No Difference in Overall Survival Between Hospital Volumes for Patients With Colorectal Cancer in The Netherlands

BACKGROUND: High-volume hospitals have been associated with improved patient outcomes for tumors with a relatively low incidence that require complex surgeries, such as esophageal and pancreatic cancer. The volume-outcome association for colorectal cancer is under debate. OBJECTIVE: This study investigated whether hospital volume for colorectal cancer is associated with surgical care characteristics and 5-year overall survival. DESIGN: This is a population-based study. SETTING: Data were gathered from the Netherlands Cancer Registry. Hospitals were grouped by volume for colon (<50, 50-74, 75-99, and >= 100 resections per year) and rectum (<20, 20-39, and >= 40 resections per year). PATIENTS: All of the patients with primary nonmetastatic colorectal cancer who underwent resection between 2005 and 2012 were included. MAIN OUTCOME MEASURES: Differences in surgical approach, anastomotic leakage, and postoperative 30-day mortality between hospital volumes were analyzed using chi(2) tests and multivariable logistic regression analyses. Cox proportional hazard models were used to investigate the effect of hospital volume on overall survival. RESULTS: This study included 61,394 patients with colorectal cancer. In 2012, 31 of the 91 hospitals performed less than 50 colon cancer resections per year, and 21 of the 90 hospitals performed less than 20 rectal cancer resections per year. No differences in anastomotic leakage rates between hospital volumes were observed. Only small differences between hospital volumes were revealed for conversion of laparoscopic to open resection (OR of less than 50 versus 100 or more resections per year = 1.25 (95% CI, 1.06-1.46)) and postoperative 30-day mortality (colon: OR of less than 50 versus 100 or more resections per year = 1.17 (95% CI, 1.02-1.35); rectum: OR of less than 20 versus 40 or more resections per year = 1.42 (95% CI, 1.09-1.84)). No differences in overall survival were found between hospital volumes. LIMITATIONS: Although we adjusted for several patient and tumour characteristics, data regarding comorbidity, surgeon volume, local recurrences, and specific postoperative complications other than anastomotic leakage were not available. CONCLUSIONS: In the Netherlands, no differences in 5-year survival rates were revealed between hospital volumes for patients with nonmetastatic colorectal cancer.