Abstract Aim and Objectives . To critically review the evidence relating to the management of agitation within the Adult Critical Care Unit (ACCU) environment, identify the risks and benefits of current management strategies. Background. Admission to an ACCU can be traumatic and potentially life altering for the patient. Patient agitation is common in ACCUs and is associated with the potential for harm. Despite the inherent safety risks, there is a paucity of evidence-based guidance underpinning the care of agitation in adult patients with critical illness. Study Design Integrative review and narrative synthesis Method A systematic procedure for searching and selecting the literature was followed and applied to databases including CINAHL, British Nursing Index, Cochrane Library, Proquest, Ovid including Embase and Medline.Selected manuscripts were analysed using a structured narrative review approach. Results 208 papers were identified and following a systematic de-selection process 24 original articles were included in the review. It was identified that agitation in the setting of ACCU is associated with high-risk events such as unplanned removal of life-supporting devices. There were consistent links to sepsis, previous high alcohol intake and certain medications, which may increase the development of agitation. 1
74
Embed
Kress et al (2000) - University of Manchester · Web viewWrist and ankle restraints being most commonly used (84.7%). Common reason for application is: Maintenance of medical devices
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Abstract
Aim and Objectives. To critically review the evidence relating to the management
of agitation within the Adult Critical Care Unit (ACCU) environment, identify the risks
and benefits of current management strategies.
Background. Admission to an ACCU can be traumatic and potentially life altering for
the patient. Patient agitation is common in ACCUs and is associated with the
potential for harm. Despite the inherent safety risks, there is a paucity of evidence-
based guidance underpinning the care of agitation in adult patients with critical
illness.
Study Design Integrative review and narrative synthesis
Method A systematic procedure for searching and selecting the literature was
followed and applied to databases including CINAHL, British Nursing Index,
Cochrane Library, Proquest, Ovid including Embase and Medline.Selected
manuscripts were analysed using a structured narrative review approach.
Results 208 papers were identified and following a systematic de-selection process
24 original articles were included in the review. It was identified that agitation in the
setting of ACCU is associated with high-risk events such as unplanned removal of
life-supporting devices. There were consistent links to sepsis, previous high alcohol
intake and certain medications, which may increase the development of agitation.
Prompt assessment and early liberation from mechanical ventilation was a major
contributing factor in the reduction of agitation. Administration of anti-deliriogenic
mediation may reduce the need for physical restraint and improve outcome. There
was repeated uncertainty about the role of physical restraint in developing agitation
and its effective management.
Conclusions
Synthesising the available evidence has unearthed that this is a poorly researched
area, of low quality evidence, in a high-risk care environment There are real
1
dilemmas for clinical teams about the effectiveness of applying physical and/or
pharmacological restraint. The review has highlighted that the risk of self-extubation
increases with the presence of agitation, reinforcing the need for constant
observation and vigilance.
Key words: agitation, intensive care, critical care, systematic narrative review,
nursing
INTRODUCTIONPatient agitation is a conditon of psychomotor disturbance, characterized by
increased motor and psychological activity (Chevrolet & Jolliet, 2007). This definition
of agitation supported by Intensive Care Society (ICS) who add that agitation is
accompanied by disorganised thought and is common among ACCU patient
(Snelson et al., 2014). An agitated patient can exhibit behaviour, such as
restlessness or thrashing, placing themselves and others at risk (Tate et al., 2012b).
The occurrence of agitation in the ACCU is frequently cited as ‘common’ across the
literature, yet research related to agitation is lacking. High risk behaviour stemming
from agitation is also reported in a highly variable way (Tate et al., 2012b). The
presence of patient agitation in adult critical care is associated with the potential for
harm including unplanned removal from organ supportive treatments, higher rates of
nosocomial infection, and increased length of both critical care and hospital stay
(Burk et al., 2014, Whitehouse et al, 2014; Hofsø and Coyer 2007; Birkett et al, 2005;
Micek et al, 2005). To facilitate intrusive and potentially distressing treatment
interventions in critical care, analgesic and sedative drugs are frequently
administered. When withdrawing patient sedation, the process of waking can expose
feelings of anxiety and acute agitation (Chevrolet & Jolliet, 2007).
It is important that the terms agitation and delirium are not confused, agitation is not
delirium and agitation without delirium is common in the ACCU (Whitehouse et al.,
2014). In healthcare literature and practice, the terms agitation, anxiety and delirium
are often used interchangeably; however compared with the definitions for agitation
above, delirium is defined as an acute, reversible organic mental syndrome with
disorders of attention and cognitive function, where psychomotor activity can be
increased or decreased, and the sleep-wake cycle is disordered (Borthwick et al.,
2014, pg 4). There are three sub types of delirium hyperactive, hypoactive and mixed
2
delirium (Van Rompaey et al, 2008) patient with mixed or hyperactive delirium potentially
displaying agitated behaviour. Whereas, the definition of anxiety is a feeling of dread or
fear and or ‘lack of control as a normal or protective response to a perceived threat to
homeostasis’ (Tate, et al, 2012, pg1).
Despite the safety risk, there is a paucity of evidence-based guidance in relation to
the risks and management of patient agitation in Adult Critical Care Units (ACCU). A
number of non-pharmacological and pharmacological strategies are referred to in
guidelines (Barr and Pandharipande., 2013) yet these fail to address how to
appropriately and safely manage a patient. Therefore, this review will focus on the
management strategies required to effectively manage risk behaviors exhibited by a
patient experiencing agitation.
AIMTo provide a critical review of the literature relating to the management of agitation
within ACCUs, and seeks to identify the risks and benefits of current management
strategies.
METHODS Search strategyA systematic search was conducted for original research manuscripts about agitation
and its management in an ACCU. The search terms used and limitations set are
summarised in Table 1. The literature search was performed in May 2017 via the
following databases: CINAHL, British Nursing Index, Cochrane Library, Proquest,
Ovid including Embase and Medline. Kress et al (2000) trial showed ventilator-
weaning time, ACCU and hospital length of stay were significantly lower in the
groups with sedation holds. These findings significantly impacted on change to
sedation practice. In light of this search publication date was limited, from 2005 to
May 2017.
[Please insert Table 1: Summary of key search terms here]
Study selection
For inclusion in this review, studies were required to have a focus on agitation
attitudes and opinions of adult intensive care unit nurses.
http://doi.org/10.1111/nicc.12197
Huang, Y.-T. (2009). Factors leading to self-extubation of endotracheal tubes in the
intensive care unit. Nursing in Critical Care, 14(2). http://doi.org/10.1111/j.1478-
5153.2008.00320.x
Jaber, S., Chanques, G., Altairac, C., Sebbane, M., Vergne, C., Perrigault, P. F., &
Eledjam, J. J. (2005). A prospective study of agitation in a medical-surgical ICU:
Incidence, risk factors, and outcomes. Chest, 128, 2749–2757.
http://doi.org/10.1378/chest.128.4.2749
Kandeel, N. A., & Attia, A. K. (2013). Physical restraints practice in adult intensive
care units in Egypt. Nursing and Health Sciences, 15, 79–85.
http://doi.org/10.1111/nhs.12000
Kress et al (2000) Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. The New England Journal of Medicine 342 (20) 1471- 1477
Langley, G., Schmollgruber, S., & Egan, A. (2011). Restraints in intensive care units--
a mixed method study. Intensive & Critical Care Nursing : The Official Journal of
the British Association of Critical Care Nurses, 27(2), 67–75.
17
http://doi.org/10.1016/j.iccn.2010.12.001
Luk, E., Sneyers, B., Rose, L., Perreault, M. M., Williamson, D. R., Mehta, S., …
Burry, L. (2014). Predictors of physical restraint use in Canadian intensive care
units. Critical Care, 18(R46), R46. http://doi.org/10.1186/cc13789
Martin, B., & Mathisen, L. (2005). Use of physical restraints in adult critical care: A
bicultural study. American Journal of Critical Care, 14, 133–142.
http://doi.org/14/2/133 [pii]
Michaud, C. J., Thomas, W. L., & McAllen, K. J. (2014). Early Pharmacological
Treatment of Delirium May Reduce Physical Restraint Use: A Retrospective
Study. Annals of Pharmacotherapy, 48, 328–334.
http://doi.org/10.1177/1060028013513559
Mion, L. C., Minnick, A. F., Leipzig, R. M., Catrambone, C. D., & Johnson, M. E.
(2007). Patient-initiated device removal in intensive care units: a national
prevalence study. Critical Care Medicine, 35(12), 2714–2720. Retrieved from
(Almeida et al., 2016)Risk Factors forAgitation inCritically Ill Patients
To evaluate theincidence ofagitation in thefirst 7 days afterICU admission,its risk factorsand itsassociation withclinical outcome
Prospective cohortstudy conducted in a single general 18 bedded ICU.Inclusion:Patients >18 yearsPredicated stay of >48 hoursExclusion:Pregnant womenPatient with previouspsychiatric conditionsPatients transferred from another ICUPatient who receivedhaloperidol, dexmedetomidine, riperidone, quetiapineprior to study
In a sample of 113 patients theincidence of agitation in thefirst 7 days was 31.8%
Agitation was more frequentamong patients with:History of smokingSever head injuryAcute neurological diseaseModerate to severe painMechanical ventilation (MV)DeliriumAgitated patent had fewer MV free days
Strengths:
All patient were visitedtwice daily in the first 7days prospectivelyassessing for agitation thus collecting data on riskfactors before agitationoccurrenceThe patient assessmentstrategy was robust
Limitations:
Small sample size in one single centreStudy team not available atthe weekends leading tomissed data and exclusionof potential patientsThe inclusion of many MV patient resulted in limitedassessment of pain anddeliriumAgitation treatment datanot collected
The need to recognise that in addition to delirium there are other independent factors foragitation in ICU.
Good care practices, sedation,analgesia and management ofMV could reduce the incidence of agitation
(Reade et al., 2016)Effect ofDexmedetomidine Added to standardcare on Ventilator-Free Time in patient
To determine the effectiveness ofDexmedetomidine when added tostandard care in patients with
Double blind, placebocontrolled, parallel grouprandomised clinical trialInclusion of all patient >18 years and with thephysicians opinion that
Results based on 39 patientreceiving dexmedetomidine and32 placebo. Dexmedetomidinegroup had a significantlyincreased number ofventilator free hours at 7 days
Strengths:
The study employed adouble blind, multicentre,randomised, placebocontrolled design
The addition ofDexmedetomidine tostandard care in patient withagitated delirium can result inmore ventilator free days
21
with AgitatedDelirium
agitated delirium the degree of agitation was so sever that reduction ofsedation and extubationwould be unsafe.Patient also need to meet all the additional criteria duringthe 4 hours prior torandomisation:The need for mechanical restraintAntipsychotic or sedative medication (or bothrestraint and medication)Presence of delirium using the CAM-ICU assessmentMotor Activity Assessment score >5Excluded:Pregnant patientsDementiaHead injury resulting inaltered mental stateAlready receivingDexmedetomidine orclonidinePreviously enrolled on studyKnown contraindication to haloperidol or a2–antagonised
compared with the placebo group (median difference between the groups 17.0hrs 95% CI , p=0.04).On several days significantly lower quantities of sedative or opioid was administered to theDexmedetomidine group thanplacebo group. Bedside nursingstaff felt patient were ready toextubate significantly early in theDexmedetomidine group (p<0.001)
The study had an objective enrolment criteriaThe protocol replicatedcurrent bedside practiceacross the selected unitsAll other therapies wereconsistent with currentconsensus recommendations
Limitations:
Small sample size.Sample size of 96 was estimated to provide 80% power. However under recruited (n=71) due to early termination of trial bythe sponsoringpharmaceutical companyDexmedetomidine oftencauses bradycardia whichmay have indicated to theclinical team therandomisation of patientsMAAS score were not collected throughoutresulting in missing data. The author felt this may be due to less attention being given to educate the bedside nurse regarding this data collection
To determine theexperiences, attitudesand opinions of adult
Questionnaire survey to all nursing staff within two large ACCU (n=192)Excluded:
non-qualified
38.9% (n=75) response rateAll respondents believed thatphysical restraint had a place
Majority of the view that the
Strengths:
There was a clear record ofthe research process.The questionnaire used was
Further research needed to:
Explore risk factors of treatment interference and the use or non-use of physical
22
And opinions of adultintensive care unitnurses
intensive care unitnurses
support staff other
professional groups
nursing staff not involved in direct patient care
reason for its application was tomaintain patient safety.
Some expressed discomfort aboutthe use of physical restraint.
Nurses were happy to discuss theuse of restraint with families
Perceived need for training andsupport for nursing staff
Need for medical staff to support the decision-making process
piloted and pre-tested priorto the study. More than onecenter included
Limitations:
Small sample size due toonly two sites accessed.Low response rateOnly captured the nursingview point rather than thatof the MDT
restraint.
Explore alternative measures toreduce ACCU delirium and aidpatient comfort
Explore the potential conflictbetween nursing and doctors inrelation to the application ofrestraint requires.
Explore the experience of thosein ACCU who wake up physically restrained and thephysiological impact this mayhave
Research needs to directly seekthe patient’s perspective
(Burk, Grap, Munro, Schubert, Sessler., 2014)
Agitation Onset, Frequency and Temporal factors in Critically Ill Adults(A*)
Paper A&B samestudy reportedfindings across two publications
To describeThe frequency, onsetand pattern ofagitation
Retrospective review of medical records
Patient sequentiallyselected over a twomonths
In one surgical and one medical ICUExclusion:ICU stay under 24hrsUnavailable medicalrecordsPrevious ICU admissionduring the study timeAdministration of paralytic agentsChronic neuromusculardisorders
200 patients included in the final analysis.Of the 200, 118 (59%) wereagitated at some point over the five days. 28.5% based on RASS scoresand the rest based on key words with the documentation.Mean number of hours to onset of agitation was 11.6 hrs. (SD 22.3, range 1-114) from ICU admission.44 (43.1%) reported on admission to ICU30 (29.1%) reported 1-4 hrs. after admission
Strengths:
A pilot study wasundertaken to support thefinal study processes.One investigator carried out the data collection to ensure constancyUsed a validated tool toassess severity of illness andpresence of agitation
Limitations:
Data accessed was notgenerated for researchpurposes, may beinaccurate, incomplete or of
Further research is needed toclarify factors and to identifystrategies both pharmacologicaland non-pharmacological toprevent, ameliorate or treatagitation.
23
Head traumaStrokeData were collected for the first five days of the ICUadmission
poor quality.
Limited generalisability of findings, one site accessed in one country.
(Burk, Grap, Munro, Schubert, & Sessler, 2014)
Predictors ofAgitation in CriticallyIll adults (B*)
To identifypredictors ofagitation byexamining demographic andclinical characteristics ofcritically ill patients
Retrospective review ofmedical records.Patient sequentiallyselected over a two-month period within one surgicaland one medical ICU.
Exclusion:
ICU stay under 24hrsUnavailable medicalrecordsPrevious ICU admissionduring the study timeAdministration of paralytic agentsChronic neuromusculardisordersHead traumaStrokeLook at two specific time points, admission and 24 hr. preceding. Data were collected for the first five days of the ICU admission
200 patients included in the finalanalysis. Of the 200, 118 (59%)were agitated at some point overthe five days.
Predicator of agitation both 24 hrs.before onset and at admissionwere previous drug use, height,use of physical restraint andseverity of illness.
Non-agitated patient were morelikely to be discharge back to theirprevious living accommodationthan the agitated group.
There were 49 adverse eventrecorded, 45 (92%) wererecorded with agitation present.
These were:Self-extubation (15%)Pulled out critical catheter (9%)Fell out of bed (3%)Tore off restraint (3%)Pulled out non critical catheter(91%)
Strengths:
A pilot study wasundertaken to support thefinal study processes. Oneinvestigator carried out thedata collection to ensureconstancy. They used avalidated tool to assessseverity of illness andpresence of agitation.
Limitations:
Data accessed was notgenerated for researchpurposes and may havebeen inaccurate, incompleteor of poor quality. Limitedgeneralisability of findingdue to only one siteaccessed in one country.
Further research is needed toidentify causes, interventionaltherapy for prevention andtreatment once agitation hasoccurred.
(Luk et al., 2014) To describe patternsof physical restraint
Secondary analysis of aprospective observational
Patients were physicallyrestrained on average 4.1 days
Strengths Recommended future studiesshould aim to determine the
24
Predictors of physicalrestraint use inCanadian intensivecare units
use in mechanicallyventilated patient(prevalence,number of daysused, number ofepisodes).
To identify patienttreatment andICU/hospital characteristics associated withphysical restraintuse and number ofdays
study carried out 2008-2009Original data set generatedby the I-CAN-SLEAPprospective observationalstudy, which describedanalog-sedation, antipsychotic, andneuromuscular blockeradministration and drugassessment or titrationpractice.Each ICU had a preset twoweek period and withinthat time patients included were:Initiation of mechanicalventilation within theinclusion periodAged >16yearsAcross51 Canadian ICU
(range 1-26 days.83% (n=311) were restrainedonceRestrained and never restrainedpatient had similar baselinecharacteristics however theirtreatment characteristics altered.
In the restrained group there was:
Increased use of sedation(benzodiazepine)
Double the number of ventilator days
Double the occurrence of adverse incidence.
Large multicenter studywith large sample size.Heterogeneous sample ofICUs and patients increasesthe generalisability, unsurethis would transfer to theUK NHS due to possibledifferences in healthcaredelivery.
As it was a prospectivestudy it did not rely onretrospective chartreviewing.
Limitations
The data collector had noworking definition ofphysical restraint unsure ifsplints, bedcovers, bed railsetc include.
Physical restraint onlyrecorded one per day and duration from initiation todiscontinuation not captured.
Unsure about theperception of sedative usebeing chemical restraint.
Not clear of eachdepartment had the sameassessment tool forsedation, delirium and painalthough they did mention a conversion of sedation
directional relationshipbetween delirium and physicalrestraint.
Future observational studies toexplore whether the use ofsedative or analgesic drugscontributed to agitationrequiring physical restrain orvice versa.
25
score from all units to one Sedation Agitation Score.
Unclear if any of the unitshad a policy on restraint usebut did mention Nationallegation to reduce its use.
Recommendations seemedweak in relation to impacton practice.
(Michaud et al., 2014)Early Pharmacological treatment of deliriummay reduce Physicalrestraint Use: Aretrospective study
To assess patientoutcome afterprompt administration ofany clinicallyaccepted delirium medication.
Patients admitted to the22-bed medical/surgicalICU from February 2009to December 2012
Screened for delirium by trained bedside nursesusing the Intensive CareDelirium ScreeningChecklist (ICDSC).
Any patient with at least 1 documented positivedelirium screen (indicated by a score of ≥4 on the ICDSC) was eligible to be included in the study.
Other inclusion criteria Included:≥18 years
Median day in physical restraint (p = <0.001)
Treatment group= 3days (range 1-12 days)
Non-treatment group 6 days (range 1-21 days).
Median time to extubation (p=0.001)
Treatment group 3 days (range 1-21 days)
Non-treatment group 6.5 days (1-99 days).
In hospital death (p=0.008) Treatment groups 8 Non-treatment group 22.
Strengths:
The cohort was recruited in an acceptable way and a power and sample size was performed as priori.Gender, primary admittingdiagnosis, APACHE II score was comparable between the groups
Limitations:
Due to the retrospectivenature of the study therewas limited data collection.
The study did not delineatebetween the deliriumsubtypes.
If the patient receivedtreatment >24 hours after the first positive deliriumscreen or did not receiveany pharmacologicaltreatment during their
Further research of aprospective, placebo-controllednature is needed. Noted thatpharmacological managementwas only one component of a multi modal; approach howeverpharmacological therapy isbeneficial when initiated soonafter a validated positive delirium score.
26
Mechanical ventilation at the time of the firstpositive delirium screen.Patients were excluded if they were pregnant.Quetiapine, olanzapine,risperidone, ziprasidone,haloperidol, anddexmedetomidine werethe medicationsconsidered clinically acceptable for delirium treatment.
If at least 1 dose of apharmacologicaltreatment medication wasadministered to a patient within 24 hours of the first positive delirium screen,the patient was placed inthe “treatment” group.
delirium course, they wereplaced in the “no-treatment” group.
Promotion of sleep/wakecycle, daily sedation breaks, reorientation of the patients was deemed to be ‘standard’ practice however notdocumented and therefore impact not assessed.
The treatment for delirium was not standardised across the group.
Duration and resolution of delirium not noted.
Kandeel N.A. and Attia A.K., 2012Physical Restraint Practice in Adult Intensive Care Units in Egypt
To investigate thepractice of physicalrestraint amongcritical care nursesin one city in Egypt
Descriptive cross sectionalDesign
Convenience sample
patent group one –patients exclude in they were in the ICU less than 12 hours
Group two nurses-excluded if less than12mths experience
Physical restraint is commonplacein the ICU accessed.
The study indicated:Poor observation/document/assessment of the restraint siteA lack of documentation andeducationLack of family/patientinvolvement
Strengths:
There was a clear statement of the aims of the study and a clear observation tool devised and pilot prior to the study.All research staffinvolved were educatedabout this.There was a clear statement of findings and ethical standards maintained.
Highlighted the need forstandard practice guidelinesand policy for physical restraintuse and the further researchwas required.
27
Limitations:
The findings are limited as to their generalisability due to the one region of Egypt
Noted by the authors that nursing staff in Egypt do not question practice but follow instruction from medical staff impacting on the practice observed
(De Jonghe et al., 2013)Physical restraint inmechanically ventilated ICUpatients: A survey of French Practice
To characterise the perceived utilisationof physical restraintin mechanicallyventilated ICUpatients and toidentify clinical andstructural factorsinfluencing PR use
Cross sectional survey
A questionnaire was hand delivered to one intensivist within all of the 130 ACCU across France
In 82% of ACCU physical restraintis used at least once duringmechanical ventilation in morethan 50% of patients.
Physical restraint used most frequently during awakening from sedation.
Physical restraint commenced without written instruction in 50%.
Only 21% of ACCU have awritten local policy on physicalrestraint use.
Strengths:
This was a large nationalsurvey with a response rateof 93%. The questionnairewas purposefully designedby three intensivist, pilotedand refined beforedistribution.
Limitations:
Possible difference betweenreported and actualpractice due to self-reporting bias.Completed by physicianwith the suggestion todiscuss with nursing staffbut this was not formalizedand not documented if itoccurred.Recruiter hand deliveredthe survey and approached
Physical restraint use could beimproved by increasingphysician awareness of thepositive and drawback of it use.
Physical restraint requires awritten order for it to becommenced and discontinued.
The study should be used as abase for developing andimplementation physicalrestraint use policy on a local ornational scale.
28
the first intensivist theymet on arrival at the unit.
The recruiter were salesrepresentatives from thepharmaceutical companyGlaxoSmithKline.
To what degree thisimpacted on response isnot clear and whatdiscussion therepresentative had inrelation to the survey not noted.
(Tate et al., 2012)Anxiety andagitationin mechanicallyventilated patients
To gain a more indepth understanding of the manifestations,interpretations and management of anxiety andagitation in themechanically ventilated patient.
Ethnographic study usingprevious generateddataset.
Data from 30 purposivelyselected patient and their family members and clinician who cared for them.
All patient had reported or experienced anxiety and/or agitation.
Data generated from Happ et al (2007) justified secondary analysis as:
The phenomena of interest frequent in data setData set extensiveMaximise the participation
The study developed an Anxiety,agitation in mechanical ventilatedpatient model, which they feltclinical staff could use to targettheir interventions. The study’sfinding support the value of‘knowing the patient’, whichenables clinical staff to interpretresponses.
Strength:
Audit trail ofmethodological memo andnotes. Multiple data sourcewere cross checked ortriangulated to confirmfinding. Credibility wasestablished via memberchecks with 5clinical participants.
Member of the researchteam were also involved inthe original study the date were generated.
Limitations:
Limited transferability to other ACCU patients.
Inclusion criteria were only
Further work is needed inrefining operational definitionsand clinical language toimprove the discriminationbetween agitation and anxiety.
Research is needed to explainthe effects of the nurse andfamily presence and clarify the effects of specific support strategies.
29
of a vulnerable population patient in the weaningphase.Study was carried out insingle center, singleinstitution.Accessing an existing dataset may affect the reliabilityof interpretations.Practice may havedeveloped over the timelapse
(van den Boogaard et al., 2011)
Incidences and short term consequences of delirium incriticallyill patients: aprospective observational cohort study
To determine theoverall incidenceand duration ofdelirium, perdelirium subtypeand per ICUadmission diagnosis.
To determine theShort-termconsequences ofdelirium
Prospective observational study.All consecutive patientsadmitted to the ICUincluded and systemicallyscreen for delirium usingCAM-ICU.
Diagnosed with one positive screen duringadmission.Delirium divided into three subtypes (hyperactive/hypoactive/mixed)
Excluded patients:Admitted for less than one daySerous visual or auditory disordersUnable to understandDutchHad a mental disabilitySuffered from serousreceptive aphasiaWere the compliance rate
1613 patients included26% (n=411) developed delirium.
The mixed subtype delirium occurring most frequently (53%).
Delirious patients were:
Mechanically ventilated for longerMore likely to remove their tubes and cathetersStayed in ICU and hospital for longer6 times higher chance of dying compared to non- delirium group
Strengths:
They used a validated screening tool to establish delirium.Well defined outcome measure and noted all variables.Had a data checker check 15% of data randomly.
Limitations:
Limited generalisability due to single site in one country.
Large number of cardiac surgery patients included which may have reduced the occurrence of delirium as these patents generally have a short ICU stay.
As a prospective cohortstudy it determineddelirium’s association withseveral outcomes but not
Nurses should be encouraged totake preventative measuressuch as:
cognitive stimulation music therapy prophylactic
haloperidol early mobilisation.
30
of delirium screening was<80% during the patients stay
necessary established acausal link.
Number of patients died andthe data for these patientswas removed from thedelirium durationcalculation
(L. C. Chang et al., 2011a)Risk factorsassociated withUnplanned Endotracheal Self-Extubation ofhospitalised intubated patients: a 3 year retrospectivecase control study
To identify the riskfactors associatedwith UnplannedEndotracheal Self-Extubation (UESE)and to comparecertain characteristics between UESE and a control group
Three year retrospective study carried out in one teaching center, which had 11 units.
Exclude patients who:
Had planned or accidental extubation (possibletypographical error inpaper)Died during hospitalizationPatient with a psychiatric historyPatient with suicidaltendencyPatient with dementiaPatient with a lifethreatening illness.
Compared thecharacteristics of patients,nurses, vital signs,duration of intubation,serum laboratory results, Glasgow Coma score (GCS)Acute Physiology and Chronis Health Evaluation
Most UPSE occurred: during the evening night shift shift with less
experienced nurses.
80% of the self-extubated patientswere physically restrainedhowever all patient in the controlgroup had physical restraintapplied.
Pulse rate and APACHE II scores were significant indicators of UESE
Strengths:Clear analysis of the data.Used a recognised outcomemeasure tool.
Case recruited and controlselected in an appropriateway.
Limitations:
Small sample size in a single site.The APACHE II scores were assessed on admission rather than at the time of UESE. Therefore these may have been different.Unable to attribute causality due to the retrospective nature of the study.
Sedative management and assessment included and unsure if consistent across the two groups.
Nurse to patient ratio also appear to be inconsistent
Recommended that nurses should reassess patient continually and routinely measure APACHE II and pulse rates.
Recommended providing continual education for nurses, increasing the number of nurses, increasing the number of nurses on duty and providing more frequent bedside visits during the evening and night shift if needed.
31
II (APACHE II) scores andthe use of physicalrestraints and sedatives.
across the care delivered.
(de Groot et al.,2011)Risk factors andoutcomes afterunplanned extubation(UE) on the ICU: aCase control study
To study theincidence, determinants andoutcomes of UE andto assess the riskfactors for re-intubation
Case control studyIncluded:All patient requiringartificial airways at threeICU were monitored for UE. From Dec 2005-June 2008
Excluded: patient who hadaccidental extubation due to nursing or medicalintervention.
For every UE four control patient were randomly selected from the total population of mechanically ventilated patient across the units.
Within the study period 74 UE in 69 patient ( 5 patients UE x2)
Determinant of UE were: male, higher BMI one unit with increased
population of surgical patients,
elevated serum sodium, reduced Ramsay
Sedation Score (anxious/agitated-awake/calm),
use of haloperidol and methadone.
47% required re-intubation with 53% not.
Those without re-intubation had significantly shorter length of stay in ICU and in hospital.
Strengths:
Pre- study support provided for staff.Had clear criteria for weaning across the three sites.Staffing, securing the ETT and sedation appear to be consistent across the three units.Control group was recruited randomly.
Limitations:
Generalisability to other region/countries as this was three centers in on region of the Netherlands
Data gathered up to 12hours after UE and thenurse interview, which mayimpact, on the re-call ofdetails.
The study has introduced additional risk factors for UE.
Introduction of weaning and/orextubation protocol and dailyevaluation of the need for MVcould reduce the length of MV and in turn reduce UE
Langley G.,Schomollgruber S.and Egan A 2010Restraints in Intensive Care Units-A Mixed Method Study
To provide a detailed description of restraint practice across three ICUs.
Quantitative datacollected:Number of patientsrestrainedType of restraint usedAverage duration ofRestraint time
Quantitative results:
219 patients were recruited48.4% (n=106) were physicallyrestrained.Patients were restrained on average for 9 days and in the case
Strengths:
It had a clear study aim andobjectives with appropriate research methodology forthe question.The paper highlights points
Further research is needed especially around patient and family opinion of restraint use.
Institutions need to have apolicy regarding the use ofrestraint and restraints
32
If the department had a policyIf so was it adhered to?
Qualitative data wasgathered around medicaland nursing staff views onthe use of restraint viaindividual interviews.
of one patient only 53 days.
In 9 cases they were inadequately or ineffectually applied.In 5 cases restraint were applied but not secured.
Qualitative findings:physical restraint had a place,balancing act re: benefit anddisadvantages; poorcommunication across the team represcription and legal directivesand with patient and families
raised in other work andadds to the overall body ofknowledge on whichfurther research could be based.
Limitations:
Limited to the generalisability of the finding due to the fact this is a small mixed method study conducted in on region within three sites. Clinical staff were purposively recruited to the qualitative phase possibly introducing bias
prescribed in line with this with a minimal duration set.
(Benbenbishty, Adam, & Endacott, 2010)Physical Restraint usin intensive careunits across Europe: The PRICE study
To explore theincidence of physicalrestraint in adult ICUacross Europe and explore the association with other factors such as staffing
Point prevalence survey used in 34 ACCU in nine countries.
Bed numbers ranged from 6-24.All patient on the unit on the observed day were included.
Exclusion criteria applied at unit rather than patient level.Specialist neurologicalsurgical unit exclude inrecognition of specificchallenges.Paediatric units excluded.
34 ACCU in nine countries providedstudy data regarding 669 patientsregarding details of physical and chemical restraint.
39% (n=219/566) were physically restrained.
Use of restraint was not related to the time of the week.
Restraint was more likely to beapplied to ventilated, sedatedpatients in a unit with a lowerdaytime nurse: patient ratio.
Commercial wrist restraint andpolicy for use was evident in the
Strengths:
This was the first study of its kind and was Europe wide.A structured data collection tool was used and translated into the relevant languages.
Limitations:
The case mix of patient was not included and therefore some patient may have been ready for discharge or high dependency patients rather than ICU.
The paper recommends the development of evidence based guidelines and educational packages to underpin restraint practice.
33
larger units.
Physical restraint use was less in unit with a policy regardingphysical restraint.
The primary reason for use was to prevent unplanned extubation or removal of infusion lines.
Data was collected by a range of collectors across 34 sites and the authors did not include if these data collectors had had any training.
Inter-country reliability testing would have ensured consistency of the data collected.
The extent to which the unit within the study are representative of others in the same country is unknown.
(Tanios et al., 2010)Can we identifypatients at high riskfor unplannedextubation? A largemultidisciplinary study
To define high riskpatient for unplanned extubation and determine clinician beliefs on perceived risk for unplanned extubation
Cross sectional survey
All critical care members of:The American Association for Respiratory CareThe America Association of Critical Care nursesThe Society of Critical Care Medicine
Total of 1976 respondents(419 respiratory therapist, 870 nurses and 605 physicians all from ACCU).
87% considered a patient tuggingat the endotracheal tube (ETT) riskof self-extubation.
71% removal of nasogastric tube increased risk.
Definition of a ‘near miss’ less clear with half the respondent believing it was dependent on the patient’s medical condition.
Strengths:Clear and transparent account of survey development with involvement of clinical experts, content validity checking and pre-testing carried out.
Clear an appropriate statistical analysis of findings.
Limitations:
The survey focus was on perceived risk, which may not equate to actual risk.
Due to the survey design
The study provided a working definition of an airway ‘near miss’ as ‘outward ETT migration of 2cm if measured at the lips or 3cm in measured at the incisor’
The relationship between unplanned extubation an the risk factor highlighted require further study via direct observational study
34
95% viewed near missed a threat to patient safety.
44% considered unplannedextubation to be a medical error.
Situations which posed a high risk of self extubation:72% felt absence of physicalrestraint.60% nurses/patient ratio59% trip out of ICU43% light sedation29% bedside portable x-ray
there may be bias in the question posed.
Recruitment was via three large organisations.
Accessing a convenience sample such as this may not represent the view held by clinical colleagues therefore limits the transferability of the findings.Those who responded may have had more interest or experience in the topic areas than non-responders skewing the results.
(Huang, 2009)Factors leading toself-extubation ofendotracheal tubesin the intensive careunit
To discuss thefactors leading toself extubation ofendotracheal tubes (ETT) and explorethe difference between the group of patient who did and who did not self remove their ETT
Case control study
Purposive sampling patient data either with a planned extubation or who self-extubated. No exclusion criteria offered
708 patients included
44 self extubated (6.4%).
63.9% of those who self extubated had been restrained
MDT stating the reason forrestraint was to prevent selfextubation.
66.7% of patients did not have sedative post operatively.
Among the 44 patients who self-extubated 28 patients met the criteria for extubation.
Strengths:
The case and control were recruited in an acceptable way with independent and controlled variable accounted for.
The paper reported a robust statistical analysis.
Limitations:
The total sample was obtained from one surgical ACCU in one region of Taiwan, reducing the generalisability of the results.
The paper recommends that the knowledge of the MDT should be improved in relation to the use of physical restraint.The psychological support and orientation of a conscious patient is vital.
Patients need to be assessed and extubated once they are ready and this should not be delayed.
35
Sedation and analgesic practice was not noted which may impact on the results.
It was noted that a significant number of patient who self extubated had been restrainedbut it was not made clear if this was at the time ofextubation and how long the patient had been in restraints.
Did not include the nurse topatient ratio, which maydiffer to UK ACCU and havea possible influence on the rate of self-extubation.
(Reade et al., 2009)Dexmedetomidine vs. haloperidol indelirious, agitated,intubated patients: a randomised openlabel trial
To establish ifdexmedetomidine would be moreeffective thanhaloperidol in thetreatment of ICUassociated deliriousagitation inmechanically ventilated patient
Pilot randomisedOpen label, parallel groupcontrolled trial.Included patientconsidered by thephysician to require ventilation only becausetheir degree of agitationrequire high dose ofsedative that extubationwas not possible.Agitation assessed using the RASS.ExcludedPatient who could not be extubatedPatient with a plannedreturn to theatre
Patients randomise to theDexmedetomidine group wereextubated significantly sooner than the haloperidol group (p=0.016)Time to extubation was shorter in the Dexmedetomidine group (19.9 vs. 49.8 hours p=0.0147)Patient who receivedDexmedetomidine weredischarged from the ICU earlierwith a short length of stay(1.5 median days vs. 6.5 p=0.0039)
Strengths:Used a validatedassessment tool to assessfor agitation which staffwere familiar with.
Limitations:As this was a pilot studythere were a number oflimitations.
Small single centre sample
Staff were not blinded to thegroup allocation and drugdelivery
Dexmedetomidine significantlyshortened time to extubationand decreased ICU length of say.
The authors felt their results justified the conduct of a larger, multi-centred blindedrandomised controlled trial
36
Those likely to require on-going airway protectionThose physiologicallyunstablePatients with previousadverse reaction tohaloperidol or a2 agonist
Patients were allocatedusing a random numbersequence 10 allocated todexmedetomidine groupand 10 to haloperidolgroup
No protocol to guide thetitration of medication
Clinicians were free toprescribe any additionalsedation or anxiolytic
The unit had no formalweaning protocol
There was no baseline data in each group regarding presence of delirium, pre-morbid cognitiveimpairment or history ofsubstance abuse
(Turgay et al., 2009)Physical restraintuse in Turkishintensive care units
To determinenurses’ reasons forthe application andremoval of physicalrestraint (PR)
Cross sectional survey
All nursing staff in thechosen sites.
ICUs in 2 universityhospitals and 5 statehospitals in West Turkeyaccessing a on-randomvolunteer sample of 190nurses
84.7% of respondent stated theyuse PR without prescription.Wrist and ankle restraints beingmost commonly used (84.7%).
Common reason for application is:Maintenance of medical devices (86.8%)Restless behaviour (86.3%)Impairment of mental status (79.5%)Performance of medical treatment (53.7%)Convenience (23.3%)Suggesting from family or medical staff (12.1%)
36.8% reported complicationsafter the application of PR withskin breakdown most common.
Strengths:
There was a clear process ofquestionnaire distributionand the relevantparticipants across 7hospitals approached.
Generated some interestingfindings to add to the smallbody of work in this area.
Limitations:Low response rate of 40%
Recruitment from oneregion of Turkey limits thegeneralisability of thefindings.This only sought the nursesview rather than the MDT.
Recommended furthereducation to standardizenursing practice and protocolrelated to the use of PR.
Further research required toidentify alternatives to PR.
Restraint used should only beimplemented when all othertherapeutic methods haveproven ineffective.
37
The sedation, analgesic practice as well as the nurse to patient ratio was not outlined which may impact on the frequency and duration of PR use.
(L.-Y. Chang et al., 2008)Influence of PhysicalRestraint onUnplanned Extubation of AdultIntensive CarePatients: A CaseControl Study
To evaluate theeffect of physicalrestraint onunplanned extubation in adultintensive carepatients
Case control study
All patients who had been admitted and intubated in the selected ICU over the 21mth period
Medical records andincident reports wereexplored on 100 patientswho had an unplannedextubation.
The cases compared to a sample of 200 other cases with matched age, sex and diagnosis with no record ofunplanned extubation.
This was to evaluate the effect of physical restraint on unplanned extubation.
Total 300 patient selected
191 had physical restraintsapplied.
This group had a higher rate of:Nosocomial infection (21.5% vs. 9.2% p=0.05)
Unplanned extubation (42.9% vs. 16.5% p=0.01)
Of the 191 patients in physical restraint 82 had unplannedextubation.
Strengths:
Recruitment was clearly stated and conducted in an acceptable way including a power calculation for sample size.
Limitations:
This was a retrospective study collecting data from medical note and incident reports of unplanned extubation.This assumed that the note were an accurate representation of events.
There was noacknowledgment of any missing data.
Some unplanned extubation may not be reported via the incident reporting system particularly if the patient did not require re-intubation.
The study findings concluded that a patient was more likely to have an unplanned extubation when they were physically restrained.
Highlighted the need for effective standard for use of physical restraint and further research of the subject wasrequired.
(Curry et al., 2008)Characteristics
To determine the characteristics of
Retrospective exploratory design in one 10 bedded
All unplanned extubation wereself-extubation.
Strengths: Recommended that the timing of weaning from the ventilator
38
associated withunplannedextubation in asurgical ICU
patients and nursesand risk factors that affects extubation
surgical ICU in America
84% (n=26) were men 86% were in PR at the time of self-extubation.
Those who needed re-intubatinghad a significantly higher meanRamsey Sedation Score.
Positive correlation between RSSand the need for re-intubation.
87% in PR mostly soft wristrestraint. 3 patients had morethan one type of restraint in placeat time of self-extubation.
89% of the attending nurses werenot at the bedside at the time ofSE. Most patient had received lowlevels of sedition within the hoursproceeding SE
The paper outlined a clear definition of self and accidental extubation.They used a validated tool for sedation soring (Ramsey Sedation Score).
Limitations:
They did not include severity of illness scoring (APACHE 11) and as this was retrospective assessment of documentation there could have been missing or inaccurate data.
This was a small sample size of surgical patients in one ICU limiting the generalisability of the findings.
support was critical.
Staffing availability duringweaning, medication management in particularly pain management is vital.
(Mion et al., 2007)Patient-InitiatedDevice Removal inIntensive Care Units:A NationalPrevalencestudy
To explore theprevalence of deviceremoval, describethe patient context,examine unit leveladjusted risk factorsand describe theconsequences.
This was a prospectiveprevalence study withdata collected for 49,482patient days by trainednurses.
1097 episodes of device removalwith no significant difference intypes of ICU noted.
181 episodes of self-extubationduring 18,308 ventilation days.
At the time of device removal:45% (n=494) patient were in physical restraints
58% had documented agitation or anxiety at the time of the episode.
Harm or injury occurs in 23%
Strengths:
This was a large multi-sited trial where 49 hospitals were selected at random across several region of the United States.
A standardised data collection tool was used which was validated and tested for inter rater reliability (>95%) before the study commenced.
Additional inter rater
Found that device removal bypatient is not uncommon but theprevalence did vary across theICUs selected.
Resources on the ICU had animpact on the number of devicesremoved by patients and theremoval of device did result insome degree of harm, mostminor in nature.
There was also a variance in there-insertion of devices once
39
4% being major (re-intubation, re insertion of central line).
Harm to staff occurred in 10% (N=110), which was mainly,unprotected exposure to bodily fluid.
checks were carried out at each site during the study.
All variables well defined and confounding factors considered with unit adjusts made.
Limitations:
However the study relied on practitioners’ documentation of device removal, which could have led to under reporting of events and the daily surveillance of staff, may have affected practices.
removed.
Further development of staffing and work load measurement is required to understand better how nursing staffing levels affect quality of patient care.
Martin B. andMathisen L., 2005
Use of PhysicalRestraint in AdultCritical Care: ABicultural Study
To examine the traditional use of restraint verses limited or no use ofrestraint.
To describe the relationship between patient characteristics, environment and actual use of physical restraint in selected critical care units in Norway and the United States
This was an observational study with a descriptive correlation design to describe the relationship between patientcharacteristics, theenvironment and the use of physical restraint in two different countries, United States and Norway.
Data collected in 3 ICU in USA and 2 in Norway.
Patient required until predetermined sample size reached.
Inclusion:Patients >18yearsLength of stay >12 hours
The use of physical restraint was associated with patient activity.
Physical restraint uses in 40% ofthe US sample- none in Norway.
US had a greater activity score from the patient observation
US patients received smallervolumes of sedative
Norway had a higher workload score per patient
Norway had a higher nurse to patient ratio (Norway 1.05:1, US 0.65:1)
7 incident of unplanned patient removal of devices all in US
Strengths:
Physical restraint wasclearly defined for theresearcher and oneresearcher collected thedata in both countriesaiding consistency inobservations.
The study had a clear inclusion and exclusion criteria.
Power calculation provided pre-set sample size and recruitment continued until this was met.
Author considered other
Both countries ICU had similarpatient groups and similartechnology yet differed on thenurse patient ratio, levels ofsedative used and restraintpractices.
The study concluded thatrestraint free practice was safefor patients yet highlighted that this was a small study.
Further research was required to assess the effectiveness ofrestraint free care and assessthe use of non-pharmacologicalinterventions, such aseducation of patients, the use ofnon-professional staff orfamilies
40
Those not undergoing cardiac surgery
Exclusion:Those awaiting diagnosis to confirm possible brain death
If active ICU treatment had been withdrawn.
sample and all has restraints applied at the time
Only statistical significant difference in across the patient population was race, all Norway sample white US sample more diverse.
confounding factors such as family visiting, ICU environment and admission policy
Limitations:
Documentation and staff response may not reflect actual practice.
Patient characteristic, which may have an impact, not assessed such as metabolic imbalance
Pain management and assessment not noted.
Limited transferability to the UK NHS system as the admission criteria was unclear to each of the departments
(Jaber et al., 2005)A Prospective Study of Agitation in a Medical-surgical ICU
To describe theincidence, riskfactors, characteristics andoutcomes ofagitation
Prospective observation study, which screened all consecutive ICU admission over an eighth month period.
Those excluded from thestudy were patient with aprior diagnosis ofdementia and those whodied before waking 182patient were recruited
52% (n=95) patient experienced at least one episode of agitation during their ICU stay.
Agitation was greater in medical patients rather than surgical.
Provided a clear definition of agitation, used a validated tool to assess agitation.
Provided a cleardocumentation trail ofrecruitment and inclusion.
Limitations:
Further research is required to better identify those with an increased risk of agitation.
More research is needed todemonstrate reducing riskfactors will decrease theincident of agitation.
41
previous alcohol abuse used of sedation fever dysnatremia psychoactive drugs.
Agitation occurred both day and night and early in the patient ICUadmission (<3-5 days).
Agitation was associated with:
Increased LOS in ICU(16+/- 19 days vs. 6 daysp=0.0001)
Higher unplanned extubation (17% vs. 2% p=0.0003)
Higher CVC removal (16% vs. 1% p=0.001)
This was a single site study in one country limiting the generalisability of the findings.The inclusion of nonventilated patients withinthe study may possiblyunderestimate the agitationwhich occur with theinvasively ventilated ICUpatient
42
Figure One: Flow of paper selection based on PRISMA
43
Table Three: A section of the synthesis matrix
Box One:
Study
Country of
Origin
44
Findings Syntactic constructs Synthesised theme Sub themes
All respondents believed that physical restraint had a placeMajority of the view that the reason for its application was to maintain patient safety.87% considered a patient tugging at the endotracheal tube (ETT) riskof self-extubation. 71% removal of nasogastric tube increased risk.Definition of a ‘near miss’ less clear with half the respondent believing it was dependent on the patient’s medical condition.95% viewed near missed a threat to patient safety.44% considered unplanned extubation to be a medical error. Situations which posed a high risk of self extubation: 72% felt absence of physical restraint.60% nurses/patient ratio, 59% trip out of ICU, 43% light sedation29% bedside portable x-ray
Sources: (Freeman et al., 2015, Tanios et al., 2010)
Physical restraint maintains safetyPrevent extubationPrevent line removal
Agitation management
strategies
Effectiveness of Physical
Restraint
Opinions on Physical Restraint
Use
Poor observation/document/ assessment of the restraint siteA lack of documentation and educationLack of family/patient involvement48.4% (n=106) were physically restrained.Patients were restrained on average for 9 days and in the case of one patient only 53 days.In 9 cases they were inadequately or ineffectually applied.In 5 cases restraint were applied but not secured.
Some expressed discomfort about the use of physical restraint.Nurses were happy to discuss the use of restraint with familiesPerceived need for training and support for nursing staffNeed for medical staff to support the decision-making process
Source: (Freeman et al., 2015, Tanios et al., 2010)
Discomfit with PR useTeam disagreement
Physical restraint had a place, balancing act re: benefit and disadvantages; poor communication across the team re prescription and legal directives and with patient and familiesSource: (Langley et al., 2011)
Physical Restraint has a place
Most UPSE occurred: during the evening, night shift and shift with less experienced nurses.80% of the self-extubated patients were physically restrained however all patient in the control group had physical restraint applied.
Pulse rate and APACHE II scores were significant indicators of UESESource: (Chang et al., 2011)
Physical restraint effective
7 America, 3 Taiwan, 2 Netherlands, 2 France, 1 Canada, 1 Egypt, 1 South Africa1 Turkey study, 1 United Kingdom, 1 Brazil, 1 Australia, 1 jointly conducted between America and Norway and 1 European study.