Which technical difficulty score can best predict postoperative outcomes after minimally invasive liver resections?

Background: To assess technical difficulty scores for laparoscopic liver resections (LLR) in a large well-characterized cohort of low to high difficulty LLR.

Methods: Patients undergoing LLR and open liver resection (OLS) (2007-2022) at Massachusetts General Hospital were included. Patients were classified according to the technical difficulty scores Ban difficulty score, IWATE criteria, Hasegawa score, IMM score, and Southhampton score (SHH) and calibration of these scores in predicting postoperative outcome parameters was assessed.

Results: 301 patients underwent LLR. Median age was 59 years and 58.5% of the patients were female. Median lesion size was 42.2 mm, median operative time was 197.7 min, and median estimated blood loss was 400.5 ml. According to the different scoring systems, 18.9% (SHH) to 52.2% (IWATE) of the LLR were high difficulty. Overall intraoperative events according to the modified Satava classification grade II (6.6%) and grade III (2.7%) were low as was postoperative 90 days major morbidity (5.3%) and mortality (1.0%). The respective scores' calibration for predicting non-textbook outcomes, intraoperative events, operative time, major postoperative morbidity, blood transfusion rates, and length of hospital stay was moderate to good for the respective scores and best for the IWATE criteria.

Discussion: Even high technical difficulty LLR can be performed with low postoperative morbidity and mortality rates. The scores evaluated performed well in predicting major liver surgery outome parameters. Among the different difficulty scoring systems, the IWATE criteria performed best.