Professional Documents
Culture Documents
Developing A Clinician-Friendly Aphasia Test: Clinical Focus
Developing A Clinician-Friendly Aphasia Test: Clinical Focus
Robert C. Marshall
University of Kentucky, Lexington
Purpose: The Kentucky Aphasia Test (KAT) is Results: Results with the KAT clearly differenti-
an objective measure of language functioning for ated the language performance of individuals
persons with aphasia. This article describes with and without aphasia. NBD participants made
materials, administration, and scoring of the KAT; few errors, and overall scores on the test for
presents the rationale for development of test individuals with aphasia were rarely within 1 SD
items; reports information from a pilot study; of the NBD group. Performance of the partici-
and discusses the role of the KAT in aphasia pants with aphasia administered KAT-1, KAT-2,
assessment. and KAT-3 suggested that the 3 versions of the
Method: The KAT has 3 parallel test batteries, test represent a hierarchy of difficulty.
KAT-1, KAT-2, and KAT-3. Each battery contains Conclusions: The KAT remains in its early
the same orientation test and 6 subtests, each stages of development. However, it does appear
with 10 items, assessing expressive and recep- to meet the requirements for a “ clinician-friendly ”
tive language functions. Subtests for KAT-1, aphasia test and, as such, offers a rapid, con-
KAT-2, and KAT-3 systematically increase in venient means of obtaining an objective score
difficulty so that it is possible to assess individ- to determine changes in language functioning
uals with severe, moderate, and mild aphasia, during the early postonset period.
respectively. The KAT was administered to
38 participants with aphasia and 31 non-brain-
damaged (NBD) participants. Key Words: aphasia, test, managed care
A
phasia test batteries have been used by clinicians to Few would dispute the need to conduct a comprehensive
assess persons with aphasia (PWA) for nearly a cen- assessment of the PWA before starting intervention. The
tury. The first test batteries used to characterize the MTDDA, PICA, BDAE, and WAB have met this need for
speech, language, and cognitive deficits of PWA were devel- decades, but using these tests to assess PWA in today’s
oped shortly before and after World War II (Eisenson, 1946; health care system is problematic for several reasons (Golper
Goldstein, 1948; Head, 1926; Weisenburg & McBride, 1935). & Cheney, 1999). The first is that clinicians have less time to
Additional measures were developed between 1960 and 1982 devote to assessment now than in the era before managed
as interest in aphasia rehabilitation grew and objective mea- care. The BDAE and MTDDA have multiple subtests and,
sures were needed to measure the effects of its treatment. Some in our clinical experience, can take up to 2 hr to give,
tests such as Examining for Aphasia (Eisenson, 1946), the particularly if the client’s aphasia is severe. The PICA and
Language Modalities Test for Aphasia (Wepman & Jones, the 1982 version of the WAB can usually be completed in
1961), and the Neurosensory Center Comprehensive Exam- less than an hour. However, 40 hr of training ( Porch, 1967)
ination for Aphasia (Spreen & Benton, 1977) are rarely used are required for a clinician to be able to use the PICA’s
today. Others, however, such as the Minnesota Test for Dif- multidimensional scoring system reliably. This may be
ferential Diagnosis of Aphasia (MTDDA; Schuell, 1972), impractical for many clinicians ( Lincoln, 1988). A second
Porch Index of Communicative Ability (PICA; Porch, 1981), problem is that clinicians are now obligated to conduct an
Boston Diagnostic Aphasia Examination (BDAE; Goodglass assessment of the PWA earlier in the poststroke course,
& Kaplan, 1983), and the Western Aphasia Battery (WAB; frequently at bedside. Both the PICA and the WAB contain
Kertesz, 1982), continue to have widespread use (Byng, Kay, materials (cards, objects, pictures) and forms that restrict
Edmundson, & Scott, 1990) and are taught to graduate students their administration in less than optimal settings, particularly
in communication disorders and sciences as part of their if the client is not medically stable. Finally, most aphasia
clinical training (Brookshire, 2003). tests in use today were designed to assess PWA in the middle
American Journal of Speech-Language Pathology • Vol. 16 • 295–315 • November 2007 • A American Speech-Language-Hearing Association 295
1058-0360/07/1604-0295
Downloaded From: http://ajslp.pubs.asha.org/ by a New York University User on 04/10/2016
Terms of Use: http://pubs.asha.org/ss/rights_and_permissions.aspx
of the severity continuum. It has been suggested that these permits the clinician to assess any client with aphasia with
lack a sufficient “ top ” (difficult tests) or “ bottom ” (easier the KAT and eliminates the need to devise hybrid protocols
tests) to adequately assess clients with mild and severe to assess clients at upper and lower ends of the severity
aphasia, respectively (Brookshire, 2003; Darley, 1983; continuum. The scoring system of the KAT requires no
Miller, Willmes, & de Belser, 2000; Raymer & LaPointe, special training to use. It combines features of the multi-
1986). When more time was available for testing, clinicians dimensional scoring system of the PICA ( Porch, 1967) and
could compensate for this by administering supplementary the communication-based system of the Communication
tests, and in some cases, devising their own measures. Activities of Daily Living, Second Edition (CADL–2;
However, managed care neither supports nor allows time Holland, Fratalli, & Fromm, 1998). Thus, it allows the
for this. examiner to record response features such as delays and
One way to compensate for the constraints on aphasia self-corrections, and it gives the client credit for respond-
assessment brought about by managed care is to develop ing correctly in modalities other than speaking (gesture,
“ clinician-friendly ” tests. Ideally, these would be measures drawing, writing, pointing).
that (a) could be given in their entirety in a short time frame, The purposes of this article are to (a) describe the ma-
( b) compensate for floor and ceiling effects and could be terials, administration, and scoring of the KAT and the
used with PWA across the severity continuum, and (c) are rationale underlying development of test items; (b) report test
convenient to administer in all patient care settings. Time scores and sensitivity information for participants with and
spent in testing PWA can be reduced by using short versions without aphasia and provide information on the scoring
of aphasia test batteries ( Disimoni, Keith, & Holt, 1975; and test–retest reliability; and (c) discuss the role of the
Goodglass, Kaplan, & Barresi, 2001; Kertesz, 2006; Schuell, KAT in present-day assessment practices of PWA.
1972) and aphasia screening tests (Crary, Haak, & Malinski,
1989; Fitch-West & Sands, 1998; Helm-Estabrooks, 1992;
Keenan & Brassell, 1975; Sklar, 1983). However, modifi- KAT Development
cations of longer aphasia tests and screening tools are rarely Orientation
standardized (Golper & Cheney, 1999) and do not meet, or To begin the test, the clinician administers the orientation
only partially meet, other desirable features of a clinician- test shown in Appendix A. The orientation test is identical
friendly test. for KAT-1, KAT-2, and KAT-3. It requires the client to per-
This article provides information on the Kentucky form 10 tasks involving reading, writing, and pointing. Items
Aphasia Test ( KAT; Marshall & Wright, 2002). The KAT for the test were constructed similarly to those used in the
is an impairment-based, objective measure of language func- Reading Comprehension Battery for Aphasia, Second Edi-
tioning for use with individuals with aphasia secondary to a tion (LaPointe & Horner, 1998), the CADL–2 (Holland et al.
stroke. It is intended to provide clinicians with a means to 1998), and the Assessment of Language-Related Functional
quantify changes in language functioning during the early Activities (Baines, Martin, & Heeringa, 1999). The orienta-
postonset period when the individual is moving from one tion test is used (a) to establish rapport with the client, (b) to
patient care setting to the next.1 In developing this exper- “settle” the individual into the testing situation, and (c) to
imental version of the KAT, the time and physical constraints provide the clinician with supplementary information about
imposed on clinicians working in managed care settings the client’s speech and language functioning.
were given paramount consideration. The KAT contains only
an orientation test, a picture description task, and six 10-item
subtests to assess expressive and receptive functions. Read- Picture Description Task
ing and writing subtests were not included in the KAT, not The client’s performance on the picture description task
because they are not important to assess, but because these guides the clinician’s decision to administer KAT-1, KAT-2,
skills are usually the most impaired in aphasia, take more or KAT-3. For this task, the client describes the divided
time to assess, and tend to receive less attention in the early attention picture shown in Figure 1. This picture was
posttreatment period when the focus is on improving developed by providing a commercial artist with several
comprehension and message exchange skills that will allow types of elicitation stimuli that have been used to elicit
the client to communicate with his or her caregivers (Holland connected speech samples from clients with communication
& Fridriksson, 2001; Marshall, 1997; Murray & Holland, disorders. Using these as guidelines, the artist constructed
1995). The KAT has three parallel test batteries: KAT-1, several drawings, one of which was eventually chosen by the
KAT-2, and KAT-3. The three batteries increase systemati- authors. For this task, the examiner places the picture in front
cally in difficulty and complexity in order to facilitate of the client and says, “ Tell me what is going on in this
assessment of individuals with severe, moderate, and mild picture. ” The client is given as much time as needed to
aphasia, respectively. This “three-in-one ” arrangement complete the task. In the development of the KAT, 63 non-
brain-damaged (NBD) adults (33 men and 30 women)
ranging from 19 to 78 years of age (M = 36.60, SD = 16.04)
1
We consider the early postonset period to encompass the first 3 months and having from 12 to 18 years of education (M = 14.78,
after a stroke causing aphasia. This is a time when the individual may be SD = 2.10) described the picture. Their narratives were
seen for speech and language assessment and treatment in a variety of
settings (acute care, general hospital ward, rehabilitation, nursing home,
transcribed verbatim, and the content units (nouns and verbs)
home health, outpatient) and when most rehabilitation services are received listed in Appendix B were identified. The number of content
in the managed care system. units produced by the NBD volunteers ranged from 6 to 24
(M = 16.64, SD = 4.62). This value falls midway between version being given to a fluent client with severe aphasia
those of younger (M = 18.0, SD = 4.7) and older NBD and good motor skills.
participants (M = 14.7, SD = 3.6) who described the Cookie
Theft picture from the BDAE in a study by Yorkston and
Beukelman (1980). We made a decision to administer KAT-1 Subtests
to any PWA who produced from 0 to 5 content units because The KAT has six 10-item subtests. Three—Picture
this was fewer than the number of content units from any Naming, Repetition Span, and Defining Words—assess
NBD participant, and we opted to administer KAT-2 and expressive abilities. Three others—Following Commands,
KAT-3 when the number of content units was 6 to 10 or ≥11, Yes/ No Questions, and Word-to-Picture Matching—assess
respectively. This decision was primarily made to be con- receptive functions. Stimuli for each of the subtests, instruc-
sistent across the three versions of the KAT, but it should tions, administration procedures, scoring, and other special
be pointed out that these values are consistent with those circumstances surrounding administration of each subtest
from Yorkston and Beukelman (1980) for participants are provided in Appendix C.
with low–moderate (M = 10.5, SD = 2.5) and mild aphasia Picture Naming. Items for the Picture Naming subtest are
(M = 16.4, SD = 3.3). black-and-white drawings approximately 2.5 in. × 3 in. in
Whereas the results of the picture description task do not size. Task difficulty was determined on the basis of frequency
contribute to the overall score for the KAT, this task does of occurrence (Kucera & Francis, 1967) of the target words.
provide information about the client’s articulation, use of Mean frequencies for words selected for KAT-1, KAT-2,
propositional language, semantic production, and syntax that and KAT-3 are 125.4 (SD = 58.3), 55.1 (SD = 14.6), and 17.4
is useful in characterizing the client’s speech and language (SD = 11.3), respectively. Frequency of occurrence indicates
abilities. Our rationale for using results of the picture de- how common the target words are; the numbers represent
scription task as an “indicator ” of severity and to deter- the average frequency of occurrence per 1 million words.
mine which version of the KAT to give was based on the Following Commands. On this subtest, the client fol-
fact that in the early postonset period, most PWA are lows spoken commands requiring the identification of body
concerned about their verbal communication status, and parts. KAT-1 involves one- and two-step commands involv-
the patient’s ability to communicate verbally is a common ing body parts only (e.g., make a fist). KAT-2 commands
standard by which early progress is judged. We are cognizant increase in difficulty by adding right-left distinctions (e.g.,
that using the number of content units produced on a picture make a fist with your left hand). Two-step commands
description task as an indicator of severity has some lim- involving right-left discriminations are also used for KAT-3,
itations. For example, in some cases it could result in a but the commands are made more difficult by having the
less difficult version of the test being given to a client with client make distinctions between the adverbs before and after
a co-occurring motor speech problem or a more difficult (e.g., after you touch your right knee, raise your hand).
Subtest M SD M SD M SD M SD M SD M SD
Orientation 22.9 13.8 23.0 13.3 36.8 10.8 38.0 9.1 31.0 21.0 34.7 22.4
Picture Naming 24.8 12.6 28.5 13.4 41.3 3.5 41.8 3.0 33.3 20.6 29.3 17.8
Commands 32.6 10.8 30.7 12.3 46.3 2.6 48.3 3.5 28.3 16.4 34.7 17.2
Repetition Span 34.0 18.4 33.2 16.3 43.5 11.1 44.0 8.1 30.0 17.6 32.3 20.1
Yes/ No 34.3 12.7 36.4 10.7 41.5 6.0 42.5 6.4 40.7 6.1 43.3 9.0
Defining Words 28.2 13.3 30.8 13.3 26.5 13.0 26.5 12.6 31.0 19.0 32.3 19.4
Word-to-Picture 35.6 12.8 34.4 9.4 42.8 3.3 48.0 2.5 33.0 20.1 34.0 15.7
Total score 212.4 80.0 217.0 75.2 278.5 37.0 289.0 34.5 227.3 114.0 240.7 119.2
Note. Maximum total score for each subtest = 50; maximum total score overall = 350.
p < .0001, and the Picture Naming, F(2, 60) = 38.55, p < .0001, this test should be the case for less severe clients. Thus,
and Word-to-Picture Matching, F(2, 60) = 22.75, p < .0001, participants completing KAT-3 would be expected to have
subtests. Planned comparisons were performed to identify the highest scores, those taking KAT-2 the next highest
group differences. The a priori p value was set at .05. Mul- scores, and those taking KAT-1 the lowest scores. Table 4
tiple comparisons were performed; thus, we controlled for shows the KAT subtest and overall means and standard
familywise error using an adjusted p of .0167. For the total deviations for the participants with aphasia. Several
score, participants yielded significantly better scores for the Kruskal–Wallis one-way ANOVA tests were conducted
KAT-1 version compared with the KAT-2 and KAT-3 ver- and revealed significant differences among groups on the
sions. Participants performed significantly worse on the orientation subtest (H = 8.55, p < .05) and naming subtest
KAT-2 version of the Picture Naming subtest compared (H = 7.87, p < .05). Planned comparisons were performed,
with KAT-1 and KAT-3 versions. Lastly, for the Word-to- and the a priori p value was set at .05 and then controlled for
Picture Matching subtest, participants yielded the lowest familywise error. Results indicated that participants who
score for the KAT-3 version compared with the KAT-1 completed the KAT-1 version performed significantly worse
and KAT-2 versions. on the orientation subtest as well as the naming subtest
Performance by participants with aphasia. Thirty-eight compared with participants who completed KAT-2 and
participants with aphasia were administered the KAT at least KAT-3 versions. No other comparisons yielded statistically
one time. We anticipated no differences among the subtests significant differences.
or overall scores for the participants completing KAT-1, Comparing participants with and without aphasia.
KAT-2, or KAT-3. The reason for this was that test items Mann–Whitney U tests were performed to compare KAT
for KAT-1, KAT-2, and KAT-3 were selected so that those performances between participants with and without aphasia.
for KAT-1 would be less difficult than those of KAT-2 and See Table 4 for groups’ performance on the KAT. Twenty
so forth, to coincide with the parallel testing concept of participants with aphasia and 31 NBD adults completed the
the KAT. However, it was anticipated that participants with KAT-1 version. The groups differed significantly for all
aphasia would differ in their performance on the orientation subtests as well as the total score (U = 618, p < .0001). In
subtest. The reason for this was that the orientation subtest all cases, the NBD group had significantly higher scores.
was the same for all participants, and better performance on Using an a priori p value of .05, similar results were found
TABLE 3. Means and standard deviations for participants without aphasia (N = 31) on KAT-1, KAT-2, and KAT-3 subtests and overall.
Subtest M SD M SD M SD
Picture Naming* KAT-2 < KAT-1 & KAT-3 48.5 3.3 42.3 5.8 48.0 4.1
Commands 49.9 0.4 49.4 1.8 48.6 2.2
Repetition Span 49.9 0.5 49.8 0.6 49.4 1.7
Yes/ No 49.5 1.2 48.6 2.1 49.1 1.5
Defining Words 46.4 3.9 48.0 3.3 46.8 2.5
Word-to-Picture Matching* KAT-1 & KAT-2 > KAT-3 49.5 2.2 49.3 2.3 43.2 7.0
Total score* KAT-1 > KAT-2 & KAT-3 343.10 5.92 336.74 9.22 334.42 11.70
Note. Maximum total score for each subtest = 50; maximum total score overall = 350.
*Statistically significant group differences.
PWA (N = 20) NBD (N = 31) PWA (N = 10) NBD (N = 31) PWA (N = 8) NBD (N = 31)
Subtest M SD M SD M SD M SD M SD M SD
Orientation 25.1 12.5 49.4 1.3 36.6 7.8 49.4 1.3 37.9 13.0 49.4 1.3
Picture Naming 26.6 14.5 48.5 3.3 39.7 3.8 42.3 5.8 38.5 13.5 48.0 4.1
Commands 33.4 9.8 49.9 0.4 40.8 9.9 49.4 1.8 31.9 13.7 48.6 2.2
Repetition Span 34.5 17.4 49.9 0.5 43.7 10.3 49.8 0.6 34.9 15.8 49.4 1.7
Yes/ No 36.6 10.5 49.5 1.2 38.9 8.8 48.6 2.1 41.9 6.6 49.1 1.5
Defining Words 29.4 12.7 46.4 3.9 30.4 10.5 48.0 3.3 35.1 13.8 46.8 2.5
Word-to-Picture 37.4 11.0 49.5 2.2 40.8 7.1 49.3 2.3 32.5 13.0 43.2 7.0
Total score 222.9 14.5 343.10 5.92 270.9 38.4 336.74 9.22 250.4 78.1 334.42 11.7
Note. Maximum total score for each subtest = 50; maximum total score overall = 350.
when comparing performances by the aphasia group (N = 10) measures. No PWA had an overall score within 1 SD of the
who completed KAT-2 with the NBD group’s (N = 31) NBD group. Thus the KAT is sensitive to aphasia when its
KAT-2 scores, as well as comparisons between participants overall score is used as a metric. However, as mentioned
with aphasia (N = 8) and without aphasia (N = 31) on the previously, 9%, 27%, and 29% of the participants with
KAT-3 version. That is, for all subtest and total score com- aphasia given KAT-1, KAT-2, and KAT-3, respectively, had
parisons, the NBD group performed significantly better than orientation, expressive, or receptive subtest scores within
the respective aphasia group. 1 SD of the NBD group. Largely, this was a result of the fact
Finally, test sensitivity for accurately differentiating that stimuli on the naming subtest of KAT-2 and the Word-to-
adults with and without aphasia was determined. Using 1 SD Picture Matching subtest of KAT-3 were more difficult for
of the NBD group’s mean as an indicator of test sensitivity, the NBD participants than anticipated. Replacing these items
we calculated the number of participants with aphasia who with easier, less abstract stimuli may improve sensitivity of
scored within this range. None of the participants with the KAT in the future, but this would be premature before
aphasia received total scores within 1 SD of the NBD group’s increasing the database for the KAT, and it does not nec-
means. However, for subtest scores within each KAT level, essarily preclude use of the test at the present time. Also,
some participants with aphasia scored within 1 SD of the Lezak and colleagues (2004) point out that judging the
NBD group’s mean. For KAT-1, 9% of scores by participants “goodness ” of a test on its diagnostic accuracy is a ques-
with aphasia across the subtests were within 1 SD; many tionable assumption because most tests have as their purpose
occurred during the Repetition Span subtest (5 out of the describing an individual’s strengths and weaknesses and
20 participants). For the KAT-2, 27% of scores by partic- monitoring the status of a disorder or disease for planning
ipants with aphasia were within 1 SD. However, most of and treatment.
these occurred during the naming subtest—8 of the 10 par- Preliminary estimates of scoring and temporal reliability
ticipants scored within 1 SD. Similar findings were found for the KAT are relatively high. To administer the KAT,
with KAT-3 scores: 29% of the scores by participants with the clinician needs only a few materials (e.g., test booklet,
aphasia were within 1 SD. The culprits of this number were scoring form, coins for making change, paper and pencil).
the naming and Word-to-Picture Matching subtests. For Other props needed to administer the test (e.g., telephone)
each of these, 4 out of 8 participants scored within 1 SD are usually available at the testing location. Perhaps the most
of the NBD group’s mean. compelling feature of the KAT is that it offers three separate
tests in one clinical package: KAT-1, KAT-2, and KAT-3.
The fact that the overall scores for participants with severe,
Discussion and Clinical Implications moderate, and mild aphasia did not differ significantly for the
The KAT is a first step in the development of a measure three batteries suggests that we are close to establishing a
for time-conscious clinicians in need of a single aphasia test reasonable hierarchy of difficulty for KAT-1, KAT-2, and
with which to obtain an objective score for any client with KAT-3, and are creating different but parallel test protocols.
aphasia. While the time to administer the KAT has not been This is confirmed, in part, by the fact that participants tested
determined empirically, our observations to date are that with KAT-1 performed significantly poorer on the only
clients with aphasia need less than 30 min to complete the test that was the same for each battery, the orientation test,
test. Although the KAT is not a diagnostic test, it appears to but unfortunately participants administered KAT-2 and
be capable of distinguishing among persons with and with- KAT-3 did not differ on the orientation test.
out aphasia. Test sensitivity refers to the probability of
accurately detecting abnormal functioning in an impaired
individual ( Keil & Kaszniak, 2002; Lezak, Howieson, & Potential Clinical Uses of the KAT
Loring, 2004). Our NBD participants made few errors on Managed care challenges clinicians to do more with
the test and performed significantly better than PWA on all less. For many clinicians, this means spending less time in
Appendix B
Content Units Produced by 63 Non-Brain-Damaged Adults in Response to the Divided Attention
Picture Shown in Figure 1
KAT-1
Cup: Drinking utensil, holds liquid, liquid container, mug, holder/container
Hot: Burn / burning, heated / with heat, matter of temperature, not cold, very warm
Large: Big, not small, humungous, great, immense, large
Sleep: Rest, quiet, go to bed, doze, not awake
Throw: Pitch, toss, sling, thrust, launch an object
Go: Move/motion, leave, get out, opposite of stop
Baby: Infant, young child, new born
Long: Measure of length, distance, not short, extended, lengthy duration
Smile: Facial expression, happy, grin
Push: Shove, move forward, opposite of pull
KAT-2
Rose: Flower, sweet smelling
Cab: Taxi, paid / hired transportation, commercial transportation
Couple: Two, two people together, pair
Journey: Trip, travel
Thief: Robber, criminal, crook
Carve: Cut, slice
Prepare: Get /make ready, to fix, plan
Rescue: Save, help, retrieve from danger
Ancient: Old, matter of age
Fake: Not real, false, a lie, not true, phony, sham
KAT-3
Salmon: Fish, food
Hilarious: Funny
Connect: Put together, join, attach two items
Complete: Finish, end, entire, whole
Beggar: Poor man, someone who wants something, soliciting funds
Companion: Friend, partner, loved one, associate, husband /wife
Monument: Statue, memorial, landmark, structure that honors a hero
Share: To give, divide with others, cut in half
Confidential: Private, secret, personal, keep to oneself
Accelerate: Speed up, go fast, increase speed, give more gas