Early Discharge and Hospital Readmission After Colectomy for Cancer

BACKGROUND: Early discharge after colectomy has been shown to be feasible in studies from specialty centers, but we hypothesized that benefits of early discharge might be offset by higher risk of readmission in the surgical community as a whole. Minimizing readmissions is a national health policy priority. OBJECTIVE: This study aimed to determine whether hospitals discharging patients early had increased readmission rates. DESIGN: Patients undergoing colectomy surgery for cancer were studied using national Medicare data (MEDPAR database). Multiple logistic regression was performed to determine whether hospitals with a pattern of early discharge (median length of stay <= 5 d after surgery) had increased readmission rates. Results were adjusted for patient comorbidity, emergency operation, laparoscopic surgery, demographic factors, and complications. A separate analysis at the patient level was conducted to determine risk factors for readmission. SETTINGS: Early discharge rates at US acute care hospitals were investigated. PATIENTS: Patients 65 and older undergoing colectomy surgery for cancer (2003-2008, n = 477,461) were included. MAIN OUTCOME MEASURE: The main outcome measure was 30-day, all hospital readmission rates. RESULTS: Hospitals with a pattern of early discharge (median length of stay <= 5 d) were not found to have a higher risk-adjusted readmission rate than hospitals with the usual median length of stay (16.3% vs 15.7%, P = .077). However, changing the cutoff for "early discharge" to <= 4 days revealed an increased risk for readmission among "very early discharge" hospitals (risk-adjusted readmission rate 21.3% vs 15.7%, P < .001). At the patient level, independent risk factors for readmission included older age, male sex, black race, lower socioeconomic status, urgent/emergent surgery, comorbidities, complications, open (vs laparoscopic) surgery, and longer length of stay for the index hospitalization. LIMITATIONS: Limitations of this study included the limitations of the administrative data and elderly population. CONCLUSIONS: Hospitals with a pattern of early discharge (median length of stay <= 5 d after surgery) do not have a higher risk-adjusted readmission rate than other hospitals. These results support the safety of early discharge programs in the Medicare population.