Learning Curve in Robot-Assisted Laparoscopic Liver Resection.

BACKGROUND: The objective of this study was to evaluate the learning curve effect on the safety and feasibility of robot-assisted liver resection (RALR).

METHODS: In 140 consecutive cases, all data about demographic, surgical procedure, postoperative course were collected prospectively and analyzed. Risk-adjusted cumulative sum model was used for determining the learning curve based on the need for conversion.

RESULTS: Among all 140 patients, no patients suffered from any organ dysfunction postoperatively and the operative mortality was 0%. The CUSUM analysis showed that at the 30th consecutive patient, the open conversion rate reached to the average value, and it further improved thereafter. In the last 70 patients, only 3 patients (4.3%) required conversion and 7 patients (10%) needed blood transfusion. Only 1 patient (1.3%) out of 79 patients with HCC had a positive resection margin. Univariate analyses showed the following risk factors associated with significantly higher risks of conversion (P < 0.05): tumor number > 1, lesions in segments 1/4a/7/8, right posterior sectionectomy, and lesions which were beyond the indications of the Louisville statement. Multivariate logistic analysis revealed that both tumor number > 1 (OR: 2.10, P < 0.05) and right posterior sectionectomy (OR: 11.19, P < 0.01) were risk factors of conversion.

CONCLUSIONS: The robotic approach for hepatectomy is safe and feasible. A learning curve effect was demonstrated in this study after the 30th consecutive patient. The long-term oncological outcomes of robotic hepatectomy still need further investigation.