Evolution of hemorrhage risk following sleeve gastrectomy amidst robotic assisted surgery adoption
*Katharine P. Playter, *Bradley Colarusso, John Kelly, Nicole Cherng, Richard Perugini
General Surgery, University of Massachusetts Chan Medical School, Worcester, MA, United States.
Background: Robotic assisted surgery (RAS) offers improved ergonomics, visualization, and stapling technology, and holds promise for improved outcomes relative to conventional laparoscopy. However, the transition to RAS also entails a learning curve that is not fully characterized in its complexity. Longitudinal sleeve gastrectomy (LSG) is the most commonly performed bariatric operation. We investigated the relationship between the adoption of RAS and the occurrence of hemorrhage, the most common complication following LSG.
Methods: This is a retrospective cohort study in an established, MBSAQIP accredited, bariatric surgery program which consists of five surgeons well beyond their learning curve for laparoscopic LSG. The incidence of hemorrhage, defined as blood transfusion within 3 days postoperatively or intervention to treat bleeding up to 30 days postoperatively, following LSG was observed from 2020 to 2023. This period was divided into an early adoption period (RAS 31% of cases) and a late adoption period (RAS 82% of cases). Demographics of laparoscopic LSG (L-LSG) and robotic assisted LSG (RA-LSG) groups were compared using t-tests and Fisher?s exact tests. A logistic regression model was used to assess differences among surgeons. Trends over time were analyzed with CUSUM, a standard method of tracking deviations in the outcome of a process.
Results: A total of 1250 patients underwent LSG over the study period. L-LSG and RA-LSG groups were similar in age, sex, and risk factors for postoperative bleeding including BMI, hypertension, and anticoagulant use. They differed in race (p<0.0001) and ASA category (p=0.007). The overall postoperative hemorrhage rate was 1.04% (n=13). Hyperlipidemia was predictive of hemorrhage on univariate analysis (p=0.045). CUSUM analysis of hemorrhage revealed a modest advantage for L-LSG over RA-LSG during the early adoption period (Figure 1). The relative risk (RR) of hemorrhage in the robotic group during this period was 4.38 (CI 0.577-3.24). In the late adoption period the risk profile inverted. There was an advantage observed in the robotic approach concerning hemorrhage risk, with the RR dropping to 0.28 (CI 0.083-0.96, p=0.063). Notably, this shift with respect to hemorrhage risk was not associated with a significant reduction of hemorrhage in the RA-LSG approach (1.5% vs 0.98%; p=0.636), but with an increase in L-LSG risk (0.35% vs 3.5%, p = 0.0245).
Conclusion: As RA-LSG became the more frequently utilized approach within a single program, the risk of hemorrhage with L-LSG significantly increased. This may represent a loss of reproducibility with the conventional approach as it is supplanted by the robotic approach. Future work will employ advanced robotic metrics to better understand the reduction in hemorrhage rate observed in the robotic approach, informing initiatives to integrate new robotic platforms into existing minimally invasive programs.
CUSUM of hemorrhage following LSG in robotic and conventional laparoscopic approaches over time
Back to 2024 Posters