Objective: To assess the utility of geriatric targeting criteria in predicting survival and health care utilization in a cohort of hospitalized older veterans.
Design: A prospective cohort study assessing geriatric targeting criteria, e.g., polypharmacy, falls, or confusion, with respect to adverse outcomes at 12 months.
Setting: A Tertiary Care VA Medical Center.
Patients: 507 acutely hospitalized male veterans aged 65 years or more.
Main outcome measures: Survival status, nursing home placement, and total hospital days during 12 months following hospital admission.
Results: Patients who had a higher number of targeting criteria at admission showed a significantly increasing trend toward death (P < or = .001), nursing home placement (P < or = .01), and longer hospital stays (P < or = .01) at 12 months. In univariate analyses, weight loss (relative hazard 3.8, 95% CI 2.4, 5.9), appetite loss (relative hazard 3.3, 95% CI 1.9, 5.8), depression (relative hazard 2.5, 95% CI 1.4, 4.5), falls (relative hazard 2.2, 95% CI 1.2, 4.1), confusion (relative hazard 2.2, 95% CI 1.2, 4.0), and socioeconomic problems (relative hazard 1.6, 95% CI 1.0, 2.5) predicted death. Polypharmacy (OR 3.4, 95% CI 1.3, 8.8), confusion (OR 4.4, 95% CI 1.5, 13.0), and prolonged bedrest (OR 7.6, 95% CI 1.5, 39.3) predicted nursing home placement. Confusion (Beta 12.0, 95% CI 2.9, 21.3), falls (Beta 14.2, 95% CI 4.2, 24.3), and prolonged bedrest (Beta 22.4, 95% CI 3.9, 41.0) predicted total hospital days. In multivariate analyses, weight loss, depression, and socioeconomic problems predicted death; confusion and polypharmacy predicted nursing home placements; and falls predicted total hospital days.
Conclusion: This prospective cohort study of hospitalized older veterans demonstrated geriatric targeting criteria as predictors of adverse hospital outcomes. Our findings suggest screening acutely hospitalized patients using chart abstracted geriatric targeting criteria is useful in identifying patients at risk for adverse outcomes of hospitalization.