The relationship between hospital volume and outcome of gastrointestinal cancer surgery in Korea

Objectives To assess the relationship between hospital volume and in-hospital mortality of patients undergoing four surgical procedures for gastrointestinal cancers in Korea. Methods Using the database of the Health Insurance Review and Assessment Service, we identified 66,201 patients who underwent the four types of gastrointestinal resection during the period 2005–2006. Participating hospitals were divided into five groups according to their surgical volume. The primary outcome was in-hospital mortality, defined as death from any cause before discharge. Multivariate logistic regression analysis was performed to determine the effect of hospital volume on risk-adjusted in-hospital mortality. Results We observed a significant relationship between hospital volume and in-hospital mortality rate for patients undergoing the four types of cancer-related gastrointestinal surgeries. The in-hospital mortality rate was lower for high-volume than for low-volume hospitals after adjusting for patient characteristics. The differences between very-high-volume and very-low-volume hospitals ranged from 0.94% to 2.77% for the four procedures, with the largest difference observed for pancreatic resection (3.75% vs. 0.98%). Conclusion High-volume hospitals had better short-term surgical outcome than low-volume hospitals. We confirmed the volume–outcome relationship for four cancer-related gastrointestinal resections in Korea. J. Surg. Oncol. 2011;104:116–123. © 2011 Wiley-Liss, Inc.