| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Neena L. Chappell, PhD, FRSC, CRC, Canada Research Chair, Social Gerontology Professor, Centre on Aging and Department of Sociology, University of Victoria, Victoria, British Columbia V8W 2Y2, Canada. E-mail: nlc{at}uvic.ca
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Home care • Cost of care • Residential care • Informed care
At the end of the 20th century, the Canadian government focused serious attention on the concern with the ability to publicly fund the costs of care for seniors as the baby-boom generation ages. In 1998, Health Canada established a one-time Health Transition Fund to support research on home care, including the research reported here. This study asks whether or not home care is a cost-effective alternative to long-term facility care for clients who are clinically eligible for institutional long-term care in two Canadian cities: Victoria, British Columbia (BC), and Winnipeg, Manitoba. The research is notable in two respects. It computes costs for informal care in both settings, and it ensures comparable groups of clients in both settings (i.e., it compares individuals at the same level of care).
Review of the Literature
In the 1980s, considerable research was conducted on home care as a cost-saving alternative to facility care. Much of this research was conducted in the United States and is based on two series of federally funded studies: 14 community-care demonstration projects funded in the late 1970s and early 1980s and an additional 10 projects funded between 1982 and 1985. These studies introduced case management, frequently with an enhanced home-care program, into the community and randomly assigned eligible clients to existing community services, to enhanced services, or to nursing homes. In general, home-care clients expressed satisfaction with enhanced services, but when the costs of the enhanced home-care program were included, the overall costs were usually greater for the enhanced home-care clients than for those receiving existing community services (Berkeley Planning Associates, 1985; Mathematica Policy Research Inc., 1986). Thus, it was noted that increased spending on home and community care was not accompanied by a commensurate reduction in spending on nursing home care.
Skellie, Favor, Tudor, and Strauss (1984) analyzed the Georgia alternative health-services project. Medicaid-eligible individuals, at least 50 years of age and certified as eligible for residential care, were enrolled by random assignment into a comprehensive range of community-based services, including screening, case management, alternative living services, adult day rehabilitation, and home-delivered services, or alternatively into a control group receiving only existing community services. After 2 years, a similar percentage from each group was admitted to long-term facility care. The costs for the experimental group were considerably higher than those for the control group and constituted add-on costs to Medicaid-reimbursed services. Vertrees, Manton, and Adler (1989) examined the Georgia and California Medicaid-waiver programs. These were enhanced programs that placed emphasis on screening to ensure that those receiving base services would be likely candidates for admission to long-term-care institutions. Although these researchers found that the monthly cost of community care was considerably less than the cost of facility care, not all community-care participants were, in fact, eligible for residential care. Such clients would not have benefited from enhanced home care as a means of keeping them out of institutions, as they did not, in fact, need institutional care. For those in the sample who were eligible for residential care, community services did not prevent admissions.
Hedrick and Inui (1986) analyzed 12 experimental or quasi-experimental cost-effectiveness studies of enhanced home-care services, all of which involved chronically ill persons. They concluded that home-care services had no impact on mortality, patient functioning, or long-term-care facility placements. In addition, home care either had no effect on hospitalization or tended to increase the number of days spent in the hospital. The cost of home care was either not affected or increased by up to 15%.
Weissert (1985) argued that it is difficult to make home- and community-based services cost effective, because (a) community care is an add-on to other services and is not a substitute for residential care; (b) community care does not reduce institutionalization rates; (c) only short long-term-care facility stays can be avoided by community-based care; (d) screening and assessment costs are high; (e) overhead costs can be relatively high when community services are small; and (f) improvements in health status are limited. Weissert and colleagues (1988) expanded this analysis, examining more than 700 citations published between 1960 and 1985 on the relative costs of community- and home-based services compared with residential long-term-care costs. Of these publications, 150 were selected for review, and the 27 most rigorous and generalizable studies were subjected to a detailed analysis. Weissert and colleagues concluded that home- and community- based long-term-care services usually raised overall health-service use and costs.
By the late 1980s, American researchers had generally concluded that enhanced home care did not reduce the demand for residential care. However, during the 1990s it was recognized that this research had tended to compare costs associated with the introduction of a new home-care service, typically enhanced services. In addition, Canadian research during the 1990s suggested that, when home care and residential care were compared directly, home care could be cost effective.
Using data from the Canadian Study of Health and Aging, Østbye and Crosse (1994) calculated the net economic costs of health care for people with dementia by using both direct costs (such as home support, physiotherapy, respite care, and day center care) and indirect costs (such as time spent by informal caregivers in assisting clients with activities of daily living). They estimated that the annual net cost of providing care for those in the community was $10,100 ($4,970 for direct costs and $5,130 for indirect costs) and of providing care in a facility was $19,100. Analyzing the same data, Hux and colleagues (1998) compared community and facility costs for those with severe Alzheimer's disease, finding that the cost of residential care was significantly higher than the cost of community care.
At this time, Weissert, Lesnick, Musliner, and Foley (1997) also published a paper reporting that home care could be cost effective when designed as a substitute for facility care. Examining the Arizona long-term-care system—the first capitated long-term Medicaid program in the United States—these authors found that the cost of home care was substantially less than that of facility care. They suggested that savings likely were associated with a payment approach that encouraged program contractors to place clients in home- and community-based services rather than in facilities for durations beyond that covered by the monthly capitated rate.
Research to date has seldom included the costs incurred by informal care providers. This has raised concerns, particularly in insurance-based systems such as those in the United States, that introducing paid home-care services may result in a process of cost shifting from the formal care system to the informal care system (Segall & Chappell, 2000; Deber et al., 1998). In Canada, where health services are primarily funded by the government, paid home care is seen as a complement to informal care. In effect, the state and the family collaborate in ensuring that appropriate and sufficient care services are provided. This differs from an insurance-based system in which formal home care is seen as an entitlement whereby one has a right to such services without commensurate filial obligations to provide care. Thus, in Canada one generally receives formal care to the extent that the informal care system is assessed as unable to care for the individual; home-care clients are less likely to have informal support systems, and similarly, those admitted to long-term-care facilities tend to be those without informal supports (Béland, 1985; Shapiro & Tate, 1985). Some might argue that informal care within long-term-care facilities is discretionary, even if it contributes to residents' quality of life. However, from a societal perspective, care from all providers is important (at least some of the care provided by informal caregivers in the home would no doubt be deemed discretionary by a panel of experts).
Max, Webber, and Fox (1995) reported that those with Alzheimer's disease living in the community received an average of 286 hours per month of unpaid care, whereas those living in a long-term-care facility received 36 hours per month of unpaid care. Using a replacement approach to estimate costs, these authors estimated a monthly cost of $34,517 for the community sample and $5,542 for the facility sample. In both instances, costs of informal care increased with the severity of the disease. Hébert and colleagues (2001) reported that, for a sample of elderly individuals using home care in Quebec, home care was less expensive than facility care when only public costs were taken into account but more expensive when informal care costs were taken into account. They estimated that family members and volunteers contributed both out-of-pocket funds (estimated to be $30–60 per day on average) and substantial amounts of time. For those living in the community who were not receiving public services, family members provided 82% of nursing care, 73% of the care related to instrumental tasks, and 80% of the supervision. For those receiving public services, family members still provided substantial care: 70% of nursing care, 57.5% of care related to instrumental tasks, and 78% of the supervision.
In summary, literature on public costs of home care compared with long-term institutional care suggests that home care might be a cost-saving alternative; however, the literature is not at all conclusive. When the costs to informal caregivers are taken into account, there is very little literature, and that which exists suggests that home care might be more expensive than long-term institutional care. In order to provide a societal perspective on the costs of home care compared with long-term residential care, both were included in this study.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
In facilities, we randomly selected clients on the basis of their level of care from a list provided by the facilities. In BC there are five levels of care: personal care, intermediate care 1 (IC 1), intermediate care 2 (IC 2), intermediate care 3 (IC 3), and extended care. There are four levels of care in Manitoba, referred to as Levels 1, 2, 3, and 4. To control for these differences, we chose Victoria clients from IC 2 and IC 3, and in Winnipeg we chose them from Levels 3 and 4; that is, we chose from those levels representing people in intermediate-care facilities, often referred to more generically as nursing homes or long-term-care facilities. We excluded extended (or chronic) care clients because a pilot study indicated that data collection from this group was not feasible; unacceptably high missing values resulted. Hollander and Pallan (1995) noted that in the 1991–1992 fiscal year there were 22,309 clients who received care in long-term-care facilities (similar to nursing homes in the United States) and 10,429 clients who received care in extended-care units (i.e., chronic care similar to skilled nursing homes in the United States).
In order to have the most direct cost comparison by level of care, we included in the study only those clients deemed stable at their level and location of care. We considered a client to be stable if he or she (a) had been receiving care from the same agency for at least the past 6 months; (b) had remained at the same level of care for at least the past 3 months; (c) had not received an increase of more than 20% in the number of hours or resources in the past month; (d) had not been admitted to an acute-care hospital more than once in the past 2 months; and, (e) if admitted, stayed in the hospital for 3 days or fewer. We selected both cognitively intact and cognitively impaired clients, but we excluded non-English-speaking individuals. According to earlier research in BC (Hollander, 2001), it is estimated that 82% of home-care clients and 59% of facility clients are stable in a given year. Unstable clients are often transferred to acute-care facilities.
In addition, we excluded those receiving more than 120 hours per month of care-aide time, because, upon examination, it was revealed that these extended hours are being used for monitoring rather than care. This occurred primarily in Winnipeg, where care aides would sometimes be used to monitor at-risk community clients. A 120-hour cut point was chosen because 120 hours is the maximum allowed in BC for home support per month and the policy in Winnipeg is to provide care up to the cost of facility care and 120 hours per month equals that cost. The cutoff, therefore, allowed the removal of program exceptions. There were four community clients who exceeded the 120-hour cutoff in Victoria compared with 80 clients who received 120 hours of care-aide or home-support-worker time, which is a 4.8% exclusion rate. The comparative numbers for Winnipeg were 25 and 41, which is a 37.5% exclusion rate.
We identified a random sample of community clients by each health region. For those who were eligible, agency personnel made first contact. In both sites, this initial contact was made directly with individuals who were considered cognitively intact and with the informal caregiver when the individual was considered cognitively impaired. Preliminary contact was made by phone, by letter, or both. For the facility sample, a senior staff member within the facility identified a potential pool of clients. We then chose a random sample from this pool by applying the inclusion–exclusion criteria noted herein. The facility representative made initial contact with the individual or, in the case of cognitive impairment, an informal caregiver, in order to obtain permission to pass the individual's name on to study personnel.
In Victoria, 5.5% of community clients and 13.5% of facility clients refused to allow their names to be passed on to study personnel. One facility client (0.4% of the sample) died before completing the study. In Winnipeg, 9.8% of community clients and 14.3% of facility clients declined to have their names passed on. Six facility clients (2.5% of the sample) died before completing the study. The final sample consisted of 580 clients, 222 community clients, and 358 facility clients. In Victoria, 281 clients participated, 121 from the community and 160 from the facilities. In Winnipeg, 299 clients participated, 101 in the community and 198 in facilities.
Of the 580 clients, 54 (9.3%) did not have a caregiver. Of the 526 caregivers to clients within the study, 25 (4.8%) refused to participate in the study. Therefore, in total 501 caregivers participated in the study, all providing care to a different client; 184 (36.7%) were providing care for community clients, and 317 (63.4%) were providing care for facility clients. Thus, 82.9% of community clients and 88.5% of facility clients who had an informal caregiver participated in this study.
Clients were approached directly by study personnel to participate in a face-to-face interview at a location convenient to the participants. Community clients were usually interviewed in their homes; facility clients were interviewed in a quiet location in the facility. Interviews took an average of 1.10 hr to complete. When clients were too ill or cognitively impaired to participate directly, proxies (usually informal caregivers) provided factual information but no subjective information regarding beliefs, feelings, or perceptions on behalf of the client. Face-to-face interviews with informal caregivers were completed in one visit and took on average.80 hr. Interviews with clients asked about sociodemographic variables, functional and cognitive abilities, use of and satisfaction with health services, and perceptions of quality of life. Interviews with caregivers collected information on sociodemographic factors, care-related stress, and satisfaction with the client's use of health-related services. Each person was interviewed separately.
The amount of time and type of assistance provided by informal and formal caregivers, and out-of-pocket expenses for both clients and informal caregivers, was provided in various diaries and through budget and operational cost data provided by the regional health authorities. Out-of-pocket payments of informal caregivers included such costs for formal services. These costs were not included as part of the costs of formal care. Diaries were kept for a 2-week period. For community clients, three diaries were left in the client's home: client-expenditure and informal caregiver diaries and a formal services community diary. Home-support agencies were asked to notify their workers regarding the purpose of the diaries and to encourage completion when they visited the client. Where possible, interviewers spoke to the home-support workers regarding the diaries and encouraged their completion during their shifts. For facility clients, the client-expenditures and informal caregiver diaries were left by the client's bed or were given to informal caregivers to keep at home. The formal services–facility diaries were kept at the nursing station within the facility. Interviewers contacted all clients regularly (every 2 or 3 days) during the 2-week period to remind them to complete the diaries. In some instances, clients dictated the relevant information to the interviewers, who recorded it on their behalf (when clients were too frail to complete them on their own or had limited time). Interviewers also checked the formal services–facility diaries every 2 or 3 days to ensure that staff were completing them.
In order to avoid a frequent problem encountered in past research, namely, failing to control for clients with similar needs, levels of care had to be determined so that clients at the same level of care receiving community care and residential care could be compared. Following the argument by Thomas, Rockwood, and McDowell (1998) that using only measures of activities of daily living (ADLs) underestimates dysfunction and disability, we assessed functional abilities by using the Functional Autonomy Measurement System (SMAF) developed by Hébert, Carrier, and Bilodeau (1988), which is based on the World Health Organization's (1980) classification of disabilities.
The SMAF consists of 29 items that measure care needs in relation to the ability to function independently in five areas: 7 ADL items, 6 mobility items, 3 communication items, 5 mental function items, and 8 instrumental ADL (IADL) items, with each item scored on a 4-point scale ranging from 0 (independent) to –3 (dependent) for a maximum score using absolute values of 87. Desrosiers, Bravo, Hébert, and Dubuc (1995) reported test–retest and interrater reliability at.95 and.96, respectively, for the total SMAF score. Correlation coefficients were over.74 for all five subscale scores for both types of reliability. The SMAF correlates.80 with the Older American Resources and Services (OARS) ADL questionnaire (McCusker, Bellavance, Cardin, & Belzile, 1999). The SMAF has also been shown to account for 85% of the variance in required nursing-care time (Hébert et al., 2001).
We collected the SMAF information from the client, a caregiver proxy if necessary, a health care professional, or a combination of these individuals. We use the total SMAF score to create four levels of care at each site, deleting from the sample any persons with outlier scores that placed them outside of the characteristics of the levels. In total, 33 cases were dropped from Victoria in order to allow comparison between home-care and facility-care clients who would be at similar levels of care, leaving 248 clients in four comparable levels of care (created from comparable scores on the SMAF). Similarly, four levels of care were created for Winnipeg residents. In this case, we excluded 56 clients, with 243 remaining for analyses (53 were excluded as they did not fall into the four levels of care and 3 facility persons were excluded for whom there was no SMAF score and therefore no level of care could be created).
An examination of the distributions revealed that the SMAF scores at the two study sites were different. More of the Victoria sample was in the lower end of the distribution; more of the Winnipeg sample was in the upper end (i.e., Victoria clients were more functional than Winnipeg clients). The two study sites were therefore analyzed separately. The levels of care included "somewhat independent" (SMAF score 14.0–22.5); "slightly independent" (SMAF score 23.0–35.0); "slightly dependent" (SMAF score 35.5–45.5); "somewhat dependent" (SMAF score 46.0–61.0); and "largely dependent" (SMAF score 61.5–68.5). In Victoria, the sample included clients in all levels except "largely dependent." In Winnipeg, the sample included all levels except "somewhat independent."
Using the total SMAF score as the dependent variable, we conducted a 2 (type of care; community versus facility) by 4 (care levels) analysis of variance (ANOVA). The analysis confirmed that there was a significant main effect for care levels, indicating that the four groups were distinctly different and that there was neither a significant main effect for type of care nor a significant interaction (see Table 1).
|
2 = 4.41, df = 1, p <.05). The nonproxy group also scored significantly lower on the SMAF (mean score = 26.21) than the proxy group (mean score = 38.62): t = 10.02, df = 244, p <.001. In the Winnipeg sample the two groups were similar with respect to gender, education, ethnic background, and marital status. However, the nonproxy group was significantly younger (mean age 81.1 years compared with 84.4 years; t = 2.56, df = 234, p <.01), had a significantly higher annual income (t = 2.19, df = 218, p <.05), was more likely to have English as their primary language (2 = 5.64, df = 1, p <.05), and had lower scores on the SMAF (M = 36.89 compared with 51.75; t = 11.54, df = 1, p <.001) than was true of the proxy group. That is, proxy and nonproxy groups were fairly similar in terms of demographic characteristics, but, as would be expected, the nonproxy group was more functional than the proxy group. Only complete diaries were utilized for estimating costs. Complete diaries were obtained in the majority of instances, varying from a low of 77.8% of informal caregiver diaries in facilities in Winnipeg to a high of 98.8% of formal services diaries in facilities in Victoria. The formal services diaries and the informal caregiver diaries were coded as to the type of task provided, who provided the task, and the amount of time spent on the task. Informal care time related to health care needs plus long-term-care needs. Tasks the informal caregiver performed as a result of a historical division of labor but for which the loved one was capable of performing were not included, such as, in some instances, cooking or laundry.
We coded types of tasks into seven major categories: clinical necessities, housekeeping, meal preparation, maintenance, psychological–social activities, respite, and companionship. We coded these tasks the same, regardless of whether an individual was in the community or in an institution. We categorized type of worker for the formal services into four groups: care aide; nurse; physician; and other health professional (occupational therapist, physiotherapist, massage therapist, and podiatrist or chiropodist). We recorded who performed the specific task by naming the relationship of the individual to the care receiver (i.e., spouse or daughter). We grouped expenditures into five categories: medical supply costs, activity costs, care-related service costs, equipment costs, and major expenditure costs, such as private home care, ambulance trips, transportation services, and dentures. We corroborated information in the diaries through questions in the client questionnaire.
We computed formal care costs by using actual care time from the diaries. We computed time spent on noncare tasks (such as dietary, housekeeping, and laundry) in facilities by using Hollander's (1994) detailed national estimates. These were derived from actual time and cost figures, including the province of BC, and revealed that, across facilities and over time, there was a consistent ratio of support costs to care costs—support costs are 50% of care costs. Therefore, we grossed up care dollars for facility clients by 50% to account for costs associated with support staff. We determined salary dollars per hour for each category of health care worker.
In the community, we made a noncare time adjustment to account for noncare time, including travel time, charting time, meeting time, staff training time, and benefits and administrative overhead because this information was not consistently recorded in the diaries. Workers focused on their own care time and, in some instances, may not have known some of these other costs. We determined this adjusted rate through consultation with those working in the system, and it varied from 40% to 60% of care costs. We then calculated the costs associated with each category of formal service worker by multiplying the amount of time (derived from the diaries) provided by that type of worker by the adjusted rate for that category of worker (specific costs available by request from the authors).
We calculated informal care costs in terms of both direct out-of-pocket expenses and total amount of informal caregiving time reported. Recreational activities, attendance at church, and attendance at adult day centers were not included. Adult day care was excluded because we were unable to determine usage. Facilities often provided in-house services that were similar to adult day care but did not refer to them as such, and the data for community-living seniors did not identify adult day care use unless a volunteer or paid worker drove the senior to it. If a family member, other than the main caregiver, drove them it would not have been recorded. As far as we ascertain, only 2.25% of respondents used adult day care. We estimated the value of informal caregiver time in three different ways: informal caregiving time valued at 0; informal caregiver time valued at minimum wage ($8.36/hr in Victoria and $6.65/hr in Winnipeg as specified by each of the provinces, including 11% benefits); and informal caregiver time valued at replacement wage, reflecting the wage rate of the appropriate category of worker(s) who would need to be hired in the absence of an informal caregiver.
To compare the cost of community care and the cost of long-term residential care, we included the following: (a) the adjusted cost of continuing care services for persons living at home or in a long-term-care facility, including home support workers, care aides, nursing staff, therapeutic staff, and support staff; (b) the cost of other health services, including physician services and acute-care hospital services; (c) purchased services for community clients and user fees for facility clients; (d) direct, informal costs, that is, out-of-pocket expenditures by clients or informal caregivers; and (e) the cost of assistance provided by informal caregivers, using the three approaches noted herein.
To ensure we were comparing types of care with comparable outcomes, analyses begin by exploring outcomes across the two sites of care (community and facility) and levels of care for two outcome measures related to client satisfaction with their quality of life. A modified version of the Terrible Delightful Scale by Andrews and Withey (1976) asked individuals to assess their satisfaction with 11 domains of life such as these: How would you rate your friendships? The kind of contact and frequency of contact you have with your friends (including personal contact, phone calls, and letters)? How would you rate your housing (the present type, atmosphere, and state of your home)? How would you rate your religion? Your spiritual fulfillment? Items are summed and then divided by the number to which the individual responded. Cronbach's alpha for the scale as used here was.76 in Victoria and.75 in Winnipeg. In addition, individuals were asked this single indicator item: "How would you describe your satisfaction with life in general at present?" Second, individuals were asked about their satisfaction with services, using questions addressing services received, worker characteristics, and care concerns as adapted from Penning and Chappell (1996). Cronbach's alpha for the scale was.70 in Victoria and.80 in Winnipeg. As subjective measures these questions could be asked only of individuals who were not cognitively impaired.
We used an ANOVA to assess cost comparisons for those living in the community and those living in the facility by level of care at each site separately. We chose the ANOVA over another multivariate statistical technique such as multiple regression analyses because it was clear that there are site differences, and our interest is focused on level of care, not other (such as demographic) characteristics. A regression analysis revealing site differences would therefore add little new knowledge. Instead, we considered the different sites to be replications of one another, and we utilized the ANOVA, incorporating type of care, level of care, and Type x Level interaction effects.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Table 4 shows the results of the two-factor ANOVAs (Type of care x Care levels) for the different cost items. The first section of Table 4 reveals that the costs of continuing care services for community clients was significantly lower than the costs of care for facility clients in both Victoria and Winnipeg. The costs of care increased as the care needs of the clients increased in the Victoria sample but not in the Winnipeg sample. (In the Winnipeg site, individuals who were "largely dependent" were excluded because of the small number of community clients at this level.) The second section of Table 4 reveals that the costs of physician and hospital services were higher for community clients than for facility clients in Winnipeg but not in Victoria. There was neither a main effect of care level nor an interaction effect between type of care and care levels in either of the sites.
|
As we noted earlier, we computed the cost of informal care time by using three different methods. The fourth section of Table 4 compares total costs when informal care time is valued at zero. In this instance, costs of care incorporate costs to government (public costs), out-of-pocket expenses for clients or caregivers, and client-purchased services–facility-user fees (the sum of columns 4, 5, and 8 in Table 3). The results of the ANOVA reveal that, for both sites, community care was less costly than facility care, and for the Victoria site, cost of care increased as levels of care increased. The fifth section of Table 4 shows the same analyses but with informal care cost determined at minimum wage (the total of columns 4, 5, 8, and 9 in Table 3). Once again, community care was significantly less costly than facility care and, in Victoria, cost of care increased as care levels increased. The results were the same when informal care was computed at replacement wages (the sixth section of Table 4).
It is to be noted that facility-user fees (often considered room and board or shelter costs) were included in the calculations here but comparable fees were not included for community-dwelling residents. This was based on the argument that the facility-user fee is a health-related cost that is paid because a client needs to be in a particular care setting as part of his or her care so that it is not comparable to non-care-related food and shelter costs. The individual in a facility does not have the option, for example, of living with a relative where they might pay nothing for food and shelter because his or her care needs require facility living. However, one can also argue that responsibility for food and shelter costs fall with the individual, irrespective of where he or she lives and should not be considered part of care costs. Although we disagree with this argument, analyses were conducted excluding the facility-user fee, for caregiver time determined with cost at zero, minimum wage, and replacement wage.
The analyses (not shown here) reveal that community care remains statistically less costly than facility care, when informal care is determined with cost at zero and at minimum wage. When it is determined with cost at replacement wage, community care still costs less but the difference does not reach statistical significance. In all cases, the cost of care increases in Victoria as levels of care increase. In all instances, the costs of care are relatively constant across care levels in Winnipeg. No interaction effects were revealed. Because government does not pay for informal caregiving, community care is a cost-effective substitute for government. From a societal perspective, it is caregiving time that assists in keeping clients in the community who may otherwise have to be institutionalized. This ensures that it is a cost-effective alternative, that is, home care is cost-effective in caring for people at home, and it is informal caregivers who often enable clients to stay at home.
Taken together, the comparative cost analyses presented here indicate that, in both Victoria and Winnipeg, costs were generally lower for community clients than for facility clients regardless of whether only costs to government were taken into account or both formal and informal costs were taken into account. When informal caregiver time was valued at either minimum wage or replacement wage, there was a substantial jump in the average annual costs for both community and facility clients relative to when informal caregiver time was valued at zero. Nevertheless, the results indicate that home care was significantly less costly than residential care.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Our findings also suggest that home care can be a cost-effective substitute for residential care. For example, even when caregiver time is determined with cost at replacement wage, home care costs $32,218 for Level C clients in Victoria and $64,715 for facility care; home care costs $27,518 for Level A clients in Winnipeg and $59,292 for facility care. Although only stable clients were included in this study and only two Canadian sites were examined, study findings clearly suggest the potential for an expansion of home-care services. This study has demonstrated that, across all levels of care, home care can be less costly, for stable clients, than residential long-term care. When only formal costs are considered, home-care costs are, on average, 50% or less of the costs of residential care. However, if governments are interested in cost effectiveness, actual substitutions of home care for residential care must be made to achieve this aim. If residential care is maintained at the current level or increased, then additional support for home care constitutes additional cost. If governments are serious about substituting care, cost-effective substitutions can be made by reducing beds or holding back the rate of expansion in facility beds and redirecting future funds to home care. Thus, one can see cost efficiencies provided home care is expanded to accommodate the increased numbers of individuals who would otherwise be placed in facility care. Using research results such as those reported in this study to cut facility beds, unaccompanied by an expansion in home-care services, is not enhancing the efficiency of care—instead, it is just depleting care.
It should also be noted that the cost comparisons presented herein are for fully developed and integrated care-delivery systems in which case managers stream clients into community and residential settings. In jurisdictions where home care and home-support services may not be fully developed, the question arises whether or not there is a "woodwork" effect in which people who previously may not have received paid home-care services come out of the woodwork and request care. This phenomenon, it is argued, could increase the overall cost of home care and be an unanticipated by-product of trying to introduce cost-effective home-care services into a less than fully developed care system. This raises the issue of comparative unit costs of home care and institutional care and aggregate costs of home care and community care. There are a number of perspectives on this matter. First, it may be that introducing additional home-care services may bring forward individuals with high-care needs whose health care costs in the community exceed what would be paid in facilities. It is a policy choice of a provincial or state government, regional health authority, managed care organization, or insurance company as to whether or not people should be allowed to remain in the community if the cost to the funder exceeds that of residential care. In our study, there were two different policy approaches to this matter. In Victoria, there were four individuals (4.8% of the overall sample) who exceeded 120 hr/month of care-aide time (the maximum amount of publicly funded care for high-needs chronic clients), whereas in Winnipeg there were 25 such individuals (37.9% of the overall sample). Thus, with regard to increased costs over and above the amount allowed for through public funds or insurance, it is up to funders to decide to what extent they will allow exceptions. That is, the funder can control excess expenditures by not allowing care in the community for people whose costs would exceed that of facility care.
Second, in an integrated system, funders may be able to obtain efficiencies by substituting community care for residential care and therefore limit, or negate, any woodwork effect. Again, funders have a policy choice in regard to maintaining splintered systems or adopting integrated systems and, in this way, they have some control over cost escalations. Finally, although it is not possible to directly estimate the magnitude of the woodwork effect in given jurisdictions, there are some guideposts as to how much home care would be provided in an integrated system of care where everyone has already come out of the woodwork. Policy makers could use these guideposts to estimate the potential impact of the woodwork effect. In terms of costs, Hollander (2001) noted that the public cost of long-term-care facilities in British Columbia in the 1992–1993 fiscal year was $434.3 million whereas the costs of home- and community-care services were $210.6 million, or less than half. Although Hollander does not provide specific data on chronic care (extended care), we can estimate the costs for chronic-care beds as being similar to the costs for long-term-care beds by using cost per bed and the number of beds. Thus, in terms of public or insurance-based expenditures in a fully developed and integrated care-delivery system, the aggregate ratio of facility (long-term care and extended care) costs compared with home- and community-care costs would be about 4 to 1; that is, home care would cost about 20% and facility care would cost about 80% of overall expenses. Naturally, the relative proportions paid by individuals (as user fees) and the state would affect these ratios and percentages.
Efforts to examine costs of home care and facility care necessitate consideration of jurisdictional differences, hence diverse policy approaches, as was the case across the two study sites selected for this investigation. Depending on the provincial or state funding approach, costs may not be linked to care level. For instance, in Winnipeg, costs were similar for all facility clients regardless of their care level. In that province, all individuals in Levels 3 and 4, the only levels included in the sample, are funded at the same level within facilities. Of relevance for the cost effectiveness of home care relative to institutional care, irrespective of the jurisdiction, is the ratio of cost categories between home care and facilities. The findings reported here will be applicable to other jurisdictions in Canada and the United States (and elsewhere), to the extent that the ratios are comparable.
It should also be noted that the Canadian jurisdictions examined here both utilize needs-based screening whereby eligibility is determined based on these assessments. Ability to buy does not facilitate receipt of these services. In addition, in Canada, residents admitted to facilities seldom leave before death; it is viewed as a last stop home. The South Carolina model in the 1980s and the Arizona model in the 1990s, which were found to be cost effective (Mathematica Policy Research, 1986), also included such needs-based screening. Improved screening was identified by Weissert (1991) as a suggestion for increasing the relative cost effectiveness of home-care services. South Carolina and Arizona also had well-integrated case-management approaches, another similarity with the Canadian systems studied here.
Aside from system and organizational differences between Canadian and American programs, the findings reported here confirming the potential cost effectiveness of home care may differ from earlier research in the United States because of study differences. For example, some of the earlier research included costs for an enhanced home-care program but reported somewhat reduced costs prior to such enhanced costs being taken into account (Berkeley Planning Associates, 1985; Mathematica Policy Research, 1986). The Canadian research reported here did not include any enhanced costs but rather compared costs within existing programs. In addition, some American research has included people with relatively low-care needs, that is, persons for whom there was a low probability that they would be admitted to long-term-care facilities. This is why we have placed importance in the current study on ensuring comparable levels of care in order to conduct the cost comparisons. Skellie and colleagues (1984), commenting on the Georgia Alternative Health Services Project, further report that cost effectiveness is difficult under a voluntary screening system when there is low demand for project services and low volume of clients, all resulting in higher administrative and direct-service costs. It is also true that, over the years, home-care programs (including but not exclusive to the two studied here) may well have learned to better target their clientele.
Conclusions
In sum, the study reported here suggests long-term chronic home care can be a cost-effective alternative to facility care. Although characteristics of the two Canadian jurisdictions included in this study suggest how this can be accomplished (single point of entry, needs-based screening, etc.), further research is required to better understand the circumstances under which cost effectiveness will result. Study findings also reveal the critical role of informal caregivers in the cost effectiveness of long-term home care. Additional research is necessary to help us understand how caregivers can best be supported in this role so as not to simply offload demands from the public purse to the private domain of families.
|
Footnotes |
|---|
1 Centre on Aging and Department of Sociology, University of Victoria, Canada.
2 Department of Community Health Sciences, University of Manitoba, Canada.
3 Hollander Analytical Services Ltd., Victoria, British Columbia, Canada.
4 School of Nursing, University of Western Ontario, Canada.
Decision Editor: Linda S. Noelker, PhD
Received for publication January 6, 2003. Accepted for publication July 18, 2003.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. E. Benjamin, R. E. Matthias, K. Kietzman, and W. Furman Retention of Paid Related Caregivers: Who Stays and Who Leaves Home Care Careers? Gerontologist, July 1, 2008; 48(suppl_1): 104 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Martikainen, E. Nihtila, and H. Moustgaard The Effects of Socioeconomic Status and Health on Transitions in Living Arrangements and Mortality: A Longitudinal Analysis of Elderly Finnish Men and Women From 1997 to 2002 J. Gerontol. B. Psychol. Sci. Soc. Sci., March 1, 2008; 63(2): S99 - S109. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Lewarski and P. C. Gay Current Issues in Home Mechanical Ventilation Chest, August 1, 2007; 132(2): 671 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Penning, M. E. Brackley, and D. E. Allan Home Care and Health Reform: Changes in Home Care Utilization in One Canadian Province, 1990-2000 Gerontologist, December 1, 2006; 46(6): 744 - 758. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
