Risk-adjusted benchmarks in laparoscopic liver surgery in a national cohort

Background This study aimed to assess the best achievable outcomes in laparoscopic liver resection (LLR) after risk adjustment based on surgical technical difficulty using a national registry. Methods LLRs registered in the Italian Group of Minimally Invasive Liver Surgery registry from November 2014 to March 2018 were considered. Benchmarks were calculated according to the Achievable Benchmark of Care (ABC (TM)). LLRs at each centre were divided into three clusters (groups I, II and III) based on the Kawaguchi classification. ABCs for overall and major morbidity were calculated in each cluster. Multivariable analysis was used to identify independent risk factors for overall and major morbidity. Significant variables were used in further risk adjustment. Results A total of 1752 of 2263 patients fulfilled the inclusion criteria: 1096 (62 center dot 6 per cent) in group I, 435 (24 center dot 8 per cent) in group II and 221 (12 center dot 6 per cent) in group III. The ABCs for overall morbidity (7 center dot 8, 14 center dot 2 and 26 center dot 4 per cent for grades I, II and II respectively) and major morbidity (1 center dot 4, 2 center dot 2 and 5 center dot 7 per cent) increased with the difficulty of LLR. Multivariable analysis showed an increased risk of overall morbidity associated with multiple LLRs (odds ratio (OR) 1 center dot 35), simultaneous intestinal resection (OR 3 center dot 76) and cirrhosis (OR 1 center dot 83), and an increased risk of major morbidity with intestinal resection (OR 4 center dot 61). ABCs for overall and major morbidity were 14 center dot 4 and 3 center dot 2 per cent respectively for multiple LLRs, 30 and 11 center dot 1 per cent for intestinal resection, and 14 center dot 9 and 4 center dot 8 per cent for cirrhosis. Conclusion Overall morbidity benchmarks for LLR ranged from 7 center dot 8 to 26 center dot 4 per cent, and those for major morbidity from 1 center dot 4 to 5 center dot 7 per cent, depending on complexity. Benchmark values should be adjusted according to multiple LLRs or simultaneous intestinal resection and cirrhosis.