Early versus late readmission after surgery among patients with employer-provided health insurance

Objectives: To define the incidence of 90-day readmission and characterize the factors associated with 90-day readmission after 10 major surgical procedures. Background: Most data on readmission focus solely on same hospital readmission (index hospitals) within 30 days of discharge. These studies may underestimate readmission, as patients may be readmitted beyond 30 days of discharge or to other non-index hospitals. Methods: Patients discharged after 10 major surgical procedures (coronary artery bypass grafting, abdominal aortic aneurysm repair, carotid endarterectomy, aortic valve replacement, esophagectomy, pancreatectomy, pulmonary resection, hepatectomy, colectomy, and cystectomy) between 2010 and 2012 were identified from the Truven Health MarketScan Commercial Claims and Encounters database. Multivariable logistic regression analysis was performed to identify determinants of early (≤30 days) and late (31-90 days) readmission. Results: A total of 158,753 patients were identified; 60.3% were male, and 42.3% had a Charlson Comorbidity Index of 2 or more. A total of 26,817 (16.9%) patients were readmitted within 90 days [early: 16,419 (10.4%) vs late: 10,398 (6.5%)]. Among readmitted patients, 38.3% were readmitted to a different hospital than the index hospital. Both early and late readmissions were more common at the index versus non-index hospital (early: 83.9% vs 16.1%; late: 75.0% vs 25.0%; both P<0.001). In-hospital mortality after early readmission and late readmission was found to be lower at index hospitals than that at non-index hospitals (early; 0.7% vs 2.5%, P = 0.04; late; 0.2% vs 2.0%, P = 0.02). Conclusions: More than one-third of readmission occurred after 30 days of index discharge. Approximately 20% of patients were readmitted to non-index hospitals. Assessment of 30-day same hospital readmissions underestimated the true incidence of readmission.