Discharge disposition as an independent predictor of readmission among patients hospitalised for community-acquired pneumonia

BACKGROUND: Community-acquired pneumonia (CAP) is the most common non-obstetrical reason for hospital admission, the leading infectious cause of death, and a target for public reporting. CAP has thus become a target of quality improvement and pay-for-performance efforts. However, the relationship between discharge disposition and readmission risk has not been investigated. METHODS: We studied CAP patients admitted to the University of Chicago from 11/2011 to 04/2015. We collected demographic information, comorbidities, laboratory values, vital signs, a modified pneumonia severity index (PSI), length of stay (LOS), clinical instabilities before discharge, discharge disposition and 30-day all-cause readmission. A multivariate logistic regression was performed, specifying readmission as the dependent variable, including as independent variables gender, ethnicity, insurance status, discharge disposition, PSI tertile, the number of clinical instabilities, LOS and comorbidities. RESULTS: Of the 2892 CAP patients identified, 14.9% were readmitted. The distribution of discharge disposition was: 43.0% home without services, 26.1% home with home health care (HHC), 16.2% to a skilled nursing or subacute rehabilitation facility and 14.8% to an acute rehabilitation or long-term acute care facility. Of patients discharged home with HHC, 20.1% were readmitted, compared to 11.5% discharged home without services. In the multivariate regression model, being discharged home with HHC was associated with a markedly greater risk of readmission (Odds ratio 1.58 [95% confidence interval 1.21-2.07]). CONCLUSIONS: Discharge home with HHC is an independent predictor of readmission risk among hospitalised CAP patients. Discharging providers should carefully consider follow-up care and social factors that may impact the risk of readmission among such patients.