Association of hospital volume with postoperative outcomes in Crohn's disease.

Abstract

Most patients diagnosed with Crohn's disease (CD) require surgery during their lifetime. While the literature has shown that certain cancer patients have superior postoperative outcomes at high-volume hospitals, there remains a paucity of data on the hospital volume-outcome relationship in CD. Given the complexities in both medical and surgical management, this study aims to determine whether patients with CD have superior postoperative outcomes at high-volume hospitals. A retrospective analysis of patients undergoing abdominal surgery for CD in New York hospitals between 2012 and 2018 was performed using data from the Statewide Planning and Research Cooperation System. Outcomes included postoperative mortality, 30-day readmission and postoperative complications. Using a penalized B-spline plot, high-volume centres were defined as those performing more than 160 abdominal surgeries for CD each year. A total of 13,221 surgeries were performed across 176 hospital centres in New York State. Of these, 73.9% of procedures occurred at low-volume centres. High-volume hospitals had lower in-hospital mortality (0.5% vs. 1.5%; p< 0.001) and 30-day readmission rates (8.3% vs. 10.4%; p< 0.001) than low-volume centres. Major postoperative complications and reoperation rates did not differ by hospital volume. On multivariate analysis, patients at high-volume hospitals had lower odds of in-hospital mortality (OR 0.54, 95% CI 0.38-0.75) and 30-day readmission (OR 0.79, 95%CI 0.64-0.98). Hospital volume remained an independent predictor of 30-day readmission for emergent admissions (OR 0.72, 95% CI 0.61-0.85) and in-hospital mortality for nonemergent admissions (OR 0.39, 95% CI 0.19-0.82). Patients undergoing abdominal surgery for CD have lower odds of postoperative mortality and 30-day readmission when the operation occurs at a high-volume hospital. These findings suggest that surgical patients with CD may benefit from care at specialized centres.