Interaction between cognitive impairment and discharge destination and its effect on rehospitalization

Abstract

OBJECTIVES: To evaluate the effect of cognitive impairment on rehospitalization in older adults. DESIGN: One-year longitudinal study. SETTING: Medical service of an urban, 340-bed public hospital in Indianapolis between July 2006 and March 2008. PARTICIPANTS: Individuals aged 65 and older admitted to the medical service (N = 976). MEASUREMENTS: Rehospitalization was defined as any hospital admission after the index admission. Participant demographics, discharge destination, Charlson Comorbidity Index, Acute Physiology Score, and prior hospitalizations were measured as the confounders. Participants were considered to have cognitive impairment if they had two or more errors on the Short Portable Mental Status Questionnaire. RESULTS: After adjusting for confounders, a significant interaction between cognitive impairment and discharge location was found to predict rehospitalization rate (P = .008) and time to 1-year rehospitalization (P = .03). Participants with cognitive impairment discharged to a facility had a longer time to rehospitalization (median 142 days) than participants with no cognitive impairment (median 98 days) (hazard ratio (HR) = 0.77, 95% confidence interval (CI) = 0.58-1.02, P = .07), whereas participants with cognitive impairment discharged to home had a slightly shorter time to rehospitalization (median 182 days) than those without cognitive impairment (median 224 days) (HR = 1.15, 95% CI = 0.92-1.43, P = .23). These two nonsignificant HRs in opposite directions were significantly different from each other (P = .03). CONCLUSION: Discharge destination modifies the association between cognitive impairment and rehospitalization. Of participants discharged to a facility, those without cognitive impairment had higher rehospitalization rates, whereas the rates were similar between cognitively impaired and intact participants discharged to the community.