Interrater Reliability of Hospital Readmission Evaluations for Surgical Patients

Abstract

Value-based purchasing initiatives have helped shift attention to the accuracy of hospital readmission information at the most clinically detailed level. The purpose of this study was to determine the interrater reliability (IRR) of surgical experts in assessing surgical inpatient readmissions for categorical causes, relation to index procedure, and potential preventability. Cases were selected from the American College of Surgeons National Surgical Quality Improvement Program local database. Of 1840 cases, 156 patients (8.5%) were readmitted within 30 days of the procedure. Surgical site infection was the most common readmission cause (32%), followed by obstruction or ileus (17%). IRR was moderate for readmission cause (60% agreement, κ = 0.51), substantial for readmission in relation to surgical procedure (92%, κ = 0.70), and lowest for potential preventability of readmissions (57%, κ = 0.18). Results suggest that readmission cause and relation to surgical procedure can be determined with moderate to high degree of IRR, while preventability of readmissions may require stricter definitions to improve IRR.