Predictors of short-term rehospitalization following discharge of patients hospitalized with community-acquired pneumonia.

ABSTRACT

BACKGROUND: Among patients hospitalized for community-acquired pneumonia (CAP), the risk factors for short-term hospital readmission after discharge are unknown. METHODS: We conducted a prospective observational study of 1,117 patients who had been discharged alive after hospitalization for CAP. We collected variables associated with CAP severity at hospital admission, in-hospital clinical evolution, clinical instability factors on hospital discharge, therapy employed during hospitalization, and diagnostic bacteriology. We assessed hospital readmission within 30 days after discharge for the index hospitalization. Risk factors independently associated with 30-day hospital readmission were identified using Cox regression models. RESULTS: Of the 81 patients (7.3%) who were readmitted to the hospital within 30 days, 29 (35.8%) were rehospitalized for pneumonia-related causes. Variables associated with pneumonia-related hospital readmission were treatment failure (hazard ratio [HR], 2.9; 95% CI, 1.2 to 6.8), and one or more instability factors on hospital discharge (HR, 2.8; 95% CI, 1.3 to 6.2). The predictive performance of these variables measured by the area under the curve (AUC) of the receiver operating characteristic was 0.65. Variables associated with pneumonia-unrelated hospital readmission were age >or= 65 years (HR, 4.5; 95% CI, 1.4 to 14.7), Charlson comorbidity index >or= 2 (HR, 1.9; 95% CI, 1.0 to 3.4), and decompensated comorbidities during in-hospital evolution (HR, 3.5; 95% CI, 2.0 to 6.3); the AUC for this model was 0.77. Patients with at least two risk factors were at significantly increased risk of 30-day hospital readmission (pneumonia-related CAP HR, 9.0; 95% CI, 3.2 to 25.3; pneumonia-unrelated CAP HR, 5.3; 95% CI, 1.6 to 18.1). CONCLUSIONS: Among patients hospitalized for CAP, different risk factors are associated with hospital readmission related to pneumonia or to other causes. The identification of two different groups of patients who were at high risk of hospital readmission raises the possibility that different management strategies could decrease the rate of hospital readmissions.