Learning process of laparoscopic liver resection and postoperative outcomes: chronological analysis of single-center 15-years' experience

Background Limited studies have reported the actual learning process of laparoscopic liver resection (LLR). This study aimed to chronologically evaluate our 15 years' experience of LLR. Methods All consecutive LLRs between 2006 to 2020 were retrospectively analyzed. The time period was divided into three groups; first (2006-2010), second (2011-2015), and third (2016-2020) period. The primary endpoint of this study was a composite of overall (Clavien-Dindo grade >= II) or major (grade >= IIIa) postoperative complications within 30 days. Using the IWATE criteria (four difficulty levels based on six indices), LLR was categorized as basic (< 7 points) and advanced (>= 7 points) one. All analyses were performed based on the intention-to-treat principles. Results During the study period, a total of 382 LLRs were gradually performed (first period, n = 54; second period, n = 114, and third period, n = 214). Low incidences of overall and major complications were maintained (9.3, 10.5, and 7.0%, p = 0.514, and 1.9, 2.6, and 2.3%, p = 1.000). Meanwhile, pure LLRs (i.e., LLRs without hand-assisted or hybrid approach) and advanced LLRs were increasingly performed in 25 (46.3%), 71 (62.3%), and 205 (95.8%) patients (p < 0.001) and 3 (5.6%), 18 (15.8%), and 58 (27.1%) patients (p < 0.001), respectively. Conclusions This study suggests that stepwise approach from basic to advanced procedures and use of hand-assisted or hybrid approach during the early phases for starting LLR practice may allow for maintaining low morbidity in specialized center.