AAC in the ICU: Critical Issues and Preliminary Research Mary Beth Happ, Ph.D., R.N. Kathryn Garrett, Ph.D., CCC-SLP Tricia Roesch, B.S.N., R.N. * * *

AAC in the ICU:AAC in the ICU: Critical Issues and Critical Issues and

Preliminary ResearchPreliminary ResearchMary Beth Happ, Ph.D., R.N.

Kathryn Garrett, Ph.D., CCC-SLP

Tricia Roesch, B.S.N., R.N.

* * * * * * * * * * * * * * *

School of Nursing University of Pittsburgh

Duquesne University, Pittsburgh PA

ASHA Convention November 2003 Chicago

OverviewOverview Part I: Literature Review Part II: Feasibility study of electronic VOCAs in the Surgical Otolaryngology

Unit and Case Example Part III: Feasibility study of electronic VOCAs in the Medical Intensive Care Unit Part IV: NIH-funded Intervention Study --

The SPEACS Project

Note:Note:

• Please refer to the Microsoft Word document by the same title for a narrative version of this presentation

• The Word document will also contain the reference list.

Part I Part I

BackgroundBackground

Descriptive reports of the Descriptive reports of the mechanical ventilation experience mechanical ventilation experience

in the ICUin the ICU

• Patients experience:• FEAR• PANIC • STRESS

• As a result of the inability to speak

Nurse-Patient communication in ICU: Brief (< 5 min), task-oriented, commands &

reassurances during physical care.

Patients typically communicate with nods, gestures, and mouthing words.

ICU interactions do NOT usually involve communication of the patient’s ideas, patient’s initiation of messages or elaboration.

Communication difficulty with mechanically ventilated (MV) patients - related to illness severity, anger

(Menzel, 1998)

Greater difficulty communicating with family than with nurses (Menzel, 1998)

Under-recognition & high levels of pain reported in MV patients (SUPPORT studies)

RNs/MDs more likely to communicate with patients who are more responsive.

Statement of the Problem

Few data-based communication intervention studies with acutely/critically ill adults have been published

(Dowden et al, 1986; Stovsky et al, 1988)

Alphabet & picture boards preferred by a critical care survivors (n=5) (Fried-Oken et al, 1991)

Clinical case reports

Introducing AAC preoperatively & word banking (Costello, 2000)

Multidisciplinary post-operative AAC plans for head and neck cancer patients (Fox & Rau, 2001)

Descriptions of AAC use in ICU (Fried-Oken, 2001)

A need exists for:A need exists for:• Specific data on communication

interventions for nonspeaking, intensive care unit patients

• Analysis of high tech versus low tech interventions

• Perceptual, qualitative, and quantitative analyses

• Comparisons between different ICU populations

• Usage data as well as interactional data

General Design of 2 General Design of 2 Feasibility StudiesFeasibility Studies

PurposePurposeExplore the feasibility of electronic

voice output communication aids (VOCAs) for use with nonvocal patients

(1) in a medical ICU and

(2) following head-neck cancer surgery.

Research QuestionsResearch Questions

What are the …What are the …

Patient characteristics (illness severity, neuromotor ability)

Usage patterns (message categories, frequency, assistance required)

Communication quality (ease, satisfaction)

Barriers to communication

…when VOCAs are used by when VOCAs are used by hospitalized adultshospitalized adults?

Complementary DesignComplementary Design

QUAL + quanQUAL + quan

No hypothesesNo hypotheses

Purposive-theoretical samplingPurposive-theoretical sampling

Morgan, 1998

Small samplesSmall samples

Settings: University of Pittsburgh Medical CenterUniversity of Pittsburgh Medical Center

- Otolaryngology surgical unit- Medical ICU – 20 beds

Entry Criteria:• Respiratory intubation Respiratory intubation • Responsive to verbal stimuliResponsive to verbal stimuli• Follows commands consistentlyFollows commands consistently• Initial Cognitive-Linguistic ScreenInitial Cognitive-Linguistic Screen**

* Dowden, Honsinger & Beukelman, 1986

ProceduresProcedures

Education & Set-UpEducation & Set-Up

Nurse Inservice (15 min) Patient Instruction (20 min) +

reinforcement Message Inventories

• What does he/she want to say?

• To whom?

• How? *Costello, 2000*Costello, 2000

Data Collection Data Collection

Enrollment Pre-test Ease of Communication Scale2

APACHE, Motor Screen1

Daily Observations (20min) Chart Review Extubation Post-test Ease of Communication Scale2

Exit Interviews

1. P. Dowden et al. (1986) 2 L.. Menzel (1998).

Part IIPart II

Pilot Research:Pilot Research:

Head and Neck Surgical Head and Neck Surgical UnitUnit

Equipment donationsEquipment donations::

DynaVox Systems, Inc.DynaVox Systems, Inc.

WordsPlus, Inc.WordsPlus, Inc.

AbleNetAbleNet

FundingFunding: : AACN/ Sigma Theta Tau AACN/ Sigma Theta Tau ONS Foundation/ OrthoBiotech ONS Foundation/ OrthoBiotech

Mentorship/Consultation:Dr. Richard Hurtig, University of IowaStephanie Williams, SLP, Dynavox Systems, Inc

MessageMate

DynaMyteTM

TM

Electronic VOCAs

Examples of Patient Message Screens

DynaMyteTM

I’m OK

AFRAID

Pain shot

HOT

MEDICINE

MOUTH CARE

SICK

MY MOUTH

TV

NURSE

BATH

CAT

NOT OK

GLASSES

HUNGRY

MUSIC DRINK

COLD

DOG

DOCTOR FAMILY

ILOVEYOU HOME TIME

2

3

4

1 Say

BackSpace

Clear

Repeat

TIRED SAD HAPPY ANGRY

PAINNAUSEA

BEDPANSUCTION

WHY?WHERE?HEAR

MessageMateTM

Basic MessagesBasic Messages

Pain Shortness of Breath Suction Help! Hot/Cold Home/Family Anxiety/Worry

pole

swivel arm

Qualitative Data AnalysisQualitative Data Analysis

Fieldnotes and interviews coded for:

1. method2. content 3. barriers 4. facilitators

Quantitative Data AnalysisQuantitative Data Analysis

Descriptive statistics (dispersion) Pattern recognition Nonparametric within case

comparison (EOC)

RESULTSRESULTS

Study #1: Exploring the Feasiblity of VOCAs with Head and Neck Cancer

Patients Following Surgery

MB. Happ1

S. Kagan2

T. Roesch1

E. Holmes1

Funding: ONS Foundation/OrthoBiotech

1 University of Pittsburgh School of Nursing2 University of Pennsylvania School of Nursing

Head & Neck SampleHead & Neck Sample(n=10)(n=10)

7 men, 3 women all Caucasian 5 MessageMate 5 DynaMyte

Observation & InterviewObservation & Interview

Observations: = 66Communication Events = 50 (75.8%)

Formal Interviews: = 9

Patient = 8Nurse = 1

CharacteristicsCharacteristics (n=10)(n=10)

Ages: 45-82 yrs (57.1+12.8)

Education: 12-20 yrs (13.5+2.9)

Computer Use: 7*

*minimal level = 3/7

Procedures: Brachytherapy 2

Laryngectomy 8

CharacteristicsCharacteristics (cont).(cont).

Days w/ device: 3-24 (9.1+ 6.2)

Post-op days prior to device: 1-6 (1.9+.1.6)

APACHE III: 5-53 (27.1+13.2)

Neuromotor Characteristics Neuromotor Characteristics

+ Motor Screen Tasks = 10

+ Write legibly = 10

+ Narcotics/sedation = 35/50 (70%)

Usage PatternsUsage Patterns• VOCAs were used by some of the post surgical patients

- some required extensive assistance, whereas others required limited or no assistance

• Other modalities were used as well-Writing- Gesture

- Mouthing Words- Head Nods

Other findingsOther findings• Of the observed communication events

in which patients utilized the VOCA, patients initiated more frequently than a historical (no-intervention) group.

• a slight increase in ease of communication was observed in the VOCA group when compared with a historical (no-intervention) group.

Novel Scenarios in which Novel Scenarios in which VOCAS were usedVOCAS were used

1. Cardiology evaluation

2. Telephone usage

What were the barriers to What were the barriers to device use?device use?

device out-of-reach upper extremity & neck

wounds blurred vision insufficient staff training in

use patient preference for writing or

other method

Message ContentMessage Content

Comfort needs (pain, thirst, suction)

Questions about home & family “I love you” Questions about tests and

condition Phone conversations

Characteristics of the head and Characteristics of the head and neck patient population that neck patient population that

may have been associated with may have been associated with successful AAC device use:successful AAC device use: All were able to write All were liberated from ventilator Voicelessness was expected More independence

Case Study

““Tim”Tim”

• 46 year old Caucasian male• S/P Total laryngectomy & tooth

extraction• No prior history of intubation and

mechanical ventilation• No significant past medical history

““Tim”Tim”

• High school graduate• Previous personal computer use• Vision corrected with eyeglasses• Right hand dominance

““Tim”Tim”

• Motor screening tasks• APACHE score = 29• Glasgow Coma Scale (GCS) = 15

EnrollmentEnrollment

• Immediate post operative phase• Transferred from Medical Intensive Care

Unit (MICU) to Head and Neck ICU• Patient appeared withdrawn

• Deferred until third post operative day• “just don’t feel like it”

• No device training prior to study enrollment

Device Set UpDevice Set Up

• Device options Message Mate- simple, smaller message

capacityDynaMyte- larger capacity, multi-level

message display

• At bedside• Duration ~1.5 hours• Initial method of communication

• Writing/Gestures

Tim’s RequestsTim’s Requests

• Voice selection

• Message deletions• “Yes/No”• “What time is it?”

• Message Additions• “Hello” & “Good-bye”

Tim’s RequestsTim’s Requests

• Icon/Message change

• Performed at bedside

• Requested by patient and/or family

• During entire enrollment period

• Affect change

Observation of Observation of Communication Events Communication Events

(OCE’s)(OCE’s)

• 7 OCE’s from 5 study days• Narcotic analgesia

• 5/7 OCE’s

• Additional non-AAC methods• Head Nods• Hand Gestures

Tim’s AAC UseTim’s AAC Use

• Most utilized mode• Keyboard feature

Utilized bilateral hands predominantly index fingers and thumb

• 6 available “pop-up” icons with additional methods• Effective navigation

General Interactions with General Interactions with AAC UseAAC Use

• Convey feelings to nurse

• Pain

• Anxiety• Establishing need for suctioning• Requesting assistance in bathing• Communication with RN’s, MD’s, family

Aspects of AAC UseAspects of AAC Use

Positive Negative

Ownership Time Consuming

Sense of Control Unfamiliarity

Connection with Others

Use of space

FeedbackFeedback

• Tim• “I can say everything I want to say right

now through typing [VOCA] and writing.”• “I am satisfied with the way I communicate

in the hospital.”

• Tim’s Sister• “Patients need this device until prosthesis

is in place. It is a great help.”

Practical ChallengesPractical Challenges

• Patient lost access to the device when he transferred off of the Head and Neck Unit (to Cardiology)

• Expensive• Nursing, Physician, Clinician unfamiliarity• Battery back up• Infection control issue -- how to keep the

device sterile• Discharge to home without device?

Tim Taught UsTim Taught Us

• Communication method needs to be customized for each patient

• Options for changes/deletions of various messages at all times

• Once a method is established, it is difficult to change or add another method

Results of this exploratory Results of this exploratory study will be submitted for study will be submitted for

publication….publication….

• Stay tuned…you will be able to access more specific data after the manuscript has been accepted to a peer-reviewed journal.

Part IIIPart III

Pilot Study #2 -- Medical Pilot Study #2 -- Medical Intensive Care Unit Intensive Care Unit (MICU)(MICU)

Exploring the Feasiblity of Exploring the Feasiblity of VOCAs withVOCAs with Nonspeaking ICU Nonspeaking ICU

PatientsPatients

M.B. Happ, PhDT. K. Roesch, BSN

MICU SampleMICU Sample(n=11)(n=11)

15 patients identified 11 participated (73%) 7 men, 4 women 10 Caucasian

Observation & InterviewObservation & Interview

Observations: = 49 Communication Events = 41 (83.7%)

Formal Interviews: = 14

Patient = 8Family = 3Clinician = 3

CharacteristicsCharacteristics (n=11)(n=11)

Ages: 20-72 yrs (45.5+16)

Education: 0-16+ yrs (13+1.9)*

Computer Use: 6

*MR patient excluded from mean

CharacteristicsCharacteristics Intubation:

Tracheostomy: 4Oral ET tube: 7

Primary Medical Dx:Pneumonia/ARDS/Sepsis 7Lung CA 1COPD 1Subglottic Stenosis 1SCI 1

CharacteristicsCharacteristics (cont).(cont).

Days w/ device: 1-14 (5.7+ 4.6)

Ventilator Days: 1-44 (15.5+12.2)

APACHE III: 10-54 (27.5+16.1)

Neuromotor CharacteristicsNeuromotor Characteristics (n=11 Study Patients)(n=11 Study Patients)

+ Motor Screen Tasks = 8- Blind, quadriplegia- Quadriplegia- Morbid Obesity

+ Write legibly = 3

Neuromotor CharacteristicsNeuromotor Characteristics(n=49 observations)(n=49 observations)

Narcotic analgesia = 13 (26.5%)

Anxiolytics/sedation = 22(44.9%)

Usage PatternsUsage Patterns

• Ventilated patients in the MICU used VOCA systems in over 1/4 of the observed communication events• However, usage patterns ranged from “limited”

to “required cues to use”.

• Most of the patients used more than one communication method

• Increased patient initiations were associated with availability of the VOCA

Observed VOCA MessagesObserved VOCA Messages “I love you” = 9 FAQs (go home, restraints, breathing

tube) = 4 Anxiety/worry/ fear = 4 Pain = 3 Comfort (thirst, position, cold) = 3 Family =1

Novel Scenarios in which MICU Novel Scenarios in which MICU patients used VOCAs to patients used VOCAs to

communicatecommunicate1. Informed consent – to participate in

research & diagnostic testing2. Semantically complex message

building3. Patient initiated messages

What is your religion? Is the house clean? I want my sister!

QualityQuality Patient ratings of “Ease of

Communication” increased significantly in the VOCA versus no VOCA (pretreatment) condition.

Anecdotal Reports of Anecdotal Reports of SatisfactionSatisfaction

That [VOCA] was a good thing there, it really helped me. (patient)

It was easier to understand what she wanted. I can’t read sign language…I’m not a good guesser. (husband)

I think it’s more complete and decisive. (RN)

SatisfactionSatisfactionSuggested Design Improvements

Larger screens Greater touch sensitivity Easier keyboard access (DynaMyte) Simplier – less expandable (DynaMyte) Realtime Tracking/Storage of Messages Backlighting (MessageMate)

BarriersBarriers poor positioning/out-of-reach UE weakness blurred vision fluctuating cognition/attention deterioration in condition

BarriersBarriers

Staff time constraints Lack of knowledge about device Device complexity

BarriersBarriers

It was easier for me to talk with him, and not have to pull out the device, because time is precious around here… Where he could get his point across to me with lip talking, it seemed to lessen the time… - RN

Partner Behaviors that Partner Behaviors that Facilitated VOCA useFacilitated VOCA use

Cueing patients in selection of messages

Repositioning patient or device Aids: glasses, hearing, access tools Patience with slow message

generation Improved condition and UE strength

What we learned about AAC…What we learned about AAC… Start simple Basic instruction card SLP support Tech support Partner training

What we learned about AAC…What we learned about AAC…

Use progressive, expandabletechniques

Capitalize on combined methods Cueing Consistency Repeat instructions

For further information and For further information and specific data from Study #2:specific data from Study #2:

• Keep an eye out for the following article:• Happ, M.B., Roesch, T.K., & Garrett, K.L.

(in press --expected 2004). Exploring the use of electronic VOCAs in the medical intensive care unit. Heart & Lung, 33, issue 2 or 3.

Part IVPart IVIntroduction to theIntroduction to theSPEACSSPEACS Project Project

Time for a large-scale study…Time for a large-scale study…• A “large n” study across multiple ICU

units • Planned prospective design with 3

patient/nurse cohorts• Treatment: A systematically designed

AAC and basic communication intervention “package” implemented by nurses and an SLP

• Quantitative analysis of the INTERACTIONS between the nonspeaking patient AND the primary nurse caregiver

SPEACS:SPEACS:SStudy of tudy of PPatient-Nurse atient-Nurse EEffectiveness with ffectiveness with AAssisted ssisted CCommunication ommunication SStrategiestrategies

Multidisciplinary Research TeamMultidisciplinary Research TeamMary Beth Happ, Ph.D., R.N.

Kathryn Garrett, Ph.D., CCC-SLPSusan Sereika, Ph.D.

Elisabeth George, Ph.D., R.N.Michael Donahoe, M.D.Judith Tate, M.S., R.N.

* * * * * * * * * * * School of Nursing University of Pittsburgh

Duquesne University University of Pittsburgh Medical Center

Expert consultants:Maria Connolly, B.S.,R.N. -- Loyola UniversityMelanie Fried-Oken, Ph.D., CCC-SLP -- OHSUNeville Strumpf, Ph.D., R.N. -- U. of Penn

5-Year Funding (2003 -- 2008)5-Year Funding (2003 -- 2008): : National Institute of Child Health and National Institute of Child Health and

Human Development (NICHHD)Human Development (NICHHD)

* * * * * * * * * ** * * * * * * * * *““Improving Communication with Improving Communication with

Nonspeaking Patients in the ICU” Nonspeaking Patients in the ICU” (R01-HD043988-01)(R01-HD043988-01)

OverviewOverview Background and Rationale Research Questions & Study Aims Research Design & Model Independent Variables: Description of 2-Phase

Intervention Packages Procedures Dependent Variables/Data Collection Data Analysis Potential Challenges Invitation to Comment

Definition of Definition of Augmentative & Augmentative & Alternative Communication Alternative Communication

(AAC):(AAC): All communication methods that supplement natural speech including unaided (signing, vocalizations) or aided (writing, typing, electronic device) techniques

- from Beukelman & Mirenda, 1998

Natural ApproachesNatural Approaches

Mouthing words

Writing

Gesture

• Natural, minimally aided communication strategies are the most frequently used by nonspeaking patients in the ICU.

• Typically, AAC devices are not available.• Problems with relying on natural communication alone

can include:• Mouthing: Patients often cannot clearly mouth words

around the endotracheal tube• Writing: Paper/pen is not made available, the patient is

illiterate, or upper extremity function is inadequate• Gestures: Patients/nurses have no consistently shared

gestural lexicon (Connolly, 1992)• Opportunities: Patients do not receive adequate

opportunities to initiate their own topics and messages (e.g., “Please find my reading glasses”)

• Rate: Message co-construction can be a slow process

Prosthetic Oral ApproachesProsthetic Oral Approaches

Electrolarynx

Tracheostomy one-way speaking valve

Aided Strategies:Aided Strategies:Low tech symbol boards/Low tech symbol boards/

direct selection direct selection

MessageMate

DynaMyteTM

TM

Electronic VOCAsElectronic VOCAs• • synthesized or digitized synthesized or digitized voice outputvoice output• • symbolized messagessymbolized messages• • multiple level optionmultiple level option• • scanning optionscanning option

ChallengesChallenges• AAC is not considered “customary care”

• Nurses do not have easy access to AAC technologies• Nurses do not receive training in their use• Natural communication strategies and/or AAC

technologies are not applied systematically to all conscious ICU patients

• Communication strategies are not individualized for specific patients

• Ongoing consultation about communication strategies typically is not available for nurses in the ICU

SPEACS:SPEACS:SStudy of tudy of PPatient-Nurse atient-Nurse EEffectiveness with ffectiveness with AAssisted ssisted CCommunication ommunication SStrategiestrategies

RQ/Specific Aim #1RQ/Specific Aim #1

What is the impact of two experimental interventions…

(1)Basic Communication Skills Training (BCST) for nurses

(2)AAC techniques and education + individualized SLP consultation

(AAC-SLP)

…on ease, quality, frequency and success of nurse-patient communication?

RQ/Specific Aim #2RQ/Specific Aim #2

How do interactions in the two communication intervention conditions (BCST and AAC-SLP) compare with those in a control (usual care) cohort?

Research ModelResearch Model

Interventions

AAC/SLPBCST

NurseCommunicationProcess

VoicelessPatient

Outcomes

SuccessEase

QualityFrequency

Happ, M.B. & Garrett, K.L. (2003)

AAC-SLP > BCST > Control on:

easequalityfrequencysuccessfulness

of nurse-patient communication interactions.

Our HYPOTHESIS:

Research DesignResearch Design

Nonconcurrent Cohort DesignNonconcurrent Cohort Designwith Repeated Measureswith Repeated Measures

Year 1Control

T1 T2 T3 T4

Year 2BCST

X1T1 T2 T3 T4

Year 3AAC-SLP

X2T1 T2 T3 T4

• Medical ICU • Cardiothoracic ICU

2 Settings2 Settings

Independent Independent VariablesVariables

Condition 1 - Usual TreatmentCondition 1 - Usual Treatment

• No specific communication training for nurses

• Communication interaction and intervention at the discretion of the patient or untrained nurses

Condition 2 -- BCSTCondition 2 -- BCST• Training for nurses in basic communication

skills prior to data collection• Delivery:

• 2 hour inservice (instruction & roleplay) with SLP <2 months prior to data collection

• Website consistently available

Sample Basic Communication Sample Basic Communication SkillsSkills

• Approach patient• Alert patient (“George…”)• Tag yes/no questions (“Yes…or No?”)• Provide auditory or written choices • Ask open-ended questions when appropriate (“Tell me what’s on

your mind.”)• Instruct patients to use specific natural modalities if they do not

initiate• Show me one of the gestures we talked about.• Write it for me.• Can you mouth the words more clearly?

• Interpret utterances/mirror gestures

Condition 3 -- AAC + SLPCondition 3 -- AAC + SLP• Incorporates basic communication skills

training• SLP also works with nurse to develop

individualized communication intervention plan for each patient.

• SLP also sets up AAC technologies, conducts message inventory, teaches patient, and trains nurse as appropriate

• SLP is available on an ongoing basis to consult with nurse about communication

Nurse Sample Nurse Sample (quasi-random selection)(quasi-random selection)

5 RNs/unit = 10 RNs x 3 phases

= 30 RNsRN Entry Criteria: 1-year critical care experience Full-time staff, not permanent night Selected from pool of volunteers

Patient SamplePatient Sample 3 pts/RN = 30 pts x 3 phases

= 90 patients

Patient Entry Criteria:• Respiratory intubation • Likely to remain intubated for a min of 48 hrs• Understand English• Glasgow Coma Scale > 13

Exclusion :• Premorbid inability to communicate verbally or

nonverbally (a score of <3 on the NOMS cognition, expressive, and receptive language subscales

• Delirium or limited movement OK

Dependent Dependent VariablesVariables

Data SourcesData Sources• Transcriptions of videorecorded

nurse-patient interactions • 3 minute segments -- 2x/day for

2 days for each nurse/patient dyad

• Observer ratings• Field Notes• Clinical record/chart

• Videotapes of the 2-minute nurse/patient interactions will be transcribed and coded for the following variables:• How frequently did the patient initiate

communication?• With which modality?• How many of the nurse-patient communication

exchanges resulted in successful message communication?

• How many breakdowns occurred? How many were successfully repaired?

• How often did the nurse demonstrate behaviors that facilitated communication?

• What was the function of the message?

Observer Ratings of Ease of Observer Ratings of Ease of CommunicationCommunication

Ease of Communication Observer Rating

1. Overall how difficult was it for the patient to communicate with the nurse?

Not difficult <1 2 3 4 5 6 7> Extremely Difficultat all

2. How difficult was it for the patient to communicate physical needs (such asbeing suctioned, being turned, etc.)?

Not difficult <1 2 3 4 5 6 7> Extremely Difficultat all

3. How difficult was it for the patient to communicate thoughts and feelings?

Not difficult <1 2 3 4 5 6 7> Extremely Difficultat all

4. Overall the nurse appeared to feel ________ at the end of the interaction

Calm/satisfied <1 2 3 4 5 6 7> Frustrated/angry

• Field Notes will also be compiled for qualitative analysis of:• Setting variables

• Topics

• Affect

• Unusual circumstances

• Presence of restraints

• Patient’s cognitive status

• Etc.

Data SampleData Sample

4 observations/pt x 30 pts/phase = 120 observations/phase x 3 phases

360 observations

CovariatesCovariates• Will specific patient or nurse variables

explain/predict patterns in the data?

• Patient Co-variates• Gender• Type of ICU• Premorbid communication ability

• Measured by subscales of the NOMS• Severity of Illness (APACHE)• Length of Intubation prior to study

enrollment • Degree of Agitation (CAM-ICU)• Degree of Sedation (RASS)• Motor Ability (Lowenstein)

• Nurse Co-variates

• Total nurse contact time with patient

• Time elapsed since training

• Critical care experience

VoicelessPatient

CommunicationProcess

Outcomes

Interventions

AAC/SLP

Nurse

BCST

Level of Consciousness

Illness Severity

Communication Fx

Motor Fx

SuccessEase

QualityFrequency

Nurse Contact Time

Time Elapsed Since Training

Data Analysis (S.S.)Data Analysis (S.S.)• Exploratory data analysis• Hierarchical generalized linear

modeling (HGLM)• Linear contrasts based on

hypotheses• Model assessment (i.e., residual

analysis and evaluation of outlier/ influential observations)

Potential Problems & Potential Problems & SolutionsSolutions• Brief ICU stays/2 day data collection period• Variable nurse scheduling/ day nurses only, request

same patient• Fluctuation in patient condition/ track delirium and

severity of illness as a co:variate• Diffusion of the intervention/ assess in 2 ICUs, use 3

separate cohorts• Measurement intrusiveness and complexity/ extra effort• Is 2 days enough time to develop an effective

communication intervention?/ oh well -- it represents the real life challenge!

Our timelineOur timeline• January 2004: Final Instrument Development &

Pilot Testing of Procedures• March 2004: Nurse/Patient enrollment for

Usual Care Condition• March 2005: Begin BCST Condition

• January 2006: Begin AAC-SLP Condition

• January 2007: Data Analysis

• July 2008: Complete Data Summarization

Questions and Comments Questions and Comments from the Audiencefrom the Audience

HandoutsHandouts• Please cite information from this

presentation as follows: ******• Correspondence:

• Mary Beth Happ, Ph.D., R.N.• University of Pittsburgh• [email protected]

• Kathryn Garrett, Ph.D., CCC-SLP• Duquesne University• [email protected]