Related Experiment Video
Updated: Jun 18, 2026
Robotics in Surgery: A Modular Robotic Platform Driven Gastric Wedge Resection
Published on: February 7, 2025
The learning curve for robotic-assisted esophagectomy: A single-surgeon cumulative sum analysis
Kenneth Meredith1, Jamie Huston2, Ravi Shridhar3
1Florida State University, College of Medicine, Sarasota, FL, United States; Sarasota Memorial Health Care System, Kolschowsky Research and Education Institute, Sarasota, FL, United States.
Background:
Minimally invasive esophagectomy has significant benefits in operative outcomes compared with open approaches. The myriad of techniques have precluded the recommendation of a standard approach. The use of the robotic approach has increased steadily. This study sought to evaluate trends and identify milestones in the learning curve for robotic-assisted esophagectomy.
Methods:
We prospectively followed all patients who underwent robotic-assisted esophagectomy performed by a single surgeon between 2010 and 2022. Clinicopathologic factors and surgical outcomes were recorded and compared across successive cohorts. To identify key inflection points in the learning curve, cumulative sum (CUSUM) analysis combined with structural change detection regression was applied. All statistical tests were 2-sided, and P <.05 was considered statistically significant.
Results:
We identified 323 patients who underwent robotic-assisted esophagectomy between 2010 and 2022. The median operative time was 340 min. CUSUM analysis demonstrated cutoff points at cases 90, 140, and 175. The learning curve was divided into phases, and cases were stratified as follows: learning/consolidation (cases 1-90), proficient (cases 91-140), second learning curve (cases 141-175), and expert (cases 176+). Operative time significantly decreased in the proficient phase (P <.001), increased significantly during the second learning curve (P <.001), and then decreased again when expertise was achieved (P <.001). The increase in operative time after case 140 coincided with an increase in Charlson-Deyo comorbidity scores (P <.001) and higher tumor stage (P =.005), indicating patients at a higher risk. Clavien-Dindo grade III to V postoperative complications also increased after case 140.
Conclusion:
For surgeons proficient in minimally invasive esophagectomy, the learning curve for a robotic-assisted procedure appears to be near proficiency after 90 cases. However, as more complex cases are undertaken, there appears to be an additional learning curve that is surpassed after 175 cases.
