The Cleveland Scale For Activities of Da

P1: FLF/FGL P2: FTK/FLK QC: FTK
Journal of Clinical Geropsychology [jcg] PP004-229800 October 30, 2000 15:7 Style file version Nov. 19th, 1999
Journal of Clinical Geropsychology, Vol. 7, No. 1, 2001
The Cleveland Scale for Activities of Daily

Living (CSADL): Its Reliability and Validity
Marian B. Patterson1,2,3,4 and James L. Mack1,2
This study evaluated the reliability and validity of the Cleveland Scale for Activities of
Daily Living (CSADL), a scale designed to measure in detail specific activities of daily
living in individuals with dementia. Administered to knowledgeable informants by trained
examiners, the CSADL demonstrated good reliability in terms of interrater agreement and
internal consistency. The validity of CSADL total scores was shown by its sensitivity to
degree of cognitive impairment: All comparisons between means of the healthy elderly
group and three groups of AD patients differing in severity were statistically significant.
The CSADL was highly correlated with the Blessed–Roth Dementia Scale (DS-ADL) and
more highly correlated with Mini-Mental State Exam scores than was the DS-ADL.
KEY WORDS: dementia; activities of daily living; CSADL; Alzheimer’s disease.
The Cleveland Scale for Activities of Daily Living (CSADL) was developed to evalu-
ate a broad range of activities of daily living (ADL), including both basic physical activities
and complex activities referred to as instrumental ADLs. Details as to development of the
CSADL are presented in Patterson et al. (1992). The scale was designed to evaluate depen-
dency of patients in a variety of settings (including home, the community, and residential
facilities) and to provide a comprehensive picture of dependency throughout the day, re-
gardless of when it occurs. To accomplish these goals, dependency ratings are obtained from
the patient’s primary caregiver, the individual most likely to have remained in proximity to
the patient over an extended period of time.
Consistent with the recommendations in a recent review of evaluating ADL by Beck
and Frank (1997), we have broken each domain of behavior, for example, “bathing,” into
specific components (“Initiates bath or shower,” “Prepares bath/shower,” “Gets in and out of
tub or shower,” and “Cleans self”). For many of the physical and instrumental domains, the
item content enables one to distinguish between the domain’s executive (initiative, planning,
monitoring, etc.) and more automatic aspects (behaviors that once initiated require little
ongoing regulation). Although items within a domain have an obvious relationship to one
another in terms of their content, no assumptions are made that items within a domain are
1 Department of Neurology, Case Western Reserve University, Cleveland, Ohio.
2 Department of Psychiatry, Case Western Reserve University, Cleveland, Ohio.
3 Department of Psychology, Case Western Reserve University, Cleveland, Ohio.
4 Alzheimer Center, University Hospitals of Cleveland, Cleveland, Ohio.
15
1079-9362/01/0100-0015$19.50/0 °
C 2001 Plenum Publishing Corporation
16 Patterson and Mack
correlated with one another or that they represent either equivalent levels of dependency
or a hierarchy of increasing dependency. One uses the same metric to identify the specific
degree of dependence for each behavior in the scale.
Our choice of an informant-based rating scale is supported in the literature. The find-
ings of Weinberger et al. (1992) indicate that there is good correspondence between in-
formant ratings and self-report for elderly medical patients with high MMSEs, but that
correspondence breaks down, particularly with respect to instrumental activities, when pa-
tient MMSEs are below 24. The breakdown in correspondence suggests the importance
of using informant-based ratings for patients with cognitive impairment. Of course, as has
been emphasized in recent reviews (Beck and Frank, 1997; Spector, 1997), it is important
that the informant be someone well informed about the patient’s behavior.
Our first study (Patterson et al., 1992) presented only information concerning individual
items from basic or “physical” ADL domains. The present study is based on all items from
the scale and provides information regarding summary scores. The goal of this study was to
evaluate the reliability and validity of the CSADL. Specifically, we examined (1) interrater
reliability of individual items and total scores; (2) internal consistency of total scores;
(3) differences in level of dependency between AD patients and comparable groups of
healthy and physically impaired elderly individuals; (4) differences in level of dependency
between AD patients with differing degrees of cognitive impairment; and (5) the relationship
between the CSADL and another measure of ADL, the modified Blessed–Roth Dementia
Scale (Blessed et al., 1968; Morris et al., 1988).
METHOD
Participants
Participants included 410 AD patients, 113 individuals reported in our previous study
(Patterson et al., 1992), and an additional 297 patients who were subsequently evaluated.
In addition, to determine the ability of the CSADL to distinguish between AD patients
and both healthy elderly individuals and those who were physically impaired, we tested
199 elderly individuals in good health and 26 with osteoarthritis/degenerative joint disease.
The AD patients and healthy elderly participants were enrolled in the research reg-
istry of the University Hospitals of Cleveland/Case Western Reserve University Alzheimer
Disease Research Center (ADRC). Written informed consent was obtained from all ADRC
registry participants (and from their caregivers when appropriate). There was no financial
compensation for participation in the registry, but patients were given annual neurologic
examinations and had access to experimental drug trials. As enrollees in the registry, partic-
ipants were evaluated by a neurologist and a neuropsychologist, using standard procedures
of the Consortium to Establish a Registry in Alzheimer Disease (CERAD, Morris et al.,
1988) supplemented by a battery of additional neuropsychological tests. All 410 patients
were diagnosed as having probable AD according to criteria outlined by the task force of
the National Institute of Neurology and Communicative Disorders and Stroke–Alzheimer’s
Disease and Related Disorders Association (Mckhann et al., 1984) and by CERAD (Morris
et al., 1988). Criteria for selection were a progressive decline in intellectual or functional
ability and deficits in memory and at least one other area of cognitive functioning. Any
patient with a current or preexisting medical condition that might account for his/her im-
pairment was excluded.
Reliability and Validity of the CSADL 17
Mean number of years since onset of symptoms for the AD patients was 3.99 (SD =
2.48, range 0.24–16.60, N = 405). Distribution of Clinical Dementia Ratings (CDR,
Hughes et al., 1982; Berg, 1988) among AD patients was as follows: CDR of 0.5, N = 6;
CDR of 1, N = 240; CDR of 2, N = 125; CDR of 3, N = 35; CDR of 4, N = 3. CDRs are
based on the examining physician’s global judgment of the patient’s cognitive and functional
ability levels; ratings are 0, “no impairment;” 5, “questionable;” 1, “mild;” 2, “moderate;”
3, “severe;” 4, “profound;” and 5, “terminal.” Each rating level is anchored with refer-
ence to functioning in six domains: memory, orientation, judgment and problem solving,
community affairs, home and hobbies, and personal care (Berg, 1988).
The physically impaired elderly patients were under care of a specialist for treatment of
osteoarthritis/degenerative joint disease. That physician provided us with names of patients
aged 60 and over, whom he judged to be cognitively normal. Each individual was contacted,
and 26 (3 men and 23 women) agreed to participate. No further demographic data with regard
to these participants was obtained.
The AD patients were divided on the basis of the severity of their cognitive impairment.
Mini-Mental State Exam (MMSE, Folstein et al., 1975) scores were used to classify AD
patients into three groups: mild AD (MMSE ≥ 20), moderate AD (MMSE 11–19), and
severe AD (MMSE ≤ 10). For the 61 patients missing MMSEs, the CDR values were used,
with CDRs of 1 or less classified as “mild,” 2 as “moderate,” and 3 or greater as “severe.”
Demographic data for the healthy elderly group and three groups of AD patients are
presented in Table I. The overall effects of age (F = 9.10, df = 3, p < .01) and education
(F = 50.92, df = 3, p < .01) were significant. Post hoc analyses of age and of education
effects between groups were carried out using Bonferroni adjustments. There was no overall
difference between groups as an effect of gender (χ 2 = 2.91, df = 3, p > .40). The overall
Table I. Basic Information for Healthy Elderly and AD Participants

AD
Variable Healthy elderly Mild Moderate Severe
Age [meana,b,c (SD)] 70.24 (6.31) 72.40 (7.77) 73.92 (8.20) 74.01 (8.24)
N 199 152 177 81
Education [meana,b,c,e (SD)] 15.46 (2.56) 13.31 (2.92) 12.50 (2.91) 11.52 (3.09)
N 197 141 169 73
MMSE [mean (SD)] 28.92 (0.98) 22.82 (2.42) 15.79 (2.39) 6.19 (2.98)
N 196 138 154 57
Gender (N )
Male 86 68 67 29
Female 113 84 110 52
Racea,b,c (N )
Caucasian 190 132 149 63
African-American 5 14 23 14
Residenceb,c,d,e (N )
Home 187 139 153 70
Residential facility 0 1 12 4
Interviewa,b,c,e,f (N )
In-person 90 69 59 21
Phone 0 14 17 26
Note. Superscripts following variable names indicate significant between-group contrasts ( p < .050 or better).
No superscript indicates that the contrast was not significant. (Note that no significance tests were carried out for
MMSE because the AD groups were constructed as nonoverlapping.) The between-group contrasts are as follows:
a Healthy Elderly vs. Mild; b Healthy Elderly vs. Moderate; c Healthy Elderly vs. Severe; d Mild vs. Moderate;
e Mild vs. Severe; f Moderate vs. Severe.
effect of race was significant (χ 2 = 20.68, df = 3, p < .01). Two-by-two group contrasts for
race were carried out using χ 2 tests. With regard to residence, low frequency of individuals
in residential facilities required use of Fisher’s exact probability test to obtain an overall
contrast between healthy elderly and AD participants (all three groups combined). That
contrast was significant ( p < .01). We then carried out an overall test for the effect of
residence within the three AD groups, which was also significant (χ 2 = 7.79, df = 2, p <
.05). Two-by-two contrasts for residence were carried out with Fisher’s exact probability
test. The overall effect of visit type was significant (χ 2 = 61.55, df = 3, p < .01). Two-
by-two group contrasts for interview type were carried out using χ 2 tests. The results of all
post hoc analyses and two-by-two contrasts are presented in Table I.
Measures
The Cleveland Scale for Activities of Daily Living
The CSADL consists of 47 items encompassing 15 domains of everyday activity,

including both physical and instrumental ADL, plus a 48th, nonspecific item inquiring
about “other” dependent behaviors not covered by the scale. Within each domain, activities
are broken down into several component behaviors. Ratings are made on a 4-point scale
(cf. Appendix).
Since the publication of our findings using the earliest version of the CSADL (Patterson
et al., 1992), several changes have been made to the scale. As a result of item analysis, num-
ber of items was reduced from 66 to 48. The wording describing the meaning of each rating
level was slightly changed in the interest of clarity, and the wording of two items was slightly
modified. One further change was made to deal with the potential problem of individuals
who might have been dependent prior to the onset of dementia. In the earliest version of the
CSADL (Patterson et al., 1992), informants were asked to rate the patient’s level of depen-
dency prior to onset of dementia. On the basis of data collected from that version, 12 items
were identified as likely to reflect prior dependency. Consequently, a follow-up question
was developed to determine if dependency on these 12 items was associated with dementia;
if the response to questioning reveals that the dependency is not dementia-related, special
ratings are used. For the remaining 36 items, no evaluation of prior dependency is made,
as these items were not found to reflect dependency before dementia onset. Finally, in the
version described by Patterson et al. (1992), some items (8 of which are included in the
present version) were rated in a special manner to account for the possibility that a person
might not be carrying out a behavior simply because he or she lacked the opportunity to do
so. In the present version, all such items are simply rated as dependent when the behavior in
question is not being carried out. The response form for the modified version of the CSADL
is included in the Appendix.
The CSADL Manual (Patterson and Mack, 1998) includes a detailed description of how
the scale is administered and scored. It usually takes between 10 and 20 min to administer the
scale. Total scores for the CSADL are based on 45 items. Valid total scores can be calculated
so long as no more than five of these items are unrated. Total scores do not include Items 23
(takes medications) and 32 (works for pay). These two items are omitted from the scores
because they require ratings of 9 (cannot rate) when they are judged not applicable to an
individual and consequently get an unusually high frequency of 9s. They are included in the
scale, however, because they concern clinically relevant dependent behaviors. Total scores
also do not include Item 48, the nonspecific item, because it does not represent a specific
type of dependent behavior. Item 48 is included in the scale to identify potentially relevant
dependent behaviors that have otherwise not been addressed.
There are two total scores for the CSADL. Total Weighted Score (TWS) is based on
actual item scores, weighted by the dependency ratings 1 through 3, and has a range from 0
to 135. Total Items Dependent (TID) is based on unweighted item scores, that is, each item
is scored dependent or independent, and it ranges from 0 to 45. When at least 40 of the
45 items comprising the total scores are neither unrated nor given special ratings, the total
scores (TWS and TID) are used. If more than 5 items are either unrated or have special
ratings, it is usually possible to use truncated scores (TWS-35 and TID-35). TWS-35 can
range from 0 to 105 and TID-35 from 0 to 35. CSADL total scores and truncated scores
may be used to evaluate both present dependency, regardless of whether it is associated with
dementia, and dementia-related dependency, that is, dependent behavior which began or
became worse following dementia onset. Truncated scores, however, are particularly useful
in determining dementia-related dependency when the informant has no knowledge of a
patient’s dependency prior to dementia onset, because they are based on the 35 items that
do not require special questioning regarding prior dependency.
The Modified Blessed–Roth Dementia Scale (DS-ADL)
The DS-ADL is a scale developed by Blessed et al. (1968) but modified by the CERAD
investigators (Morris et al., 1988) to evaluate functional impairment in activities of daily
living. It consists of 11 items ranging from basic ADLs, such as toileting, to more instru-
mental activities, such as performing household tasks. The total score ranges from 0 to 17.
The Mini-Mental State Exam
The MMSE is a mental status examination that includes questions regarding orien-
tation, attention and memory, language, and constructional praxis. The total score ranges
from 0 to 30 (Folstein et al., 1975).
Procedure
A trained interviewer administered the CSADL to each AD patient’s primary caregiver

during the initial evaluation of the patient at the ADRC. Raters being trained first study the
CSADL Manual (Patterson and Mack, 1998) and next complete a didactic session with a
trainer. They then observe an experienced rater administer the scale. Then they observe
and rate several patients being interviewed by an experienced rater, subsequently compare
their ratings with those of the experienced rater, and discuss any discrepancies. Finally,
occasional meetings are held by the senior author for all raters to discuss rating procedures.
The 12 raters included research assistants, nurses, graduate students, and a social worker.
For healthy and physically impaired elderly participants, the interviewer administered
the CSADL directly to the participant rather than to an informed observer. Initially inter-
views were conducted in person. As the project continued, however, some interviews were
conducted by telephone, and all interviews of the 26 physically impaired elderly partici-
pants were by telephone. For the 296 participants for whom the interview type was noted,
the number of in-person versus telephone interviews are presented in Table I. For 31 of
the AD patients, in addition to the interviewer, a trained observer was present to com-
plete CSADL ratings to be used to evaluate interrater agreement. The DS-ADL and MMSE
were administered as part of the neurological and neuropsychological evaluation previously
described.
RESULTS
Reliability of the CSADL was assessed by evaluating interrater agreement and internal
consistency. Interrater agreements for the 31 AD patients rated by both an interviewer and an
observer were evaluated by Spearman rank-order correlations. In these analyses, 47 CSADL
items were used (all but the nonspecific Item 48). For any item, if either one or both raters
used a rating of 9 (“cannot rate”) for a particular subject, the participant was omitted from
the analysis. For Item 32 (Works for pay) there were 22 ratings of 9 (test instructions
required a rating of 9 if the individual was retired). Consequently, the interrater correlation
for that item (Rho = 1.00) was based on only 9 participants and considered invalid. Of the
remaining 46 items, only 16 had ratings of 9, and no item had more than two participants
who were not rated. For those items, 41 correlations were .92 or greater (27 were actually
.99 or greater). The remaining 5 items had correlations ranging from .84 to .89. We also
calculated correlations for the TWS and TWS-35. Both were greater than .99.
Internal consistencies of CSADL total and truncated scores were quite high. Based
on the results of the AD patients and healthy elderly participants combined, the values of
Cronbach’s α for the four measures (TWS, TID, TWS-35, TID-35) were, respectively, .97
(N = 502), .97 (N = 502), .96 (N = 521), and .96 (N = 521).
Validity of the CSADL was evaluated by determining the extent to which it distin-
guished among healthy elderly participants, physically impaired elderly participants, and
the three groups of AD patients with differing levels of cognitive impairment. CSADL scores
for each of the groups are presented in Table II. Between-group differences scores were
evaluated by analysis of variance (ANOVA). The effect of group on TWS was significant
Table II. Statistics for CSADL Scores by Group

AD
Healthy elderly Physically impaired Mild Moderate Severe All
Score (N = 199) (N = 26) (N = 162) (N = 177) (N = 81) (N = 410)
TWS
Mean .56 10.77 25.57 45.24 75.31 43.89
SD 1.95 11.19 16.98 22.76 27.75 28.30
Range 0–18 0–46 0–91 0–112 18–131 0–131
TID
Mean .30 5.19 12.93 20.20 30.16 19.47
SD 1.00 5.17 7.16 8.39 9.34 10.25
Range 0–8 0–19 0–34 0–41 8–45 0–45
TWS-35
Mean .11 4.31 12.69 25.20 50.67 25.59
SD .62 6.42 10.66 16.89 23.30 21.38
Range 0–7 0–28 0–61 0–82 12–101 0–101
TID-35
Mean .07 1.96 6.95 11.72 20.51 11.69
SD .30 2.82 4.63 6.46 8.01 7.88
Range 0–3 0–10 0–24 0–30 5–34 0–34
(F = 307.40, df = 4, p < .001). Post hoc analysis indicated that all group differences
were significant (all ps < .002) except that between the healthy and physically impaired el-
derly groups ( p < .064). For TID (F = 381.88, df = 4, p < .001) the results were nearly
identical, except that the healthy and physically impaired elderly groups were also sig-
nificantly different from one another ( p < .017). The results for TWS-35 (F = 239.60,
df = 4, p < .001) and TID-35 (F = 281.42, df = 4, p < .001) were essentially the same
as those for TWS, except that the contrasts between the healthy and physically impaired
elderly groups were not significant.
As noted above, healthy elderly and AD groups differed with respect to age, education,
race, and residence (data that was not available for the physically impaired elderly group),
as well as interview type. We repeated the ANOVAs with effects of these variables removed
by covariance analysis, and the results were unchanged. Because of possible concern about
use of telephone interview data, we carried out separate ANOVAs restricted to those par-
ticipants whose data were based on telephone interviews, a restriction which eliminated
healthy elderly but included the 26 physically impaired elderly participants. The results
paralleled those for the overall group: the effect of group on CSADL scores was consis-
tently significant for all four CSADL scores, and the post hoc contrasts were all significant
with one exception, the difference between the physically impaired and mild AD groups
was not significant for TWS-35.
Actual numbers of individuals in each group showing functional impairment on the
CSADL is presented in Table III. TID scores of 0 (all items rated independent) were common
only among the healthy elderly. Only 5 of 26 physically impaired elderly participants showed
no impairment. Among AD patients, scores of 0 were even more rare, and all severely
impaired AD patients showed some degree of impairment.
We assessed concurrent validity by comparing CSADL results with those of the
DS-ADL. We used only TWS and TWS-35, as the DS-ADL score is also a weighted score.
Using both AD and healthy elderly participants (N = 592), the correlations of DS-ADL
with TWS and TWS-35 were, respectively, .90 and .87. Using only AD patients (N = 394),
the equivalent correlations were .82 and .81.
Finally, we examined the correlation between cognitive impairment, as measured by
the MMSE, and functional impairment, as measured by both the CSADL and the DS-ADL,
in 337 AD patients who had all three measures. The highest correlations were observed
between CSADL weighted scores and the MMSE (rTWS = −.63, rTWS-35 = −.63). Corre-
lations of the unweighted CSADL scores with MMSE were somewhat lower (rTID = −.60,
rTID-35 = −.60). The correlation between the DS-ADL and the MMSE was lowest of all
(r = −.47). All correlations were significant ( p < .001). Correlations of the four CSADL
Table III. Number of Participants by Group with Specific Numbers of Total Items
Dependent for 45-Item Scale
Number of Items Dependent (TID)
Group 0 1–10 11–20 21–30 31–40 41–45
Healthy elderly 167 32 — — — —

Physically impaired 5 18 3 — — —
Mild AD 3 62 70 11 4 —
Moderate AD 1 24 79 54 19 —
Severe AD — 2 13 28 26 11
All AD 4 88 162 93 49 11
total scores with MMSE were not significantly different (using Fisher’s r to z transforma-
tion) from one another, but the DS-ADL-MMSE correlation was significantly lower than
the correlation of any of the four CSADL measures with the MMSE ( p < .018 or better).
DISCUSSION
Results of our study indicate that the CSADL is a reliable measure of functional deficits
in individuals with AD. The interrater correlations for individual items and total scores were
consistently high, as were the internal consistency measures. Reliability can be affected by
insufficient clarity of test instructions and item wording. In both the scale and manual, we
paid particular attention to defining terms precisely and providing clear instructions. In
this regard Spector (1997) has raised questions about the use of terms such as “initiate,”
“appropriate,” and “acceptable.” Although we used such words, in the CSADL Manual
(Patterson et al., 1992) we provide explicit instructions for dealing with them, for example:
“When ratings call for assessment of “appropriateness,” the rater is to use the informant’s judgment.
Standards of propriety may vary widely from one family or social setting to the next, and only the
informant is in a position to determine if the behavior is appropriate in the context in which the
subject lives. The rater must not impose his or her standards to determine the appropriateness of
the behavior in question” (Patterson and Mack, 1998, p. 4).
The fact that our interrater reliabilities for individual items were high indicates that items
containing such terms are being interpreted consistently by raters. Our reliability findings
are limited in that we have not evaluated the consistency with which informants interpret
these terms. We plan to examine test–retest reliability in future studies. It would also be
useful to evaluate consistency between informants, but obtaining two equally knowledgeable
informants would be very difficult.
Validity of the CSADL total and truncated scores was supported by our results. Con-
trasts between AD groups and both control groups were consistently significant, even for
the AD group with only mild cognitive impairment. All four scores from the CSADL appear
effective in identifying functional impairment. Concurrent validity was demonstrated by the
high correlations between the CSADL weighted scores and the DS-ADL.
Furthermore, the CSADL appears to be sensitive to increases in functional disability
across a broad range of dementia severity: the contrasts between the three AD groups differ-
ing in degree of cognitive impairment were all significant. Although these findings suggest
that the scale may be sensitive to changes in functional impairment over time, they are based
on cross-sectional data. We are now conducting a longitudinal study of CSADL measures.
It should be noted that our sample of AD patients, although relatively large, did not
include very severely impaired patients—only 3 patients with CDRs of 4 and none with
CDRs of 5 were included. Relative absence of very severely impaired patients does not affect
our conclusions regarding the validity of the CSADL, as all contrasts between the three AD
groups with different levels of cognitive impairment were significant. Further research is
needed to determine whether more severely impaired patients demonstrate even lower levels
of functional impairment. Given that no AD patient, even in the group with the greatest
cognitive impairment, reaches the ceiling TWS score of 135, the scale seems capable of
reflecting a greater degree of functional impairment than was present in our sample.
Functional impairment in activities of daily living, as measured by the CSADL, was
extensively present in AD patients, regardless of degree of cognitive impairment. In a review
of both longitudinal and cross-sectional studies on the relationship between cognitive and
functional impairment in AD, Gélinas and Auer (1996) concluded that although there is a
general relationship between the two, the relationship is not perfect, and functional loss may
not be entirely dependent on cognitive loss. Our findings are consistent with this conclusion,
although CSADL scores indicate that functional impairment is highly prevalent, even among
mildly impaired AD patients. The finding that the MMSE was more highly correlated with
the CSADL than with the DS-ADL indicates that the CSADL is more sensitive to cognitive
impairment than is the DS-ADL.
Because the CSADL was developed to evaluate the effect of dementia on functional
impairments in activities of daily living, it includes many items that focus on the potential
effects of cognitive deficits rather than physical infirmities. To validate efficacy of the scale
in this regard, physically impaired elderly individuals were used as a comparison group in
the present study. The finding that all AD groups, even the mildest, showed significantly
more functional impairment than did the physically impaired group demonstrates that the
CSADL is selectively sensitive to the effects of cognitive deficits. Results of the physically
impaired group, however, were significantly worse than those of the healthy elderly group
on TID. This finding is consistent with those of Kiely et al. (1997), whose group of elderly
arthritic patients showed mildly impaired functioning in activities of daily living. The precise
nature of the functional deficits of our physically impaired participants was not addressed
in the present study.
As CSADL scores show equivalent reliability and validity, a user should select a score
appropriate to his or her particular needs. The total scores, TWS and TID, are based on the
greatest number of scale items and thus provide the most comprehensive evaluation of an
individual’s functional impairment. The truncated scores, TWS-35 and TID-35, are to be
used only when the number of items rated is too small to allow the use of the total scores. The
extent of dependency omitted by the scores is illustrated in the following comparison: the
mean number of items rated dependent (TID) in the severe AD group was 30.16, whereas
the equivalent value for TID-35 was only 20.51. Nearly ten dependent items on average
were omitted by the truncated score. Because the weighted scores, TWS and TWS-35,
include the actual dependency rating for each item, they provide a better measure of gross
dependency and have a greater range than the unweighted scores. If, however, one wishes
to evaluate the diversity of dependent behaviors, TID and TID-35 are more appropriate,
since they are not influenced by the degree of dependency of individual items. For each of
the four scores, the manual (Patterson and Mack, 1998) provides normative information,
based on the present sample of 410 AD patients.
The present study deals only with summary scores. A number of important questions,
such as the relationship between dependent behavior and specific cognitive deficits, may
require an examination of individual scale items or subsets of those items. We believe that
the CSADL provides a particularly fruitful avenue for investigating such questions because
CSADL items represent task components quite independently. The items were not based on
assumptions about the hierarchical relationships between components of tasks, assumptions
made by many existing ADL scales. Even when such relationships have been empirically
derived (e.g., using Guttman scaling techniques), rating tasks in such a fashion makes it
impossible to examine the contribution of individual behaviors to task performance. Our ap-
proach enables one to evaluate possible hierarchical relationships between subsets of items
empirically, including the possibility that patients with different types of dementia or differ-
ent patterns of cognitive decline might reveal different relationships among items. The next
step in our research will be to examine item interrelationships to construct subscales based
on subsets of items. We plan to compare such subscales with previously used approaches
to categorizing dependent behavior, for example, basic and instrumental ADL, and then
to examine the robustness of those subscales in various groups of demented patients, for
example, those with different levels of severity.
APPENDIX
Cleveland Scale for Activities of Daily Living (CSADL)
Name or ID of Subject Date / / Rater

mm dd yy
Name of Informant
Relation of Informant to Subject (Circle one.) Contact with Subject Interview Type
1 Spouse 4 Friend or other family 1 2 days/week 1 Visit
2 Child 5 Professional: 2 3-4 days/week 2 Telephone
3 Sibling 6 Other: 3 5 or more days/week
To administer this scale, the rater must be thoroughly familiar with the Manual, which
includes the full instructions. Place rating in blank after each item number. Several items
have specific rating instructions. In particular, some require special questioning if the
subject is rated as dependent (rating of 1, 2, or 3).
Rating Meaning of Rating

0 Behavior is carried out effectively, quite independently, without direction or help
1 Usually independent but sometimes or in some situations needs direction or help
2 Usually requires some direction or help, but sometimes or in some situations
is independent
3 Always requires direction or help—behavior is never carried out independently
9 Cannot rate because of insufficient information
Bathing
1. Initiates bath or shower with appropriate frequency and at appropriate times
2. Prepares bath/shower (draws water of proper temperature, ensures soap and
towel are present, etc.)
3. Gets in and out of tub or shower
4. Cleans self
Toileting
5. Able to physically control timing of urination
6. Able to physically control timing of bowel movements
7. Recognizes need to eliminate
8. After toileting, cleans and re-clothes self appropriately
Copyright °
c 1994 by University Alzheimer Centre
Personal hygiene and appearance

9. Initiates personal grooming with appropriate frequency and at appropriate times
10. Washes hands and face
11. Brushes teeth
12. Combs hair, shaves (as appropriate)
Dressing
13. Initiates dressing at appropriate time
14. Selects clothes
15. Puts on garments, footwear, etc.
16. Fastens clothing (buttons, shoelaces, zippers, etc.)
Eating
17. Initiates eating at appropriate times of day and with appropriate frequency.
18. Carries out physical acts of eating (including using utensils)
19. Eats with acceptable manners, e.g., with appropriate speed, does not speak
with food in mouth, etc.
20. Prepares own meals (includes cooking on stove) This item requires special
questioning.
Mobility
21. Initiates actively moving about the environment, as opposed to sitting, not
attempting to get about, etc.
22. Actively moves about environment (with or without assisting device)
22a. Does subject have physical limitations of mobility. (Circle one of following
codes.)
0 No physical limitations of mobility
1 Yes, there are physical limitations of mobility. (Circle all that apply.)
Needs assistance of Trouble getting in or out of bed Other Mobility Problems (describe):
other persons to walk Trouble getting in or out of chair
Needs cane Trouble getting on or off toilet
Needs walker Trouble climbing or descending stairs
Needs wheelchair
Medications
23. Takes medications as scheduled and in correct dosages If subject has taken no
medications during prior year, rate item as 9. This item requires special
questioning.
Shopping
24. Does necessary grocery shopping, buying appropriate items and quantities. This
item requires special questioning.
25. Does necessary clothes shopping, buying appropriate items and quantities. This
Travel
26. Finds way about in familiar surroundings
27. Orients to unfamiliar surroundings without undue difficulty
28. Travels beyond walking distance (i.e., driving own vehicle or using public
transportation)
29. Drives motor vehicle. This item requires special questioning.
Hobbies, personal interests, employment

30. Initiates activities of personal interest (e.g., card playing, woodworking, others)
This item requires special questioning.
31. Carries out such activities This item requires special questioning.
32. Does subject work for pay? If subject does not work because of having reached
an age appropriate to retirement from his or her occupation, rate 9. This item
requires special questioning.
Housework/home maintenance (as appropriate to individual situation)

33. Initiates work around house as needed This item requires special questioning.
34. Carries out work effectively, e.g. cleanly, neatly, accurately, efficiently This
Types of work done (Don’t score, just circle)
Dish washing Vacuuming Mowing lawn
Sweeping Scrubbing floors Gardening
Personal laundry Small home repairs Minor car care
Other types of work (Describe):
Telephone
35. Looks up numbers
36. Dials numbers
37. Answers phone
38. Takes messages
Money Management
39. Pays for purchases (selecting appropriate amount and determining correct
change) This item requires special questioning.
40. Manages financial responsibilities beyond paying for immediate purchases (e.g.,
paying monthly bills, managing checking or savings account, etc.) This item
requires special questioning.
Communication skills
41. Spontaneously expresses thoughts and needs to others
42. Responds accurately to spoken instructions and conversation
43. Reads and understands single words and short phrases (signs, lists, etc.)
44. Reads and understands complex material (books, newspapers, etc.)
45. Writes short phrases (lists, brief messages)
46. Writes complex material (letters, diary, etc.)
Social Behavior
47. Behaves in a socially appropriate manner. Socially inappropriate behaviors
encompass a wide range of behavior, including but not limited to such things as
making rude remarks, belching, touching private parts, showing little regard for
personal privacy, etc. For this item, dependency refers to the extent to which
other people must direct or manage the subject to ensure that he or she behaves
in a socially appropriate fashion.
Other Problems—Are there any situations in which patient does not behave in an
independent and responsible fashion that have not been covered by these
questions? (Circle one of following codes.)
48. 0 No other dependent behaviors
1 Yes, there are other dependent behaviors. (Please provide details below.)
QUALITY OF INTERVIEW (Rater’s Judgment)
Interview appeared valid 0

Some questions about interview, but it is probably acceptable 1
Information from interview is of doubtful validity. 2
Rater should record the basis for judging the interview of questionable or doubtful
validity.
Comments:
(V 97.08)
ACKNOWLEDGMENTS
This study was supported in part by grant AG08012 from the National Institute of
Aging, a grant from the Eli Lilly Company, Grant NH4344-01 from the National Institute
of Mental Health, an Alzheimer Center grant from the Ohio Department of Aging, and
a grant from Phillip Morris, Inc., USA. We would like to acknowledge the contribution
to this project of Jon Stuckey, Ph.D., Case Western Reserve University, who played an
important role in data management and analysis, and to the staff of the Case Western
Reserve University/University Hospitals of Cleveland Alzheimer’s Disease Research Center
who assisted in recruitment and diagnosis of research participants, data collection, data
management, and scale development.
REFERENCES
Beck, C. K., and Frank, L. B. (1997). Assessing functioning and self-care abilities in Alzheimer disease research.
Alzheimer Dis. Assoc. Disord. 11(Suppl. 6): 73–80.
Berg, L. (1988). Clinical Dementia Rating (CDR). Psychopharmacol. Bull. 24: 637–639.
Blessed, G., Tomlinson, B. E., and Roth, M. (1968). The association between quantitative measures of dementia
and of senile change in the cerebral gray matter of elderly subjects. Br. J. Psychiat. 114: 797–811.
Folstein, M. F., Folstein, S. E., and McHugh, P. R. (1975). Mini-Mental State. J. Psychiat. Res. 12: 189–198.
Gélinas, I., and Auer, S. (1996). Functional autonomy. In Gauthier, S. (ed.), Clinical Diagnosis and Management
of Alzheimer’s Disease, Martin Dunitz, London, pp. 191–202.
Hughes, C. P., Berg, L., Danziger, W. L., Coben, L. A., and Martin, R. L. (1982). A new clinical scale for the
staging of dementia. Br. J. Psychiat. 140: 566–572.
Kiely, D. K., Morris, J. N., Morris, S. A., Cupples, L. A., Ooi, W. L., and Sherwood, S. (1997). The effect of
specific medical conditions on functional decline. J. Am. Geriat. Soc. 45: 1459–1463.
McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., and Stadlan, E. M. (1984). Clinical diagnosis
of Alzheimer’s disease. Report of the NINCDS-ADRDA Work Group under the auspices of Department of
Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34: 939–944.
Morris, J. C., Mohs, R. C., Rogers, H., Fillenbaum, G., and Heyman, A. (1988). Consortium to Establish a
Registry for Alzheimer’s Disease (CERAD) Clinical and Neuropsychological Assessment of Alzheimer’s
disease. Psychopharmacol. Bull. 24: 641–652.
Patterson, M. B., and Mack, J. L. (1998). Cleveland Scale for Activities of Daily Living (CSADL): Manual,
University Alzheimer Center, Cleveland, OH.
Patterson, M. B., Mack, J. L., Neundorfer, M. M., Martin, R. M., Smyth, K. A., and Whitehouse, P. J. (1992).
Assessment of functional ability in Alzheimer disease: A review and a preliminary report on the Cleveland
Scale for Activities of Daily Living. Alzheimer Dis. Assoc. Disord. 6: 145–163.
Spector, W. D. (1997). Measuring functioning in daily activities for persons with dementia. Alzheimer Dis. Assoc.
Disord. 11(Supp 6): 81–90.
Weinberger, M., Samsa, G. P., Schmader, K., Greenberg, S. M., Carr, D. B., and Wildman, D. S. (1992). Comparing
proxy and patients’ perceptions of patients’ functional status: Results from an outpatient geriatric clinic. J.
Am. Geriat. Soc. 40: 585–588.

The Cleveland Scale For Activities of Da

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Cleveland Scale For Activities of Da

Uploaded by

Copyright:

Available Formats

P1: FLF/FGL P2: FTK/FLK QC: FTK

Journal of Clinical Geropsychology, Vol. 7, No. 1, 2001

The Cleveland Scale for Activities of Daily

16 Patterson and Mack

Reliability and Validity of the CSADL 17

Table I. Basic Information for Healthy Elderly and AD Participants

18 Patterson and Mack

The Cleveland Scale for Activities of Daily Living

The CSADL consists of 47 items encompassing 15 domains of everyday activity,

Reliability and Validity of the CSADL 19

The Modified Blessed–Roth Dementia Scale (DS-ADL)

The Mini-Mental State Exam

A trained interviewer administered the CSADL to each AD patient’s primary caregiver

20 Patterson and Mack

Table II. Statistics for CSADL Scores by Group

Reliability and Validity of the CSADL 21

Healthy elderly 167 32 — — — —

22 Patterson and Mack

Reliability and Validity of the CSADL 23

24 Patterson and Mack

Cleveland Scale for Activities of Daily Living (CSADL)

Name or ID of Subject Date / / Rater

Rating Meaning of Rating

Reliability and Validity of the CSADL 25

Personal hygiene and appearance

26 Patterson and Mack

Hobbies, personal interests, employment

Housework/home maintenance (as appropriate to individual situation)

Reliability and Validity of the CSADL 27

QUALITY OF INTERVIEW (Rater’s Judgment)

Interview appeared valid 0

28 Patterson and Mack

You might also like