Learning curves, safety, and experiences of a tertiary surgical center in the introduction of robotic-assisted surgery in gynecologic oncology

The dynamic development towards robotic-assisted surgery particularly affects operative gynecology. The analysis of operative data from robotic-assisted procedures since the first application at a surgical center provides valuable insights into the introduction phase and integration of the DaVinci system into routine clinical operations, as well as their impact on patient care. The aim of this work was to specifically examine the learning curve progression and to present the trend of the professionalization process in implementing the methodology in gynecologic oncology. A retrospective data analysis was conducted of the first n = 107 patients who underwent surgery for a gynecological malignancy with the DaVinci surgical system at the University Medical Center Freiburg between 2020 and 2022. Classic operative parameters were evaluated, including preparation time, skin-to-skin time, console time, and the resulting learning curves of the surgeons and the operative team (including CUSUM analysis and linear regression models). Additionally, perioperative patient characteristics were recorded (e.g., blood loss, length of hospitalization, conversion rate). The average operative preparation time is 26.11 ± 8.13 min. The maximum value (CUSUM peak) is at approximately 20 performed procedures, indicating that the processes of operative preparation were mastered after this number of operations. The average skin-to-skin time is 172.84 ± 71.68 min (range 43–387 min), whereby after an initial reduction in skin-to-skin time within the first 30 cases, there was a slight increase in the further course with renewed reduction from approximately 65 procedures. The average console time for all tumor entities is 131.98 ± 63.74 min; for the most common operative indication (endometrial cancer, n = 61), it is 109.89 ± 52.04 min (range 48–221 min). In the surgeons' learning curves, the two surgeons with the most procedures show a CUSUM peak after 11 and 22 procedures, respectively. The average length of stay is 5.00 days (± 2.30). A total of two conversions occurred (conversion rate = 1.9%). Upon evaluation of the first n = 107 DaVinci operations, rapid learnability of robotic-assisted operations was demonstrated. The conversion rate was low at 1.9%. A positive effect on the learning curve of individual surgeons was evident after approximately 20 procedures. Both the preparation time and the skin-to-skin time could be rapidly reduced, so that integration into routine clinical operations was possible without problems.