The learning curve in robotic-assisted thoracic phase of oesophagectomy

Abstract Background Robotic oesophagectomy has gained increasing prominence in the management of oesophageal cancer, offering potential benefits over open procedures including improved visualisation, enhanced dexterity, and reduced surgical trauma. The adoption of this technique, however, requires navigating a learning curve. As oesophagectomy typically involves two stages, a staggered approach to training and implementation of robotic assistance can be employed. This approach enables surgeons to focus on mastering each phase individually, whilst allowing the completion of the procedure using an established technique. This study investigates the learning curve associated with the thoracic phase during oesophagectomy for a single high-volume surgeon. Methods A prospectively maintained database was used to retrospectively analyse the thoracic phases of the first 40 robotic-assisted oesophagectomies and compare them to 20 open thoracic phase procedures. These cases were sequentially performed by an experienced surgeon at a UK high-volume upper GI tertiary referral centre who had already achieved competence for the robotic abdominal phase. Operating time, blood loss, nodal count, R0 rate and length of stay were reviewed to determine the number of cases required to reach parity with the open procedure. Results There were no significant differences between the two regarding median age, BMI or anaerobic threshold. Operation time (210 vs. 247 minutes) and one lung ventilation time (175 minutes vs. 235 minutes) were significantly longer with robotic surgery, but reported blood loss was significantly less with robotic surgery (350 vs. 100ml) and there was no difference in median node count (34 vs. 35), R0 rate (95 vs. 95%) or length of stay (10 vs. 9 days). No Parameter differences were seen between early cases and later cases. Conclusions The findings of this study demonstrates that although operative time was longer there were comparable outcomes between the robotic and open thoracic phases of oesophagectomy for a single high-volume surgeon. Although progressive improvement in parameters for the robotic-assisted chest phase would be expected as the surgeon acclimated to the procedure, no such changes seen in this dataset. This may indicate that the learning curve in the thoracic phase is flattened in those who have already achieved competence in the robotically-assisted abdominal phase.