Full text of DOI:10.1093/geront/41.6.723

The Gerontologist
Vol. 41, No. 6, 723–732
Copyright 2001 by The Gerontological Society of America
How Much Is Enough? Perspectives of
Care Recipients and Professionals on the
Sufficiency of In-Home Care
Nancy Morrow-Howell, PhD,1 Enola Proctor, PhD,1 and Philip Rozario, MSW1
Purpose of the Study: This study sought to increase
knowledge about the impact of rater role on the assess-
ment of in-home supportive care. The authors compared
the perspectives of care recipients and professionals on
one aspect of the broad concept of quality of care in
home care—the sufficiency of the amount of care pro-
vided by informal and formal caregivers. Design and
Method: Sufficiency of home care was assessed through
concurrent elderly persons’ self-report through telephone
interview and nurse clinical report based on in-home in-
terviews with the elder. Care was assessed in terms of the
sufficiency of the amount of informal and formal assis-
tance received to meet functional dependency needs. Sta-
tistical analyses compared the ratings of elder and
professional. Results: Professional ratings of the suffi-
ciency of care were significantly lower than those of the
elderly care recipients. From the perspective of both care
recipients and professionals, sufficiency of care was sig-
nificantly related to coresidence of elder and caregiver,
and to caregiver health. Implications: Consistent with
previous literature, rater role was found to influence the
assessment of the sufficiency of in-home care. Research-
ers and providers should recognize that care recipient
and professional ratings are not interchangeable.
baum, 1995). Although family and friends provide
considerably more care than formal providers (Kos-
berg & Cairl, 1992; Morrow-Howell, Proctor, Dore,
& Kaplan, 1998; Strawbridge & Wallhagen, 1992),
relatively little attention has focused on assessing and
improving the quality of in-home care provided by
the informal sector.
As with the assessment of quality of care in gen-
eral, research on the quality of home care is chal-
lenged by definitional issues (Capitman, Abrahams,
& Ritter, 1997). Reflecting the complexity of the
concept “quality of care,” different criteria are used
to define and measure quality (McGlynn, 1997). Ac-
cordingly, there is lack of consensus as to what con-
stitutes quality home care. Further challenges sur-
round the issue of who should rate quality. Which
home care consumer should be considered in assess-
ing quality (Applebaum et al., 1997–1998), and
which stakeholders should rate quality?
This study seeks to increase knowledge about the
variance in the assessment of quality of home care re-
lated to rater role. There are many roles assumed by
people related to the home care industry—care recip-
ient, informal caregiver, professional caregivers, clin-
ical case managers, financial case managers, etc. We
seek to understand if people systematically view and
thus rate home care quality differently, depending on
their role. Specifically, we focus on two raters: care
recipients and professionals. Among the many poten-
tial domains of quality of care, we focus on one: the
sufficiency of the amount of care provided by infor-
mal and formal caregivers to meet the dependency
needs of frail elders in activities of daily living (ADL).
Key Words: Quality of home care, Rater role,
Measurement of quality of care
As the home-care industry has grown, so too have
concerns about the quality of in-home care (Apple-
baum, Mollica, & Tilley, 1997–1998). In response,
the health and human service sectors have stepped up
efforts to assess and improve quality of care (Apple-
Related Literature
Conceptions and Indicators of Quality.—Various at-
tributes are used to conceptualize and evaluate the
quality of care. With respect to health care in gen-
eral, Donabedian (1992) identified six attributes of
quality: effectiveness, efficiency, cost-effectiveness,
acceptability, legitimacy, and equity. The aspect of
health care addressed in this study, home care, has
been typically evaluated in terms of prevented nurs-
ing home placements, costs, and such elder outcomes
This work was funded by the Agency for Health Care Policy and Re-
search Grant R01 HS-6406-01 as well as a Faculty Award from the
George Warren Brown School of Social Work, Washington University in
St. Louis, Missouri.
Address correspondence to Nancy Morrow-Howell, Campus Box
1196, George Warren Brown School of Social Work, Washington Univer-
sity, St. Louis, MO 63130. E-mail: nancymh@gwbmail.wustl.edu
1George Warren Brown School of Social Work, Washington Univer-
sity, St. Louis, MO.
Vol. 41, No. 6, 2001
723

as functional status and mortality. Attention to more
proximal indicators of quality, particularly features
of the in-home care per se, is more recent. For exam-
ple, Ory and Duncker (1992) called for reconceptu-
alizing outcomes to include the adequacy of care.
Morrow-Howell, Proctor, and Berg-Weger (1993)
addressed the adequacy of informal care for elderly
patients discharged home from the hospital, docu-
menting concerns that discharge planners expressed
about informal caregivers’ abilities to perform the
tasks expected of them in posthospital care.
tening to the consumer and designing systems of care
that are more responsive to meeting needs defined by
the consumer, more effective—and accountable—
delivery systems are likely” (p. 117). Yet particularly
in gerontology, in which elderly subjects may be too
sick or cognitively impaired to participate in inter-
views, researchers may depend on proxy informants.
Thus, differences introduced by rater role may
threaten the validity or interpretation of quality of
care measurement (Tennstedt, Skinner, Sullivan, &
McKinlay, 1992).
Health care has also been evaluated in terms of the
extent to which services meet needs for assistance, an
aspect of quality that may be particularly germane to
care recipients. McGlynn (1997) noted that patients
tend to evaluate care in terms of its responsiveness to
their individual needs. Morrow-Howell, Proctor, and
Dore (1998) conceptualized adequacy in terms of the
extent to which services meet needs for assistance.
Consistent with this approach, Skinner, Steinwachs,
Handley, and colleagues (1999) postulated that out-
comes of care (for persons with severe and persistent
mental illness) would be better when services meet
needs for assistance. Particularly in postacute care,
meeting needs for functional assistance is an impor-
tant component of quality. Over the past 2 decades
there have been trends for patients leaving the hospi-
tal after shorter stays at higher levels of dependency
and with increased need for care (Coe et al., 1986;
Kosekoff et al., 1990). Mamon, Steinwachs, Fahey,
and colleagues (1992) found that 97% of patients
discharged home after hospitalization had need for
care at 2 weeks postdischarge.
Related Work on Rater Perspectives.—Incongruence
between professional and lay perspectives of service
needs and preferences is well-documented (Avant &
Dressel, 1980; Keith, 1975; Riesenfield, Newcomer,
Berlant, & Dempsey, 1972; Wilson & Netting, 1987).
For example, Wilson and Netting (1987) found that
elders identified fewer problems and needs than
health professionals did, whereas professionals per-
ceived that the elderly adults had more access to ser-
vices (transportation, specifically) than did the el-
derly adults themselves. Epstein, Hall, Tognetti, and
colleagues (1989) reported low correlations between
patient and proxy ratings of satisfaction with health
care. Factors associated with satisfaction with hospi-
tal discharge plans are different for patient and fam-
ily raters (Proctor, Morrow-Howell, Albaz, & Weir,
1992). Patient ratings were associated with involve-
ment in the decision-making process, social support,
and physical condition; family ratings were associ-
ated with length of hospital stay, discharge planning
processes, and discharge destination. Efforts to im-
prove satisfaction ratings may be different depending
on which consumer group is being considered.
Stakeholder Perspectives on Quality.—As McGlynn
(1997) has noted, “quality is in the eye of the be-
holder; that is, expectations and the value associated
with different aspects of care are likely to vary
among different stakeholders” (p. 9). In assessing the
quality of home care, the relevant perspectives in-
clude those of the elder-care recipient, any one of a
number of family members (some providing care and
some not), and the formal care providers (the home
care aides themselves, their supervisors, or the care
managers; Wagner, 1988). Additionally, funders and
regulators are key stakeholders. Each of the various
constituencies or stakeholders in the home-care in-
dustry has a legitimate and perhaps unique perspec-
tive; their motivations, values, and criteria may be
truly different (Kaufert, 1983). Each perspective may
be differentially important in planning or evaluating
programs and services. For example, discrepancy be-
tween the need for a service (as assessed by provid-
ers) and the demand for that service (as expressed by
consumers) is critical in program planning; that is,
there is a risk that services will be structured by pro-
viders for hypothetical target populations (Wilson &
Netting, 1987). Patient perspective, particularly pa-
tient satisfaction, has been regarded as an “indis-
pensable” reflection of the quality of care (Donabe-
dian, 1988, p. 1746); and Skinner, Steinwachs,
Handley, and colleagues (1999) noted that by “lis-
The study of systematic differences between vari-
ous raters of health status has a somewhat longer
history. Moderate to weak associations have been
documented between self-rated health and physician-
rated health, with older adults generally rating their
health more positively than physicians do (Friedsam
& Martin, 1963; Kivinen, Halonen, Eronen, &
Nissinen, 1998; Rothman, Hedrick, Bulcroft, Hickam,
& Rubenstein, 1991; Streib, Suchman, & Phillips,
1958; Maddox, 1964). Additionally, self-rated health
has been found to be a better predictor of future
health-related outcomes (Maddox & Douglas, 1973)
and health service use (Markides, 1979).
Similarly, systematic differences are observed in
various raters’ assessments of elders’ functional abil-
ity (Rubenstein, Schairer, Wieland, & Kane, 1984;
Dodge, Janz, & Clark, 1994). As with health status,
older adults tend to report higher levels of function-
ing, with family members reporting the lowest levels,
and professionals (nurses) assigning intermediate rat-
ings (Rubenstein et al., 1984). Further, the closer the
relationship and the more frequent the contact, the
lower the family members’ perception of the elder’s
functional and health status is (Rubenstein et al.,
1984; Rothman et al., 1991).
Several researchers have tested explanations for
differences between ratings of health and function.
724
The Gerontologist

Kivinen and colleagues (1998) postulated that it is
the differential perception of the effect of age that ex-
plains the difference in rating of health status be-
tween patients and physicians. After controlling for
medical condition, they found that age of the patient
did not relate to self-rated health whereas it did re-
late to assessments by physician, with older patients
assessed as being in poorer health. Other researchers
have suggested that characteristics of the elders as
well as the family proxies affect ratings of the elder’s
health and functional status (Kiyak, Teri, & Borson,
1994; Magaziner, Simonsick, Kashner, et al., 1988;
Rothman, Hedrick, Bulcroft, Hickam, & Ruben-
stein, 1991; Poulshock & Deimling, 1984; Hooy-
man, Gonyea, & Montgomery, 1985). Better agree-
ment between patient and family raters has been
associated with better mental health and cognitive
ability of the patients (Magaziner et al., 1988; Kiyak,
Teri, & Borson, 1994). Family proxies experiencing
greater burden or psychological distress report greater
impairment in health and functional ability of the
care recipient. This suggests that family members
perceive higher disability to correspond with burden
they feel or to justify nursing home placement or for-
mal service use they may feel they need (Rothman et
al., 1991). Finally, it has been suggested that the dis-
abled elder may seek to conceal disability so as not to
burden caregivers, deny disability as a coping mecha-
nisms, or use a different time frame in the evaluation
of their ability (Rubenstein et al., 1984).
Although previous literature consistently demon-
strates that older adults rate their functional ability
higher than either family members or health care
professionals do, we sought to replicate this finding
in a sample of chronically ill elders recently dis-
charged home from the hospital. The last three ques-
tions have the potential to add new knowledge re-
garding rater role.
Methods
Parent Study
The data for this study were derived from a larger,
prospective study of postacute care for older adults
with chronic heart disease. To accomplish the aims of
the parent study, the researchers recruited 253 older
adults between June, 1990, and October, 1991, while
the participants were hospitalized at a large Midwest-
ern urban teaching hospital. Subjects met the follow-
ing criteria: they were 65 years of age or older, were
diagnosed with congestive heart failure (CHF), were
served by a discharge planner, and were discharged to
a home setting. CHF is a particularly appropriate di-
agnosis for studying postacute home care, given its
prevalence among elderly persons, high public health
costs, and the importance of supportive home care to
maintain gains achieved during hospitalization (Bon-
neux, Baredregt, Meeter, Bonsel, & van der Mass,
1994). Given the aims of the parent study, we se-
lected patients served by discharge planners to ensure
a sample with need for home care and one that re-
ceived a mix of informal and formal services.
Study Focus and Conceptualization
Discharge planners identified patients who met
study criteria, whom research assistants approached
to explain the study; 82% of patients contacted con-
sented to participate. We approached a family mem-
ber at the instruction of the discharge planner when
patients were too ill or cognitively impaired to par-
ticipate. In 35% of the cases, this family collateral
provided consent and completed data collection for
purposes of the larger study.
This study addressed the issue of stakeholder per-
spectives on the quality of in-home supportive care.
Quality is conceptualized and measured in terms of
the sufficiency of the amount of care received to meet
specific functional dependency needs. Who says how
much is enough help with activities of daily living—
the care recipient, the informal or formal care pro-
vider, the professional care manager, or the public
payer? In this study, we focused on the care-recipient
perspective versus the professional perspective of a
nurse researcher. Both formal and informal care are
considered when assessing sufficiency of care, which
is very important in the context of community care
(Capitman et al., 1997).
Data Collection
At the time of hospital discharge, a research assis-
tant abstracted the medical records. At 2 weeks post-
discharge, a research assistant completed a telephone
survey with the study participant or the collateral.
Within 48 hours of this telephone interview, a regis-
tered nurse (RN) research assistant completed an in-
home interview. Both the telephone and in-home in-
terview focused on functional ability and informal
and formal assistance received to meet the patient’s
ADL and IADL needs. We recognize that two data
collection methods are used (telephone and in-home
interviews) and that this may confound our findings
in that differences attributed to rater role may result
from differing data collection methods. Yet the ad-
vantages of this approach outweigh this risk, as we
describe below.
This study addressed four research questions:
1. Do care recipients and professionals vary in their
assessment of physical functioning ability, and if
so, what specific functional areas (ADL and in-
strumental activities of daily living [IADL] areas)
contribute to the variation?
2. Do care recipients and professionals vary in their
assessment of the sufficiency of care? That is, do
care recipients and professionals have different
perspectives on whether enough help is provided?
3. What specific functional need areas (ADL and
IADL areas) contribute to the variation in raters’
assessment of sufficiency of care?
During the telephone interview, the research assis-
tant relied on the study participant’s self-report. We
4. What factors are associated with sufficiency of
care from these two different perspectives?
Vol. 41, No. 6, 2001
725

viewed the telephone interview as the preferred ap-
proach to data collection in self-report data because,
compared with an in-person interview, it minimizes
cues and the influence of the interviewer. In contrast,
we viewed the in-person interview as conducive to
obtaining a professional clinical assessment. During
the in-home interview, the nurse research assistant
had access to a broader range of information she
could draw from the study participant, the family
members present at the interview, and clinical obser-
vations to make a professional assessment of func-
tional ability and sufficiency of care provided. Thus,
we obtained concurrent assessment of physical func-
tioning ability and the sufficiency of care through the
telephone interview (self-report from the study par-
ticipant) and the in-home interview (clinical report of
nurse research assistant).
(as mentioned above, 209 study participants com-
pleted the first follow-up interview). There are no
gender, race, or socioeconomic status (SES) differ-
ences between the two groups; however, the 85 el-
ders in this study were more functional (t ꢀ 9.39,
p ꢁ .01) and younger (t ꢀ ꢂ3.79, p ꢁ .01) than el-
ders who did not participate. This difference results
from the elimination of study participants whose
family proxies provided the information because the
elder was not able, most often by virtue of illness or
dependency. This sample bias limits generalizability,
which is discussed below.
As noted above, achieving a valid subsample to
address differences in rater perspective is associated
with limits to both analytic power and generalizabil-
ity. In the functional area-by-area analysis reported
below, four cell sizes were too small for valid tests of
statistical difference (transfer, toileting, walking, and
eating). Thus we eliminated these areas from the
comparison of self-report and professional assess-
ment. In regard to generalizability, those care recipi-
ents too ill or too cognitively impaired to participate
in the telephone interview are different from those
who can participate, and indeed the amount of care
they receive and their perceptions of the sufficiency
of this care could be different. Thus we can general-
ize only to care recipients who can participate in tele-
phone interviews regarding their conditions and their
care arrangements. This generalizability issue would
plague any such study no matter how big the sample
size because one can never measure and compare
self-report perspective from study participants inca-
pable of self-report.
Sample Subset
Addressing the research questions posed in this ar-
ticle required that (a) the care recipient be capable of
providing self-report and (b) each subject provide
both a telephone interview and concurrent in-home
interview. Although the requisite sample for these
analyses is select and, in our case, small (n ꢀ 85),
such a sample is necessary to answer questions about
rater perspectives that are important both substan-
tively and methodologically. Despite limits to statisti-
cal power, the subsetted sample provides a scarce
opportunity to examine issues in quality of care mea-
surement that are critical in both gerontology and
health services research.
Accordingly, we subset the sample from the larger
study as follows. Of the 253 eligible patients at the
time of discharge, 209 were available to participate
in a telephone survey 2 weeks post discharge. In this
2-week period, 39 patients died, were readmitted to
the hospital, or entered a nursing home before infor-
mation could be gathered; this number reflects the
unstable condition of CHF and the high readmission
and mortality rate associated with this condition
(Bonneux et al., 1994). Five patients refused to par-
ticipate in the 2-week interview. Of the 209 study
participants completing the 2-week postdischarge
telephone interview, in-home interviews were com-
pleted on 149. In-home interviews were not com-
pleted for two reasons: (a) the participant lived more
than 2 hours from the study site and in-home inter-
views were not requested because of the excessive
travel time and (b) the participant consented to the
telephone interview but not to the in-home assess-
ment. Thus, 149 observations had concurrent tele-
phone and in-home interviews at the 2-week postdis-
charge observation point. Finally, to compare care
recipients with professionals, we include only those
85 observations for which the older adult completed
the telephone interview rather than a family proxy.
Bivariate analyses were completed to compare the
85 elders in this study sample with the 124 elders
who were available for the 2-week follow-up inter-
view but were not included in this particular analysis
Dependent Variables
The two dependent variables in this study, physi-
cal functioning ability and sufficiency of care, were
obtained from both the care-recipient and the profes-
sional perspectives. For the care-recipient perspec-
tive, research assistants (Masters in social work stu-
dents) conducted telephone interviews with each care
recipient, capturing the dependent variables through
self-report. To obtain a professional perspective, we
trained geriatric nurse practitioners to conduct the
in-home assessments using all relevant information
available in the interview as well as observation to
inform their ratings of functional ability and suffi-
ciency of care. To increase reliability, we used only
two nurses for the study. Both nurses had RN de-
grees and were active and experienced geriatric prac-
titioners.
Physical functioning ability was assessed using the
OARS (Older Americans Resources and Services
Multidimensional Functional Assesssment Question-
naire) instrument (Duke University Center for Study
of Aging and Human Development, 1978), which
has seven ADLs (transfer, walking, toileting, bathing,
grooming, dressing, and eating), and six IADLs
(meal preparation, shopping, money management,
travel, housekeeping, and medicine administration).
In each area, functional ability is captured with a
726
The Gerontologist

three-level response: able to perform activity with no
help (2), able to perform with some help (1), or com-
pletely unable to perform (0). The summative mea-
sure ranges from 0 to 26, with higher scores reflect-
ing greater functional ability. Telephone interviews
captured the care recipients’ self-report of functional
ability, whereas the nurses assessed functional ability
through clinical interviewing during the in-home in-
terview.
would be an average of four sufficiency ratings. Each
summative sufficiency rating—one for care recipient
and one for professional—ranged from 1 (never
enough help) to 4 (always enough help). Scores were
thus comparable across rater, regardless of differ-
ences between the raters in assessing the patient’s
functional ability.
Independent Variables
Sufficiency of care reflects the extent to which the
informal and formal assistance is seen as enough to
meet subjects’ ADL and IADL needs. The measure-
ment tool used to capture the sufficiency of care was
developed by the authors for studying posthospital
care; reliability was established through tests on 20
subjects (Morrow-Howell, Proctor, and Dore, 1998).
Test–retest reliability, established on subject tele-
phone self-report on two consecutive days, was .87,
using a weighted kappa. Interrater reliability of the
instrument across two professionals, who conducted
simultaneous in-home assessments, was .81, again
using a weighted kappa (Morrow-Howell, Proctor,
& Dore, 1998).
A variety of independent variables were measured
for purposes of exploring their contribution to the
dependent variable sufficiency of care from care re-
cipient and professional perspectives. Demographic
characteristics were extracted from the medical record
and included age in years, gender, and race recorded
dichotomously as White and African American.
Marital status was recorded as currently married and
not currently married. Socioeconomic status was as-
sessed from patient response to questions for the
Hollingshead Index of Social Class; this measure uses
education and occupation to assign subjects a rating
of 11 to 77, with lower numbers representing higher
social status (Hollingshead, 1957).
Procedures for scoring sufficiency of care were as
follows. In areas of the OARS instrument in which
function in an ADL or IADL area was rated as inde-
pendent, we asked no further questions about that
specific function. In ADL or IADL areas with any de-
pendency, further information was captured about
who provided assistance with the task, how often as-
sistance was received from formal and informal
sources, and how sufficient this help was. “Suffi-
ciency of care” was rated, specific to each area of
functional need, using a 4-point scale on which 1 ꢀ
without help at most times; 2 ꢀ usually does not
have enough help; 3 ꢀ usually has enough, but on
occasion help is not available when needed; and 4 ꢀ
always has enough help. As with the functional-abil-
ity ratings, care-recipient ratings of sufficiency were
self-reports obtained during the telephone inter-
views; professional ratings of sufficiency of care were
nurse assessments from the in-home interviews.
To compare care recipient and professional rat-
ings, we calculated two summative scores, averaging
the rated sufficiency of care across areas of func-
tional dependence—one score reflecting the care re-
cipient’s overall sufficiency of care and one reflecting
the professional (nurse) assessment of the overall suf-
ficiency of care. The quantity ratings (1–4) in each
area of need were summed and divided by the num-
ber of identified dependency areas, producing an av-
erage sufficiency rating. These scores were therefore
able to control for difference in care-recipient and
professional perspective in functional ability, as fol-
lows. If a care recipient reported functional depen-
dency in three areas (transportation, meal prepara-
tion, and housekeeping, for example), the three
sufficiency ratings in each need area were summed
and divided by three. If, however, the nurse assessed
more areas of dependency (for example, four areas,
including transportation, meal preparation, house-
keeping, and medicine administration), her score
The care recipient’s health characteristics were as-
certained by the nurse during the in-home assess-
ment. Cognitive ability was assessed using the Short
Blessed Mental Status exam, which yields a score of
0 to 28, with higher scores indicating more cognitive
impairment (Blessed et al., 1968). The Chronic Con-
ditions Checklist from the OARS instrument (Fillen-
baum, 1988) captured the existence of 32 chronic
medical conditions, weighted by the extent to which
the condition interferes with activity (from 0 [no lim-
itation] to 3 [severe limitation]).
The informal care situation was captured through
four variables obtained as part of the in-home inter-
view. The identity of primary caregiver was recorded
as spouse, adult child, or other. If other caregivers
assisted in any way, a dichotomous variable captured
the involvement of secondary caregiver. Co-residence
of the caregiver and care recipient was also recorded
dichotomously. The study participant provided in-
formation about the health of primary caregiver, us-
ing a 4-point rating scale of excellent to poor. During
the in-home interview, the nurse also queried about
the level of formal care. Formal care was measured
through the number of hours of in-home care pro-
vided by agency staff, volunteers, or any paid helper
during the previous month.
Statistical Analysis
The univariate distributions of the independent
and dependent variables were examined and descrip-
tive statistics are presented in Table 1. The extreme
skew on the number of hours of formal assistance
per month warranted attention. This variable was
log transformed, resulting in a more symmetric dis-
tribution, and the logged variable was used in re-
gression analysis. Dependent t tests were used to
compare care recipient and professional ratings on
Vol. 41, No. 6, 2001
727

Table 1. Description of Sample (n ꢀ 85)
Percentage Mean (SD)
Items Measured
Independent Variables
Range
65–94
Skew
0.18
Age
76.9 years (6.5)
Gender
Race
66% female
61% African
American
Marital status
Socioeconomic status
Cognitive status
Interference from chronic conditions
Identity of caregiver
71% not married
56.4 (18.0)
6.8 (5.1)
9.8 (4.5)
11–77
0–19
2–27
ꢂ0.70
0.49
0.72
Spouse 17%
Child 43%
Other 40%
Presence of 2ndary
Co-residence
Health of caregiver
72% had other helper
52% co reside
Poor 6%
Fair 26%
Good 52%
Excellent 16%
Hours of formal help
Logged hours of formal help
48.6 hrs (144)
1.76 (1.0)
0–906
0–4.2
4.9
0.35
Dependent Variables
Functioning (care recipient)
Functioning (professional)
Sufficiency (care recipient)
Sufficiency (professional)
19.32 (4.2)
17.51 (4.1)
3.21 (0.84)
2.73 (0.76)
4–25
8–24
1–4
ꢂ0.81
ꢂ0.21
ꢂ0.93
ꢂ0.48
1–4
functional ability and sufficiency of care. Similarly,
we used dependent t tests to compare the functional
ability ratings and the sufficiency of care ratings on
each ADL and IADL area to determine the sources of
variation in the summative scores. Thus, functional
ratings as well as sufficiency ratings were compared
between care recipients and professionals in bathing,
grooming, toileting, and so forth. In cases in which
sufficiency ratings were compared within each ADL
and IADL area, we compared care recipients to pro-
fessionals only when both raters assessed a func-
tional dependency in the area and therefore provided
a sufficiency of care rating in that area (i.e., suffi-
ciency ratings in the area of bathing could only be
compared if both care recipient and professional re-
ported a dependency in that area and therefore pro-
vided a sufficiency of care rating in bathing). Al-
though this strategy results in the loss of sample size
within each cell, it enables the comparison of suffi-
ciency ratings that are freed from the rater bias that
is known to exist in the assessment of functional
ability.
We used analysis of variance techniques through
the SAS general linear model procedure to test the re-
lationship of the independent variables to the suffi-
ciency of care measures (Cody & Smith, 1997). Main
effect and repeated measure models were both used.
We used repeated measures analysis of variance to
test the interactions of the independent variables and
rater identify on sufficiency of care. Given the num-
ber of interactions to be tested, we tested each inter-
action separately, after controlling for all indepen-
dent variables and rater role. The caregiver identity
variable was dummy coded. Because of the substan-
tial correlation between spousal caregiver and mari-
tal status in this sample (r ꢀ .71) as well as the lack
of a statistically significant association with the suffi-
ciency of care ratings, we eliminated caregiver iden-
tity from the regressions to reduce the number of in-
dependent variables being tested. The only model
with a significant interaction term is presented.
Results
Do Care Recipients and Professionals Vary in Their
Assessment of Physical Functioning Ability? If so, What
Specific Functional Areas Are the Sources of the Vari-
ation?—As would be predicted from previous work,
the older adults receiving in-home supportive care
rated their functional ability higher than nurse re-
searcher assistants did, t(84) ꢀ ꢂ5.57, p ꢀ .0001.
As seen in Table 1, care recipients rated their func-
tional status approximately 2 points higher on the
26-point scale of physical functioning (19.32 vs.
17.51). Ratings from each of the ADL and IADL ar-
eas are compared in Table 2. For each functional
area, in which scores range from 0 to 2, the mean
care recipient rating was compared with the mean
professional rating. As can be seen, care recipients
rated their functioning higher in all but three ADL
and IADL areas. Nurses rated the older adults as
more functional in the activity of bathing than did
the care recipients themselves. In two areas—meal
preparation and housekeeping—professional and care-
recipient ratings of functional ability were not signif-
icantly different.
728
The Gerontologist

Table 2. Physical Functioning Ability: Differences Between Care
Recipient and Professional Ratings by Functional Area (n ꢀ 85)
medications is included, given a p value of .06 and a
small sample size). Nurses rated assistance in the
area of transportation as more sufficient than did
care recipients.
Significant difference
Functional area
between raters?
Higher score
What Factors Are Associated With Sufficiency of
Care From the Two Different Perspectives?—Analysis of
variance reveals that three variables have main ef-
fects on sufficiency of care: coresidence of care recip-
ient and care receiver, health of caregiver, and rater
identity (see Table 4). Two of these variables relate
to the informal care system. Care was rated as more
sufficient when care recipients lived with their care-
givers and when caregivers were healthier. As re-
flected in earlier analyses, identity of the rater is re-
lated to sufficiency of care, with professional ratings
being lower than care-recipient ratings. When the
two informal support variables were tested in inter-
actions with rater identity, the interaction effects
were not significant. Thus, these variables are related
to the sufficiency ratings, regardless of whether the
rater was a care recipient or a professional. Of the 11
interaction terms tested, only 1 interaction was signif-
icantly associated with sufficiency of care. We found
no main effect for amount of formal services and no
relationship between formal services and sufficiency
of care from the perspective of care recipients. How-
ever, for professionals, higher amounts of formal care
were associated with lower sufficiency ratings.
Medications
Transfer
Toileting
Walking
Bathing
Grooming
Dressing
Eating
Meal preparation
Shopping
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .01)
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .00)
No (p ꢀ .29)
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .01)
No (p ꢀ .22)
Care recipient
Care recipient
Care recipient
Care recipient
Nurse
Care recipient
Care recipient
Care recipient
Not applicable
Care recipient
Care recipient
Care recipient
Not applicable
Money management
Transportation
Housekeeping
Do Care Recipients and Professionals Vary in Their
Assessment of the Sufficiency of Care?—As seen in
Table 1, professionals tended to rate the amount of
care as less sufficient than the care recipients did
(2.73 vs. 3.21); (t(84) ꢀ ꢂ4.9, p ꢀ .0001). On aver-
age, the nurses were almost one half a point lower on
the 4-point scale.
What Specific Functional Need Areas Contribute to
the Variation in Raters’ Assessment of Sufficiency of
Care?—Ratings in each of nine ADL and IADL areas
are compared in Table 3. Four functional areas
(transfer, toileting, walking, and eating) were elimi-
nated because the number of pairs of ratings was less
than five. For each functional area in which both the
care recipient and nurse agreed that there was a de-
pendency, the mean care recipient rating of suffi-
ciency of care was compared with the mean profes-
sional rating. As can be seen, the sufficiency ratings
did not vary in three functional need areas. However,
it should be noted that sample size drops quite low in
two of these comparisons, and lack of statistical
power could explain the null finding. Care recipients
rated the sufficiency of care higher in five need areas
(note that the marginally significant difference in
Discussion
We acknowledge both weaknesses and strengths
in our conceptualization and measurement of the de-
pendent variable, sufficiency of in-home care. The
construct, sufficiency of care to meet elders’ func-
tional dependency needs, is but one aspect of the
broader construct of quality of care. Our focus on
the amount or quantity of care does not capture the
important aspect of quality, or how good that care
Table 4. Factors Associated With Sufficiency of Care
(Dependent Variable)
Variable
Estimate St. Error Probability
Table 3. Sufficiency of Care: Differences Between Care Recipient
and Professional Ratings by Functional Need Area
Age
Nonwhite
Female
SES
.02
ꢂ.19
.09
.001
.15
.01
.16
.15
.004
.18
.07
.27
.56
.74
.39
Number
of pairs
of ratings between raters?
Significant
difference
Functional
need area
Married
Higher score
Interference from
chronic conditions
Cognitive status
Coresidence
Presence of secondary caregiver
Caregiver health
Hours of formal services
Rater identity
(care recipient/professional)
Rater identity ꢃ formal
services
ꢂ.01
ꢂ.02
.36
.18
.26
.02
.02
.17
.16
.09
.07
.48
.33
.04
.24
.01
.18
Medications
Bathing
Grooming
Dressing
23
41
11
8
Yes (p ꢀ .06)
Yes (p ꢀ .00)
No (p ꢀ .66)
No (p ꢀ .41)
Care recipient
Care recipient
Not applicable
Not applicable
ꢂ.10
Meal
preparation
Shopping
24
69
No (p ꢀ .11)
Yes (p ꢀ .00)
Not applicable
Care recipient
.54
.10
.08
.0001
.02
Money
ꢂ.19
management
Transportation
Housekeeping
21
56
77
Yes (p ꢀ .00)
Yes (p ꢀ .00)
Yes (p ꢀ .01)
Care recipient
Nurse
Care recipient
Notes: Sample Size ꢀ 79; Model F ꢀ 3.224, p ꢀ .001; Model
R2 ꢀ .54. SES ꢀ socioeconomic status.
Vol. 41, No. 6, 2001
729

is. Yet quantity seems to be a minimal component of
quality; how can care be construed as “quality”
when there isn’t enough care?
reveal their opinions for fear of alienating their med-
ical attendants” (1988, p. 1746). Particularly with el-
derly adults, Woodruff and Applebaum (1996) sug-
gested that consumers’ perspectives may be influenced
by their strong desire to remain independent and in
their current living situations. Care recipients may
lower their own standard of acceptability in regard
to meal preparation, bathing, housekeeping, and so
forth when faced with the reality of their needs and
their available resources. Second, care recipients may
not be willing to criticize their care providers, espe-
cially family members, by assigning low sufficiency
ratings. On the other hand, professionals may main-
tain their standards of acceptable quantity of care
across care recipients, with less threat (certainly less
personal repercussions) about the consequences of
the deficiencies. Additionally, they could be less con-
cerned about judging care providers as deficient in
some way. Our findings point to the need for further
research on motivations leading to differences in rat-
ings between different stakeholders in home care.
There are some similarities in factors associated
with sufficiency ratings between the two perspec-
tives. Two variables are important predictors from
both perspectives: coresidence of caregiver and care
recipient, and health of the caregiver. These two vari-
ables had main effects but no interaction effects in
the regression analyses, reflecting the salience of
these independent variables for both care recipient
and professional. Remarkably, only characteristics of
the informal caregiving network were related to the
sufficiency of care ratings irrespective of rater. These
findings speak to the critical role of informal care-
givers in the community long-term care system. The
proximity and the health of the primary helper affect
the perceived sufficiency of the amount of assistance
received. These findings suggest that if we are to in-
crease the quality of in-home supportive care, we
should focus on programs and policies that shore up
the informal system.
In addition to caregiver proximity and caregiver
health, one other independent variable affected pro-
fessionals’ ratings of the sufficiency of care: amount
of formal service. (In this sample of chronically ill el-
ders in the post-acute period, there are comparatively
high levels of formal service, which was planned as
part of the parent study.) Interestingly, nurses rated
care as less sufficient in situations with higher levels
of formal care. Although counterintuitive, this find-
ing is not infrequent in services research and often in-
dicates limitations in measurement—specifically the
inability to adequately capture and control for sick-
ness or need level. That is, sicker or more dependent
elders usually receive more formal care, yet despite
higher levels of care, their care arrangements remain
less adequate because of the high demands of the
care situation. In this particular study, we might in-
terpret this association to mean that, in the profes-
sional assessment, the addition of formal services
does not bring the amount of care up to a sufficient
level (or in the few cases in which there are only for-
mal services involved, formal services alone are not
Consistent with a few other examples in the qual-
ity of care literature (e.g., Skinner et al., 1999), we
chose to measure our dependent variable, sufficiency
of care, within specific areas of functional needs,
with two important consequences. First, our data
can reveal variations in sufficiency of care across the
various ADL and IADL areas for which care is
needed and provided. Findings thus have the poten-
tial to pinpoint specific areas of challenge or diffi-
culty in providing enough care, as well as highlight
areas of care for which the care provider and care re-
cipient have different perspectives. Second, the re-
sultant aggregate ratings of sufficiency should be
more reliable than global ratings that are not tied to
specifics.
This study’s findings replicate and build on prior
work demonstrating that older adults rate their func-
tional abilities higher than professionals. The study
adds new knowledge regarding differences in per-
spectives by documenting that the assessments of suf-
ficiency of care also vary by rater role. In general,
care recipients rate amount of care as more sufficient
compared with how professionals rate it. These find-
ings are very consistent with related work that sug-
gests older adults rate their functional ability and
health higher than professionals do, as well as iden-
tify fewer problems and needs.
These data suggest that variance between nurse
ratings and care recipient ratings of physical function
derives from both ADL and IADL areas. In 10 out of
13 activities, care recipients rated their functional
ability higher than nurses did. The nurses rated func-
tional ability higher in only one area—bathing. Per-
haps dependent elders are more conservative about
their abilities in the area of bathing out of fear of
falling in the bathroom.
When considering the source of variance from
ADL and IADL areas in the sufficiency ratings, the
pattern is less clear than in functional assessments.
Care recipients rate their care as more sufficient in
five out of nine need areas, with no statistical differ-
ence between the two ratings in three areas. In regard
to transportation, nurses assessed the amount of help
to be more sufficient. It may be that nurses view the
amount of transportation as sufficient if it is enough
to get the care recipient to necessary destinations,
like medical appointments. The care recipient may
view other trips as important as well, such as those
for social visits or pleasure outings; the care recipient
may assess help as less than enough if these trips are
not occurring, whereas the nurse may view these
types of trips as more discretionary.
Although this study does contribute to under-
standing how raters’ assessment of care varies, we
did not capture information about why ratings are
different. It is likely that care recipients are moti-
vated to report sufficient levels of assistance for two
reasons. First, as Donabedian noted, “unless special
precautions are taken, patients may be reluctant to
730
The Gerontologist

Dodge, J., Janz, N., & Clark, N. (1994). Self-management of the health
care regimen: A comparison of nurses’ and cardiac patients’ percep-
tions. Patient Education and Counseling, 23, 73–82.
Donabedian, A. (1988). The quality of care: How can it be assessed? Jour-
nal of the American Medical Association, 260, 1743–1748.
Donabedian, A. (1992). The role of outcomes in quality assessment and
assurance. Quality Review Bulletin, 18, 356.
Duke University Center for Study of Aging and Human Development.
(1978). Multi-dimensional functional assessment: The OARS method-
ology (2nd edition). Durham, North Carolina: Author.
Epstein, A. M., Hall, J. A., Tognetti, M., Son, L., & Conant, L. (1989).
Using proxies to evaluate quality of life. Medical Care, 27, S91–S98.
Fillenbaum, G. (1988). Multidimensional functional assessment of older
adults. Hillsdale, NJ: Erlbaum.
Friedsam, H. J., & Martin, H. W. (1963). A comparison of self- and phy-
sician’s health ratings in an older population. Journal of Health and
Social Behavior, 4, 173–183.
Hollingshead, A. (1957). A two-factor index of social position. Unpub-
lished manuscript.
Hooyman, N., Gonyea, J., & Montgomery, R. (1985). The impact of in-
home services termination for caregivers. The Gerontologist, 25, 141–
145.
Kaufert, J. M. (1983). Functional ability indices: Measurement problems
in assessing their validity. Archives of Physical Medicine and Rehabili-
tation, 64, 260–267.
Keith, P. (1975). Evaluation of services for the aged by professionals and
the elderly. Social Service Review, 49, 271–278.
Kivinen, P., Halonen, P., Eronen, M., & Nissinen, A. (1998). Self-rated
health, physician-rated health and associated factors among elderly
men: The Finnish cohorts of the Seven Countries Study. Age and Age-
ing, 27, 41–47.
Kiyak, H. A., Teri, L., & Borson, S. (1994). Physical and functional health
assessment in normal aging and in Alzheimer’s disease: Self-reports vs
family reports. The Gerontologist, 34, 324–330.
Kosberg, J., & Cairl, R. (1992). Burden and competence in caregivers of
Alzheimer’s disease patients. Journal of Gerontological Social Work,
18, 85–97.
Kosekoff, J., Kahn, K. L., Rogers, W. H., Reinisch, E.J., Sherwood, M. J.,
Rubenstein, L. V., Draper, D., Roth, C. P., Chew, C., & Brook, R. H.
(1990). Prospective payment system and impairment at discharge.
Journal of American Medical Association, 265, 1980–1983.
Lenert, L., Ziegler, J., Lee, T., Sommi, R., & Mahmoud, R. (2000). Differ-
ences in health values among patients, families members, and providers
for outcomes in schizophrenia. Medical Care, 38, 1011–1021.
Maddox, G. L. (1964). Self-assessment of health status: A longitudinal
study of selected elderly subjects. Journal of Chronic Disease, 17, 449–
460.
Maddox, G. L., & Douglas, E. B. (1973). Self-assessment of health: A lon-
gitudinal study of elderly subjects. Journal of Health and Social
Behavior, 14, 87–93.
Magaziner, J., Simonsick, E. M., Kashner, T. M., & Hebel, J. (1988). Pa-
tient-proxy response comparability on measure of patient health and
functional status. Journal of Clinical Epidemiology, 41, 1065–1074.
Mamon, J., Steinwachs, D. M., Fahey, M., Bone, L. R., Oktay, J., &
Klein, L. (1992). Impact of hospital discharge planning on meeting pa-
tient needs after returning home. Home Services Research, 27, 155–
175.
enough). These interpretations are consistent with
our finding that caregiver proximity is a critical fac-
tor; often, formal providers are not as accessible or
proximate as informal providers. In short, these find-
ings suggest that a focus on formal care at expense of
informal care is not sufficient to improve the quality
of in-home supportive care.
Ratings of the sufficiency of care did not vary sig-
nificantly by elder age, gender, race, or SES, whether
from the perspective of care recipient or of the pro-
fessional. This failure to find disparities in the suffi-
ciency of care for various demographic groups, in-
cluding those at risk for poorer health outcomes,
may be reassuring from a social justice perspective.
Important implications from this study are that
raters of sufficiency of in-home supportive care are
not interchangeable, and that information source is a
critical factor in quality of care studies. Both care
recipient and professional ratings are valid because
sufficiency perspectives may be based on different
standards and different values—but they are not in-
terchangeable. Program evaluators and researchers
need to determine which perspective is most appro-
priate for the situation under study. Professional rat-
ings may be more related to medical outcomes, such
as readmission or morbidity, whereas care recipient
ratings may be more related to quality of life or satis-
faction measures. Program developers and funders
also need to pay close attention to the source of in-
formation in needs assessments. As Rosow (1964)
stated many years ago, problems of old age can be
divided into two groups: those that they actually
have or think they have and those the professionals
think they have. More recently, Lenert, Ziegler, Lee,
Sommi, & Mahmoud, (2000) pointed out that sub-
stituting providers’ views for those of patients might
lead to overtreatment of patients. In sum, these find-
ings confirm that rater perspective is likely to have
consequences for health service planning, delivery,
and evaluations, and that various stakeholders define
quality differently, which translates into different ex-
pectations of the health care system and thus differ-
ing evaluations of quality (McGlynn, 1997).
Markides, M. (1979). Self-rated health. Research on Aging, 1, 98–112.
McGlynn, E. A. (1997). Six challenges in measuring the quality of health
care. Health Affairs, 16, 7–21.
Morrow-Howell, N., Proctor, E., & Berg-Weger, M. (1993). Adequacy of
informal care for elderly patients going home from the hospital: Dis-
charge planner perspectives. Journal of Applied Gerontology, 12(2),
188–205.
Morrow-Howell, N., Proctor, E. K., & Dore, P. (1998). Adequacy of care:
The concept and its measurement. Research on Social Work Practice,
8, 86–102.
References
Applebaum, R. (1995). Quality and quality assurance. In G. Maddox
(Ed.), Encyclopedia of aging (pp. 786–787). New York: Springer.
Applebaum, R., Mollica, R., & Tilley, J. (1997–1998). Assuring homecare
quality: A case study of state strategies. Generations, Winter, 57–63.
Avant, W., & Dressel, P. (1980). Perceiving needs by staff and elderly cli-
ents: The impact of training and client contact. The Gerontologist, 20,
71–77.
Blessed, G., Tomlinson, B., & Roth, M. (1968). The association between
quantitative measures of dementia and of senile change in the cerebral
grey matter of elderly subjects. British Journal of Psychiatry, 114,
797–811.
Bonneux, L., Baredregt, J., Meeter, K., Bonsel, G., & van der Mass, P.
(1994). Estimating clinical morbidity due to ischemic heart disease and
congestive heart failure. American Journal of Public Health, 84, 20–28.
Capitman, J., Abrahams, R., & Ritter, G. (1997). Measuring the adequacy
of home care for frail elders. The Gerontologist, 37, 303–313.
Cody, R., & Smith, J. K. (1997). Applied Statistics and the SAS Program-
ming Language (4th ed.) [Computer software]. Upper Saddle River,
NJ: Prentice Hall.
Morrow-Howell, N., Proctor, E., Dore, R., & Kaplan, S. (1998). Post-
acute services to older patients with heart disease. Journal of Applied
Gerontology, 17, 150–171.
Ory, M. G., & Duncker, A. P. (1992). In-home care for older people:
Health and supportive services. Newbury Park, CA: Sage.
Poulshock, S. W., & Deimling, G. T. (1984). Families caring for elders in
residence: Issues in the measurement of burden. Journal of Gerontol-
ogy, 39, 230–239.
Proctor, E., Morrow-Howell, N., Albaz, R., & Weir, C. (1992). Patient and
family satisfaction with discharge plans. Medical Care, 30, 262–275.
Riesenfield, M., Newcomer, R., Berlant, P., & Dempsey, W. (1972). Per-
ceptions of public service needs: The urban elderly and the public
agency. The Gerontologist, 12, 185–190.
Rosow, I. (1964). Housing and social integration of the aged. Cleveland,
OH: Case Western Reserve University.
Coe, M., Wilkenson, A., & Patterson, P. (1986). Preliminary evidence on
the impact of DRG’s: Dependency at Discharge (Final report). Beaver-
ton, OR: Northwest Oregon Health Systems.
Vol. 41, No. 6, 2001
731

Rothman, M. L., Hedrick, S. C., Bulcroft, K. A., Hickam, D. H., &
Rubenstein, L. Z. (1991). The validity of proxy-generated scores in
measures of patient health status. Medical Care, 29, 115–124.
Rubenstein, L., Schairer, C., Wieland, D., & Kane, R. (1984). Systematic
biases in functional status assessments of elderly adults: Effects of dif-
ferent data sources. Journal of Gerontology, 39, 686–691.
Skinner, E., Steinwachs, D., Handley, K., Lehman, A., Fahey, M. & Lyles,
C. (1999). Met and unmet needs for assistance and quality of life for
people with severe and persistent mental disorders. Mental Health Ser-
vices Research, 1, 109–118.
(1992). Response comparability of family and staff proxies for nursing
home residents. American Journal of Public Health, 82, 747–749.
Wagner, D. (1988). Who defines quality: Consumers or professionals?
Caring, 7, 26–28.
Wilson, C. C., & Netting, F. E. (1987). Comparison of self- and health
professionals’ ratings of the health of community-based elderly. Inter-
national Journal of Aging and Human Development, 25, 11–25.
Woodruff, L., & Applebaum, R. (1996). Assuring the quality of in-home
supportive services. Journal of Aging Studies, 10, 157–169.
Strawbridge, W., & Wallhagen, M. (1992). Is all in the family always
best? Journal of Aging Studies, 6, 81–92.
Streib, G. F., Suchman, E. A., & Phillips, B. S. (1958). An analysis of the
validity of health questionnaires. Social Forces, 36, 223–232.
Tennstedt, S. L., Skinner, K. M., Sullivan, L. M., & McKinlay, J. B.
Received February 2, 2001
Accepted July 31, 2001
Decision Editor: Laurence G. Branch, PhD
732
The Gerontologist