Full text of DOI:10.1177/0164027500221004

RP e
E
n
Sr oE
d
A eR t Ca lH
.
/ OI NN F
A
O
G
RI MN GA L CARE
Effects of Posthospital Informal
Care on Nursing Home Discharge
JOAN D. PENROD
University of Nebraska Medical Center
ROSALIE A. KANE
ROBERT L. KANE
University of Minnesota
This study examines the effect of family caregiving on the probability that nursing
home residents would be discharged to the community. The effect of the number of
hours of informal care on the probability of nursing home discharge was estimated
using a logistic regression of a 6-week postadmission location (home or institution)
on the number of hours of informal care in the first 2 weeks in the nursing home, of
caregiver visits, and other patient factors. The odds of being discharged to their home
were higher for those who received more care that is informal. Informal care may
increase the quality and the amount of care that residents receive, thus, influencing
rehabilitation outcomes and returns to home. Caregiving families may advocate for
residents and signal to nursing home staff that the resident has a well-functioning sup-
port system.
When older people move to nursing homes, they usually are not aban-
doned by their families. In fact, the norm is for families to maintain
their ties to the older person through visiting and through continued
involvement in care activities (Naleppa 1996). Much of the research
on family involvement in nursing home care focuses on the role that
families and, to a lesser extent, friends play in providing and receiving
emotional support through their continued contact with the older
AUTHORS’NOTE: This study uses data from the Post Acute Care (PAC) study funded un-
der a cooperative agreement with the Office of the Assistant Secretary for Planning and Evalua-
tion (ASPE) and the Health Care Financing Administration (HCFA Grant No. 17-C98891) in the
U.S. Department of Health and Human Services. Funding was also provided by the National In-
stitute on Aging (Grant No. NIH/5R01-AG11133-02). Joan Penrod was funded in part by the
Brookdale Foundation as a Brookdale National Fellow.
RESEARCH ON AGING, Vol. 22 No. 1, January 2000 66-82
©
2000 Sage Publications, Inc.
6
6
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
67
person (Bitzan and Kruzich 1990; Dobrof and Litwak 1977; Mont-
gomery 1982). This study examines the effect of informal care from
family on the probability that the older person will be discharged from
the nursing home to the community within 6 weeks of admission.
Dobrof and Litwak (1977) suggested that families should be
encouraged to maintain ties with the institutionalized older person
through a shared responsibility with the nursing home staff for the
resident’s care. A survey of residents and staff by Schwartz and Vogel
(1990) revealed some discrepancy between staff and family views
about roles and responsibilities. Although the staff believes that per-
sonal care is their responsibility, families have indicated an interest
and willingness to provide at least some of the care.
A qualitative study suggested that families, in fact, do provide both
activities of daily living (ADL) and instrumental activities of daily liv-
ing (IADL) assistance to nursing home residents. Tickle and Hull
(1995) observed and interviewed the families of 30 residents while
they were visiting a nursing home. Some family members fed, shaved,
and bathed residents. In addition, they did laundry, shopped, and
changed bedlinens. Unfortunately, the studydidnot report the propor-
tion or frequency of family involvement in personal care activities.
Family members noted that they provided care to improve the quality
of care; as an expression of their love and commitment to their rela-
tive; and to provide continuity, sometimes in the form of familiar
foods from home, with the resident’s life before the nursing home. In
another qualitative study, Bowers (1988) found that many family
members of nursing home residents perceive a particular need for
them to help the staff gain a sense of who their family member is and
has been as a person. This perceived need is expressed by relatives of
cognitively intact, as well as cognitively impaired, residents.
Professional wisdom suggests that families have an important reha-
bilitation role to play. Geriatric physicians, nurses, and rehabilitation
therapists emphasize the importance of including the family as part of
the rehabilitation team (Brody and Pawlson 1990). Family caregivers
may directly influence the quality of care that the resident receives by
adding to the rehabilitation efforts under the direction of the profes-
sionals in the nursing home.
These few studies have suggested reasons to expect a positive rela-
tionship between the amount of family care residents received during
the postacute period and the probability of discharge from the nursing
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

6
8
RESEARCH ON AGING
home. First, informal care may increase the quality of care that the
resident receives from the paid caregivers in the nursing home or the
total amount of care received from all sources, thus influencing reha-
bilitation outcomes and returns to home. The influence on rehabilita-
tion may be direct. Families may actually help their family members
to become more independent in ADLs by encouraging them to do
self-care. Alternatively, the effect may act indirectly through staff
efforts to engage available and willing families in rehabilitation
activities.
In addition, families who deliver care in the nursing home may sig-
nal to nursing home staff that this resident has a well-functioning sup-
port and care system who are willing and able to provide care at home
just as they are providing it during the nursing home stay. The experi-
ence of caring may also increase resident, family, and staff confidence
in the resident’s ability to manage at home. Moreover, caregiving
families may serve as advocates for the resident with the staff. Conse-
quently, those residents may get more care from the staff. The addi-
tional care may improve outcomes, including being discharged back
home.
In summary, family involvement in nursing home care has been
examined primarily with respect to emotional and psychological
effects on residents and their families. The effects of family care on the
probability that the patient will be discharged to the community have
rarely been examined. An exception is Lewis and colleagues’ (1985)
discovery that visits by family or friends predicted short-term survival
in a nursing home, but that they were not associated with being dis-
charged back home or long-term survival. This study did not measure
actual care, just visits. Nevertheless, most studies of predictors of
nursing home discharge to a home or community setting invariably
fail to include a measure of family involvement in the nursing home
care (Berg and Mor 1995; Engle and Graney 1993; Garrard et al. 1990;
Gillen et al. 1996; Greene and Ondrich 1990; Kiel et al. 1994; Liang
et al. 1996; Liu, Coughlin, and McBride 1991; Murtaugh 1994).
This analysis uses data from a study of postacute care that followed
a cohort of Medicare patients who were discharged from a hospital for
up to a year (Kane et al. 1996). The current analysis examines the
effect of family caregiving on the probability that the resident will be
discharged to the community by 6 weeks after admission to the nurs-
ing home. In contrast to previous studies of predictors of nursing home
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
69
discharge, a measure of the amount of informal care that the older per-
son receives in the nursing home is included as a predictor. In addition,
a separate measure of the time spent visiting the resident in the nursing
home was included, and its effect on the discharge probability is dis-
tinguished from the effect of the time spent caregiving.
Methods
SAMPLE
The analytic sample was obtained from a study that examined
short-term (6 weeks) and long-term (up to 12 months) postacute care
outcomes for Medicare beneficiaries who were discharged from hos-
pitals after a treatment of one of five conditions common among the
elderly and known to require varying levels of postacute care: stroke,
chronic obstructive pulmonary disease (COPD), congestive heart fail-
ure (CHF), hip fractures, and hip procedures (Kane et al. 1996). Thus,
the sample provides an opportunity to observe variation in postacute
care and outcomes.
Participants were enrolled in the study between March 1988 and
February 1989 from hospitals in three metropolitan areas: Pittsburgh,
Pennsylvania (18 of 20 eligible hospitals participated); Houston,
Texas (15 of 31 eligible hospitals participated); and Minneapolis/St.
Paul, Minnesota (19 of 19 eligible hospitals participated).
Patients were interviewed just before their discharge and again at 6
weeks postdischarge. The patient’s primary caregiver, defined as the
person identified by the patient as providing the most assistance with
care or arranging care in the 6 weeks after the hospital discharge, was
interviewed 6 weeks postdischarge. Additional data collected
included the patient’s functional and cognitive status, and living situa-
tion. Finally, patient’s degree of illness at discharge was measured
with a composite sickness score adapted from the RAND study (Kee-
ler et al. 1990). A score of 1 indicates that a patient had at least one of
the following indicators of sickness at discharge: temperature, use of
catheter, shortness of breath, abnormal heart rate, abnormal respira-
tory rate, elevated blood pressure, and cardiac arrhythmia.
Because this analysis focuses on the effect of postacute informal
care on patients discharged from the hospital to a nursing home,
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

7
0
RESEARCH ON AGING
patients discharged to their home (n = 1,560) or to rehabilitation facili-
ties (n = 184) were excluded from the analyses. Of the 451 patients
discharged to a nursing home, 72 (16%) reported that they did not
have a caregiver. An additional 97 patients (22%) were not included
in the analysis because the caregiver was not interviewed at 6 weeks
because of logistical problems with data collection that were unre-
1
lated to the characteristics of patients or caregivers. In summary, 282
patients and their primary caregiver met the following criteria for
inclusion in this study:
1
2
. The patient was discharged to a nursing home from the hospital.
. The patient interviews at discharge and at 6 weeks were available for
analysis.
3
. The patient had a caregiver, and the 6-week caregiver interview was
available for analysis.
Differences between the patients included in the analytic sample
and those excluded because their caregiver was not interviewed were
examined. Specifically, the probability that there was a caregiver in-
terview for the patient was regressed on the patient’s characteristics
measured at discharge and on the patient’s location at 6 weeks post-
hospital. The two groups did not differ on the probability of nursing
home discharge at 6 weeks, with regard to gender, diagnosis related
group (DRG), hospital discharge functional or cognitive status, sick-
ness, or patient age.
MODEL SPECIFICATION AND ESTIMATION
The effect of informal care on the probability of nursing home dis-
charge by 6 weeks posthospital was estimated using a logistic regres-
sion of the 6-week location (home or institution) on the amount of
informal care given by the caregiver in the first 2 weeks posthospital
discharge to nursing home, frequency of visits by the caregiver, DRG,
illness severity, patient gender, and the functional and cognitive status
at the time of discharge from the hospital. Because visiting is a prereq-
uisite for giving informal care, the two concepts are inevitably linked.
To eliminate problems with collinearity, we developed separate
regression models using the amount of informal care and the extent of
visiting separately.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
71
Several recent studies of factors associated with nursing discharge
and length of stay (LOS) suggest that empirical duration models are
appropriate for examining questions about the timing of events (time
until death or discharge) and handling the right censoring of these data
(the arbitrary cutoff of observation of the LOS) (Greene and Ondrich
1
990; Liang et al. 1996; Liu et al. 1991). Unfortunately, the patient’s
LOS in the nursing home was not available in the data set. Hence, the
research question addresses the effect of informal care on discharge
from nursing home within a specified time rather than on the specific
timing or duration of nursing home care. The logistic model estimated
that using transition by 6 weeks as the dependent variable is reasona-
bly equivalent to a duration model for a 6-week period.
MEASURES
Dependent variable. Discharge from a nursing home is measured
with a dichotomous variable indicating whether the patient was dis-
charged home from the nursing home by the 6-week posthospital
interview. As indicated in Table 1, residents discharged home by 6
weeks are coded as a 1, and those who are still in the nursing home or
returned to the hospital are coded as a 0.
Independent variables. The caregivers’self-report of the number of
hours of care that they provided at the nursing home in the first 2
weeks following hospital discharge was used as the measure of infor-
mal care. It is an estimate of the combined time spent on helping with
personal care such as bathing, dressing, and toileting; with IADL
activities such as housekeeping, cooking, and laundry, if it is required
because of the illness; with arranging services and managing business
affairs; and with general supervision and watching to make sure that
the patient was safe. The amount of care was coded with three catego-
ries: 0 hours of care, 1 to 13 hours over 2 weeks, and 14 to 224 hours.
Data on the hours of care provided by other informal caregivers were
not available in the data set. As such, the variable covers the majority,
but not the total hours, of the informal care received.
A separate measure of the time spent visiting the resident in the
nursing home was included. Unfortunately, the interview question did
not provide a category for caregivers who did not visit at all. Conse-
quently, nonvisiting caregivers make up an unknown proportion of
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

7
2
RESEARCH ON AGING
those who visited weekly or less. The variable is categorical as fol-
lows: caregiver visits daily, several timesa week, orweeklyandless.
The patient’s functional status is measured with a weighted sum of
seven functions: incontinence, bathing, dressing, toileting, transfer-
ring, feeding, and walking. The weighted ADL measure was devel-
oped specifically for the PAC study (for a detailed discussion, see
Finch, Kane, and Philp 1995). It was developed through a magnitude
estimation procedure to produce the weights (generated by a panel of
experts in geriatrics) for each of the functional areas and for each level
of impairment within the areas.
The weighted score is highly correlated with the conventional, sim-
ple count of ADL limitations. However, the weighted measure
addresses problems inherent in other ADL and IADL measures. The
weighted score does not assume that each area of limitation is equiva-
lent to every other area. For example, the inability to use the toilet has a
relatively high score of 848, compared to a little help with feeding
(
score of 424) and a little help with transferring (score of 401), which
has nearly the same score as the person requiring help with toileting
score of 825). The weighted score for each function was summed to
(
produce a total dependency score. Higher scores reflect higher levels
of disability. Values on the scale range from 0 to 5,431 (patient needs
complete assistance with all seven ADLs).
Cognitive status was measured with the Short Portable Mental
Status Questionnaire (SPMSQ) developed by Pfeiffer (1975) as the
number of errors on the test. Higher scores reflect higher levels of dis-
ability. Values range from 0 to 10 wrong.
Table 1 summarizes all the variables included in the analysis. The
table indicates how the variable is measured, its data source, and the
reference category for categorical variables in the analysis.
Results
OVERVIEW OF SAMPLE
Table 2 summarizes the characteristics of patients and caregivers in
the sample. The majority of patients discharged from a hospital to a
nursing home were still in the nursing home at 6 weeks posthospital.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

TABLE 1
Variables Specification
Variable
Measurement
Source
Reference Category
Probability of discharge from nursing
home by 6 weeks postdischarge
Patient functional disability score at
hospital discharge
1 = patient is discharged home
by 6 weeks postdischarge
Points on the weighted ADL
measure
In-person 6-week patient interview
In-person discharge patient interview
0 = patient is in nursing home
Patient cognitive status at discharge
Patient gender
Patient age
Patient illness is stroke
Patient illness is CHF
Patient illness is hip procedure
Composite sickness score at discharge
Caregiver visits nursing home
weekly or less
several times a week
daily
Caregiver hours of care over 2 weeks
zero
Number wrong on SPMSQ
1 = female
Year
1 = DRGs 14, 1, 5
1 = DRG 127
1 = DRGs 209, 210, 211
1 = at least one indicator
In-person discharge patient interview
In-person discharge patient interview
In-person discharge patient interview
Patient medical record
Patient medical record
Patient medical record
0 = male
Patient illness is hip fracture
Patient illness is hip fracture
Patient illness is hip fracture
0 = no sickness
Patient medical record
1 = Yes
1 = Yes
6-week telephone caregiver interview
6-week telephone caregiver interview
Caregiver visits weekly or less
Caregiver provides 0 care hours
1
to 35
over 224
NOTE: ADL = activitiesofdailyliving; SPMSQ = Short Portable Mental Status Questionnaire; CHF = congestive heart failure; DRG = diagnosis related group.

7
4
RESEARCH ON AGING
TABLE 2
Characteristics of Study Sample:
Means and Proportions (n = 282)
Percentage
Patient characteristics
Patient is home by 6 weeks postdischarge from hospital
Female
White
34
77
97
Age
Mean
83
SD
6.9
Cognitive status score at discharge (number wrong on SPMSQ)
0
13.5
33.7
52.8
41
1
3
to 2
or more
Patients with at least one sickness indicator at discharge
DRG
Stroke
COPD
23
0
CHF
6
Hip procedures
Hip fractures
10
61
Functional status score at discharge
Mean
SD
3,097.7
944.3
Caregiver characteristics
Female
69
Age
Mean
59
SD
14.4
44
Caregiver works outside the home
Relationship to patient
Spouse
Adult child
Other family or friend
Missing
16
50
30
4
Has at least one minor dependent at home
Amount of caregiving in the nursing home over 2 weeks
13
0
1
3
hours
to 34 hours
5 to 224 hours
8.5
31.9
58.2
1.4
Missing
Caregiver visits
Daily
Several times a week
Weekly or less
Missing
53
25
13
9
NOTE: DRG = diagnosis related group; COPD = chronic obstructive pulmonary disease; CHF =
congestive heart failure.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
75
The average age of the patients was 83 years, and they were predomi-
nately female and White (97%).
At discharge, patients were relatively cognitively impaired, with
5
2% making 3 or more errors out of 10 on the SPMSQ. With respect to
functional disability, patients were quite impaired at their discharge to
a nursing home (mean = 3,097.7 on a scale in which a score of 5,431
indicates that the patient needs complete assistance with all seven
ADLs).
Table 2 additionally summarizes the characteristics of the primary
informal caregivers. The majority were female, with an average age of
5
9 years. The majority visited the patient daily or several times a
week. One half of the caregivers were the patient’s progeny. A smaller
group were friends or other relatives (e.g., nieces, nephews, siblings)
of the older person. A minority of caregivers were spouses. This is not
surprising given the average age of the patients, many of whom are
widows. A minority of caregivers (13%) had a dependent under age 18
at home. Close to 50% were employed outside the home.
Caregivers were asked to distinguish between providing ADL and
IADL assistance and simply visiting the resident. About 91% of the
residents received some informal care from the primary caregiver
while in the nursing home. The majority of caregivers provided over
3
5 hours of care over 2 weeks. The majority of the caregivers also vis-
ited the resident daily.
As indicated in Table 3, the frequency of visits and the amount of
care provided by caregivers are not independent (chi-square = 61.1, df =
4
, p < .001). As indicated in the table, caregivers who visit more often
also give more care. The Spearman rank order correlation of .47 (p =
01) confirms the positive direction of the relationship. The relatively
.
high correlation suggests that, although caregiving and visiting are not
the same activity, they overlap. However, at least 4.7% of caregivers
who visit do not provide the older person with any personal care,
IADL assistance, or supervision.
THE EFFECTS OF INFORMAL
CARE ON NURSING HOME DISCHARGE
Approximately 34% of the residents in the sample were home by 6
weeks posthospital discharge. Table 4 summarizes the results of the
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

7
6
RESEARCH ON AGING
TABLE 3
The Relationship Between Visiting and Caregiving
Amount of Care Provided (%)
1
to 35 hours
More than 35 hours
over 2 weeks
Visits
None
over 2 weeks
Total
a
Weekly or less
9 (3.5)
8 (3.1)
4 (1.6)
21 (8.2)
26 (10.2)
56 (22)
74 (29)
1 (0.4)
6 (2.4)
71 (27.8)
78 (30.6)
36 (14.1)
70 (27.5)
149 (58.4)
255 (100)
Several times a week
Daily
Total
156 (61.2)
NOTE: Chi-square = 61.1, df = 4, p < .001.
a. Data were coded such that caregivers who did not visit at all are indistinguishable from those
who visited less than weekly.
logistic regression model to estimate the effect of each independent
variable on the probability of the residents being home by the 6-week
interview. Caregiver characteristics were not included in the final
models presented here because none were significantly different from
zero, and they contributed to large standard errors for the estimates of
effects of other variables. Thus, we concluded that the deletion of
caregiver characteristics yielded a better specified model with more
stable estimates.
We first estimated a logistic regression of the effects of resident
characteristics, caregiving, and visiting by the caregiver on the resi-
dent’s probability of being discharged from the nursing home. Visit-
ing did not influence the odds of discharge (p = .68). In this model, the
odds ratio for low caregiving (compared to none) was 17.9 and the
confidence interval ran from 1.82 to 177.39. The results were similar
for high levels of caregiving. In addition, when visiting and caregiving
are in the same model, the confidence intervals for other resident vari-
ables widen. Consequently, we concludedthat collinearitywasa prob-
lem, and we estimated separate models, one including visiting and the
other including caregiving. The results are presented in Table 4.
The first regression results in Table 4 show the effect of visits by the
caregiver on the probability of the resident being discharged, control-
ling for patient factors. Residents who were visited daily were no
more or less likely to be discharged than those who were visited less
frequently. The second regression, with three levels of caregiving
(including 0 hours as the reference), indicates that residents who
received between 1 and 35 hours of care from a family caregiver were
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
77
TABLE 4
Predictors of Nursing Home Discharge by 6 Weeks:
a
Odds Ratios and 95% CI (n = 278)
With Visiting Included
Odds Ratio 95% CI
With Amount of Care Included
p
Odds Ratio 95% CI
p
Patient is female
Patient age
Discharge cognitive status
Sickness score
DRG is stroke
DRG is CHF
1.05 0.461, 2.37 ns
0.88
0.96
0.87
0.43
0.77
0.20
6.5
0.43, 1.8
ns
0.96 0.92, 1.01
0.83 0.73, 0.94
0.48 0.24, 0.93
0.60 0.25, 1.42
0.26 0.05, 1.36
ns
0.92, 1.00 ns
0.77, 0.97 .01
0.23, 0.82 .01
0.35, 1.69 ns
0.04, 1.04 ns
0.004
0.03
ns
ns
DRG is hip procedure
Patient discharge
functional status
5.3
1.73, 16.4
0.003
2.15, 18.68 .0009
0.98 0.96, 1.00
1.31 0.53, 3.2
1.54 0.60, 3.98
ns
ns
ns
0.98
0.96, 0.99 .04
b
Visits daily
Visits several times a week
c
1
to 35 hours of care
5.02
7.23
1.17, 21.50 .03
1.58, 33.03 .01
More than 35 hours of care
2
NOTE: With visiting included, Cox and Snell pseudo R = .22; with amount of care included, Cox
2
and Snell pseudo R = .25. CI = confidence intervals; DRG = diagnosis related group; CHF =
congestive heart failure.
a. Twenty-five cases deleted due to missing data.
b. Visits weekly or less is reference category.
c. No caregiving is the reference.
five times as likely to be discharged by 6 weeks as those who did not.
Those who received over 35 hours of care were seven times as likely to
be discharged. In addition, the odds of being discharged by 6 weeks
were lower for sicker (odds ratio = .43, p = .01), and more cognitively
and functionally impaired residents. Finally, residents admitted to a
nursing home after hospital treatment for hip procedures were 6.5
times as likely to be discharged as those with other diagnoses.
Discussion
This study examined the effect of informal care from the family on
the probability that the older person would be discharged from the
nursing home to the community within 6 weeks of admission. Care-
givers were, on average, providing 10 hours a week of care to their
family member. After taking account of the resident’s health and
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

7
8
RESEARCH ON AGING
functional status, those who received informal care from family mem-
bers during the nursing home stay were much more likely to be dis-
charged back home by 6 weeks. Visiting by the caregiver did not influ-
ence the likelihood of discharge. Thus, care itself, and not the simple
presence of a caregiver visiting daily, affects discharge from the nurs-
ing home.
This study shows that families provide hands-on care even when a
family member is in a nursing home. Moreover, the care matters. Fam-
ily care in the nursing home greatly increases the likelihood of being
discharged back home within a relatively short time. As the amount of
care increased, the odds of discharge increased, suggesting a dose
effect.
This study cannot definitively explain the relationship between
informal care and discharge. To examine this relationship more
closely, we categorized residents by whether their functional status
score at hospital discharge, when compared to their functional status
score at 6 weeks posthospital, showed functional decline, no change,
or improvement; and whether they received family care or not (table
not shown). There was no association between the receipt of family
care and the change in functional status over the 6 weeks (chi-square =
2
.3, df = 2, p = .311). However, the results from the multivariate logis-
tic regression show that functional status has an independent effect
independent of family care) on discharge: More disabled people are
(
less likely to be discharged. In addition, family care affects discharge
independent of functional status: More family care increases the prob-
ability of discharge. Therefore, family care influences discharge, but
not by improving functional status.
Although this study was not designed to identify how family care
influences discharge, there are several possibilities for further study.
In particular, caregiving families are an obvious indicator to nursing
home staff that the resident has an active, functioning care situation
available in the community. Moreover, the experience of caregiving in
the nursing home may result in greater confidence on the part of the
older person, the family, and the nursing home staff about the resi-
dent’s safety and ability to manage at home.
The research literature offers few insights into how nursing home
staff interacts with family members and gives little reason to think that
any purposive harnessing of family energy occurs. In an ethnographic
study, Foner (1995) found that the nursing staff has negative
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
79
perceptions of family members who visit frequently, even when the
staff takes for granted the help with ADL and IADL activities (e.g.,
feeding, laundry, mobility assistance) performed by those family
members.
Moreover, aides’ attitudes toward family members tended to be
negative regardless of what family members did. Family members
who disrupted the aides’ schedules with requests were resented. As
Foner writes, “A wet bed, a resident who has slipped down in a chair,
food that has spilled on the floor—these to the aide are the realities of
nursing home life.” (p. 73) As such, they resent being summoned for
what they regard as a minor matter. Oddly enough, however, fre-
quently visiting family members who make no demands and who uni-
formly provide actual care to their relatives and perhaps even to their
relative’s roommates were also resented. A common complaint is that
they are spoiling the resident by raising expectations for a high stan-
dard of attention. If these are the prevalent attitudes, much conscious
attention would be needed to create a truly synergistic relationship
between paidcaregivers andcaregivingfamilies inthe nursinghome.
In addition to residents with informal care, hip procedure patients
were also more likely to be home by 6 weeks. This finding is consis-
tent with the clinical course of the illness. These patients are healthier
(the procedure is elective), and they tend to recover more quickly as a
result. Consistent with other studies of nursing home discharge,
higher levels of functional and cognitive impairment decrease the
odds of being discharged back home.
This study has several limitations. The findings are conditioned on
two important factors that should be considered when generalizing to
other groups. First, all the older people in this sample identified a pri-
mary caregiver. Although the vast majority of older people with func-
tional impairments have one or more family members involved in their
care, a subgroup is without informal support. It is possible that the
effect of the number of hours of care would be different if older people
without family caregivers were added to the group of caregivers (8%)
who did not provide any care in the nursing home.
Second, the analyses are restricted to patients discharged from a
hospital directly to a nursing home. Patients who enter nursing homes
from community settings are not considered. The sample describes
the majority, but excludes an important subgroup of older people.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

8
0
RESEARCH ON AGING
This study adds to the large body of literature on effects of informal
care and extends findings to the effects on nursing home residents.
The next step is to determine what family caregivers do that improves
the odds of nursing home discharge and what nursing home staff and
other health care providers—those managing care from outside the
nursing home—can do to help enhance those outcomes.
NOTE
1
. Caregiver interviews were more likely to be missing when the patient discharge interview
was done by a research firm that was later terminated for nonperformance of the contract. Spe-
cifically, the firm was not getting interviews with patients or caregivers completed in a timely
manner.
REFERENCES
Berg, Katherine and Vincent Mor. 1995. “Medicare Nursing Home Residents with a Stroke:
Characteristics and 90-Day Outcomes of Care.” Journal of Aging and Health 7:384-401.
Bitzan, Janet E. and Jean M. Kruzich. 1990. “Interpersonal Relationships of Nursing Home
Residents.” The Gerontologist 30:385-90.
Bowers, B. J. 1988. “Family Perceptions of Care in a Nursing Home.” The Gerontologist
2
8:361-26.
Brody, Stanley J. and Gregory L. Pawlson. 1990. Aging and Rehabilitation II: The State of the
Practice. New York: Springer.
Dobrof, Rose and Eugene Litwak. 1977. Maintenance of Family Ties of Long-Term Care
Patients:Theory andGuide to Practice. Rockville, MD: National Institute of Mental Health.
Engle, Veronica and Marshall Graney. 1993. “Predicting Outcomes of Nursing Home Residents:
Death and Discharge Home.” Journal of Gerontology 48:269-75.
Finch, Michael D., Robert L. Kane, and Ian Philp. 1995. “Developing a New Metric for ADLs.”
Journal of the American Geriatrics Society 43:877-84.
Foner, N. 1995. “Relatives as Trouble: Nurses Aides and Patients’Families.” Pp. 165-178 in The
Culture of Long-Term Care: Nursing Home Ethnography, edited by J. Neil Henderson and
Maria D. Vesper. Westport, CT: Begin & Harvey.
Garrard, Judith, Robert L. Kane, David M. Radosevich, Carol L. Skay, S. Arnold, Loyd Kepfer-
lie, Susan McDermott, and Joan L. Buchanan. 1990. “Impact of Geriatric Nurse Practi-
tioners on Nursing Home Residents’ Functional Status, Satisfaction, and Discharge Out-
comes.” Medical Care 28:271-83.
Gillen, Patrick, Diana Spore, Vincent Mor, and Walter Freiberger. 1996. “Functional and Resi-
dential Status Transitions Among Nursing Home Residents.” Journal of Gerontology
5
1A:29-36.
Greene, Vernon L. and Jan I. Ondrich. 1990. “Risk Factors for Nursing Home Admissions and
Exits: A Discrete-time Hazard Function Approach.” Journal of Gerontology 45:250-58.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

Penrod et al. / INFORMAL CARE
81
Kahn, L. Katherine, Emmett B. Keeler, Marjorie J. Sherwood, William H. Rogers, David
Draper, Stanley S. Bentow, Ellen J. Reinisch, Lisa V. Rubenstein, Jacqueline Kosecoff, and
Robert H. Brook. 1990. “Comparing Outcomes of Care Before and After Implementation of
the DRG-based Prospective Payment System.” Journal of the American Medical Associa-
tion 264:1984-8.
Kane, Robert L., Michael Finch, Lynn Blewett, Qing Chen, Risa Burns, and Mark Moskowitz.
1
996. “Use of Post-hospital Care by Medicare Patients.” Journal of the American Geriatrics
Society 44:242-50.
Keeler, Emmett B., Katherine L. Kahn, David Draper, Marjorie J. Sherwood, Lisa V. Rubenstein,
Ellen J. Reinisch, Jacqueline Kosecoff, and Robert H. Brook. 1990. “Changes in Sickness at
Admission Following the Introduction of the Prospective Payment System.” Journal of the
American Medical Association 264: 1962-68.
Kiel, Douglas P., Ann Eichorn, Orna Intrator, Rebecca A. Silliman, and Vincent Mor. 1994. “The
Outcomes of Patients Newly Admitted to Nursing Homes After Hip Fracture.” American
Journal of Public Health 84:1281-6.
Lewis, Mary A., Robert L. Kane, Shan Cretin, and Virginia Clark. 1985. “The Immediate and
Subsequent Outcomes of Nursing Home Care.” American Journal of Public Health
7
5:758-62.
Liang, Jersey, Xian Liu, Edward Tu, and Nancy Whitelaw. 1996. “Probabilities and Lifetime
Durations of Short-stay Hospital and Nursing Home Use in the United States, 1985.” Medi-
cal Care 34:1018-1036.
Liu, Korbin, Teresa Coughlin, and Tim McBride. 1991. “Predicting Nursing Home Admission
and Length of Stay.” Medical Care 29:125-41.
Montgomery, Rhonda J. V. 1982. “Impact of Institutional Care Policies on Family Integration.”
The Gerontologist 22:54-8.
Murtaugh, Christopher. 1994. “Discharge Planning in Nursing Homes.” Health Services
Research 28:751-69.
Naleppa, Matthias J. 1996. “Families and the Institutionalized Elderly: A Review.” Journal of
Gerontological Social Work 27:87-111.
Pfeiffer, E. 1975. “A Short Portable Mental Status Questionnaire for the Assessment of Organic
Brain Deficit in Elderly Patients.” Journal of the American Geriatrics Society 23:433-41.
Schwartz, Arthur N. and Mark E. Vogel. 1990. “Nursing Home Staff and Residents’ Families
Role Expectations.” The Gerontologist 30:49-53.
Tickle, Eugenia H. and Kathleen V. Hull. 1995. “Family Members’ Roles in Long-term Care.”
Medical Surgical Nursing 4:300-304.
Joan D. Penrod, Ph.D., is an assistant professor at the University of Nebraska Medi-
cal Center. She is a 1997 recipient of a Brookdale Foundation National Fellowship.
Her research interests include aging, long-term care and rural health policies and
programs.
Rosalie A. Kane, D.S.W., is a professor of health services research and policy at the
University of Minnesota School of Public Health. She is a member of the graduate fac-
ulties of social work and biomedical ethics, and she directs a national long-term care
resource center. Her research deals with long-term care practice and policy in multi-
ple settings, including home care, assisted living, and nursing homes.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015

8
2
RESEARCH ON AGING
Robert L. Kane, M.D., is a professor of health services research and policy at the Uni-
versity of Minnesota School of Public Health, where he holds an endowed chair in
long-term care and aging. He directs the university’s center on aging and its Clinical
Outcomes Research Center. His research interests lie in the intersection of long-term
care and clinical outcomes research that concentrates on the effects of various types
of care on the lives of older people.
Downloaded from roa.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 29, 2015