Relationship between hospital volume and operative mortality for liver resection: Data from the Japanese Diagnosis Procedure Combination database

Aim: The present study aimed to conduct a nationwide investigation on the relationship between hospital volume and outcomes following liver resection in Japan. We also discuss health policy implications of the results. Methods: Using the Japanese Diagnosis Procedure Combination database, we identified 18 046 patients who underwent hepatic resection between July and December 2007–2009. Patients were subdivided into hospital‐volume quartiles: very low‐ (<18/year), low‐ (18–35), high‐ (36–70) and very high‐volume groups (>70). Multivariate logistic regression analysis for in‐hospital mortality within 30 days of surgery was performed to analyze adjusted effects of various factors. Results: Patients in the very high‐volume group had a higher Charlson Comorbidity Index (P < 0.001) than those in the very low‐volume group. Very low‐volume hospitals were significantly less likely to perform extended lobectomy than very high‐volume hospitals (5.4% vs 17.6%, P < 0.001). Crude in‐hospital mortality within 30 days of surgery was 1.1% (0.6%, 0.8%, 1.9% and 3.0% for limited resection, segmentectomy, lobectomy and extended lobectomy, respectively). With reference to the very low‐volume group, risk‐adjusted odds ratios (95% confidence intervals) of low‐, high‐ and very high‐volume groups for overall mortality were 0.70 (0.48–1.02; P = 0.060), 0.52 (0.34–0.81; P = 0.004) and 0.16 (0.09–0.30; P < 0.001), respectively. Conclusion: There is a linear trend between higher hospital volume and lower in‐hospital mortality of liver resection in Japan, particularly for lobectomy and extended lobectomy. Based on these results, regionalization of lobectomy and extended lobectomy in high‐volume centers could be effective for reducing postoperative mortality.