Patterns and predictors of nursing home placement over 14 years: Dubbo study of elderly Australians

Objective: To describe patterns and predictors of nursing home placement. Methods: This is a longitudinal study of an elderly cohort living in Dubbo, New South Wales, Australia. A total of 1233 men and 1572 women aged 60 years and older living in the community were examined in 1988 and followed for 14 years. The incidence of nursing home placement and underlying clinical diagnoses were assessed. Results were modelled for baseline predictors of placement using Cox proportional hazards regression. Results: From 2805 subjects in the study there were 244 (8.7%) nursing home placements, 95 in men (7.7%) and 149 in women (9.5%). A total of 44% placements were primarily due to dementia, but dementia was a secondary diagnosis in another 20% of cases. The other principal diagnoses were stroke (16%) and coronary heart disease (14%). Median time to death in a nursing home was 10.5 months in men and 9.0 months in women. In a Cox proportional hazards model using baseline characteristics, the hazard of nursing home placement increased significantly with age (hazard ratio = 1.13), urinary incontinence (1.66), impaired peak expiratory flow (3.19), physical disability (1.59) and depression (1.85). The hazard of placement was significantly reduced by alcohol intake (0.55) and female gender (0.55). Socioeconomic factors were not significant. Conclusions: Dementia and disability‐related care burdens rather than other health conditions are the dominant causes of nursing home placement in an unselected elderly community. Risk factors such as incontinence, impaired respiratory function and depression have been identified and these are potentially amenable to intervention.