Impact of hospital volume on clinical outcomes of endoscopic biliary drainage for acute cholangitis based on the Japanese administrative database associated with the diagnosis procedure combination system

Abstract

We aimed to determine the relationship between hospital volume and the clinical outcomes of endoscopic biliary drainage for acute cholangitis, using the Japanese administrative database associated with the diagnosis procedure combination (DPC) system. A total of 8698 patients with endoscopic biliary drainage were referred to 654 hospitals. We corrected patients' data from the database to compare risk-adjusted length of stay (LOS) and drainage-related complications in relation to the hospital volume. Hospital volume was categorized into three groups based on number of cases during the study period: low-volume hospitals (LVHs; <16 cases), medium-volume hospitals (MVHs; 16-32 cases), and high-volume hospitals (HVHs; >32 cases). Significant variation in mean LOS was observed between hospital volume categories (26.8±22.6days in LVHs vs. 23.3±21.5days in MVHs vs. 19.7±17.2days in HVHs, P<0.001). There was a significant difference with regard to complications of endoscopic biliary drainage (5.6% in LVHs vs. 4.3% in MVHs vs. 3.2% in HVHs, P<0.001). Multiple linear regression analysis revealed that hospital volume was most significantly associated with a decrease in risk-adjusted LOS. The standardized coefficient of MVHs was -0.155, whereas that of HVHs was -0.802. Multiple logistic regression analysis showed that hospital volume decreased the relative risk of drainage-related complications. The odds ratio (OR) of MVHs was 0.764 [95% confidence interval (CI), 0.604-0.965], whereas the OR of HVHs was 0.561 (95% CI, 0.434-0.725). There was a significant association between hospital volume and the clinical outcomes of endoscopic biliary drainage for acute cholangitis.