Pattern of Length of Stay in Swedish Residential Care among Residents with and without Cognitive Impairment between 2015 and 2023: A Longitudinal Population-Based Study

Context: Length of stay (LOS) is a metric for evaluating the utilization of residential care and informing future long-term care policies. Decreases in LOS have previously been observed in the Swedish setting. However, whether this trend has continued is unexplored. Objectives: To examine changes in residential care LOS between 2015 and 2023 among persons with and without cognitive impairment, and to assess the effects of socio-demographic factors, cognitive impairment, and activities of daily living (ADL) on LOS. Methods: Persons ≥65 years who moved into residential care between 2015 and 2023 in Stockholm municipality (N = 18,805). Data on residents’ characteristics, moving-in date, and date of death were retrieved from the longitudinal SNAC Stockholm Eldercare study. Hazard regression was used to assess factors associated with LOS, and Laplace regression was used to examine changes in LOS from the 10th to the 50th percentiles. Findings: Most residents moving in were women, aged 86 years, and requiring extensive daily support. Cox regression showed that older age, male gender, lower income, greater ADL needs, and no cognitive impairment predicted shorter stays and higher death risk. LOS declined for cognitively impaired residents in the 40th and 50th percentiles. Limitations: Cognitive status was assessed before moving in and not measured during stay. The definitions of LOS vary across studies, which complicates comparisons. Implications: Most people who moved into residential care were cognitively impaired, leading to longer stays. This should be considered for future care resources allocation. Population ageing and rising cognitive impairment will increase the need for residential care.