Learning curve analysis of single-port thoracoscopic combined subsegmental resections

Abstract

Background: Combined subsegmental surgery (CSS) is considered to be a safe and effective resection modality for early-stage lung cancer. However, there is a lack of a clear definition of the technical difficulty classification of this surgical case, as well as a lack of reported analyzes of the learning curve of this technically demanding surgical approach.

Methods: We performed a retrospective study of single-port thoracoscopic CSS performed by the same surgeon between April 2016 and September 2019. The combined subsegmental resections were divided into simple and complex groups according to the difference in the number of arteries or bronchi which need to be dissected. The operative time, bleeding and complications were analyzed in both groups. Learning curves were obtained using the cumulative sum (CUSUM) method and divided into different phases to assess changes in the surgical characteristics of the entire case cohort at each phase.

Results: The study included 149 cases, including 79 in the simple group and 70 in the complex group. The median operative time in the two groups was 179 min (IQR, 159-209) and 235 min (IQR, 219-247) p < 0.001, respectively. And the median postoperative drainage was 435 mL (IQR, 279-573) and 476 mL (IQR, 330-750), respectively, with significant differences in postoperative extubation time and postoperative length of stay. According to the CUSUM analysis, the learning curve for the simple group was divided by the inflection point into 3 phases: Phase I, learning phase (1st to 13th operation); Phase II, consolidation phase (14th to 27th operation), and Phase III, experience phase (28th to 79th operation), with differences in operative time, intraoperative bleeding, and length of hospital stay in each phase. The curve inflection points of the learning curve for the complex group were located in the 17th and 44th cases, with significant differences in operative time and postoperative drainage between the stages.

Conclusion: The technical difficulties of the simple group of single-port thoracoscopic CSS could be overcome after 27 cases, while the technical ability of the complex group of CSS to ensure feasible perioperative outcomes was achieved after 44 operations.

Keywords: combined dimensionality reduction method; learning curve; subsegmental resection; three-dimensional reconstruction and simulation; video-assisted thoracoscopy.