Changes in Performance of More Than 1000 Minimally Invasive Liver Resections

This cohort study examines changes in laparoscopic liver resection procedures in more than 1000 patients in a single center over a period of 17 years.


Changes in Performance of More Than 1000 Minimally Invasive Liver Resections
Laparoscopic liver resection (LLR) is gaining in popularity. The purpose of this study is to report our performance with minimally invasive liver resections.
Methods | We report our experience with LLRs in 1062 patients at a single center from calendar years 2001 to 2017. This study received expedited approval from the University of Pittsburgh institutional review board. The approval included waivers of the Health Insurance Portability and Accountability Act and the requirement of documented consent because the study used data previously deidentified. We divided performance There were 664 female patients (62.5%) and 398 male patients (37.5%), with a mean age of 58 years (range, 17-94 years), a mean body mass index of 29 (range, 16-61; calculated as weight in kilograms divided by height in meters squared), and an American Society of Anesthesiologists mean (SD) score of 2.6 (0.6). The approach was purely laparoscopic in 724 patients (68.2%), hand-assisted in 134 patients (12.6%), hybrid in 130 patients (12.2%), and robotic in 74 patients (7.0%). Laparoscopic major hepatectomy (right or left lobectomy) was done in 91 of 1062 cases (8.5%), and this did not change across study periods.  abdominal abscess, 5; thromboembolism, 6; and ileus, 8). The unplanned open conversion rate was 2.5% and did not change across periods. The median hospital length of stay was 2 (range, 1-3) days. Thirty-day and 90-day mortality rates were 2 of 1062 patients (0.2%) and 4 of 1062 patients (0.4%), respectively.
Discussion | Surgical performance for LLR improved and perioperative morbidity decreased with greater experience. To our knowledge, this is one of the largest single-center series of LLR reported worldwide. Looking at LLR in more than 9500 patients, a meta-analysis 1 showed comparable mortality and significantly fewer complications, transfusions, blood loss, and hospital stays in LLR vs open liver resection.
The common learning curve parameters are operating room time, conversions, blood loss, and morbidity. 2 Most laparoscopic learning curve studies report on 1 of these parameters but not others. Performance and learning curves can vary with the degree of difficulty. The learning curve is a moving target and is different in the self-taught era vs the masterapprentice era.
In assessing the learning curve for laparoscopic minor hepatectomy, Vigano et al 3 reported that conversion rate, operating room time, blood loss, use of the Pringle maneuver, and morbidity improved over 3 periods. When examining laparoscopic major hepatectomy in 173 patients, Nomi et al 4 identified 3 phases (at 45, 30, and 98 cases) in the learning curve, using cumulative sum chart (CUSUM) analysis for operating time. Operating room time, pedicle clamping, blood loss, and conversions all improved in phase 3 vs phase 1. In 159 cases, van der Poel et al 5 reported a conversion rate of 11% and a learning curve of 55 cases for laparoscopic hemihepatectomies. When looking at performance in 150 consecutive LLRs, Villani et al 6 reported 5 groups of 30 consecutive cases per group. Operative complexity for laparoscopic major hepatectomy increased from 3% to 23% (in group 1 vs group 5). Complications decreased from 20% to 3% (in group 1 vs 2) but increased as more complex procedures were performed (in group 2 vs 5).
Based on our data and analysis of other published experiences, 3-6 the evolution of LLR learning can be divided into 3 phases defined by 45 and 70 cases (Figure). The learning curve parameters may have periods of improvement and regression as more difficult cases are done until mastery is achieved. The goal is for training programs to shorten the learning curve and improve performance compared with surgeons who are