Abstract
Objectives: Evaluate the influence of social factors on admission to aged residential care (ARC) facilities using a national comprehensive geriatric assessment database in New Zealand.
Design: Time-to-event analysis of a continuously recruited national cohort.
Participants and Setting: An anonymized data extract from a large national database for home care assessments (June 2012-December 2015) was matched with data on mortality and admissions into ARC. Methods: Four key components of psychosocial risk in relation to ARC admission were used for analysis: living alone, negative social interactions, perceived loneliness, and carer stress. Exploratory data analysis was conducted for each of the variables of interest and demographics. Unadjusted and adjusted competing risk regressions were then performed with admission into ARC being the primary outcome, death the competing risk, and remaining at home the survival case.
Results: After data cleaning, matching, and applying exclusions, the study population consisted of 54,345 eligible participants. Mean age of participants was 81.9 years (standard deviation 7.4), 62.1% were female, and 88.7% identified as European ethnicity. In the adjusted model, all 4 social factors remained significantly associated with ARC admission, namely: living alone [subhazard ratio (SHR) = 1.43 95% confidence interval (CI) 1.37-1.50]; negative social interactions (SHR = 1.22, 95% CI 1.15-1.30); perceived loneliness (SHR = 1.18, 95% CI 1.13-1.24); and carer stress (SHR = 1.28, 95% CI 1.23-1.34).
Conclusions and Implications: Interventions targeted at social factors in the context of delaying ARC admission merit further development and evaluation.