| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
Correspondence: Address correspondence to Margaret J. Penning, PhD, Department of Sociology and Centre on Aging, University of Victoria, PO Box 1700, STN CSC Victoria, British Columbia, Canada V8W 2Y2. E-mail: mpenning{at}uvic.ca
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Health reform • Health service utilization • Home care
Since the 1950s, Canada's health care system (Medicare) has been characterized by universal and publicly funded coverage. Although health care is primarily a provincial government responsibility, the provinces must comply with five major conditions (universality, accessibility, comprehensiveness, portability, and public administration) in order to qualify for federal funding. Consequently, as one author notes, "there is no Canadian system; instead there are a set of publicly financed, provincially run insurance plans covering all legal residents for specified service categories" (Deber, 2003, p. 20). Canada's Health Act guarantees citizens universal access to medically necessary services, defined as physician and hospital care. To date, similar coverage has not been extended to home- and community-based care. Instead, home care falls completely under provincial jurisdiction. In most Canadian provinces, the services include home support (e.g., assistance with activities of daily living, meal programs), home nursing care, and rehabilitation services (physical and occupational therapy), and they are provided by a mix of public, nonprofit, and for-profit organizations. Within home care, home nursing care and rehabilitative services are often deemed medically necessary and consequently are publicly provided and fully funded, whereas home support services are likely to be contracted out, provided by a mix of for-profit and nonprofit organizations, and have user fees and stringent eligibility criteria applied. Consequently, differences in the financing and delivery of medical (physician and hospital care) versus community-based (home) care during the 1990s suggest that a shift away from a medically focused system of care also entails a move away from a system that is comparatively well resourced, offers full public coverage, and has universal accessibility toward a system of care that has few resources, provides limited public coverage, and is not governed by requirements for universal access.
Concern has been expressed regarding the implications of regionalization and associated health reforms for home care and other community-based health services as well as those requiring such care (e.g., Williams, Deber, Baranek, & Gildiner, 2001). According to Church and Barker (1998), for example, the primary focus of current restructuring efforts is in fact cost reduction. From their perspective, the shift in emphasis from institutional to community-based settings for service delivery reflects both a desire to contain costs in the institutional sector and an increasing recognition that institutionally delivered services are not as cost effective as community-based services. Others note a coincident trend toward medicalization of community-based care, reflecting the insertion of a medical model of care into the community sector or governmental interests in shifting the costs of medical and particularly hospital care to the community. According to McDaniel and Chappell (1999), for example, earlier hospital discharges and the increasing use of outpatient surgeries have generated increased demand for postacute home care, which is threatening resources allocated to nonmedical supportive services. A related though somewhat different perspective links shifts in the locus of care from hospitals to community settings to broader political forces, the primary objectives of which are to support profitization, which is "the provision of health services as a profit-making business" (Williams et al., p. 11). From this view, cost reduction is not the primary objective but rather a justification for state retrenchment; the real objective is to restructure health services in ways that undermine publicly provided care in favor of the private market. According to Williams and colleagues, reforms of this type not only change the locus of care but also "fundamentally change the conditions under which care is provided" (p. 10).
The extent to which the policy and structural changes introduced in the 1990s shifted foci consistent with a community- and population-health-focused model of care remains unclear. On the one hand, federal and provincial governments report reductions in hospital budgets, hospital beds, and time spent in the hospital, together with increases in outpatient surgeries (Merlis, 2000). For example, in British Columbia, from 1991–1992 to 1996–1997, short-stay hospital beds declined by 30%, acute-care beds per 1,000 population declined by 28.8%, and the average length of stay declined by 12.9%; among older adults specifically, acute hospital days per 1,000 population declined by 27.3% whereas extended hospital days declined by 14.4% (Sheps et al., 2000). In addition, resources allocated to home care services appear to have increased (Health Canada, 1998; Hollander & Chappell, 2002). On the other hand, reductions in hospital and institutional care, together with increases in resources to home-based care, do not, on their own, indicate a shift from a medically focused and institutionally based model of care to one based on broader notions of health, incorporating prevention and community-based care. For example, it may be that increased resources are targeted to salaries, administrative costs, or to providing post-hospital medical care in the home setting.
It should be noted that it is often difficult to determine where increased expenditures were directed. For example, with regard to the province of British Columbia, the Canadian Institute of Health Information (CIHI; see Ballinger, Zhang, & Hicks, 2001) noted the following:
[O]nce the province regionalized, financial reports no longer reported detail greater than provincial funding for overall Continuing Care. Continuing Care in British Columbia includes home care, public health and residential care. ... CIHI has been informed that regional expenditures on continuing care aggregated to a provincial total cannot be broken down into sufficient detail to identify either the components of home care nor total home care expenditure. (p. 24)
Cost-cutting, medicalization, and profitization lead to somewhat different expectations regarding trends in home care and other health services utilization. To the extent that health reform is indeed effecting transformation toward a community-based population health model of care, one would expect to see enhanced resources going to support such services, reflected in increased access to and utilization of community-based services. In addition, insofar as both socially and medically defined health needs are recognized, one would expect to see increased access to and utilization of services supporting continued community living as well as to those delaying requirements for hospitalization and institutionalization. Trends toward cost reduction would seem to suggest the opposite—that is, fewer resources going to support community-based care and, consequently, decreases in access to such care by the population. To the extent that cost-cutting is applied across the board, one would expect to see similar trends evident across services, whether medical and institutional or social and community based. Alternatively, if community care is being used as a vehicle for cost-cutting in comparatively more expensive acute medical and hospital care sectors, one would expect to see a redirection of such services toward post-hospital nonresidential care to fill the gap created by cost-reduction strategies (delayed admissions, reductions in lengths of stay) in the hospital and institutional care sectors. This includes redirection with regard to who receives services, what services are available, and the extent of services available. For example, one might expect to see increased resources going to those individuals requiring medical and related services (e.g., nursing care, rehabilitation) rather than those requiring more preventive and supportive types of care (e.g., housekeeping, help with meals). Finally, if the objective is to transfer responsibility for health services from a universally accessible publicly funded system (which, in the Canadian context, is centered on providing medically focused, acute, and hospital services) toward a nonobligatory (i.e., no universal entitlement), privately funded, and market-based system, once again, one might expect to see substitution of hospital care by community-based care. However, in this case, the expectation might also be for less governmentally supported and more market-driven services within the community-care sector (i.e., home support services) to be preferred over fully publicly funded and supported services (i.e., home nursing care, rehabilitation services).
To address whether health services restructuring is consistent with trends toward enhanced access to community-based care or, alternatively, with trends toward cost-cutting, medicalization, or profitization, in this study we examine the extent to which a shift toward community-based care is taking place in one Canadian province, British Columbia, where health reform, originally predicated on a community-based population-health model, has been under way since the early 1990s. Here, structural change was based on recommendations of the British Columbia Royal Commission on Health Care and Costs (1991) in its report titled Closer to Home and introduced into policy in 1993. The commission endorsed the basic principles of Medicare but also noted a need for restructuring; it recommended changing the traditional focus of the health care system from curing illness to preventing illness and on protecting health, and that "services ... be provided in, or as near to, the patient's place of residence as is consistent with quality and cost-effective health care" (p. A6). The report described home care as "one of the most critical elements in our health care system" (p. C154), suggested that the government commit a realistic portion of its funding to care in the community, and recommended that, over time, funds be transferred from the acute hospital sector to home care and other forms of less intensive institutional and residential care. Subsequent policy documents emphasized the improving of health services rather than population health per se (see McPhail, 1996). However, restructuring through regionalization was formally introduced in April 1997, when authority to manage and fund community-based health services was transferred from the province to 18 geographically defined Regional Health Authorities.
We address three research questions. First, has access to community-based home care services changed over time in a manner consistent with the timing and intent of health reform? Specifically, have the likelihood of receiving home care services and the extent of care provided to the population changed? Second, has the intensity of the care provided to those individuals receiving care changed over time? Third, and finally, has there been a change in trends associated with joint home care–hospital use?
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Study Population
We drew data from 10% annual independent random cross-sectional samples of health service users (derived from a registry of users of all publicly funded health services) who resided within the province each year from 1990–2000. Because medical and hospital services in Canada are publicly insured benefits and almost all residents draw on health services of one form or another at some point, this registry represents, as closely as possible, the entire population of the province: 3.29 million people in 1990, and just over 4.0 million people in 2000 (British Columbia Ministry of Management Services, 2004). Individuals aged 65 and older constituted 12.6% of the total population in 1990; this increased to 13.1% in 2000.
Using these data, we isolated claims for services received through the publicly funded home care program during each year and subsequently, identified individuals (unique claims) receiving these services. Here we focus on two components of the provincial home care program, home support services and home nursing care. We do not include rehabilitation services because of the relatively small number of people who receive such services. Although individuals younger than 65 years of age represent a minority of those receiving home care services, we include them in the sample in order to facilitate a comparison of trends over time across age groups.
Within our sample, the overall number of claims for home support service stood at 34,793 in 1990 and 39,911 in 2000. The number of claims by any one individual ranged from a low of 1 to a high of 66 in a given year (M = 9.36–9.99 claims/individual from 1991 to 2000). Thus, the total number of individuals receiving home support services ranged from 4,578 in 1990 to 3,995 in 2000 (representing 1.39% and 0.98% of the British Columbia population, respectively). It should be noted that data on home support claims for 3 months in 1990 only (January–March; i.e., the period prior to the first full fiscal year) were missing from the dataset. This caused in an estimated loss of 25% of all claims and 6.4% of unique claims for home support service in 1990, which one should keep in mind when reviewing the results.
With regard to home nursing care, the overall number of claims stood at 3,115 in 1990 and 4,068 in 2000, whereas the number of claims by any single individual ranged from 1 to 8 per year across the years studied (M = 1.13–1.18). Thus, the number of unique claims ranged from 2,726 in 1990 to 3,614 in 2000 (0.83% and 0.89% of the British Columbia population, respectively; see Table 1).
|
Measures
These analyses address population-based changes in publicly subsidized home support and home nursing care service utilization (including receipt of service, extent of care received, intensity of care received, and pattern of joint home–hospital care) over time.
Initially, we determined the number of discrete individuals receiving each of these services in a given calendar year. Subsequent analyses draw on actual counts as well as on age–gender standardized utilization rates. We calculated the latter by using the Manitoba SAS rates macro (Burchill et al., 2000) and by using British Columbia population datasets from 1990 to 2000 (British Columbia Ministry of Finance and Corporate Relations, 1999), with the 1991 provincial population as a standard.
To measure the extent of care received, we used the total number of hours of home support services and the number of home nursing care visits received by all clients sampled during a given year. Reliance on different measures to assess home support and home nursing care reflect the limitations on the availability of suitable measures within the dataset. Once again, our analyses draw on actual counts as well as on age–gender standardized utilization rates (total hours or visits per 1,000 population).
We created measures of the average intensity of home support and home nursing care services by dividing the extent of care received (i.e., total number of hours or visits received in a given year by all clients) by the total number of unique clients receiving each service.
Finally, by using calendar year home care files (July through December only) and full fiscal year hospitalization files (April of the corresponding year through March of each subsequent year), we identified three patterns of joint home–hospital care: the receipt of home support (or home nursing care) for the first time during a given year (a) 1 or more months prior to the first hospitalization; (b) 0–2 months following the first hospitalization; or (c) 3 or more months following the first hospitalization during a given year. By separating individuals who began receiving home care services 0–2 months from those who began 3 or more months after hospitalization, we attempt to isolate home care that is initiated directly in conjunction with a hospital visit. Again, analyses concerning joint use of home care and hospitalization use actual counts as well as age–gender standardized utilization rates.
Variables in the analyses include gender, age group (<40, 40–49, 50–54, 55–59, 60–64, 65–69, 70–74, 75–79, 80–84, and 85+ years of age), and health region of residence. We determined health region by using local health area codes derived from residence as of June 30 of each year of the study period. For the purpose of analysis, we subsequently combined several of the regions (i.e., their populations and number of service users were relatively small), resulting in a variable with 14 categories.
Analysis
To assess temporal changes in service utilization at the population level, we analyzed trends in age–gender adjusted rates of claims, visits or hours of care, and joint home and hospital care. We also examined temporal trends in unadjusted average intensity of care. We utilized joinpoint regression, a nonlinear regression procedure that fits a model based on the minimum number of joinpoints (points of change) evident within a series of rates over time (Kim, Fay, Fever, & Midthune, 2000; National Cancer Institute, 2005). Joinpoint (JP) analyses assumed a log-linear model and assessed whether trends in utilization rates over time could best be described by use of a straight line (JP = 0) or whether one (JP = 1) or two (JP = 2) points of change in trend signified the best-fit model.
We then conducted multivariate analyses by using generalized linear modeling (SAS PROC GENMOD). We aggregated data into clusters defined by age group, gender, region of residence, and year of measurement. Given 14 health regions and 11 years of data for each age–gender grouping, we defined 3,080 clusters (14 x 11 x 10 x 2) for analyses conducted with the sample as a whole. Subgroup analyses (by age group, categories of joint home–hospital care) involved somewhat fewer clusters. Chi-square tests contrast counts (i.e., unique claims; total hours or visits; hours or visits per individual receiving service) offset by population count for each calendar year to those in the following year as well as to 2000, the final year for which data were available. We applied Bonferroni adjustment to analyses involving multiple contrasts. We report only GENMOD results in this text; details are available upon request.
Minor differences in the results obtained by using JP and generalized linear modeling procedures can be expected because of the differences in control variables and in the number of data points included in the analyses (i.e., a single rate per year for the regression analyses vs 3,080 counts in the generalized linear models).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
Figure 1 also portrays changes over time within selected age groups (also see Table 2). Claims are lowest among those individuals under the age of 65, and they increase steadily with age. Several trends are evident. There is a significant increase in claims from 1990 to 1992 (slope = 0.186, p =.025), followed by a gradual decline (slope = –0.068, p =.000) among individuals under the age of 65. There is a stationary trend during the first few years of the decade, which is followed, once again, by a period of decline during the middle to later years of the decade among individuals aged 65–74 (slope = –0.137, p =.000) and 75–84 (slope = –0.110, p =.000). Finally, there is a significant increase in claims from 1990 to 1992 (slope = 0.063, p =.013), followed by a period of decline (slope = –0.034, p =.014), which accelerates from 1998 on (slope = –0.095, p =.021) among those individuals aged 85 years or older.
GENMOD results indicate that, for all age groups, claims rates are significantly lower in 2000 than for almost all years prior to 1999.
Figure 2 reports trends in home support hours of care (also see Table 2). As was the case with claims for service, hours of care were lowest among those under 65 years of age and increased steadily with age. They increased during the early years of the decade in almost all groups (e.g., slope = 0.230, p =.004 for total sample); no change was evident among individuals aged 65–74. For the sample as a whole and for age groups <65, 65–64, and 75–84, rates then declined (p =.000 to p =.007). As a result, significant changes in trend were evident for the sample as a whole, as well as for each of these three age groups. Among those individuals in the 85+ age group, in contrast, an increasing trend during the first few years of the decade (slope = 0.234, p =.033) was followed by two significant changes in trend (in 1992 and 1998, toward stationary and then declining hours of care).
GENMOD analyses also show a significant impact of year and an increase in hours of care from 1990 to 1991 for the sample as a whole. However, no subsequent year-to-year changes emerge. As a result, hours of care used in 2000 are significantly greater than those in 1990 but are comparable with most other years of the decade. These analyses also point to an increase in hours of care early in the decade in all age groups, followed by little subsequent year-to-year change. As a result, for most age groups (<65, 75–84, 85+), utilization is significantly greater in 2000 than in 1990 but does not differ from 1991 through 1998 (for those aged 85+) to 2000 (for those <65, 75–84). For individuals in the 65–74 age group, however, utilization does not differ when we compare 2000 to 1990; however, in this case, hours of care received in 2000 are significantly lower than several intervening years of the decade.
Home Nursing Claims and Extent of Care
With regard to home nursing care, in 1990 the age–gender standardized rate of unique claims stood at 8.05 per 1000 population for the province as a whole. It reached 9.04 in 1993 and 8.20 in 2000. JP regression results indicate a flat slope and no change in trend from 1990 through 2000 for the sample as a whole and for all four age groups (Figure 1 and Table 2).
GENMOD results do indicate a significant main effect of year, but only for those individuals aged 75–84 years and for the sample as a whole. For both groups, annual contrasts suggest fluctuations in service use from 1990 to 1993, but no subsequent year-to-year change. In addition, parameter estimates emerge as significantly lower in 2000 than in some (i.e., 1991, 1993, and 1998), but not all, of the preceding years.
In 1990, rates for home nursing visits for the population of British Columbia stood at 168.29 per 1,000 population; in 2000, the rate was 157.59 visits per 1,000 population. JP regression results indicate no significant slopes or changes in trend within the sample as a whole or in the <65 and 65–74 subgroups (see Figure 2 and Table 2). Among the 75–84 and 85+ subgroups, in contrast, although slopes are once again not significantly different from zero, borderline significant and downward changes in trend occur in 1998.
GENMOD analyses indicate no significant impact of year on home nursing care visits for the sample as a whole or for any of the age groups.
Intensity of Home Support and Home Nursing Care
Figure 3 and Table 3 assess hours of home support or visits of home nursing care received per individual per year.
Home support hours per claim stood at 95.59 in 1990 and 178.08 in 2000. Correspondingly, regression findings point to a significant impact of year for the sample as a whole, with the average number of hours of service per claim increasing over the decade (slope = 0.055, p =.000). In general, hours of home support received per claim were greatest among those individuals under the age of 65, followed by those aged 85 and older. A trend toward increasing intensity of care over time is also evident within each of the four age groups. However, the extent and significance of the change involved varies. Among individuals in the youngest age group, intensity of care increased significantly during the first few years of the decade (slope = 0.172, p =.039); it continued to increase in subsequent years but at a somewhat slower pace (slope = 0.024, p =.016). Among individuals in the 65–74, 75–84, and 85+ age groups, in contrast, intensity of care appeared to increase at a steady pace throughout the decade (slope = 0.063, p =.000 for 65–74; slope = 0.058, p =.000 for 75–84; slope = 0.040, p =.000 for 85+ ); none show a significant change in trend over time.
GENMOD results support these findings, indicating a significant and positive impact of year. Consequently, total hours of home support provided per individual were greater in 2000 than in most prior years of the decade. These results also indicate a significant impact of year on hours per claim in all age groups. In the youngest and oldest groups, intensity of care is greater at the end than the beginning of the decade but does not differ from the intervening years, whereas in the two intermediate age groups, intensity of care is greater at the end of the decade than during most of the preceding years.
Figure 3 and Table 3 also report trends for intensity of home nursing care. The number of visits received per individual claimant within the sample as a whole stood at 21.10 in 1990 and 19.23 in 2000. JP regression analyses reveal no significant slopes and no changes in trend with regard to visits per claim over the course of the decade for the sample as a whole. In general, the number of home nursing care visits per individual was lowest among those individuals under the age of 65 and greatest among those aged 85 and older. JP analyses conducted within specific age groups reveal that, for the <65 group, declining intensity of care characterizes the decade (slope = –0.014; p =.039). For the 65+ age groups, no significant trend or change in trend is evident over the decade.
With GENMOD, for the sample as a whole, the findings indicate that the average number of home nursing care visits per individual was lower in 2000 than in the years 1990 through 1995. Similar results are evident among the <65, 75–84, and 85+ age groups, where intensity of care is significantly lower in 2000 compared with almost all years from 1990 through 1995. No significant trend is evident among the 65–74 age group.
Joint Home and Hospital Care
Our final analyses assess whether changes have taken place linking provision of home support and home nursing services to hospital care (see Figure 4 and Table 3). In both analyses, we included only those individuals who made use of both services.
In 1990, rates of joint home support–hospital care stood at 1.29 per 1,000 population. The highest rate (0.71/1,000 population) occurred among those who received home support services 0–2 months following hospitalization; the lowest rate (0.18/1,000 population) was for those receiving home support prior to hospitalization. In 2000, the overall rate of joint home support–hospital care was lower (0.77/1,000 population) but, once again, the subgroup with the highest rates was that in which use was initiated within 2 months following hospital discharge (0.43/1,000 population). The lowest rate was for receipt of home support prior to hospital care (0.14/1,000 population). JP regression analyses indicate that overall joint use declined significantly over time (slope = –0.053, p =.000), with no discontinuity of trend over the course of the decade. The same pattern was evident within each of the three subgroups.
GENMOD results also point to a significant impact of year on joint home support–hospital utilization overall, as well as within each of the three subgroups: Although little year-to-year change was evident, rates were significantly lower in 2000 than in most of the early years of the decade.
Joint home nursing–hospital care rates were higher than those for home support–hospital care. In 1990, the overall rate was 3.54 per 1,000 population. Among subgroups, the highest rate (2.44/1,000 population) was again observed among those receiving home nursing services 0–2 months following hospitalization and the lowest rates (0.32/1,000 population) were for those receiving home nursing care prior to hospitalization. In 2000, overall rates were slightly lower (3.15/1,000 population) than in 1990. However, once again, the subgroup with the highest utilization rates was that in which home nursing care was initiated within 2 months following hospital discharge (2.15/1,000 population), and the lowest rates were for those receiving home nursing care prior to hospital care (0.33/1,000 population). For the most part, analyses of trends in rates of home nursing–hospital care indicate no significant change over time. However, for those individuals using home nursing care 0–2 months after hospitalization, GENMOD results indicate a significant impact of year, with rates in 2000 lower than those in the early years of the decade (1990–1993).
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Overall, these results point to reduction rather than expansion of access to community-based care over the course of the 1990s. In this way, they appear consistent with a cost-reduction hypothesis. However, to what extent do they also support the view that community-based care is being used as a vehicle for shifting the costs of medical, and particularly hospital, care to the community, as suggested by some proponents of a medicalization thesis? On the one hand, support for this view would seem limited. If home support and home nursing care were indeed being used to substitute for more expensive hospital or other medical services, one would expect increases, rather than either decreases or relative stability, in access to such care over time. This is particularly true with regard to home nursing care, a service that could directly substitute for medical and hospital care. In addition, one would expect to see services increasingly targeted to those designated as requiring more intensive services. Our findings did show that intensity of home support services increased over time (i.e., more hours of care provided to fewer clients), but the results are weaker and, where evident, in the opposite direction for home nursing care. Finally, although one might also expect to see an increasing link between home and hospital care such that trends toward post-hospital care would increase, such a trend was not found. Instead, joint home and hospital care utilization rates have either declined (home support–hospital care) or remained stable (home nursing–hospital care) in recent years, along with other home care utilization rates.
Although these results do not appear consistent with a view of community-based care as a vehicle for shifting the costs of medical and particularly hospital care to the community, they do suggest fairly consistent declines over time in access by the population to nonmedical home support services; this is less apparent with regard to more medically focused home nursing services. This suggests the possibility that service reduction (cost-cutting) is being directed primarily toward social components of the home care program, while attempts are made to preserve more medical elements of community-based care. Notably, home support services represent the least costly components of such care, with clinical services and hospital care shown to be major cost drivers within home care (see Chappell, Havens, Hollander, Miller, & McWilliam, 2004).
Finally, what support is there for the view that a gradual refocusing of health care away from publicly insured services toward the private market is taking place? Generalized reductions in access to publicly funded health care services would seem to support this trend. Those individuals unable to obtain access to publicly funded services will have little choice but to utilize private services or do without care. However, if one assumes that this should also imply a move, within the public system, toward an increased use of home support services (insofar as they are the least publicly supported and therefore most closely connected to profitization of the services studied), once again, our findings do not appear to provide much support. Access to home support services revealed consistent declines over much of the decade.
Overall, therefore, our findings would seem to point to reduced access to health services in general and home support services in particular, rather than a shift toward a community-based model of health care. Some redistribution of services also appears evident. Specifically, home support services were increasingly concentrated among fewer service users while home nursing care became somewhat less concentrated over time. As well, in a few cases, differential patterns were observed with respect to specific age groups, suggesting possible redistribution along age lines. With regard to home support claims, for example, those in younger age groups experienced steadily declining rates from 1992 or 1993 through to the end of the decade, whereas those aged 85 and older experienced two changes in trend—a somewhat slower decline that began in 1993 and then accelerated in 1998. Somewhat similar patterns were found with regard to home support hours and home nursing care visits, suggesting that service reductions may have directed to the oldest age group somewhat later in the decade than was the case with regard to younger age groups.
To what extent can we attribute these findings to specific reform initiatives, including the regional restructuring, introduced during the 1990s? The answer is unclear. In a few instances, changes in trend were evident, yet these changes appear at odds with the objectives of health reform with regard to enhanced access to community-based care. Furthermore, for the most part, our findings provide little evidence of sudden change in service-utilization trends that corresponds with actual restructuring efforts. Instead, they suggest an ongoing process of change. Observed reductions in home care not only coincided with but also, in many cases, preceded formal restructuring within the province, suggesting that restructuring may have played a fairly limited role in directing the types of changes taking place within this service sector. Consequently, it could be that changes were initiated in response to policy documents and in anticipation of, rather than in response to, actual restructuring. In 1993, the initial legislation required to permit restructuring was introduced; restructuring itself followed in 1997 and again in 2001. In addition, British Columbia was not the first or only province in Canada to undertake such reform; several other provinces began the process of reform before this province did (Mhatre & Deber, 1992).
This study has several limitations. For example, as noted, it cannot conclusively establish that changes in utilization during the 1990s are directly attributable to the specific health reform initiatives that were underway during that time. The lack of data on home support claims for the months of January through March, 1990 is also problematic. Nevertheless, the major finding with regard to home support services over the decade was not the increased utilization evident during the first year or two of the decade, but rather the steadily decreasing utilization observed during subsequent years. In addition, relatively few years of postrestructuring data were assessed, yet changes associated with restructuring might be observed only over a longer period of time. Finally, generalizing from these findings to other jurisdictions also remains tentative. Insofar as community-based reforms are intended to ensure services better suited to meeting local needs, implications for access to services at different points in time in different areas may vary.
Despite these limitations and the consequent need for further research to address changes in community-based care occurring in conjunction with health care restructuring and reform initiatives, our findings nevertheless have potentially important implications for national and international policy discussions in areas of home- and community-based care and health reform. In particular, they suggest that reforms that rely on the restructuring of health care delivery on a regional basis in order to try to enhance access to community-based care may be unlikely to achieve this goal.
Although clearly relevant to the Canadian context, findings of this nature are of potential importance to other countries as well. Models of health care reform adopted in the Canadian context are similar to those pursued in other countries. Since the late 1980s, the need to strengthen long-term-care services has been widely acknowledged, with OECD (Organisation for Economic Co-operation and Development ) countries (e.g., the United Kingdom and the United States) more or less uniformly promoting policies of deinstitutionalization and enhanced community-based care (Merlis, 2000). At the same time, reforms designed to shift from national to local government decision making (Muramatsu & Campbell, 2002; Wiener & Cuellar, 1999) and take advantage of assumed market efficiencies have been adopted within these countries (Twaddle, 2003). The effects of specific policy reforms are likely to vary; a crucial factor accounting for these differences lies in the character of the initial health care system to which the policy is applied (Cabiedes & Guillén, 2001). Canada provides a setting in which reforms are being pursued within the context of a health care system that offers universal public coverage of medical (physician and hospital) care but does not make comparable provisions for community-based care. This is similar to the systems of care currently in place in numerous other countries and populations, including the older adult population in the United States (Feder, Komisar, & Niefeld, 2000). As Feder and colleagues note,
elderly Americans are just about the only group of U.S. citizens whose health care is universally insured as an entitlement. However, elders who need long-term care have much less protection. Medicare ... covers many of the costs of acute medical care but only tangentially covers some long-term-care services. Medicaid ... covers long-term care but only for people who are poor or become poor paying for long-term care or medical care. (p. 40)
Overall, our findings support concerns regarding the implications of contemporary health reform initiatives for community-based care, particularly home support services. Reductions and reallocation of such services have important consequences for those whose health care needs are most dependent on such care, including older adults. From one perspective, safeguarding and targeting services to those requiring more intensive, medical, and nursing care might be considered a more efficient and equitable system, with care increasingly delivered to those most in need (see, e.g., Brownell, Roos, & Roos, 2001). However, this conclusion appears premised on a definition of need associated with acute rather than longer-term chronic conditions. Defining needs for care primarily or exclusively on the basis of medical criteria for acute conditions will therefore disadvantage those whose needs are broader in scope and reflect social criteria associated with a community-based population health model.
|
Footnotes |
|---|
1 Department of Sociology and Centre on Aging, University of Victoria, British Columbia, Canada.
2 Department of Anthropology and Centre on Aging, University of Victoria, British Columbia, Canada.
3 Centre on Aging, University of Victoria, British Columbia, Canada.
Decision Editor: Linda S. Noelker, PhD
Received for publication February 22, 2005. Accepted for publication May 17, 2006.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
