Scale in the cardiac surgical population. Journal of WoundOstomy & Continence Nursing. 2000;27(1):36-41.

Lincoln R, Roberts R, Maddox A, Levine S, Patterson C. Use ofessment scoring system with

elderly patients in acute care. Journal of Enterostomal Therapy.

Lindgren M, Unosson M, Krantz AM, Ek AC. A risk assessmentscale for the prediction of pressure sore development: reliability

nal of advanced nursing. 2002;38(2):190-9.

Lothian P. Identifying and protecting patients who may getpressure sores. Nursing Standard. 1989;18(4):26-9.

Lyder CH, Yu C, Emerling J, Mangat R, Stevenson D, Empleo-r pressure ulcer risk:

evaluating the predictive validity in Black and Latino/Hispanicelders. Applied Nursing Research. 1999;12(2):60-8.

Ongoma C, Schmollgruber S, Langley GC. Predictive validity ofpressure risk assessment scales in a private sector traumaintensive care unit. Southern African Journal of Critical Care.

Page KN, Barker AL, Kamar J. Development and validation of apressure ulcer risk assessment tool for acute hospital patients.

1):31-7.

43. Pang SM, Wong TK. Predicting pressure sore risk with the Norton,Braden, and Waterlow scales in a Hong Kong rehabilitationhospital. Nursing Research. 1998;47(3):147

44. Perneger TV, Rae AC, Gaspoz JM, Borst F, Vitek O, Heliot C.Screening for pressure ulcer risk in an acute care hospital:development of a brief bedside scale. Journal of ClinicalEpidemiology. 2002;55(5):498

45. Ramundo JM. Reliability and validity of the Braden Scale in thehome care setting. Journal of Wound Ostomy &Nursing. 1995;22(3):128-34.

46. Salvadalena GD, Snyder ML, Brogdon KE. Clinical trial of theBraden Scale on an acute care medical unit. Journal ofEnterostomal Therapy Nursing. 1992;19(5):160

47. Schoonhoven L, Haalboom JRE, Bousema MT, AlgDE, Grypdonck MH, et al. Prospective cohort study of routine useof risk assessment scales for prediction of pressure ulcers. BMJ.2002;325(7368):797.

48. Seongsook RNJ, Ihnsook RNJ, Younghee RNL. Validity ofpressure ulcer risk assessment scBraden, and Douglas scale. International Journal of NursingStudies. 2004;41(2):199-204.

49. Serpa LF, de Gouveia Santos VL, Gomboski G, Rosado SM.Predictive validity of Waterlow Scale for pressure ulcerdevelopment risk in hospitalized patients. Journal of wound,ostomy, and continence nursing. 2009;36(6):640

50. Smith I. Waterlow/Norton scoring system: a ward view. CAREScience and Practice. 1989;7(4):93

51. Stotts NA. Predicting pressure ulcer development in surgicalpatients. Heart & Lung. 1988;17(6 Pt 1):641

52. Suriadi, Sanada H, Sugama J, Thigpen B, Kitagawa A, Kinosita S,et al. A new instrument for predicting pressure ulcer risk in anintensive care unit. Journal of tissue viability. 2006;16(3):21

53. Suriadi, Sanada H, Sugama J, Thigpen B, Subuh M. Developmentof a new risk assessment scale for predicting pressure ulcers in anintensive care unit. Nursing in Critical Care. 2008;13(1):34

521

Pang SM, Wong TK. Predicting pressure sore risk with the Norton,Braden, and Waterlow scales in a Hong Kong rehabilitationhospital. Nursing Research. 1998;47(3):147-53.

Perneger TV, Rae AC, Gaspoz JM, Borst F, Vitek O, Heliot C.for pressure ulcer risk in an acute care hospital:

development of a brief bedside scale. Journal of ClinicalEpidemiology. 2002;55(5):498-504.

Ramundo JM. Reliability and validity of the Braden Scale in thehome care setting. Journal of Wound Ostomy & Continence

34.

Salvadalena GD, Snyder ML, Brogdon KE. Clinical trial of theBraden Scale on an acute care medical unit. Journal ofEnterostomal Therapy Nursing. 1992;19(5):160-5.

Schoonhoven L, Haalboom JRE, Bousema MT, Algra A, GrobbeeDE, Grypdonck MH, et al. Prospective cohort study of routine useof risk assessment scales for prediction of pressure ulcers. BMJ.

Seongsook RNJ, Ihnsook RNJ, Younghee RNL. Validity ofpressure ulcer risk assessment scales; Cubbin and Jackson,Braden, and Douglas scale. International Journal of Nursing

204.

Serpa LF, de Gouveia Santos VL, Gomboski G, Rosado SM.Predictive validity of Waterlow Scale for pressure ulcer

talized patients. Journal of wound,ostomy, and continence nursing. 2009;36(6):640-6.

Smith I. Waterlow/Norton scoring system: a ward view. CARE-Science and Practice. 1989;7(4):93-5.

Stotts NA. Predicting pressure ulcer development in surgicalents. Heart & Lung. 1988;17(6 Pt 1):641-7.

Suriadi, Sanada H, Sugama J, Thigpen B, Kitagawa A, Kinosita S,et al. A new instrument for predicting pressure ulcer risk in anintensive care unit. Journal of tissue viability. 2006;16(3):21-6.

Sanada H, Sugama J, Thigpen B, Subuh M. Developmentof a new risk assessment scale for predicting pressure ulcers in anintensive care unit. Nursing in Critical Care. 2008;13(1):34-43.

522

54. Towey AP, Erland SM. Validity and reliability of an assessmenttool for pressure ulcer risk. Decubitus. 1988;1(2):40

55. VandenBosch T, Montoye C, Satwicz M, DurkeeBoylan-Lewis B. Predictive validity of the Braden Scale and nurseperception in identifying pressure ulcer risk. Applied NursingResearch. 1996;9(2):80-6.

56. Wai-Han C, Kit-Wai C, French P, Yim-Sheung L, LaiWhich pressure sore risk calculator? A study of the effectivenessof the Norton scale in Hong Kong. International Journal of NursingStudies. 1997;34(2):165-9.

57. Weststrate JT, Hop WC, Aalbers AG, Vreeling AW, Bruining HA.The clinical relevance of the Waterlow pressure sore risk scale inthe ICU. Intensive Care Medicine. 1998;24(8):815

58. Serpa LF, Santos VL, Campanili TC, Queiroz M. Predictive validityof the Braden scale for pressure ulcer risk in critical care patients.Revista Latino-Americana de Enfermagem. 2011;19(1):50

59. Vanderwee K, Grypdonck M, Defloor T. Nonas an indicator for the need for pressure ulcer prevention: arandomized-controlled trial. Journal of Clinical Nursing.2007;16(2):325-35.

60. Konishi C, Sugama J, Sanada H, Okuwa M, Konya C, NishizawaT, et al. A prospective study of blanchable erythema amonguniversity hospital patients. International wound journal.2008;5(3):470-5.

61. Newman P, Davis NH. Thermography as a predictor of sacralpressure sores. Age and Ageing. 1981;10(1):14

62. Nixon J, Cranny G, Bond S. Skin alterations of intact skin and riskfactors associated with pressure ulcer development in surgicalpatients: a cohort study. International Journal of Nursing Studies.2007;44(5):655-63.

63. Duimel-Peeters IG, Halfens JG, Ambergen AW, Houwing RH,Berger PF, Snoeckx LH. The effectiveness of massage with andwithout dimethyl sulfoxide in preventing pressure ulcers: arandomized, double-blind cross-over trial in patients prone to


Towey AP, Erland SM. Validity and reliability of an assessmentfor pressure ulcer risk. Decubitus. 1988;1(2):40-8.

VandenBosch T, Montoye C, Satwicz M, Durkee-Leonard K,Lewis B. Predictive validity of the Braden Scale and nurse

perception in identifying pressure ulcer risk. Applied Nursing

Sheung L, Lai-Kwan T.Which pressure sore risk calculator? A study of the effectivenessof the Norton scale in Hong Kong. International Journal of Nursing

WC, Aalbers AG, Vreeling AW, Bruining HA.The clinical relevance of the Waterlow pressure sore risk scale inthe ICU. Intensive Care Medicine. 1998;24(8):815-20.

Serpa LF, Santos VL, Campanili TC, Queiroz M. Predictive validitypressure ulcer risk in critical care patients.

Americana de Enfermagem. 2011;19(1):50-7.

Vanderwee K, Grypdonck M, Defloor T. Non-blanchable erythemaas an indicator for the need for pressure ulcer prevention: a

al. Journal of Clinical Nursing.

Konishi C, Sugama J, Sanada H, Okuwa M, Konya C, NishizawaT, et al. A prospective study of blanchable erythema amonguniversity hospital patients. International wound journal.

wman P, Davis NH. Thermography as a predictor of sacralpressure sores. Age and Ageing. 1981;10(1):14-8.

Nixon J, Cranny G, Bond S. Skin alterations of intact skin and riskfactors associated with pressure ulcer development in surgical

rt study. International Journal of Nursing Studies.

Peeters IG, Halfens JG, Ambergen AW, Houwing RH,Berger PF, Snoeckx LH. The effectiveness of massage with andwithout dimethyl sulfoxide in preventing pressure ulcers: a

over trial in patients prone to

pressure ulcers. International Journal of Nursing Studies.2007;44(8):1285-95.

64. Defloor T, De BD, Grypdonck MH. The effect of variouscombinations of turning and pressure reducing devices on theincidence of pressure ulcers. International Journal of NursingStudies. 2005;42(1):37-46.

65. Moore Z, Cowman S, Conroy RM. A randomised controlled clinicaltrial of repositioning, using the 30degrees tilt, for the prevention ofpressure ulcers. Journal of Clinical Nursing. 2011;20(1744.

66. Vanderwee K, Grypdonck MHF, De Bacquer D, Defloor T.Effectiveness of turning with unequal time intervals on theincidence of pressure ulcer lesions. Journal of advanced nursing.2007;57(1):59-68.

67. Fineman LD, LaBrecque MA, Shih MC, Curley MAQ. Pronepositioning can be safely performed in critically ill infants andchildren. Pediatric Critical Care Medicine. 2006;7(5):413

68. Smith AM, Malone JA. Preventing pressure ulcers ininstitutionalized elders: assessing the effects of small,unscheduled shifts in body position.

69. van Nieuwenhoven CA, VandenbrouckeJoore HC, van Schijndel RJ, van dTI, et al.of the semirecumbent position tpneumonia: a randomized study. Critical Care Medicine.2006;34(2):396-402.

70. Gentilello L, Thompson DA, Tonnesen AS, Hernandez D, KapadiaAS, Allen SJ, et al. Effect of a rotating bed on the incidence ofpulmonary complications in critically ill patients. Critical CareMedicine. 1988;16(8):783-6.

71. Summer WR, Curry P, Haponikm EF, Nelson S, Elston R.Continuous mechanical turning of intensive care unit patientsshortens length of stay in some diagnosticof Critical Care. 1989;4:45-53.

72. Young T. The 30 degree tilt position vs the 90 degree lateral andsupine positions in reducing the incidence of non

KCE Report 193S

pressure ulcers. International Journal of Nursing Studies.

Defloor T, De BD, Grypdonck MH. The effect of variouscombinations of turning and pressure reducing devices on theincidence of pressure ulcers. International Journal of Nursing

Moore Z, Cowman S, Conroy RM. A randomised controlled clinicaltrial of repositioning, using the 30degrees tilt, for the prevention of

f Clinical Nursing. 2011;20(17-18):2633-

Vanderwee K, Grypdonck MHF, De Bacquer D, Defloor T.Effectiveness of turning with unequal time intervals on theincidence of pressure ulcer lesions. Journal of advanced nursing.

LD, LaBrecque MA, Shih MC, Curley MAQ. Pronepositioning can be safely performed in critically ill infants andchildren. Pediatric Critical Care Medicine. 2006;7(5):413-22.

Smith AM, Malone JA. Preventing pressure ulcers inssessing the effects of small,

unscheduled shifts in body position. Decubitus. 1990;3(4):20-4.

van Nieuwenhoven CA, Vandenbroucke-Grauls C, van Tiel FH,Joore HC, van Schijndel RJ, van dTI, et al. Feasibility and effectsof the semirecumbent position to prevent ventilator-associatedpneumonia: a randomized study. Critical Care Medicine.

Gentilello L, Thompson DA, Tonnesen AS, Hernandez D, KapadiaAS, Allen SJ, et al. Effect of a rotating bed on the incidence of

tions in critically ill patients. Critical Care6.

Summer WR, Curry P, Haponikm EF, Nelson S, Elston R.Continuous mechanical turning of intensive care unit patientsshortens length of stay in some diagnostic-related groups. Journal

53.

Young T. The 30 degree tilt position vs the 90 degree lateral andsupine positions in reducing the incidence of non-blanching

KCE Report 193S

erythema in a hospital inpatient population: a randomisedcontrolled trial. Journal of tissue viability. 2004;14(3):88, 90, 290, 6.

73. McInnes E, Dumville JC, Jammali-Blasi A, BellSupport surfaces for treating pressure ulcers. Cochrane Databaseof Systematic Reviews. 2011;Issue 12:CD009490.

74. Cadue JF, Karolewicz S, Tardy C, Barrault C, Robert R, Pourrat O.[Prevention of heel pressure sores with a foam bodydevice. A randomized controlled trial in a medical intensive careunit]. Presse Medicale. 2008;37(1 Pt 1):30

75. Ewing MR, Garrow C, Presley TA, Ashley Cexperiences in the use of sheep skins as an aid in nursing.Australian Nurses Journal. 1964:215-9.

76. Gilcreast DM, Warren JB, Yoder LH, Clark JJ, Wilson JA, MaysMZ. Research comparing three heel ulcerJournal of Wound Ostomy and Continence Nursing.2005;32(2):112-20.

77. Tymec AC, Pieper B, Vollman K. A comparison of two pressurerelieving devices on the prevention of heel pressure ulcers.Advances in Wound Care. 1997;10(1):39-

78. Economides NG, Skoutakis VA, Carter CA, Smith VH. Evaluationof the effectiveness of two support surfaces followingmyocutaneous flap surgery. Advances in Wound Care.1995;8(1):49-53.

79. Brienza D, Kelsey S, Karg P, Allegretti A, Olson M, Schmeler M, etal. A randomized clinical trial on preventing pressure ulcers withwheelchair seat cushions. Journal of the American GeriatricsSociety. 2010;58(12):2308-14.

80. Demarre L, Beeckman D, Vanderwee K, Defloor T, Grypdonck M,Verhaeghe S. Multi-stage versus single-stage inflation anddeflation cycle for alternating low pressure air mattresses toprevent pressure ulcers in hospitalised patients: A randomisedcontrolled clinical trial. International Journal of Nursing Studies.2012;49(4):416-26.


erythema in a hospital inpatient population: a randomisedissue viability. 2004;14(3):88, 90, 2-88,

Blasi A, Bell-Syer Sally EM.Support surfaces for treating pressure ulcers. Cochrane Databaseof Systematic Reviews. 2011;Issue 12:CD009490.

rdy C, Barrault C, Robert R, Pourrat O.[Prevention of heel pressure sores with a foam body-supportdevice. A randomized controlled trial in a medical intensive careunit]. Presse Medicale. 2008;37(1 Pt 1):30-6.

Ewing MR, Garrow C, Presley TA, Ashley C, Kisella NM. Furtherexperiences in the use of sheep skins as an aid in nursing.

Gilcreast DM, Warren JB, Yoder LH, Clark JJ, Wilson JA, MaysMZ. Research comparing three heel ulcer-prevention devices.

Wound Ostomy and Continence Nursing.

Tymec AC, Pieper B, Vollman K. A comparison of two pressure-relieving devices on the prevention of heel pressure ulcers.

-44.

Carter CA, Smith VH. Evaluationof the effectiveness of two support surfaces followingmyocutaneous flap surgery. Advances in Wound Care.

Brienza D, Kelsey S, Karg P, Allegretti A, Olson M, Schmeler M, etal on preventing pressure ulcers with

wheelchair seat cushions. Journal of the American Geriatrics

Demarre L, Beeckman D, Vanderwee K, Defloor T, Grypdonck M,stage inflation and

ation cycle for alternating low pressure air mattresses toprevent pressure ulcers in hospitalised patients: A randomised-controlled clinical trial. International Journal of Nursing Studies.

81. Grisell M, Place HM. Face tissue pressurecomparison of three face pillows while in the prone position forspinal surgery. Spine. 2008;33(26):2938

82. Malbrain M, Hendriks B, Wijnands P, Denie D, Jans A,Vanpellicom J, et al. A pilot randomised controlled trial comparinreactive air and active alternating pressure mattresses in theprevention and treatment of pressure ulcers among medical ICUpatients. Journal of tissue viability. 2010;19(1):7

83. van Leen M, Hovius S, Neyens J, Halfens R, Schols J. Pressurerelief, cold foam or static air? A single center, prospective,controlled randomized clinical trial in a Dutch nursing home.Journal of tissue viability. 2011;20(1):30

84. Bennett RG, Baran PJ, DeVone LV, Bacetti H, Kristo B, TaybackM, et al. Low airloss hydrotherapy versus standard care forincontinent hospitalized patients. Journal of the AmericanGeriatrics Society. 1998;46(5):569

85. Aronovitch SA, Wilber M, Slezak S, Martin T, Utter D. Acomparative study of an alternating air mattress for the preventof pressure ulcers in surgical patients. Ostomy/WoundManagement. 1999;45(3):34

86. Cavicchioli A, Carella G. Clinical effectiveness of a lowhigh-tech pressure-redistributing mattress. Journal of Wound Care.2007;16(7):285-9.

87. Cobb GA, Yoder LH, Warren JB. Pressure ulcers: patientoutcomes on a KinAir bed or EHOB waffle mattress. Bethesda,Maryland, USA: TriService Nursing Research Program (TSNRP);1997. Available from:http://www.ntis.gov/search/product.aspx?ABBR=PB2002107655

88. Collier ME. Pressure-reducing mattresses. Journal of Wound Care.1996;5(5):207-11.

89. Conine TA, Daechsel D, Lau MS.Silicore overlays in preventing decuJournal of Rehabilitation Research. 1990;13(1):57

90. Conine TA, Daechsel D, Hershler C. Pressure sore prophylaxis inelderly patients using slab foam or customised contoured foam

523

Grisell M, Place HM. Face tissue pressure in prone positioning: acomparison of three face pillows while in the prone position for

Spine. 2008;33(26):2938-41.

Malbrain M, Hendriks B, Wijnands P, Denie D, Jans A,A pilot randomised controlled trial comparing

reactive air and active alternating pressure mattresses in theprevention and treatment of pressure ulcers among medical ICUpatients. Journal of tissue viability. 2010;19(1):7-15.

van Leen M, Hovius S, Neyens J, Halfens R, Schols J. Pressurecold foam or static air? A single center, prospective,

controlled randomized clinical trial in a Dutch nursing home.Journal of tissue viability. 2011;20(1):30-4.

Bennett RG, Baran PJ, DeVone LV, Bacetti H, Kristo B, Taybackotherapy versus standard care for

incontinent hospitalized patients. Journal of the AmericanGeriatrics Society. 1998;46(5):569-76.

Aronovitch SA, Wilber M, Slezak S, Martin T, Utter D. Acomparative study of an alternating air mattress for the prevent ionof pressure ulcers in surgical patients. Ostomy/WoundManagement. 1999;45(3):34-.

Cavicchioli A, Carella G. Clinical effectiveness of a low-tech versusredistributing mattress. Journal of Wound Care.

, Yoder LH, Warren JB. Pressure ulcers: patientoutcomes on a KinAir bed or EHOB waffle mattress. Bethesda,Maryland, USA: TriService Nursing Research Program (TSNRP);

http://www.ntis.gov/search/product.aspx?ABBR=PB2002107655

reducing mattresses. Journal of Wound Care.

Conine TA, Daechsel D, Lau MS. The role of alternating air andSilicore overlays in preventing decubitus ulcers. InternationalJournal of Rehabilitation Research. 1990;13(1):57-65.

Conine TA, Daechsel D, Hershler C. Pressure sore prophylaxis inelderly patients using slab foam or customised contoured foam

524

wheelchair cushions. Occupational Therapy Jo1993;13(2):101-16.

91. Conine TA, Hershler C, Daechsel D, Peel C, Pearson A. Pressureulcer prophylaxis in elderly patients using polyurethane foam orJay wheelchair cushions. International Journal of RehabilitationResearch. 1994;17(2):123-37.

92. Cooper PJ, Gray DG, Mollison J. A randomised controlled trial oftwo pressure-reducing surfaces. Journal of Wound Care.1998;7(8):374-6.

93. Daechsel D, Conine TA. Special mattresses: effectiveness inpreventing decubitus ulcers in chronic neurArchives of Physical Medicine and Rehabilitation. 1985;66(4):2468.

94. Exton-Smith AN, Overstall PW, Wedgwood J, Wallace G. Use ofthe 'air wave system' to prevent pressure sores in hospital. Lancet.1982;1(8284):1288-90.

95. Feuchtinger J, de BR, Dassen T, Halfens R. A 4viscoelastic foam pad on the operating room table to preventpressure ulcer during cardiac surgery. Journal of Clinical Nursing.2006;15(2):162-7.

96. Geyer MJ, Brienza DM, Karg P, Trefler E, Kelsey S. A rcontrol trial to evaluate pressure-reducing seat cushions for elderlywheelchair users. Advances in skin and wound care.2001;14(3):120-9.

97. Goldstone LA, Norris M, O'Reilly M, White J. A clinical trial of abead bed system for the prevention of pressure sores in elderlyorthopaedic patients. Journal of advanced nursing. 1982;7(6):5458.

98. Gray DG, Campbell M. A randomized clinical trial of two types offoam mattresses. Journal of tissue viability. 1994;4(4):128

99. Gunningberg L, Lindholm C, Carlsson M, Sjoden PO. Effect ofvisco-elastic foam mattresses on the development of pressureulcers in patients with hip fractures. Journal of Wound Care.2000;9(10):455-60.


wheelchair cushions. Occupational Therapy Journal of Research.

Conine TA, Hershler C, Daechsel D, Peel C, Pearson A. Pressureulcer prophylaxis in elderly patients using polyurethane foam orJay wheelchair cushions. International Journal of Rehabilitation

Cooper PJ, Gray DG, Mollison J. A randomised controlled trial ofreducing surfaces. Journal of Wound Care.

Daechsel D, Conine TA. Special mattresses: effectiveness inpreventing decubitus ulcers in chronic neurologic patients.Archives of Physical Medicine and Rehabilitation. 1985;66(4):246-

Smith AN, Overstall PW, Wedgwood J, Wallace G. Use ofthe 'air wave system' to prevent pressure sores in hospital. Lancet.

J, de BR, Dassen T, Halfens R. A 4-cm thermoactiveviscoelastic foam pad on the operating room table to preventpressure ulcer during cardiac surgery. Journal of Clinical Nursing.

Geyer MJ, Brienza DM, Karg P, Trefler E, Kelsey S. A randomizedreducing seat cushions for elderly

wheelchair users. Advances in skin and wound care.

Goldstone LA, Norris M, O'Reilly M, White J. A clinical trial of af pressure sores in elderly

orthopaedic patients. Journal of advanced nursing. 1982;7(6):545-

Gray DG, Campbell M. A randomized clinical trial of two types offoam mattresses. Journal of tissue viability. 1994;4(4):128-32.

C, Carlsson M, Sjoden PO. Effect ofelastic foam mattresses on the development of pressure

ulcers in patients with hip fractures. Journal of Wound Care.

100. Hampton S. Evaluation of the new Cairwave Therapy System inone hospital trust. British Journal of Nursing. 1997;6(3):167

101. Hofman A, Geelkerken RH, Wille J, Hamming JJ, Hermans J,Breslau PJ. Pressure sores and pressurecontrolled clinical trial. Lancet. 1994;343(8897):568

102. Inman KJ, Sibbald WJ, Rutledge FS. Clinical utility and costeffectiveness of an air suspension bed in the prevention ofpressure ulcers. Journal of the American Medical Association.1993;269:1139-43.

103. Jolley DJ, Wright R, McGowan S, Hickey MB, Campbell DA,Sinclair RD, et al. Preventing pressure ulcers with the AustralianMedical Sheepskin: an openMedical Journal of Australia. 2004;180(7):324

104. Kemp MG, Kopanke D, Tordecilla L, Fogg L, Shott S, MatthiesenV, et al. The role of support surfaces and patient attributes inpreventing pressure ulcers in elderly patients. Research in Nursingand Health. 1993;16(2):89-96.

105. Keogh A, Dealey C. Profiling beds versus standard hospital beds:effects on pressure ulcer incidence outcomCare. 2001;10(2):15-9.

106. Laurent S. Effectiveness of pressure decreasing mattresses incardiovascular surgery patients: a controlled clinical trial. 3rdEuropean Conference for Nurse Managers; Brussels, Belgium.1998.

107. Lazzara DJ, Buschmann MT. Prevention of pressure ulcers inelderly nursing home residents: are special support surfaces theanswer? Decubitus. 1991;4(4):42

108. Lim R, Sirett R, Conine TA, Daechsel D. Clinical trial of foamcushions in the prevention of deJournal of Rehabilitation Research and Development.1988;25(2):19-26.

109. McGowan S, Montgomery K, Jolley D, Wright R. The role ofsheepskins in preventing pressure ulcers in elderly orthopaedicpatients. Primary Intention. 2000;8(4):1

KCE Report 193S

Hampton S. Evaluation of the new Cairwave Therapy System intrust. British Journal of Nursing. 1997;6(3):167-70.

Hofman A, Geelkerken RH, Wille J, Hamming JJ, Hermans J,Breslau PJ. Pressure sores and pressure-decreasing mattresses:controlled clinical trial. Lancet. 1994;343(8897):568-71.

ald WJ, Rutledge FS. Clinical utility and cost-effectiveness of an air suspension bed in the prevention ofpressure ulcers. Journal of the American Medical Association.

Jolley DJ, Wright R, McGowan S, Hickey MB, Campbell DA,RD, et al. Preventing pressure ulcers with the Australian

Medical Sheepskin: an open-label randomised controlled trial.Medical Journal of Australia. 2004;180(7):324-7.

Kemp MG, Kopanke D, Tordecilla L, Fogg L, Shott S, Matthiesenof support surfaces and patient attributes in

preventing pressure ulcers in elderly patients. Research in Nursing96.

Keogh A, Dealey C. Profiling beds versus standard hospital beds:effects on pressure ulcer incidence outcomes. Journal of Wound

Laurent S. Effectiveness of pressure decreasing mattresses incardiovascular surgery patients: a controlled clinical trial. 3rdEuropean Conference for Nurse Managers; Brussels, Belgium.

J, Buschmann MT. Prevention of pressure ulcers inelderly nursing home residents: are special support surfaces theanswer? Decubitus. 1991;4(4):42-4, 6, 8.

Lim R, Sirett R, Conine TA, Daechsel D. Clinical trial of foamcushions in the prevention of decubitis ulcers in elderly patients.Journal of Rehabilitation Research and Development.

McGowan S, Montgomery K, Jolley D, Wright R. The role ofsheepskins in preventing pressure ulcers in elderly orthopaedic

ion. 2000;8(4):1-8.

KCE Report 193S

110. Mistiaen P, Achterberg W, Ament A, Halfens R, Huizinga J,Montgomery K, et al. The effectiveness of the Australian MedicalSheepskin for the prevention of pressure ulcers in somatic nursinghome patients: A prospective multicentertrial (ISRCTN17553857). Wound Repair and Regeneration.2010;18(6):572-9.

111. Nixon J, McElvenny D, Mason S, Brown J, Bond S. A sequentialrandomised controlled trial comparing a dry viscopad and standard operating table mattress in the prevention ofpost-operative pressure sores. International Journal of NursingStudies. 1998;35(4):193-203.

112. Nixon J, Cranny G, Iglesias C, Nelson EA, Hawkins K, Phillips A,et al. Randomised, controlled trial of alternating pressuremattresses compared with alternating pressure overlays for theprevention of pressure ulcers: PRESSURE (pressure relievingsupport surfaces) trial. BMJ. 2006;332(7555):1413.

113. Price P, Bale S, Newcombe R, Harding K. Challenging thepressure sore paradigm. Journal of Wound Care. 1999;8(4):18790.

114. Russell L, Reynolds TM, Carr J, Evans A, Holmes M. Randomisedcontrolled trial of two pressure-relieving systems. Journal ofWound Care. 2000;9(2):52-5.

115. Russell JA, Lichtenstein SL. Randomized controlldetermine the safety and efficacy of a multimattress system in the prevention of pressure ulcers in patientsundergoing cardiovascular surgery. Ostomy/Wound Management.2000;46(2):46-5.

116. Sanada H, Sugama J, Matsui Y, Konya C, Kitagawa A, Okuwa M,et al. Randomised controlled trial to evaluate a new doubleair-cell overlay for elderly patients requiring head elevation.Journal of tissue viability. 2003;13(3):112-

117. Santy JE, Butler MK, Whyman JD. A comof hospital mattress to determine which most effectively reducesthe incidence of pressure sores in elderly patients with hipfractures in a District General Hospital. Report to Northern &Yorkshire Regional Health Authority. 1994.


Mistiaen P, Achterberg W, Ament A, Halfens R, Huizinga J,Montgomery K, et al. The effectiveness of the Australian MedicalSheepskin for the prevention of pressure ulcers in somatic nursinghome patients: A prospective multicenter randomized-controlledtrial (ISRCTN17553857). Wound Repair and Regeneration.

Nixon J, McElvenny D, Mason S, Brown J, Bond S. A sequentialrandomised controlled trial comparing a dry visco-elastic polymer

ble mattress in the prevention ofoperative pressure sores. International Journal of Nursing

Nixon J, Cranny G, Iglesias C, Nelson EA, Hawkins K, Phillips A,et al. Randomised, controlled trial of alternating pressuremattresses compared with alternating pressure overlays for theprevention of pressure ulcers: PRESSURE (pressure relievingsupport surfaces) trial. BMJ. 2006;332(7555):1413.

Price P, Bale S, Newcombe R, Harding K. Challenging thegm. Journal of Wound Care. 1999;8(4):187-

Russell L, Reynolds TM, Carr J, Evans A, Holmes M. Randomisedrelieving systems. Journal of

Russell JA, Lichtenstein SL. Randomized controlled trial todetermine the safety and efficacy of a multi-cell pulsating dynamicmattress system in the prevention of pressure ulcers in patientsundergoing cardiovascular surgery. Ostomy/Wound Management.

Konya C, Kitagawa A, Okuwa M,et al. Randomised controlled trial to evaluate a new double-layer

cell overlay for elderly patients requiring head elevation.-4, 6, 8.

Santy JE, Butler MK, Whyman JD. A comparison study of 6 typesof hospital mattress to determine which most effectively reducesthe incidence of pressure sores in elderly patients with hipfractures in a District General Hospital. Report to Northern &Yorkshire Regional Health Authority. 1994.

118. Sideranko S, Quinn A, Burns K, Froman RD. Effects of positionand mattress overlay on sacral and heel pressures in a clinicalpopulation. Research in Nursing and Health. 1992;15(4):245

119. Stapleton M. Preventing pressure soresproducts. Geriatric Nursing. 1986;6(2):23

120. Takala J, Varmavuo S, Soppi E. Prevention of pressure sores inacute respiratory failure: a randomised controlled trial. ClinicalIntensive Care. 1996;7(5):228

121. Taylor L. Evaluating the Pegapproach. British Journal of Nursing. 1999;8(12):771

122. Theaker C, Kuper M, Soni N. Pressure ulcer prevention inintensive care – a randomised control trial of two pressurerelieving devices. Anaesthesia. 2005;60(4):395

123. Vyhlidal SK, Moxness D, Bosak KS, Van Meter FG, Bergstrom N.Mattress replacement or foam overlay? A prospective study on theincidence of pressure ulcers. Applied Nursing Research.1997;10(3):111-20.

124. Whitney JD, Fellows BJ, Larson E. Do mattrdifference? Journal of Gerontological Nursing. 1984;10(9):20

125. Gebhardt KS, Bliss MR, Winwright PL, Thomas J. Pressurerelieving supports in an ICU. Journal of Wound Care.1996;5(3):116-21.

126. Vanderwee K, Grypdonck MH, Defloor T. Effealternating pressure air mattress for the prevention of pressureulcers. Age and Ageing. 2005;34(3):261

127. Gray D, Smith M. A randomized controlled trial of two pressurereducing foam mattresses. European Wound ManagementAssociation Conference; Harrogate, UK. 1998.

128. National Pressure Ulcer Advisory P. Terms and definitions relatedto support surfaces (Ver. 01/29/2007); 2007.

129. Andersen KE, Jensen O, Kvorning SA, Bach E. Decubitusprophylaxis: a prospective trial on the efficiencpressure air-mattresses and waterVenereologica. 1983;63(3):227

525

Sideranko S, Quinn A, Burns K, Froman RD. Effects of positionand mattress overlay on sacral and heel pressures in a clinicalpopulation. Research in Nursing and Health. 1992;15(4):245-51.

Stapleton M. Preventing pressure sores--an evaluation of threeproducts. Geriatric Nursing. 1986;6(2):23-5.

Takala J, Varmavuo S, Soppi E. Prevention of pressure sores inacute respiratory failure: a randomised controlled trial. ClinicalIntensive Care. 1996;7(5):228-35.

Taylor L. Evaluating the Pegasus Trinova: a data hierarchyapproach. British Journal of Nursing. 1999;8(12):771-8.

Theaker C, Kuper M, Soni N. Pressure ulcer prevention ina randomised control trial of two pressure-

relieving devices. Anaesthesia. 2005;60(4):395-9.

Vyhlidal SK, Moxness D, Bosak KS, Van Meter FG, Bergstrom N.Mattress replacement or foam overlay? A prospective study on theincidence of pressure ulcers. Applied Nursing Research.

Whitney JD, Fellows BJ, Larson E. Do mattresses make adifference? Journal of Gerontological Nursing. 1984;10(9):20-5.

Gebhardt KS, Bliss MR, Winwright PL, Thomas J. Pressure-relieving supports in an ICU. Journal of Wound Care.

Vanderwee K, Grypdonck MH, Defloor T. Effectiveness of analternating pressure air mattress for the prevention of pressureulcers. Age and Ageing. 2005;34(3):261-7.

Gray D, Smith M. A randomized controlled trial of two pressure-reducing foam mattresses. European Wound Management

onference; Harrogate, UK. 1998.

National Pressure Ulcer Advisory P. Terms and definitions relatedto support surfaces (Ver. 01/29/2007); 2007.

Andersen KE, Jensen O, Kvorning SA, Bach E. Decubitusprophylaxis: a prospective trial on the efficiency of alternating-

mattresses and water-mattresses. Acta Dermato-Venereologica. 1983;63(3):227-30.

526

130. Nixon J, Nelson EA, Cranny G, Iglesias CP, Hawkins K, CullumNA, et al. Pressure relieving support surfaces: a randomisedevaluation. Health Technology Assessment. 2006;10(22):1

131. Russell L, Reynolds TM, Towns A, Worth W, Greenman A, TurnerR. Randomized comparison trial of the RIK and the Nimbus 3mattresses. British Journal of Nursing. 2003;12(4):254, 6

132. Schultz A, Bien M, Dumond K, Brown K, Myers A. Etiology andincidence of pressure ulcers in surgical patients. AORN Journal.1999;70(3):434, 7-40, 43-49.

133. Takala JVSSE. Prevention of pressure sores in acute respiratoryfailure: a randomised controlled trial. Clinical Inten1996;7(5):228-35.

134. Donnelly J, Winder J, Kernohan WG, Stevenson M. An RCT todetermine the effect of a heel elevation device in pressure ulcerprevention post-hip fracture. Journal of Wound Care.2011;20(7):309-8.

135. Torra IBJE, Rueda LJ, Camanes G, Herrero NE, Blanco BJ,Balleste TJ, et al. Preventing pressure ulcers on the heel: aCanadian cost study. Dermatology Nursing. 2009;21(5):268

136. Langer G, Schloemer G, Knerr A, Kuss O, Behrens J. Nutritionalinterventions for preventing and treating pressure ulcers. Cochranedatabase of systematic reviews (Online). 2003;Issue 4:CD003216.

137. Dennis MS, Lewis SC, Warlow C. Routine oral nutritionalsupplementation for stroke patients in hospital (FOOD): amulticentre randomised controlled trial. Lancet.2005;365(755):763.

138. Craig LD, Nicholson S, Silverstone FA, Kennedy RD. Use of areduced-carbohydrate,modified-fat enteral formula for improvingmetabolic control and clinical outcomes in longwith type 2 diabetes: results of a pilot trial. Nutrition.1998;14(6):529-34.

139. Theilla M, Singer P, Cohen J, Dekeyser F. A diet enriched ineicosapentanoic acid, gamma-linolenic acid and antioxidants in theprevention of new pressure ulcer formation in critically ill patients


Nixon J, Nelson EA, Cranny G, Iglesias CP, Hawkins K, CullumNA, et al. Pressure relieving support surfaces: a randomised

Technology Assessment. 2006;10(22):1-163.

Russell L, Reynolds TM, Towns A, Worth W, Greenman A, TurnerR. Randomized comparison trial of the RIK and the Nimbus 3mattresses. British Journal of Nursing. 2003;12(4):254, 6-, 9.

mond K, Brown K, Myers A. Etiology andincidence of pressure ulcers in surgical patients. AORN Journal.

Takala JVSSE. Prevention of pressure sores in acute respiratoryfailure: a randomised controlled trial. Clinical Intensive Care.

Donnelly J, Winder J, Kernohan WG, Stevenson M. An RCT todetermine the effect of a heel elevation device in pressure ulcer

hip fracture. Journal of Wound Care.

Camanes G, Herrero NE, Blanco BJ,Balleste TJ, et al. Preventing pressure ulcers on the heel: aCanadian cost study. Dermatology Nursing. 2009;21(5):268-72.

Langer G, Schloemer G, Knerr A, Kuss O, Behrens J. Nutritionald treating pressure ulcers. Cochrane

database of systematic reviews (Online). 2003;Issue 4:CD003216.

Dennis MS, Lewis SC, Warlow C. Routine oral nutritionalsupplementation for stroke patients in hospital (FOOD): a

ial. Lancet.

Craig LD, Nicholson S, Silverstone FA, Kennedy RD. Use of afat enteral formula for improving

metabolic control and clinical outcomes in long-term care residentsts of a pilot trial. Nutrition.

Theilla M, Singer P, Cohen J, Dekeyser F. A diet enriched inlinolenic acid and antioxidants in the

prevention of new pressure ulcer formation in critically ill patients

with acute lung injury: A randomized, prospective, controlled study.Clinical Nutrition. 2007;26(6):752

140. Olofsson B, Stenvall M, Lundstrom M, Svensson O, Gustafson Y.Malnutrition in hip fracture patients: An intervention study.of Clinical Nursing. 2007;16(11):2027

141. Bourdel-Marchasson I, Barateau M, Rondeau V, DequaeMerchadou L, Salles-Montaudon N, Emeriau JP, et al.center trial of the effects of oral nutritional supplementation incritically ill older inpatients. GAGE Group.Geriatrique d'Evaluation. Nutrition. 2000;16(1):1

142. Delmi M, Rapin CH, Bengoa JM, Delmas PD, Vasey H, BonjourJP. Dietary supplementation in elderly patients with fractured neckof the femur. Lancet. 1990;335(8696):1013

143. Houwing RH, Rozendaal M, WoutersBuskens E, Haalboom JR. A randomised, doubleassessment of the effect of nutritional supplementation on theprevention of pressure ulcers in hipNutrition. 2003;22(4):401-5.

144. Hartgrink HH, Wille J, Konig P, Hermans J, Breslau PJ.sores and tube feeding in patients with a fracture of the hip: arandomized clinical trial. Clinical Nutrition. 1998;17(6):287

145. Ncgc. Venous thromboembolic diseases: the manvenous thromboembolic diseases and the role of thrombophiliatesting, Clinical Guideline, Methods, evidence andrecommendations. London: Royal College of Physicians; 2009.

146. Consortium for Spinal Cord M. Pressure Ulcer Prevention andTreatment Following Spinal Cord Injury: A Clinical PracticeGuideline for Health-Care Professionals. Washington DC:Paralyzed Veterans of America; 2000.

147. National Institute for H, Clinical E. Guidelines on pressure ulcerrisk management and preventionInstitute for Clinical Excellence; 2001.

148. National Institute for H, Clinical E. Pressure ulcer prevention:pressure ulcer risk assessment and prevention, including the useof pressure-relieving devices (beds, mattresses and over

KCE Report 193S

th acute lung injury: A randomized, prospective, controlled study.Clinical Nutrition. 2007;26(6):752-7.

Olofsson B, Stenvall M, Lundstrom M, Svensson O, Gustafson Y.Malnutrition in hip fracture patients: An intervention study. Journal

rsing. 2007;16(11):2027-38.

Marchasson I, Barateau M, Rondeau V, Dequae-Montaudon N, Emeriau JP, et al. A multi-

center trial of the effects of oral nutritional supplementation incritically ill older inpatients. GAGE Group. Groupe AquitainGeriatrique d'Evaluation. Nutrition. 2000;16(1):1-5.

Delmi M, Rapin CH, Bengoa JM, Delmas PD, Vasey H, BonjourJP. Dietary supplementation in elderly patients with fractured neck

Lancet. 1990;335(8696):1013-6.

ng RH, Rozendaal M, Wouters-Wesseling W, Beulens JWJ,A randomised, double-blind

assessment of the effect of nutritional supplementation on theprevention of pressure ulcers in hip-fracture patients. Clinical

5.

Hartgrink HH, Wille J, Konig P, Hermans J, Breslau PJ. Pressuresores and tube feeding in patients with a fracture of the hip: arandomized clinical trial. Clinical Nutrition. 1998;17(6):287-92.

Ncgc. Venous thromboembolic diseases: the management ofvenous thromboembolic diseases and the role of thrombophiliatesting, Clinical Guideline, Methods, evidence andrecommendations. London: Royal College of Physicians; 2009.

Consortium for Spinal Cord M. Pressure Ulcer Prevention andt Following Spinal Cord Injury: A Clinical Practice

Care Professionals. Washington DC:Paralyzed Veterans of America; 2000.

National Institute for H, Clinical E. Guidelines on pressure ulcerrisk management and prevention – guideline. London: NationalInstitute for Clinical Excellence; 2001.

National Institute for H, Clinical E. Pressure ulcer prevention:pressure ulcer risk assessment and prevention, including the use

relieving devices (beds, mattresses and overlays) for

KCE Report 193S

the prevention of pressure ulcers in primary and secondary care.London: National Institute for Clinical Excellence; 2003.

149. National Institute for H, Clinical E. Pressure ulcers: Themanagement of pressure ulcers in primary and secondary careLondon: National Institute for Clinical Excellence; 2005.

150. Australian Wound Management A. Clinical Practice Guidelines forthe Prediction and Prevention of Pressure Ulcers.WA: Cambridge Publishing; 2001.

151. Kwaliteitsinstituut voor de G. Decubitus (2de ed.). Alphen aan denRijn: Van Zuiden Communications B.V; 2002.

152. Registered Nurses Association of O. Risk Assessment &Prevention of Pressure Ulcers. Ontario: RNAO; 2005.

153. Defloor T, Herremans A, Grypdonck M, De SchuijmSchoonhoven L, et al. Belgische richtlijn decubituspreventie 2004.Brussel: Belgisch Federaal Ministerie van Sociale Zaken,Volksgezondheid en Leefmilieu; 2004.

154. Whitney J, Phillips L, Aslam R, Barbul A, Gottrup F, Gould L, et al.Guidelines for the treatment of pressure ulcers. Wound Repair andRegeneration. 2006;14(6):663-79.

155. Stechmiller JK, Cowan L, Whitney JD, Phillips L, Aslam R, BarbulA, et al. Guidelines for the prevention of pressure ulcers. WoundRepair and Regeneration. 2008;16(2):151

156. Stockton L, Gebhardt KS, Clark M. Seating and pressure ulcers:clinical practice guideline. Journal of tissue viability. 2009;18(4):98108.

157. Association for the Advancement of Wound C. Association for theAdvancement of Wound Care guideline of pressure ulcerguidelines. Malvern (PA): Association for the Advancement ofWound Care (AAWC); 2010.

158. Deutsches Netzwerk f r Qualit„tsentwicklung in der P.Expertenstandard Dekubitusprophylaxe in der Pflege.

159. Verpleegkundigen, Verzorgenden N. Landelijke multidisciplinairerichtlijn Decubitus preventie en behandeling.


the prevention of pressure ulcers in primary and secondary care.London: National Institute for Clinical Excellence; 2003.

National Institute for H, Clinical E. Pressure ulcers: Themanagement of pressure ulcers in primary and secondary care.London: National Institute for Clinical Excellence; 2005.

Australian Wound Management A. Clinical Practice Guidelines forthe Prediction and Prevention of Pressure Ulcers. West Leederville

oor de G. Decubitus (2de ed.). Alphen aan denRijn: Van Zuiden Communications B.V; 2002.

Registered Nurses Association of O. Risk Assessment &Ontario: RNAO; 2005.

Defloor T, Herremans A, Grypdonck M, De Schuijmer J, Paquay L,Belgische richtlijn decubituspreventie 2004.

Brussel: Belgisch Federaal Ministerie van Sociale Zaken,

Whitney J, Phillips L, Aslam R, Barbul A, Gottrup F, Gould L, et al.delines for the treatment of pressure ulcers. Wound Repair and

Stechmiller JK, Cowan L, Whitney JD, Phillips L, Aslam R, BarbulA, et al. Guidelines for the prevention of pressure ulcers. Wound

2008;16(2):151-68.

Stockton L, Gebhardt KS, Clark M. Seating and pressure ulcers:clinical practice guideline. Journal of tissue viability. 2009;18(4):98-

Association for the Advancement of Wound C. Association for theare guideline of pressure ulcer

guidelines. Malvern (PA): Association for the Advancement of

r Qualit„tsentwicklung in der P.Expertenstandard Dekubitusprophylaxe in der Pflege. 2010.

igen, Verzorgenden N. Landelijke multidisciplinairerichtlijn Decubitus preventie en behandeling. 2011.

160. Wound OaCNS. Guideline for prevention and management ofpressure ulcers. Mount Laurel (NJ): Wound,Ostomy,andContinence Nurses Society (WOCN); 20

161. European Pressure Ulcer Advisory PPUAP. Prevention andtreatment of pressure ulcers: Clinical Practice Guideline.Washington (DC): 2009.

527

Wound OaCNS. Guideline for prevention and management ofpressure ulcers. Mount Laurel (NJ): Wound,Ostomy,andContinence Nurses Society (WOCN); 2010.

European Pressure Ulcer Advisory PPUAP. Prevention andtreatment of pressure ulcers: Clinical Practice Guideline.

· COLOPHON Title: A National Guideline for the prevention of pressure ulcers Authors: Dimitri (KULeuven), Geert Vanwalleghem ((CNC vzw), Hilde Heyman (WCS), Christian Thyse ((KCE),

Documents