This study’s findings provide some explanations for the inconsistent research findings in the literature regarding the association between older adults’ HCBS use and their future residential transitions. In the literature, HCBS have been examined either as a unified category or as individual services. Previous research studies have not examined how outcomes from one service might be supported by use of other services, or how using different combinations of services might result in different outcomes regarding residential transitions. Our findings show the possibility that older adults’ use of different HCBS, and different patterns of HCBS use, might interact with older adults’ residential transitions. These findings have some value for future research and policy making for older adults living in communities.
Nondiscretionary services and residential transitions
Among the nondiscretionary services, there were significant differences among the four groups in the use of skilled nursing care, physical therapy, and occupational therapy. Figure 2 shows that the older adults in the CIC group were the most likely to use these three services. It is likely that the older adults in this group were introduced to these services in an early stage of disability and therefore knew how to access the services again once institutionalization was no longer required. Knowledge of service accessibility may have supported these adults in returning to their communities . Our study findings provide some empirical support to Bradley and colleagues’ expanded Health Behavioral Model, in that previous experiences of using long-term care services could increase older adults’ knowledge and perceived control of accessing the services when they next need them . Such knowledge and attitude could empower older adults to return to communities from institutions.
On the other hand, we cannot ignore the possibility that older adults might have these services ordered for them upon leaving an institution. The literature suggests that older adults who go to nursing homes for short-term rehabilitation after an acute event may use more occupational therapy or physical therapy after returning to their communities . However, both scenarios point out that using nondiscretionary services, no matter whether before or after institutionalization, may help older adults to return to communities. Fewer older adults in the CII group used these nondiscretionary services between T1 and T2, compared to the CIC group, and these older adults tended to use institutional services again between T2 and T3. Research studies found that formal community services, when appropriately targeted to certain subgroups of older adults with functional limitations, appear to be significantly associated with reduced risk of nursing home use [10, 36]. The findings from this study provide further insights: (a) skilled nursing care, physical therapy, occupational therapy, and speech therapy may be helpful in helping older adults to return to communities from institutions, and thus may be more closely related to reducing nursing home use; and (b) there may be ideal times to provide services to older adults in order to achieve these results. For example, early introduction to occupational services may enhance older adults’ ability to return to community after institutionalization, as in the CIC group. This requires further investigation.
Discretionary services and residential transitions
The four groups of older adults used significantly different types of discretionary services. Different key services, and services that support these key services, were identified for each of the four groups. Older adults in the CCC, CIC, and CCI groups commonly used senior centers. Older adults in the CIC group commonly used both senior centers and PCS. Older adults in the CII group commonly used meal services (both Meals On Wheels and meals at senior centers) as well as other discretionary services, such as PCS. We further discuss these service use patterns below.
Although the use of senior centers was not significantly different among the four transition groups, senior centers were commonly used by older adults in the CCC, CIC, and CCI groups. It is possible that use of senior centers, rather than directly facilitating the ability to stay in community, is instead a characteristic of older adults who are more outgoing and more willing to connect with others. These personal characteristics in themselves may eventually enable older adults to continue to stay in their communities (like those in the CCC group), to community after being institutionalized (like those in the CIC group), or to stay in community longer (like those in the CCI group). Personal characteristics can influence the type of care received , while using senior centers can moderate the effects of stress on distress and has been found to positively associate with older adults’ physical and psychological well-being [38–40]. It may be a combination of these effects that provides older adults with functional limitations who use senior centers the strength to stay healthy and to either continue to live in their communities or live in community longer. The mechanism that underlies these adults’ consequent ability to remain living in communities merits further investigation.
Older adults’ use of senior centers in combination with different other services may associate with different future residential transitions as well. Older adults who used senior centers but not other services (like the older adults in the CCC group) tended to continue to stay in communities. It is possible that this group used senior centers for recreational purposes only, which is good for maintaining psychological health, and they did not need to use other services [38–40]. It could be that when these adults develop needs for and begin to use other services (like the older adults in the CCI group), future nursing home admission becomes possible. Older adults in the CCI group used all other types of discretionary service equally and only slightly less commonly than senior centers. This could indicate that senior center services used along with many other HCBS, as seen in the CCI group, helps older adults manage to stay in communities longer when they developed needs.
If older adults learn to access PCS (like the older adults in the CIC group) early on, then use of senior centers in combination with PCS may enable them to return to communities after being institutionalized (like the older adults in the CIC group). PCS stands out as the most commonly used service by the older adults in the CIC group. Studies have found that providing PCS to older adults reduced their use of nursing facilities and supported them to remain in communities [35, 41]. Findings in the current study provide a potential explanation regarding the mechanism by which using PCS supports older adults to remain in communities: that using PCS along with senior centers, skilled nursing care, physical therapy, occupational therapy, and speech therapy, like the CIC group did, might enable older adults to move back to communities from institutions.
In the CII group, PCS was also used by many older adults; however, the pattern of HCBS use in this group was quite different from that in the CIC and CCI groups. In the CII group, Meals On Wheels was the most common service used, followed by PCS and meals at senior centers/facilities. This pattern seemed to be associated with possible transition to an institution in the future. It could be that older adults in the CII group, whether due to physical limitations or lack of access to appropriate kitchen facilities, had difficulties preparing food for themselves, and that this characteristic is associated with future long-term or frequent use of institutional care services. Case managers or health care professionals who notice this pattern of HCBS use might consider recommending institutional care, instead of trying to keep these older adults in communities.
What kind of services to provide and when has always been of great interest to policy makers [42–44]. The current study findings not only shed some light on HCBS service provision and policy development, but also point out the importance of rethinking the relationships between different HCBS and older adults’ residential transitions. With the use of HCBS, older adults may be able to continue staying in communities, move back to their communities from institutions, or stay in their communities longer before moving into an institution. The U.S. government has been supporting older adults living in communities, and also supporting those who live in nursing homes but could live in communities to transfer back to communities (Mor, 2007). Our study findings could provide insight for helping both the older population and the U.S. government to achieve their goals.
Although the mechanism for how different service use or service use in combinations interact with older adults’ residential transitions will require further study, our current findings provide some information for policy makers and case managers as well as reason to pay attention to older adults’ patterns of HCBS use. Our study findings also provide a potential explanation for the inconsistent findings in the literature regarding the effects of HCBS use: It is possible that using HCBS in different combinations may result in different outcomes. Past studies had proposed that different HCBS may have different effects on nursing home use for different subgroups of older adults . Our study findings further pointed out that different combinations of HCBS use are related to different residential transitions. Further study to identify the characteristics of the older adults in these four different residential transition groups will be important for future policy making and practice.
Another contribution of the current study is the acknowledgment of older adults’ residential transitions from a longitudinal perspective while studying the effect of older adults’ HCBS use. A review of trends in the quantitative analysis of social science data on aging during the past half century shows that cross-sectional analysis remains the single most frequent type of study design , particularly in studies examining the effect of HCBS [17, 31, 32, 46–48]. This type of study design does not allow researchers to study the dynamics of older adults’ residential transitions and use of HCBS and may have contributed to the inconsistent research findings in research studies that have examined the effect of HCBS. Thus, longitudinal study design is recommended for future research in aging services. In addition, our study findings recommend that future research in HCBS use among older adults include at least three time points, to study the transitions. As HCBS have become increasingly available, older adults have been shown to transit through different residential statuses over time. Most previous research studies have failed to acknowledge this societal change. Understanding the interplay between older adults’ use of HCBS and their residential transition patterns could be key to developing an effective community-based long-term care system. Examining the associations between older adults’ residential transitions and HCBS use from a longitudinal perspective could provide further insight into the inconsistent findings in the literature and policy and practice implications [10–15][10, 16, 17].
Several study limitations warrant discussion. First, the results reported here are subject to the limitations of variable availability in the data set. Service utilization was one of the most important outcome measures in this study. Receipt of HCBS is related to service availability and depends on the distribution system, not on the mobility of the user. A service that is not available cannot be used . However, the variable of service availability was not recorded in the national data set used in this study and therefore could not be studied. Other service variables not included in the data set also could not be studied. For example, adult day care service, which was a common HCBS available in communities, was not surveyed at the T2 interview. Therefore, we were not able to study how this service might influence older adults’ residential transitions. We would also like to note that the four residential transition patterns included as variables in the current study do not capture all of the possible transitions between observation points, and thus in some sense these patterns are artificially determined by the design of the LSOA II.
We also questioned whether the older adults in the CIC group were those who were hospitalized for an acute condition and sent to a nursing home for a few weeks of recovery, with high probability of return to the community. We ran some analyses to compare the number of days in institutions and the frequency of institutional service use for the older adults in the CIC and CII groups. The CIC and CII groups had a similar ratio of total institutional days per institution admission during the period from T1 and T2 (1.08 ± 0.66 days per admission to an institution vs. 1.44 ± 1.78 days per admission), but very different ratios of days in institutions per admission during the period from T2 and T3 (0 days per admission to an institution vs. 16.66 ± 13.34 days per admission). It is intriguing to consider whether these two groups’ use of different HCBS (or different combinations of HCBS) between T1 and T2 could be associated with their differential use of institutional services at a later time. However, many factors may play a role in whether a group of older adults changes from short-term users to nonusers of institutional services (CIC), or from short-term to frequent/long-term users of institutional services. Studies to further investigate these factors are recommended.
We noted that individuals in the study sample tended to have more functional limitations and ADL and IADL disabilities than the overall LSOA II population. The functional status of the older adults in the study sample was similar to, but a little less disabled than, that of those LSOA II respondents who were not able to participate in the T2 and T3 surveys: our study sample had more Nagi’s functional limitations, which are considered light disabilities, and less ADL and IADL disabilities, which are considered more severe disabilities. Therefore, the current study’s findings may be generalizable only to older adults with a light to moderate level of functional disabilities. Further investigation is merited for service use patterns in severely disabled older populations as well as service use patterns among respondents who died or dropped out prior to completion of the LSOA II study, because they may have very different service use patterns than did the older adults included in this study.