Predictors of nursing home placement in a hospital-based cohort

Introduction

Nursing home (NH) placement among older adults is influenced by numerous factors, including prognostic disease indicators, functional and cognitive status, and one's social environment, and is associated with multiple problematic outcomes. Interventions to prevent or delay NH placement have mainly been researched in the community. However, identification of predictive factors during hospitalization could provide opportunities to shape interventions for high-risk patients. The purpose of this investigation was to identify predictors of NH placement in a hospital-based cohort.

Materials and Methods

Information about the study population (n=6,009 discharged persons) was obtained from the Scott and White Memorial Hospital's billing and electronic medical records. Descriptive characteristics were compared between community discharges and NH discharges using the ÷2 statistic for categorical data and t-test for continuous variables. Multiple logistic regression models were performed to identify the most salient predictors of NH placement.

Results/Discussion

Compared to community-discharged persons, NH-discharged persons were more likely to be male, without a caregiver, have more hospitalizations, be severely ill, need help with ADLs (activities of daily living), and not understand their illness. Significant predictors of being discharged to the NH included not understanding one's illness, being female, living alone, being more severely ill, having longer hospitalizations, and having a fall risk. Significant interactions between needing help with ADLs and number of prior hospitalizations and the presence of a family caregiver, but not a spousal caregiver, were observed (p<0.0001). In conclusion, these results could help develop hospital-based interventions to postpone or prevent NH placement among high-risk patients.