Geriatric Heart Failure, Depression, and Nursing Home Admission: An Observational Study Using Propensity Score Analysis

Abstract

Heart failure (HF) and depression are both common in older adults, and the presence of depression is known to worsen HF outcomes. For community-dwelling older adults, admission to a nursing home (NH) is associated with loss of independent living and poor outcomes. The objective of this study was to examine the effect of depression on NH admission for older adults with HF. Using the 2001-2003 National Hospital Discharge Survey datasets, the authors identified all community-dwelling older adults who were discharged alive with a primary discharge diagnosis of HF. The authors then identified those with a secondary diagnosis of depression. Using a multivariable logistic regression model, the authors then determined probability or propensity to have depression for each patient. The authors used propensity scores for depression to match all 680 depressed patients with 2,040 nondepressed patients. Finally, the authors estimated the association between depression and NH admission using bivariate and multivariable logistic regression analyses. Patients had a mean (+/- standard deviation) age of 79 (+/- 8) years, 72% were women, and 9% were blacks. Compared with 17% nondepressed patients, 25% depressed patients were discharged to a NH. Depression was associated with 50% increased risk of NH admission (unadjusted relative risk [RR]: 1.50; 95% confidence interval [CI]: 1.28-1.74). The association became somewhat stronger after multivariable adjustment for various demographic and care covariates (adjusted RR: 1.60; 95% CI: 1.35-1.68). In ambulatory older adults hospitalized with HF, a secondary diagnosis of depression was associated with a significant increased risk of NH admission.