Predictors of long-term care placement in persons with dementia: a systematic review and meta-analysis

Objective: The main objective of this study was to summarize the effects of various individual, caregiver, and system-related factors on the risk of long-term care (LTC) placement for persons with dementia. Methods: We searched electronic databases for longitudinal studies reporting on predictors of LTC placement for persons with dementia residing in the community or supportive care settings. We performed meta-analyses with hazard ratios (HRs) of various predictors using random effects models and stratified the HRs with several study variables. Data on predictors not included in the meta-analyses were summarized descriptively. Results: Full-text reviews of 360 papers were performed with data from 37 papers used to calculate pooled HRs for LTC placement of select person with dementia (age, sex, race, marital status, type of dementia, living arrangement, and relationship to caregiver) and caregiver (age, sex, and depressive symptoms) characteristics. White race [HR = 1.67, 95% confidence intervals (CI): 1.41-1.99], greater dementia severity (HR = 1.05, 95% CI: 1.03-1.06), and older age (HR = 1.02, 95% CI: 1.01-1.03) increased the risk of LTC placement. Married persons with dementia (HR = 0.38, 95% CI: 0.16-0.86) and living with their caregiver (HR = 0.72, 95% CI: 0.56-0.92) had a lower risk. Behavioral and psychological symptoms of dementia, the degree of functional impairment, and caregiver burden had a consistent effect on the risk of LTC placement in our descriptive review. Conclusion: We quantified the predictive effect of several risk factors for LTC placement. These estimates could be used to more precisely categorize the risk of institutionalization and potentially link those at higher risk to appropriate services. Copyright (C) 2016 John Wiley & Sons, Ltd.