Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
592� September/October 2009, Volume 63, Number 5
Kettle Test—A Brief Measure of Cognitive Functional Performance: Reliability and Validity in Stroke Rehabilitation
KEY WORDS• cognition• reproducibility of results• stroke• task performance and analysis
Adina Hartman-Maeir, PhD, OT, is Lecturer, School of Occupational Therapy, Hadassah and Hebrew University, Mount Scopus, POB 24026, Jerusalem 91240 Israel; [email protected]
Hagit Harel, MSc, OT, is Clinician, Rehabilitation Department, Sheba Medical Center, Tel Aviv, Israel.
Noomi Katz, PhD, OTR, is Director, Research Institute for Health and Medical Professions, Ono Academic College, Israel, and Professor Emeritus, School of Occupational Therapy, Hadassah and Hebrew University, Jerusalem, Israel.
OBJECTIVES. We examined the reliability and validity of the Kettle Test, a brief performance measure based on a complex everyday task designed to tap into basic and higher level cognitive processes.
METHOD. Participants included 21 people attending stroke rehabilitation and 4 occupational therapists for the reliability analysis, 36 people at discharge from stroke rehabilitation, and 36 age-matched healthy control participants for the validity analyses. Instruments included a battery of conventional cognitive measures and functional outcomes.
RESULTS. Interrater reliability was found to be high. Stroke survivors at discharge from rehabilitation were found to require significantly more assistance on the Kettle Test than control participants (p < .000); their scores on the Kettle Test were significantly and moderately correlated with the conventional cognitive and functional outcome measures.
CONCLUSIONS. The results support the reliability and validity of the Kettle Test as a top-down measure of cognition-in-function in people at discharge from stroke rehabilitation.
Hartman-Maeir, A., Harel, H., & Katz, N. (2009). Kettle Test—A brief measure of cognitive functional performance: Reliability and validity in stroke rehabilitation. American Journal of Occupational Therapy, 64, 592–599.
Adina Hartman-Maeir, Hagit Harel, Noomi Katz
People�with�acquired�brain�injury�are�at�high�risk�for�cognitive�impairments�that�have�a�detrimental�impact�on�occupational�performance�(Cicerone�et�al.,�2005;�
Donnovan�et�al.,�2008;�Hershkovitz�&�Brill,�2007;�Zinn�et�al.,�2004).�Occupational�therapy�has�developed� a�unique�body�of�knowledge� regarding� the� relationship�between�cognition�and�occupation� involving�occupation-based,� client-centered�intervention� (Gillen,� 2009;�Katz,� 2005).�Conventional� “tabletop”�measures� of�cognition�are�valuable�diagnostically�but�have�limited�ecological�validity�and�do�not�fully�address�the�functional�implications�of�cognitive�deficits.�Moreover,�the�interdisciplinary�rehabilitation�community�has�emphasized�the�need�for�ecologi-cally�valid�measures�of�cognition�in�function�(Burgess�et�al.,�2006).�This�current�emphasis�is�an�important�milestone�in�the�conceptualization�of�measurement�in�health-related�professions.�Instruments�that�follow�these�directives�are�essential�and�are� integral� to� the� occupational� therapy� philosophy� and� practice� (American�Occupational�Therapy�Association,�2008).�Occupational�therapy�is�in�a�unique�position�to�fulfill�a�central�role�in�this�development�and�is�establishing�its�expertise�in�the�development�of�top-down�assessments�(Hartman-Maeir,�Katz,�&�Baum,�2009;�Law,�Baum,�&�Dunn,�2005).
Evaluation�of�outcomes�in�stroke�rehabilitation,�beyond�basic�activities�of�daily�living�(BADLs),�has�become�extremely�important�as�more�people�survive�stroke�and�return�to�their�homes�in�the�community.�One�primary�concern�at�discharge�from�rehabilitation�is�predicting�the�ability�to�function�independently�in�the�community�or�the�degree�of�assistance�required�to�do�so.�Instrumental�activities�of�daily�living�
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
The American Journal of Occupational Therapy� 593
(IADLs)�are�more�complex�than�BADLs�and�thus�require�basic�and�higher-level�cognitive�functions�such�as�executive�functioning�(Cahn-Weiner�et�al.,�2007;�Coster,�Haley,�Jette,�Tao,�&�Seibens,� 2007;�MacNeill�&�Lichtenberg,� 1997).�Thus,�to�address�this�important�outcome�during�stroke�reha-bilitation,�occupational�therapists�have�a�need�for�top-down,�performance-based�assessments�that�will�incorporate�the�cog-nitive�complexity� involved� in� IADLs.�A� small�number�of�performance-based�measures�in�occupational�therapy�fulfill�these�requirements�and�incorporate�cognitive�challenges�in�functional� IADL� contexts;� these� measures� include� the�Cognitive� Performance� Test� (CPT;� Burns,� 2006),� the�Assessment� of� Motor� and� Process� Skill� (AMPS;� Fisher,�2006a,�2006b),� and� the�Executive�Function�Performance�Test� (EFPT;�Baum,�Morrison,�Hahn,�&�Edwards,�2003;�Baum�et� al.,� 2008).�These� instruments� are� valuable� top-down�measures�that�address�cognitive�and�process�skills�in�IADLs.�The�CPT�and�EFPT�each�have�four�or�five�IADL�tasks�that�provide�a�wide�range�of�activities�in�this�domain.�The� scores� on� the�CPT�measure� the� six� cognitive� levels�according� to� the� Cognitive� Disabilities� Model� (Allen,�Earhart,�&�Blue,�1992;�Levy�&�Burns,�2005),�and�the�scores�of� the�EFPT�measure� executive� functioning�components.�The�AMPS�is�an�extensively�researched,�reliable,�and�valid�measure�of�motor�and�process�skills�in�IADLs;�however,�it�requires�intensive�and�expensive�training,�which�is�not�avail-able�for�all�occupational�therapists.
The�focus�of�this�study�is�on�the�development�of�the�Kettle�Test�(Hartman-Maeir,�Armon,�&�Katz,�2005),�a�brief�performance�measure�that�is�based�on�a�complex�everyday�task�and�is�designed�to�tap�into�basic�and�higher-level�cogni-tive�processes.�The�aim�of�the�test�is�to�evaluate�the�ability�for�independent�community�living�of�people�with�identified�or�suspected�cognitive�disabilities.
Development of the Kettle TestThe�Kettle�Test�(Hartman-Maeir�et�al.,�2005)�was�developed�to�provide�a�brief�performance-based�assessment�of�an�IADL�task�that�can�be�conducted�in�a�clinical�setting�or�at�home.�The�test�was�designed�to�tap�into�a�broad�range�of�cognitive�skills�within�a�functional�context�to�assist�in�the�clinical�deci-sion-making�process�regarding�the�need�for�assistance�in�daily�living�skills�of�adults�with�suspected�cognitive�disabilities.�The� basic� task� of� preparing� a� hot� beverage� was� chosen�because�of� its� functional� significance,�broad� cultural� rele-vance,�and�feasibility�(in�terms�of� tools�and�time�require-ments)� in�multiple� settings� (e.g.,� clinic,�home).�The� task�complexity�was�elevated�to�enable�the�assessment�of�basic�and�higher-level�cognitive–functional�skills�that�are�required�for�IADLs,�with�the�following�elaborations:
1.� The�essential�task�of�preparing�oneself�a�hot�beverage�was�expanded�to�include�preparing�an�additional�cup�of�hot�beverage�for�the�therapist;�the�additional�cup�differs�in� two� ingredients� from� that� requested�by� the� client,�providing�additional�load�on�working�memory.
2.� The�electric�kettle�is�emptied�and�disassembled�(lid�and�electric� cable� are�disconnected� from� the�body�of� the�kettle)� to� challenge�problem-solving� skills� and� safety�judgment�regarding�the�use�of�electricity.
3.� Additional�kitchen�utensils�and�ingredients�are�placed�in�the�immediate�task�environment�(on�a�tray)�as�dis-tracters�to�increase�attention�demands.This�design�creates�a�brief�task�using�familiar�objects�and�
settings�that�are�presented�in�a�manner�that�targets�cognitive�skills�underlying�complex�ADLs.�The�observation�is�struc-tured�so�that�the�rater�is�required�to�score�the�performance�on�13�discrete�steps�of�the�task�(e.g.,�turning�on�the�faucet,�filling�the�kettle�with�2�cups�of�water).�Clear�guidelines�for�cueing�are�provided,�and�the�rater�scores�each�step�according�to�the�degree�of�cueing�that�was�necessary�to�complete�the�step�(0�through�4).�Total�scores�range�from�0�to�52�(higher�scores�indicate�more�assistance)�and�can�be�transformed�to�clinically�meaningful�categories�of�independence�on�the�task�(independent,�mild�assistance,�considerable�assistance).
Initial�research�to�examine�convergent�and�ecological�validity�was�conducted�on�the�Kettle�Test�in�a�sample�of�41�elderly� clients� (mean� age� =� 75.2)� referred� to� a� geriatric�assessment�clinic�because�of�suspected�cognitive�disabilities�(Hartman-Maeir,�Katz,�&�Armon,�2004).�Small�to�moder-ate�significant�correlations�were�found�between�Kettle�Test�scores�with�conventional�measures�of�cognition�(r�=�.56�with�Mini-Mental� Status� Evaluation� [MMSE;� Folstein� &�Folstein,�1975];�r�=�.59�with�Clock�Drawing�Test�[CDT;�Freedman�et�al.,�1994];�and�r�=�.32�with�the�visual�attention�Star�Cancellation�subtest�of�the�Behavioral�Inattention�Test�[BIT;�Wilson,�Cockburn,�&�Halligan,�1987),�and�moderate�correlations�were�found�with�caregiver�ratings�of�ADLs�(r�=�.53)�and�IADLs�(r�=�.58).�This�initial�study�provided�sup-port�for�the�validity�of�the�instrument.�Therefore,�the�pur-pose�of�this�study�was�to�further�examine�the�reliability�and�validity�of�the�Kettle�Test�in�a�geriatric�stroke�population�faced�with�similar�concerns�regarding�the�impact�of�cogni-tive� deficits� on� daily� living� abilities� at� discharge� from�rehabilitation.
The�study�was�divided�into�two�stages:�(1)�The�first�consisted�of�the�study�of�interrater�reliability�of�the�Kettle�Test�between�two�sets�of�raters�in�two�rehabilitation�set-tings,�and�(2)�the�second�consisted�of�the�study�of�different�aspects�of�validity�in�a�sample�of�people�at�discharge�from�stroke� rehabilitation� and� healthy� elderly� control� partici-pants.�The�objective�of�the�first�stage�was�to�examine�the�
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
594� September/October 2009, Volume 63, Number 5
interrater�reliability�of�the�Kettle�Test�scores�as�administered�by� certified� occupational� therapists� to� stroke� patients� in�rehabilitation.�The�objectives�of�the�second�stage�were�to�examine•� The�construct�validity�of�the�Kettle�Test,�in�terms�of�its�
ability� to� differentiate� between� scores� of� people� after�stroke�at�discharge�from�rehabilitation�and�scores�of�age-matched�healthy�control�participants�who�live�indepen-dently�in�the�community;
•� The�convergent�validity�of� the�Kettle�Test,� calculating�relationships�between�Kettle�Test� scores�with� conven-tional�measures�of�cognition;�and
•� The�ecological�validity�of�the�Kettle�Test,�examining�the�relationship�of�safety�level�and�BADL�status�at�discharge�and�IADL�status�at�home,�1�month�after�discharge
Method
Stage 1: Interrater Reliability Study
Participants�were�recruited�from�two�centrally�located�geri-atric� rehabilitation� hospitals� in� Jerusalem� and� Tel� Aviv.�Inclusion�criteria�required�participants�to�be�admitted�to�a�rehabilitation�hospital�within�1�month�after�stroke,�be�ages�60�or�older,�be�alert�and�without�receptive�aphasia,�and�live�independently�in�the�community�before�stroke.�Twenty-one�patients�consecutively�admitted�to�each�rehabilitation�hos-pital�who�met�inclusion�criteria�participated�in�the�reliability�study�(Hospital�1:�n�=�10;�4�men,�6�women;�mean�age�=�79.3,�standard�deviation�[SD]�=�5.8.�Hospital�2:�n�=�11;�6�men,�5�women;�mean�age�=�77.82,�SD�=�5.1).�The�study�was�approved�by�the�Human�Rights�Helsinki�Committee�of�each�hospital,�and�all�participants�signed�informed�con-sent.�Four�certified�and�experienced�occupational�therapists�(two� in� each�hospital)� rated�patient�performance�on� the�Kettle�Test.�The�test�was�administered�by�one�of�the�raters�while�the�other�observed�the�assessment.�Independent�rat-ings�were�then�conducted�immediately�after�the�assessment.�Interrater�reliability�(Spearman�correlation�coefficient)�was�found�to�be�high�for�the�Kettle�Test�total�scores�for�both�sets�of�raters�(Hospital�1:�r�=.851,�p�=�.001;�Hospital�2:�r�=�.916,�p�=�.000).
Stage 2: Validity Study
Participants. Thirty-six� stroke�patients� (18�with� right�hemisphere�stroke�and�18�with�left�hemisphere�stroke)�were�recruited�from�a�geriatric�rehabilitation�hospital�before�their�discharge� (mean� length�of� stay�=�63.1�days,�SD� =�29.2).�Inclusion�criteria�were�the�same�as�delineated�in�the�reliabil-ity�study.�In�addition,�36�healthy�control�participants�were�recruited�from�a�convenience�sample�of�healthy�elderly�vol-
unteers.�The�inclusion�criteria�for�the�control�participants�were�age�60�or�older,�living�independently�in�the�commu-nity,�and�scoring�>23�on�the�MMSE;�see�Table�1�for�demo-graphic�data).�The�study�was�approved�by�the�Human�Rights�Helsinki�Committee�of� the�hospital,� and� all� participants�signed�informed�consent.
Instruments. A�short�battery�of�conventional�standard-ized�measures�of�cognition�recommended�in�geriatric�stroke�rehabilitation�(Adunsky,�Fleissig,�Levenkrohn,�Arad,�&�Noy,�2002)�was�used�in�this�study.�The�battery�included�the�fol-lowing�measures:•� The�MMSE�is�a�universal�measure�of�cognitive�status�that�
assesses�a�broad�range�of�basic�cognitive�abilities�including�orientation,� attention,� recall,�working�memory,� spatial�abilities,�and�language.�Scores�range�from�0�to�30,�and�scores� >23� are� considered� the� cutoff� for� cognitive�impairment.
•� The�CDT�is�widely�used�in�cognitive�screening�for�many�neurological�conditions�(Freedman�et�al.,�1994).�In�the�stroke�population,�several�scoring�systems�of�a�free-style�drawing�of�a�clock�with�hands�at�10�past�11�have�been�shown�to�have�construct�validity�and�correlate�with�mul-tiple�cognitive�domains,�such�as�abstract�thinking,�execu-tive� functioning,� and� visuospatial� construction� (Suhr,�Grace,�Allen,�Nadler,�&�McKenna,�1998).�In�this�study,�we�used�the�scoring�system�of�Rouleau,�Salmon,�Butters,�Kennedy,�and�McGuire�(1992),�with�a�score�range�from�0�to�10,�which�was�found�to�be�reliable�and�valid�in�stroke�rehabilitation�(Suhr�et�al.,�1998).
•� The�Star�Cancellation� subtest� of� the�BIT�was�used� to�measure� visual� attention,� as� recommended� for� use� in�stroke�rehabilitation�practice�(Edwards�et�al.,�2006).
•� The�Cognitive�scale�of�the�FIM™�(CognFIM;�Granger,�1998)�was�used�as�a�measure�of�cognitive–functional�sta-tus�as�expressed�in�daily�activities�(based�on�general�obser-vation)�in�five�areas�of�cognition�(expressive�and�receptive�language,�problem�solving,�social�interaction,�and�mem-ory).�Each� area� is� rated�on� a�7-point� scale;� total� score�ranges�from�5�to�35.�Reliability�and�validity�have�been�extensively�established�in�the�stroke�population�(Heruti,�Lusky,�&�Dankner,�2002).
•� The� functional� outcome�battery� included� the� areas� of�BADLs,� IADLs,� and� safety—outcomes� necessary� for�independent�community�living.
•� BADLs�were�measured�with�the�Motor�scale�of�the�FIM�(Granger,�1998)�because� it� is� a�universal,� reliable,� and�valid�measure�of�rehabilitation�outcome.�The�FIM�Motor�scale�includes�13�items�of�basic�self-care�rated�on�a�7-point�scale�(total�scale�score�range�=�13–91).
•� IADLs�were�measured�with� the� IADL�scale� (Lawton�&�Brody,�1969;�Lawton,�Moss,�Fulcomer,�&�Kleban,�1982),�
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
The American Journal of Occupational Therapy� 595
a�widely�used�measure�for�independence�in�extended�areas�of�ADLs,�including�telephone�use,�shopping,�food�prepara-tion,�housekeeping,�laundry,�transportation,�taking�medica-tion,�and�financial�management�(total�score�range�=�0–23).
•� Safety�was�measured�with�the�Safety�Rating�scale,�which�is�part�of�the�Routine�Task�Inventory�(RTI–E),�a�struc-tured�observation�used� to� rate� cognitive� levels� in�daily�routine�activities�(Allen�et�al.,�1992;�Katz,�2006).�The�4-point�scale�ranges�from�3�(unable to recognize the need for safety precautions)�to�6�(anticipates hazards and plans safety procedures).�The�reliability�and�validity�of�the�RTI–E�has�been� studied� in�populations�with� cognitive�disabilities�(Katz,�2006).
•� The�Fugl-Meyer�Motor�Assessment�(FMA;�Fugl-Meyer,�Jääsko,�Leyman,�Olsson,�&�Steglind,�1975;�Rabadi�&�Rabadi,�2006)�Upper-Extremity�scale�was�used�to�mea-sure�motor�status�to�evaluate�the�possible�confounding�relationship�between�motor� function� after� stroke�with�performance�on�the�Kettle�Test.
Procedure
The� assessment�battery�was� administered�within� the� last�week�before�discharge�from�the�rehabilitation�hospital.�The�Kettle�Test�was� administered�by�Hagit�Harel.�The�other�assessments�were�administered�by�experienced�occupational�therapists�and�rehabilitation�professionals�as�part�of�the�rou-tine�discharge�assessment�battery�of�the�rehabilitation�depart-ment.�In�addition,�to�examine�IADL�status�1�month�after�discharge,�study�group�follow�up�included�a�telephone�inter-
view�administered�to�caregivers.�The�follow-up�assessment�was� completed�on�29�participants,�because�7�participants�could�not�be�reached.�The�Kettle�Test�scores�of�the�7�partici-pants� that� were� lost� to� follow-up� were� not� significantly��different�from�those�of�the�other�participants�(t[34]�=�.892,�p�=�.403)
Data Analysis
Data�were�analyzed�with�SPSS�(Version�15.0;�SPSS,�Inc.,�Chicago).�Descriptive� statistics� and� analysis�of� covariance�(ANCOVA)�were�used�to�examine�the�group�effect�on�the�Kettle�Test�while�controlling�for�years�of�education,�which�were�found�to�differ�significantly�between�groups.�Pearson�correlation�analyses�were�used�to�examine�the�relationships�among�the�cognitive�and�functional�measures�and�motor�and�demographic�variables.�No�significant�differences�were�found�between�participants�with�right-�and�left-hemisphere�stroke�on�the�Kettle�Test�(t[34]�=�0.178,�p�=�.860);�therefore,�the�analyses�were�conducted�on�the�entire�stroke�group.
ResultsBefore�studying�validity,�we�examined�the�relationships�of�the�Kettle�Test�with�demographic�and�motor�variables.�The�correlations�between�the�Kettle�Test�scores�and�age,�years�of�education,�and�FMA�were�all�low�and�nonsignificant�(rs�=�.04,�–.14,�and�–.19,�respectively).�Regarding�gender,�no�sig-nificant�differences�were�found�between�men�and�women�on�the�Kettle�Test�(t�[34]�=�.435,�p�=�.67).
Table 1. Demographic Variables and Performance on the Kettle Test by Group
Stroke Control
Demographics n (%) n (%) Results and Significance
Gender Male 22 (61) 7 (19) χ2(1) = 12.99
p = .000 Female 14 (39) 29 (81)Total 36 (100) 36 (100)
Range M (SD) Range M (SD)
Age (years) Education (years)
60–89 0–19
74.81 (7.32) 10.44 (5.28)
60–84 0–19
72.67 (6.59) 10.44 (5.28)
t(70) = 1.30, ns t(70) = –2.31 p = .02413.10 (4.05)
Kettle Test scoresTotal scorea 1–29 9.34 (5.79) 0–3 0.42 (0.91) F (1,60)b = 63.53
p = .000
Assistance level n (%) n (%)
Independent 5 (17) 35 (97) χ2(3) = 43.53 and 3.53t = .000Mild assistance 13 (45) 1 (3)
Considerable assistance 11 (38) 0 (0)
Note. m = mean; SD = standard deviation; ns = not significant.aHigher scores represent more assistance needed.bAnalysis of covariance, controlling for years of education.
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
596� September/October 2009, Volume 63, Number 5
Construct Validity
As�seen�in�Table�1,�the�Kettle�Test�scores�of�the�stroke�group�were�significantly�higher�than�those�of�the�control�group,�showing�the�need�for�more�assistance�on�the�test.�ANCOVA,�controlling�for�years�of�education�(because�a�significant�dif-ference�was�found�between�groups�on�this�variable),�demon-strated�a�large�significant�group�effect�on�the�test�(F [1,�60]�=�63.53,�p�=�.000).�The�study�group�showed�a�wide�range�of�performance�on�the�Kettle�Test�(1–29),�whereas�the�control�group� showed� a�narrow� range� (0–3),� and�only�1� control�participant�required�mild�assistance,�demonstrating�a�floor�effect�on�the�test�for�this�group.
Convergent Validity
The�mean�scores�of� the� study�group�on�the�conventional�battery� of� cognitive� measures� are� presented� in� Table� 2.�Results�show�a�wide�range�of�performance�on�all�measures.�The�Kettle�Test,�which�targets�cognitive�abilities�in�a�func-tional�context,�was�found�to�moderately�significantly�corre-late�with�these�conventional�cognitive�measures�(Table�2).�The�correlations�of�the�Kettle�Test�with�the�four�cognitive�measures�ranged�from�.478�to�.659�(all�significant�at�p�<�.01),�the�highest�being�with�the�CognFIM,�which�is�a�measure�based�on�observation�in�ADLs.
Ecological Validity
The�mean�scores�of�the�stroke�group�on�the�functional�out-comes�were�as�follows:�76.09�(SD�=12.19)�on�the�FIM�Motor�scale,�4.94�(SD�=�0.75)�on�the�Safety�Rating�scale,�and�10.70�(SD�=�5.07)�on�the�IADL�scale�at�follow-up.�The�correlations�of� the�Kettle�Test� and� the�other� cognitive�measures�with�these� functional�outcomes� are�presented� in�Table�3.�The�Kettle�Test�scores�were�significantly�correlated�with�all�three�outcome�measures,� FIM�Motor� (r� =� –.759),� Safety� (r� =��–.571),� and� IADL�1�month�after�discharge� at�home� (r�=��–.505).�The�correlations�between�the�other�cognitive�mea-sures� and� functional�outcomes�were� lower,� and�no� single�measure�correlated�significantly�with�all�three�outcomes.
DiscussionThe�findings�of� this� study�provide� initial� support� for� the�reliability�and�validity�of�the�Kettle�Test�in�stroke�rehabilita-tion.�We�demonstrated�that�the�Kettle�Test—a�structured�observation�of�performance�in�a�complex�task—can�be�scored�consistently�among�professional�occupational�therapists.�In�addition,�participants�after�stroke�required�significantly�more�assistance� than� healthy� control� participants,� who� scored�almost�perfectly�on�the�test.�Moreover,�the�test�performance�of�participants�after�stroke�was�not�related�to�their�motor�status� or� educational� background,�but� it�was�moderately�correlated� with� their� conventional� cognitive� test� scores.�Finally,�Kettle�Test� scores�were� found� to�be� significantly�correlated�with�functional�outcomes�at�the�time�of�discharge�from�the�rehabilitation�hospital�and�at�home.
The�results�concerning�interrater�reliability�were�encour-aging,�considering�the�difficulty�of�reliably�assessing�complex�functional�tasks.�This�positive�finding�can�be�attributed�to�the�structured�scoring�scale�of�the�observation�(divided�into�discrete�steps)�and�the�expertise�of�the�raters,�who�were�occu-pational�therapists�trained�in�observations�of�functional�per-formance.�Further� reliability� analysis� is� in�progress�using�videotape� recordings� that�will� enable�comparisons�among�multiple�raters.�Another�aspect�of�reliability�pertaining�to�the�stability� of� performance� (test–retest� reliability)� was� not�deemed�relevant�for�the�test.�Because�the�test�incorporates�an�element�of�novel�everyday�problem�solving,�we�hypothe-sized�that�immediate�learning�would�occur�from�the�expo-sure�to�the�task�and�examiner�cueing.�Thus,�similar�to�other�tests�that�incorporate�novel�problem�solving,�performance�is�not�expected�to�be�identical�in�a�subsequent�immediate�retest�(Wilson,�Alderman,�Burgess,�Emslie,�&�Evans,�1996).
The�construct�validity�of�the�test�was�highly�supported�by�the�significant�differences�that�were�found�between�the�stroke�and�control�groups.�This�group�effect�could�not�be�attributed�to�the�difference�in�education�because�the�effect�
Table 2. Conventional Cognitive Measures: Descriptive Statistics and Pearson Correlation Coefficients With Kettle Test Scores
Cognitive MeasuresMinimum–Maximum M (SD)
r With Kettle Test
Mini-Mental Status Evaluation 15–30 26.11 (3.96) –.478*Clock Drawing Test 2–10 8.16 (1.85) –.566*BIT Star Cancellation subtest 44–54 52.94 (2.33) –.578*FIM Cognitive scale 15–35 30.38 (4.44) –.659*
Note. M = mean; SD = standard deviation; BIT = Behavioral Inattention Test.*p < .01.
Table 3. Pearson Correlation Coefficients of Functional Outcomes With the Kettle Test and Conventional Cognitive Measures
BADLs(FIM Motor
Scale)
Safety Level(RTI Safety
Rating Scale)a
IADLs(IADLs Scale)
Kettle Test –.759** –.571** –.505**Mini-Mental Status Evaluation .261 .122 .401*Clock Drawing Test .365* .112 .182Star Cancellation subtest of BIT .462** .102 .237FIM Cognitive scale .435** .446** .287
Note. BADLs = basic activities of daily living; RTI = Routine Task Inventory; BIT = Behavioral Inattention Test; IADLs = instrumental activities of daily living.aSpearman correlation analysis was performed on this variable because of its ordinal 4-point scale.*p < .05. **p < .01.
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
The American Journal of Occupational Therapy� 597
remained�when�controlling�for�years�of�education.�Moreover,�education�was�not�found�to�be�significantly�related�to�per-formance�on�the�Kettle�Test,�and�no�difference�was�found�between�the�performance�of�men�and�women�on�the�test.�Therefore,�the�differences�between�the�groups�support�the�construct�validity�of�the�test�to�differentiate�between�known�groups�with� identified�differences� in� cognitive� functional�status.�We�expected�that�healthy�adults�who�are�living�inde-pendently�in�the�community�would�succeed�on�the�Kettle�Test;� this� expectation�was� indeed� confirmed�by� the�floor�effect�among�the�control�group,�in�which�only�1�participant�needed�mild�assistance�on�the�test.�The�findings�suggest�that�the� test� can�detect� cognitive–functional�deficits� in� adults�with�neurological�dysfunction;�however,�the�test�is�not�sensi-tive�to�possible�variance�in�cognitive–functional�performance�of�normative�healthy�adults.�Further�research�is�required�on�larger,�more�representative�samples�of�different�diagnostic�groups�with�identified�cognitive�disabilities.
The�significant�moderate�correlations�that�were�found�with�conventional�established�measures�of�cognition�support�the�convergent�validity�of�the�Kettle�Test,�demonstrating�the�expected�strength�of�association�(r�=�.50)�between�measures�with�overlap�of�construct�(Gregory,�2000).�The�correlations�with�the�“table-top”�tests�(MMSE,�CDT,�and�Star�Cancella-tion)� suggest� that� common�underlying� cognitive� abilities,�including�attention,�memory,�praxis,�and�executive�functions,�are�being�tapped�by�these�measures.�However,�the�moderate�degree�of�association�supports�the�contention�that�the�Kettle�Test�is�also�tapping�into�additional�aspects�of�performance�not�addressed�by�these�conventional�cognitive�measures.
These�findings�are�similar�to�the�initial�findings�regarding�the�Kettle�Test�in�adults�with�suspected�dementia�(Hartman-Maeir�et�al.,�2004)�and�to�those�of�other�studies�of�top-down�assessments�such�as�the�AMPS�and�the�EFPT,�where�moder-ate�associations�were�found�with�table-top�measures�of�cogni-tion�(Baum�et�al.,�2008;�Fisher,�2006a).�The�significant�cor-relation�that�was�found�between�the�Kettle�Test�scores�and�the�FIM�supports�the�validity�of�the�test�as�a�measure�of�cognition�in� function.�The�FIM�Cognitive� scale�provides� a�valuable�measure�of�functional�cognition�in�stroke�rehabilitation;�how-ever,�it�relies�on�the�day-to-day�observations�of�clinicians�in�the�rehabilitation�setting�and�has�a�high�focus�on�BADLs.�This�setting�does�not�necessarily�provide�an�opportunity�to�observe�the�client� in� the�more�complex� tasks� that� are� essential� for�independent� living� in� the�community�after�discharge.�The�moderate�association�that�was�found�may�shed�light�on�the�possible�unique�contribution�of�a�short�IADL�performance�test,�beyond�a�general�observation�in�daily�life.
The�ecological�validity�of�the�Kettle�Test�was�substan-tially�supported�by�the�significant�correlations�with�all�func-tional�outcomes�(ADLs,�safety,�and�IADLs).�Conversely,�the�
pattern�of�correlations�between�these�outcomes�and�the�con-ventional� cognitive� measures� showed� fewer� and� smaller�effects,�and�no�single�measure�correlated�significantly�with�all�outcomes.�The�outcome�measures�that�were�chosen�for�this�study�represent�functional�areas�that�are�pertinent�issues�for�discharge�planning�after�stroke�and�for�independent�com-munity�living.�The�findings�clearly�support�the�advantage�of�the�Kettle�Test�in�its�relation�to�these�outcomes.
From�our� clinical� experience,� the� ecological� and� face�validity�of�the�Kettle�Test�for�discharge�planning�has�been�shown� to�be�valuable� for� clients� and� their� caregivers.�On�discharge�from�hospital,�the�need�for�assistance�in�ADLs�and�safety�status�is�of�major�concern�in�planning�for�community�living.�Questions� such� as� “Can� I� return�home�alone?”�or�“What�type�of�assistance�will�I�need�to�live�independently�in�the�community?”�are�of�vital�importance.�For�example,�study�participant�A.�R.�is�a�73-year-old�married�man�who�was�liv-ing�independently�in�the�community�before�his�stroke.�He�did�not�have�any�residual�motor�deficits�at�this�stage�of�reha-bilitation�and�was�independent�in�BADLs,�yet�his�conven-tional� cognitive� test� scores� revealed� some�mild�deficits� in�measures�of�clock�drawing�and�visual�attention.�A.�R.’s�per-formance�on�the�Kettle�Test�demonstrated�a�significant�need�for�assistance�on�multiple�steps:�He�was�baffled�by�the�empty�kettle,�had�difficulty�connecting�the�electrical�cord,�and�only�prepared�one�cup�of�beverage�(instead�of�two),�using�cold�water.�The�face�validity�of�this�assessment�for�the�client�and�his�wife�was�useful�in�accepting�the�current�need�for�assis-tance�on�going�home.
Conclusions, Limitations, and Future RecommendationsThe�Kettle�Test� is� a�brief,� cognitive–functional,� top-down�measure�that�assesses�actual�performance�on�a�familiar�IADL�task�with�built-in�complexity.�The�results�of�this�study�support�the�reliability�and�validity�of�this�measure�in�stroke�rehabilita-tion.�The�Kettle�Test�can�be�used�in�diverse�settings;�it�is�short,�easy�to�learn�and�administer,�and�provides�meaningful�infor-mation�pertaining�to�independent�community�living.
The�study�was�limited�by�a�relatively�small�sample�size�that�did�not�enable�multiple�regression�analysis�for�examin-ing�the�unique�prediction�of�the�Kettle�Test�to�the�explained�variance�of�functional�outcomes�beyond�other�conventional�measures�commonly�used�in�geriatric�stroke�rehabilitation.�In�addition,�the�data�regarding�IADLs�were�collected�by�a�telephone� interview�with� caregivers,�which�may�not�have�fully�captured�this�area�of�functioning.�Further�studies�should�include�home�visits� to� confirm�and� replicate� the� current�findings�in�the�geriatric�stroke�population�and�in�other�diag-nostic�groups�with�cognitive�disabilities.� s
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
598� September/October 2009, Volume 63, Number 5
ReferencesAdunsky,�A.,�Fleissig,�Y.,�Levenkrohn,�S.,�Arad,�M.,�&�Noy,�S.�
(2002).�Clock�drawing�task,�Mini-Mental�State�Examination�and�cognitive–functional�independence�measure:�Relation�to�functional�outcome�of�stroke�patients.�Archives of Gerontology and Geriatrics, 35,�153–160
Allen,�C.�K.,�Earhart,�C.�A.,�&�Blue,�T.� (1992).�Occupational therapy treatment goals for the physically and cognitively dis-abled.� Bethesda,� MD:� American� Occupational� Therapy�Association.
American�Occupational�Therapy�Association.�(2008).�Occupational�therapy�practice�framework:�Domain�and�process�(2nd�ed.).�American Journal of Occupational Therapy, 62(6),�625–683.
Baum,�C.�M.,�Connor,�L.�T.,�Morrison,�T.,�Hahn,�M.,�Dromerick,�A.�W.,�&�Edwards,�D.�F.�(2008).�Reliability,�validity,�and�clinical�utility�of�the�Executive�Function�Performance�Test:�A�measure�of�executive�function�in�a�sample�of�people�with�stroke.� American Journal of Occupational Therapy,� 62(4),�446–455.
Baum,� C.,� Morrison,� T.,� Hahn,� M.,� &� Edwards,� D.� (2003).�Executive Function Performance Test: Test protocol booklet.�St.�Louis,�MO:�Program�in�Occupational�Therapy,�Washington�University�School�of�Medicine.
Burgess,�P.�W.,�Alderman,�N.,�Forbes,�C.,�Costello,�A.,�Coates,�L.�M.-A.,�Dawson,�D.�R.,� et� al.� (2006).�The� case� for� the�development� and� use� of� “ecologically� valid”� measures� of�executive� function� in� experimental� and� clinical�neuropsy-chology.�Journal of the International Neuropsychological Society, 12,�194–209.
Burns,�T.�(2006).�Cognitive Performance Test (CPT).�Pequannock,�NJ:�Maddak.
Cahn-Weiner,�D.,�Tomaszewski�Farias,�S.�T.,�Julian,�L.,�Harver,�D.�J.,�Kramer,�J.�H.,�Reed,�B.�R.,�et�al.�(2007).�Cognitive�and�neuroimaging�predictors�of� instrumental� activities� of�daily�living.�Journal of the International Neuropsychological Society, 13,�747–757.
Cicerone�K.�D.,�Dahlberg,�C.,�Malec,�J.�F.,�Langenbahn,�D.�M.,�Felicetti,�T.,�Kneipp,�S.,�et�al.�(2005).�Evidence-based�cog-nitive�rehabilitation:�Updated�review�of�the�literature�from�1998–2002.�Archives of Physical Medicine and Rehabilitation, 86,�1681–1692.
Coster,� W.,� Haley,� S.� N.,� Jette,� A.,� Tao,� W.,� &� Seibens,� H.�(2007).� Predictors� of� basic� and� instrumental� activities� of�daily�living�performance�in�persons�receiving�rehabilitation�services.�Archives of Physical Medicine and Rehabilitation, 88, 928–935.
Donnovan,�N.�J.,�Kendall,�D.�L.,�Heaton,�S.�C.,�Kwon,�S.,�Velozo,�C.�A.,�&�Duncan,�P.�W.�(2008).�Conceptualizing�functional�cognition�in�stroke.�Neurorehabilitation and Neural Repair, 22,�122–135.
Edwards,�D.�F.,�Hahn,�M.�G.,�Baum,�C.�M.,�Perlmutter,�M.�S.,�Sheedy,�C.,�&�Dromerick,�A.�W.�(2006).�Screening�patients�with�stroke�for�rehabilitation�needs:�Validation�of�the�post-stroke� rehabilitation� guidelines.� Neurorehabilitation and Neural Repair,�20,�42–48.
Fisher,�A.�G.�(2006a).�Assessment of motor and process skills. Vol. 1: Development, standardization, and administration manual�(6th�ed.).�Fort�Collins,�CO:�Three�Star�Press.
Fisher,�A.�G.�(2006b).�Assessment of motor and process skills. Vol. 2: User manual�(6th�ed.).�Fort�Collins,�CO:�Three�Star�Press.
Folstein,�M.�F.,�&�Folstein,�S.�E.�(1975).�Mini�Mental�State:�A� practical� method� for� grading� the� cognitive� state� of�patients� for� the� clinician.� Journal of Psychiatric Research, 12,�189–198.
Freedman,�M.,�Leach,�L.,�Kaplan,�E.,�Winocur,�G.,�Shulman,�K.,�&�Delis,�D.�C.�(1994).�Clock drawing: A neuropsychological analysis.�New�York:�Oxford�University�Press.
Fugl-Meyer,�A.�R.,�Jääsko,�L.,�Leyman,�I.,�Olsson,�S.,�&�Steglind,�S.�(1975).�The�post-stroke�hemiplegic�patient:�A�method�for�evaluation�of�physical�performance.�Scandinavian Journal of Rehabilitation Medicine, 7,�13–31.
Gillen,�G.�(2009).�Managing�executive�function�impairments�to�optimize� function.� In�Cognitive and perceptual rehabilita-tion: Optimizing function (pp.�245–283). New�York:�Mosby/�Elsevier.
Granger,�C.�V.�(1998).�The�emerging�science�of�functional�assess-ment:�Our� tool� for�outcomes� analysis.�Archives of Physical Medicine and Rehabilitation,�79,�235–240.
Gregory,�R.� J.� (2000).�Psychological testing: History, principles, and applications�(3rd�ed.). Needham�Heights,�MA:�Allyn�&�Bacon.
Hartman-Maeir,� A.,� Armon,� N.,� &� Katz,� N.� (2005).� Kettle Test protocol.� Jerusalem:�School�of�Occupational�Therapy,�Hadassah�and�Hebrew�University�of�Jerusalem.
Hartman-Maeir,�A.,�Katz,�N.,�&�Armon,�N.�(2004,�July).�Validity of a cognitive–functional observation (the “Kettle Test”) in an elderly sample with suspected dementia.�Paper�presented�at�the�Israeli�Society�for�Occupational�Therapy�Annual�Conference,�Haifa.
Hartman-Maeir,�A.,�Katz,�N.,�&�Baum,�C.� (2009).�Cognitive�Functional�Evaluation�(CFE)�process�for�individuals�with�sus-pected�cognitive�disabilities.�Occupational Therapy in Health Care, 23, 1–23.
Hershkovitz,� A.,� &� Brill,� S.� (2007).� The� association� between�patients’� cognitive� status� and� rehabilitation�outcome� in� a�geriatric� day� hospital.� Disability and Rehabilitation,� 29,�333–337.
Heruti,�R.� J.,�Lusky,�A.,�&�Dankner,�R.� (2002)�Rehabilitation�outcome�of�elderly�patients�after�a�first�stroke:�Effect�of�cogni-tive�status�at�admission�on�the�functional�outcome.�Archives of Physical Medicine and Rehabilitation, 83,�742–749.
Katz,�N.� (2005).�Cognition and occupation across the life span: Models for intervention in occupational therapy.�Bethesda,�MD:�AOTA�Press.
Katz,�N.�(2006).�Routine Task Inventory–RTI–E manual. Retrieved�March�11,� 2009,� from�www.allen-cognitive-network.org/images/stories/pdf_files/rtimanual2006.pdf
Law,�M.,�Baum,�C.,�&�Dunn,�W.�(2005).�Measuring occupational performance: Supporting best practice in occupational therapy (2nd�ed.).�Thorofare,�NJ:�Slack.
Lawton,�M.�P.,�&�Brody,�E.�M.� (1969).�Assessment� of� older�people:�Self-maintaining�and�instrumental�activities�of�daily�living.�Gerontologist,�9,�179–186.
Lawton,�M.�P.,�Moss,�M.,�Fulcomer,�M.,�&�Kleban,�M.�(1982).�A�research-�and�service-oriented�multilevel�assessment.�Journal of Gerontology, 37,�91–99.
Levy,�L.�L.,�&�Burns,�T.�(2005). Cognitive�disabilities�reconsid-ered:�Rehabilitation�of�older�adults�with�dementia. In�N.�Katz�
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms
The American Journal of Occupational Therapy� 599
(Ed.),�Cognition and occupation across the life span: Models for intervention in occupational therapy�(pp.�347–388).�Bethesda,�MD:�AOTA�Press.
MacNeill,�S.�E.,�&�Lichtenberg,�P.A.�(1997).�Home�alone:�The�role�of�cognition�in�return�to�independent�living.�Archives of Physical Medicine and Rehabilitation,�78,�755–758.
Rabadi,�M.�H.,�&�Rabadi,�F.�M.� (2006).�Comparison�of� the�Action� Research� Arm� Test� and� the� Fugl-Meyer� assess-ments�as�measures�of�upper�extremity�motor�weakness�after�stroke.�Archives of Physical Medicine and Rehabilitation,�87,�962–966.
Rouleau,�I.,�Salmon,�D.�P.,�Butters,�N.,�Kennedy,�C.,�&�McGuire,�K.� (1992).�Quantitative� and�qualitative� analyses� of� clock�drawings�in�Alzheimer’s�and�Huntington’s�disease.�Brain and Cognition, 18,�70–87.
Suhr,�J.,�Grace,�J.,�Allen,�J.,�Nadler,�J.,�&�McKenna,�M.�(1998).�Quantitative�and�qualitative�performance�of�stroke�versus�nor-mal�elderly�on�six�clock�drawing�systems.�Archives of Clinical Neuropsychology,�13,�495–502.
Wilson,�B.�A.,�Alderman,�N.,�Burgess,�P.�W.,�Emslie,�H.,�&�Evans,�J.�J.�(1996).�Behavioral Assessment for Dysexecutive Syndrome.St.�Edmunds,�England:�Thames�Valley�Test�Company.
Wilson,� B.,� Cockburn,� J.,� &� Halligan,� P.� (1987).� Behavioral Inattention Test: Manual.�Suffolk,�England:�Thames�Valley�Test�Company.
Zinn,�S.,�Dudley,�T.�K.,�Bosworth,�H.�B.,�Hoenig,�H.�M.,�Duncan,�P.�W.,�&�Horner�R.�D.�(2004).�The�effect�of�poststroke�cog-nitive� impairment�on�rehabilitation�process�and�functional�outcome.�Archives of Physical Medicine and Rehabilitation,85,�1084–1090.
Essential Practice Guideline From AOTA Press!
Call 877-404-AOTA Shop Online at http://store.aota.org
By Joyce Sabari, PhD, OTR, FAOTA
This�evidence-based�guideline�details�the�signifi�cant�contribution�of�occupational�therapy�in�treating�adults�dealing�with�the�limitations�of�having�had�a�stroke.�Appendixes�include�valuable�resources�such�as�CPTTM codes�related�to�occupational�therapy�for�stroke�survivors.
Practice Guidelines can
■��Help�clarify�occupational�therapy�(OT)�services�for�external�audiences
■��Help�determine�appropriateness�of�OT�referrals
■�� Assist�third-party�payers�in�determining�medical�necessity�for�OT
■���Assist�OT�educators�in�design�of�appropriate�curricula. ORDER #2211
AOTA Members: $49.00Nonmembers: $69.50
ISBN: 978-1-56900-263-6
Occupational Therapy Practice Guidelines for
Adults With Stroke
BK-130
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930156/ on 10/09/2018 Terms of Use: http://AOTA.org/terms