Factors associated with accessing aged care services in Australia after approval for services: Findings from the historical cohort of the Registry of Senior Australians

Abstract Objective To evaluate the access of approved aged care services and factors associated with accessing these services. Methods A retrospective cohort study was conducted (1/7/2003‐30/6/2013). The incidence of accessing permanent residential, home and respite care services within one year or transition care within 28 days of approval was evaluated. The association of participants’ socio‐demographic characteristics, limitations, health conditions and assessment characteristics with service use was evaluated. Results In 799 750 older Australians, the incidence of accessing approved permanent residential care within one year was 70.9% (95% confidence interval [CI] 70.8%‐71.0%), home care 49.5% (95% CI 49.3%‐49.7%) and respite 41.8% (95% CI 41.7%‐41.9%). The incidence of accessing transition care at 28 days was 78.5% (95% CI 78.2%‐78.7%). Aged care seekers’, assessments’ and assessors’ characteristics are associated with service access. Conclusion Monitoring the use of aged care service approvals is necessary for the identification of service access barriers to support evidence‐based policy changes.


| INTRODUCTION
Australia has an ageing population, putting significant stress on its aged care and health-care systems. 1 Over 10% of Australians over 65 years old use government-subsidised aged care services every year, costing the Government $18 billion/year. 2 These subsidies are granted after an aged care eligibility assessment in which the needs of older people are reviewed. In the 2017/2018 financial year, 186 128 new eligibility assessments were performed and historically most (83%) result in service approvals. 2,3 However, according to the most recent estimates (2009/2010) only 75% of individuals with approvals accessed these services. 3 While there is significant literature on factors that lead to use of long-term care, [4][5][6][7][8][9] less is understood about factors that lead to use of home, respite and transition care. Even less is understood about individuals who obtained approvals for support and did not access these services. Understanding the patterns of service access and characteristics of people accessing, or not accessing, these services can be useful, to identify groups that may potentially experience barriers to entry into care, or possibly groups where access is less prioritised. Additionally, a historical understanding of the access to these services is important for appropriate resource allocation and as a base for comparison for recent changes in the aged care sector (eg, Living Longer Living Better Act). 10 This study's aims were to (a) characterise individuals with a first-time aged care eligibility assessment and approvals for permanent residential, home, respite and transition care services and (b) investigate the factors associated with accessing an approved service.

| METHODS
A retrospective cohort study was conducted using data from the Historical National Cohort of the Registry of Senior Australians (ROSA). 11,12 This data set is comprised of linked de-identified data obtained from the Australian Institute of Health and Welfare's (AIHW) National Aged Care Data Clearinghouse. [13][14][15] In brief, these data sets contain information on aged care services eligibility assessments performed in Australia, the aged care seekers' socio-demographic characteristics, activity limitations, health problems, living conditions, carer availability, assessors, approvals for service, service utilisation and mortality.
The study cohort includes individuals ≥65 years old, not identified as being Aboriginal or Torres Strait Islander, who had a first-time aged care eligibility assessment by an Aged Care Assessment Team (ACAT), and were approved for permanent residential, home, respite or transition care between 1/7/2003 and 30/6/2013 (N = 799 750). Individuals not approved for services were not included.
The cumulative incidence of accessing approved permanent residential, home or respite care within one year, or receiving transition care within 28 days of approval, were the outcomes interest. The cohort was followed until 30/6/2014 for one-year follow-up. A one-year follow-up period was chosen for permanent residential, home and respite care because i) the median time for service access ranges from 6 to 136 days, ii) prior to 2009, the approvals were valid for one year, 3 and iii) this is generally agreed to be a long wait for services. Twenty-eight days was used for transition care as approvals expire after this period. 14 Factors associated with access to service that were evaluated included: age, sex, country of birth, Department of Veterans' Affairs (DVA) Card Status, 16 living arrangements, usual accommodation, carer availability, remoteness location, State, activity limitations and health conditions. Assessor/ assessment variables evaluated were as follows: assessors' professional training, year, other approved services, priority category and emergency care.
Individuals were grouped by approval types for analysis, and the four types of approvals (non-exclusive) are permanent residential, home, respite and transition care. To identify factors associated with accessing an approved service within the specified time period, Fine-Gray subdistribution hazard regression models that accounted for the competing risk of death were employed. Time to accessing services was the difference between the dates of approval and entry to services. Cases were censored if they entered another service or reached the end of the study period. A stepwise variable selection approach was used, and Akaike information criterion was used for model selection. Only cases with complete information were analysed (<3.4% of cases had missing data). Survival estimates, cumulative incidence function plots, and adjusted subdistribution hazard ratios (asHR) and 95% confidence intervals (CI) are provided. All tests were two-sided and alpha = 0.0125 was considered statistically significant, which accounts for multiple hypothesis testing. R version 3.5.1 was used for analysis.
As a sensitivity analysis, for comparison to the estimates of a model treating death as a competing risk, a Cox regression model censoring individuals at the time of death was performed (Table S1).

Policy Impact
This study evaluated the access of aged care services in the Australian population. Monitoring the use of aged care services approvals is necessary for the identification of service access barriers to support evidence-based policy changes.  Figure 1). The cumulative incidence of accessing transition care at 28 days was 78.5% (95% CI 78.2%-78.7%) (Table 2, Figure 2).
Individuals who accessed permanent residential care were more likely to (asHR, 95%CI): be older, with a 1.11 (1.10-1.11) increase in risk per 10 Individuals with a history of falls (1.14, 1.13-1.15), delirium (1.14, 1.11-1.17), dementia (1.31, 1.30-1.32) and incontinence (1.07, 1.06-1.08) were more likely to access their approved service. People with approvals prior to 2013 were more likely to access care than those in 2013, as were those whose assessment team included a medical practitioner, nursing professional and/ or social welfare professional (Table 3).

| Use of respite care
Most characteristics associated with permanent residential care access were similar for respite care access. The main differences (asHR, 95% CI) were that: those with gold (1.09, 1.08-1.11) DVA cards were more likely to use service approvals; people outside of major cities were more likely (1.14, 1.13-1.15); people with limitations with moving around (0.98, 0.97-0.99), hypertension (0.98, 0.97-0.999), arthritis (0.97, 0.96-0.98), diabetes (0.97, 0.96-0.98) and eye diseases (0.98, 0.97-0.99) were less likely; finally, those with approvals from a team with a social welfare professional were less likely to use respite services, while those with additional approvals for permanent residential care (1.30, 1.29-1.32) were more likely to do so (Table 3).   Table 3).

| DISCUSSION
Twenty-eight per cent of individuals with approvals for permanent residential, home or respite aged care services did not access services within one year, and 22% with approvals for transition care did not access this service within 28 days. The cumulative incidence of accessing services varied between 42% and 79%. Service access depended an individuals' age, sex, living arrangements, state, remoteness and their specific limitations and health conditions (ie, pain, falls, cancer, delirium, dementia, diabetes, depression, fractures and incontinence) as well as external factors, including the eligibility assessment team involved and other service approvals.
Our study highlights common factors associated with the use of aged care services but also factors important for specific types of service access.
Use of respite (42%) and home (50%) care approvals was lower than for permanent (71%) and transition (79%) care. Our estimates of access are lower for transition care (89%) and higher for permanent residential (49%), respite (25%) and home care (42%), than 2009 estimates. 3 Our analysis included individuals from a 10-year period, censored individuals that received other services and treated mortality as a competing risk during the follow-up period, therefore reflecting the cumulative incidence of people with approvals at one year (or 28 days for transition care) who had not accessed other services or died, which is different from evaluating crude service access. 3 As expected, approvals for transition care were the most often used following approval, as these services are designed to assist hospitalised individuals in need of short-term care to recover. The second most used approval type was for permanent residential care, which is in line with this being the more commonly used longer term care solution in Australia (>72 000 entries in 2016). 2,17 The less frequent use respite approvals could reflect the reported attitudes and concerns of older people about accessing these services. [18][19][20] Finally, although not examined here, the lower use of home care services could be influenced by difficulties in accessing services (eg limited availability). 21 The likelihood of using approved permanent residential care services was significantly higher for individuals born in Australia, living alone, living in houses that they did not own, living in major cities, living with activity limitations and with a history of falls, delirium, dementia and incontinence. These findings confirm prior strong evidence that age, functional limitations, home ownership status and having dementia are associated with entering residential care. 8 Other factors, including incontinence, 7,8 have also been reported in prior literature, but less consistently. Australian-specific factors T A B L E 2 Cumulative incidence of accessing aged care services after first aged care eligibility assessment approval a were identified, including being born in Australia, living in a city and state there was variation by State. These factors are consistent with several reports 22 that highlight that entry into care varies by race and cultural background. 5,7,8 Functional limitations, living with dementia, incontinence and a history of falls were also factors associated with an individuals' use of home care approvals. Additionally, being younger, female, living with family or others or in institutions, having a carer and living with arthritis, depression and eye diseases were specifically associated with accessing home care services. However, having a history of cancer or diabetes, having approvals for all other services and being assessed by a team with a medical practitioner made people less likely to access home care. These findings highlight that while staying at home is reportedly more desirable amongst older Australians 23 and increasingly prevalent, 17 the individuals entering these services are different to those entering residential care. Some of these differences suggest that these are individuals with health conditions manageable at home and with more carer or family support.
Factors associated with the use of respite service approvals were similar to those for permanent residential care approvals. These similarities are due to the overlap in the population accessing these services. 5,6 However, we also identified factors associated with the use of respite approvals only, including having a DVA card, living outside major cities and having hypertension, arthritis, diabetes, eye diseases and fractures. The differences in the effect of certain diseases could be related to the demanding nature of certain conditions (ie, fractures). Unlike in the other approval groups, those with F I G U R E 2 Cumulative incidence of accessing approved transition care and incidence of death within 28 days. DVA benefits were more likely to access respite services, a reflection of veterans' options regarding services for longer term and transition care. Contrary to people with approvals for home care and permanent residential care, those with respite approvals were more likely to use them if they were from outside major cities. Individuals accessing transition care were more likely to have pain, arthritis, falls, fractures, eye diseases and delirium and have been assessed by teams that included a nursing professional and to have had additional approvals for home care. However, this is a service that men, individuals born overseas, and those who have support, are less likely to access. Individuals were also less likely to access this type of care if they had the assessment done by a team that included a welfare professional, live in South Australia or have approvals for permanent residential care. The high incidence of accessing these services and the conditions that characterise this cohort are consistent with our understanding that these individuals are in need of co-ordinated and supportive posthospital discharge care. 24 This was an observational study that relied on linked existing data sources, limiting our ability to comment on factors reported to be associated with aged care services use (eg personal preferences, policy and availability) not in our data sets. For example, the national target provision rate of subsided aged care places likely influences the use of services and clustering effects related to these geographical allocations could exist, but we were unable to account for these in this analysis. 2 However, a comprehensive set of characteristics of a national cohort was examined and several factors associated with entry into approved services were identified. Our study focused only on aged care services that required an aged care eligibility assessment by an ACAT and therefore does not evaluate other aged care services provided by the government or carer/family. Additionally, we did not examine the association of hospitalisations before the use of the service approval, which is also likely associated with use of the services, particularly with transition care. However, we do not feel that the use of other minor aged care and hospital services would change the associations between the factors and services evaluated. Another limitation of our study is some weak but statistically significant associations; therefore, the strength of the associations reported should be considered. We focused on the access to services within one year of approval, which underestimates the overall accessing of some services. It is possible individuals had additional eligibility assessments before 2009 or accessed services with their longer term approvals after 2009 and these were not captured in this study.
The strengths of this study include the national capture of people who underwent an aged care eligibility assessment and the use of a systematic approach that relies on trained assessors to collect its data, therefore increasing the internal validity of the information collected. Additionally, the data collection process was implemented in 2003 and remained the same during the study period. Ours is a population-based study, and our findings are generalisable to the entire Australian population seeking and accessing aged care. Our analysis also addressed the people who died or entered a different service, to obtain accurate estimates of service use by the end of the period, which had not been previously done.
Twenty-two to twenty-eight per cent of individuals approved for aged care services did not access them, and the cumulative incidence of accessing the four services varied between 42% and 79%. Aged care seekers', assessments' and assessors' characteristics were associated with people accessing services, and some may be considered barriers to services. For example, while being female and having a history of dementia, falls, depression and incontinence is ubiquitously associated with individuals accessing aged care services, older individuals with less support and more functional limitations and health conditions were more likely to use residential care. This suggests that while home care is increasingly desirable for older individuals it does not seem to be accessed by those with major functional limitations, highlighting potential unmet needs for individuals with higher care requirements who may wish to remain at home. This information is a first step to characterise the people still in need of care and potential barriers to aged care access.