Mortality, morbidity, and failure to rescue in hepatopancreatoduodenectomy: An analysis of patients registered in the National Clinical Database in Japan.

BACKGROUND: The high operative mortality rate after hepatopancreatoduodenectomy (HPD) is still a major issue. The present study explored why operative mortality differs significantly due to hospital volume. METHOD: Surgical case data were extracted from the National Clinical Database (NCD) in Japan from 2011 to 2014. Surgical procedures were categorized as major (≥2 sections) and minor (<2 sections) hepatectomy. Hospitals were categorized according to the certification system by the Japanese Society of Hepato-Biliary-Pancreatic Surgery (JSHBPS) based on the number of major hepato-biliary-pancreatic surgeries performed per year. The FTR rate was defined as death in a patient with at least one postoperative complication. RESULTS: A total of 422 patients who underwent HPD were analyzed. The operative mortality rates in board-certified A training institutions, board-certified B training institutions, and non-certified institution were 7.2%, 11.6%, and 21.4%, respectively. Multiple logistic regression showed that certified A institutions, major hepatectomy, and blood transfusion were the predictors of operative mortality. Failure to rescue rates were lowest in certified A institutions (9.3%, 17.0%, and 33.3% in certified A, certified B, and non-certified, respectively). CONCLUSIONS: To reduce operative mortality after HPD, further centralization of this procedure is desirable. Future studies should clarify specific ways to improve the failure-to-rescue rates in certified institutions.