The perceived unmet need for home care and impact on frailty related health outcomes among community-dwelling middle-aged and older adults in Canada

Abstract

With the aging population, there is an increased need for the development of interventions and prevention programs to mitigate the impact of frailty on the health and well-being of older Canadians. The association between frailty and adverse health outcomes has been well established, however little is known about the impact of contextual factors that may mitigate or moderate this relationship. Home care services play an important part in providing necessary assistance to seniors within their homes and potentially preventing functional decline. Improved availability and/or accessibility to health care within the community are important target areas for potential prevention and policy reforms. To better understand the significance of unmet home care needs, and potential contributing factors, among Canadians with varying levels of frailty, this thesis sought to explore the role of unmet home care need as a moderating factor of the association between frailty and key health outcomes (i.e., inpatient acute care hospitalization and mortality) among community-dwelling adults aged 45 years and older. This research utilized data from the 2008/09 Canadian Community Health Survey-Healthy Aging (CCHS-HA) cycle 4.2, linked to administrative health databases, specifically the Discharge Abstract Database (DAD) 2007 through 2011 and Canadian Mortality Database (CMDB) December 2008 through 2011. The study population consisted of all participants aged 45+ years residing in all provinces (excluding Quebec). A previously validated frailty index (FI) was derived from survey items and utilized to determine those who were robust (≤0.1), pre-frail (>0.1 to ≤0.21) and frail (>0.21), within the sample. Descriptive analyses were conducted to determine the prevalence of frailty (robust 52.5%, pre-frail 33.5%, frail 14%) and unmet home care need (2.4%) overall, and by key respondent characteristics. Multivariable logistic regression models adjusted for relevant covariates were utilized to examine the associations between frailty and first-event inpatient hospitalization and mortality over a 2-year follow-up period, as well as possible effect modification of these associations by the presence/absence of unmet home care need. This study builds on previous frailty research, which largely focused on more impaired populations, by utilizing population-based data to derive a FI to assess prevalence and outcome associations. The exploration of unmet home care need as a modifier of frailty-outcome associations also builds on research evaluating the role of contextual factors in frailty trajectories. Bivariate results were largely consistent with past frailty research. Respondents who were frail (FI >0.21) were significantly more likely to be older, female, widowed, and to report low household income, low education and low social support availability (SSA). The same characteristics were associated with higher prevalence of unmet home care need. Previous research has suggested greater unmet health care need among younger (<65 years) cohorts. Studies investigating unmet home care need have found higher prevalence in older age groups, while the absolute number of those reporting unmet home care need is highest among 35-49 year olds. The prevalence of unmet home care need was higher in frail (10.5%) compared to robust (<0.7%) respondents, Those reporting unmet home care need were also more likely to report receiving either formal (21.3%) or informal (65.9%) home care support compared to respondents without such a need (4.9% and 10.7%, respectively). Multivariable analysis showed significant associations between frailty and both hospitalization (OR=3.18, 95% CI 2.35-4.29) and death (OR=4.06, 95% CI 2.39-6.88) after adjusting for key covariates, with the odds of hospitalization for frail respondents much higher in this population than has been found in more impaired populations. Although unmet need was a significant independent predictor of hospitalization in select models, it was not found to significantly modify the association between frailty and hospitalization. Unmet home care need was significantly associated with death in age and sex adjusted models only, with no statistically significant effect modification found for this outcome. Secondary analyses exploring SSA as an effect modifier were also conducted due to high correlations with both frailty and unmet home care need. SSA was also not found to modify observed frailty-outcome associations in this population. These results provide evidence of the predictive ability of frailty indices for community-dwelling populations, including middle-aged Canadians (aged 45-64 years). The findings also suggest that, although unmet home care need was not an effect modifier in the current study (possibly reflecting low power given its low prevalence in this relatively healthy survey sample), it was an independent predictor of hospitalization in select multivariable models, and therefore, may represent an important contextual factor requiring further exploration. Given the preliminary nature of the investigation of SSA as an effect modifier of frailty-related outcomes in the present study, further research of its relevance in other middle-aged and older populations is also warranted. The identification of differences in outcomes of those with met versus unmet home, health and psychosocial needs among more vulnerable or frail community-dwelling Canadians will lead to a broader understanding of where services, policy and prevention measures should be targeted.