The learning curve of robotic coronary arterial bypass surgery: A report from the STS database

Background There is limited data to inform minimum case requirements for training in robotically assisted coronary artery bypass grafting (RA‐CABG). Current recommendations rely on nonclinical endpoints and expert opinion. Objectives To determine the minimum number of RA‐CABG procedures required to achieve stable clinical outcomes. Methods We included isolated RA‐CABG in the Society of Thoracic Surgeons (STS) registry performed between 2014 and 2019 by surgeons without prior RA‐CABG experience. Outcomes were approach conversion, reoperation, major morbidity or mortality, and procedural success. Case sequence number was used as a continuous variable in logistic regression with restricted cubic splines with fixed effects. Outcomes were compared between operations performed earlier versus later in case sequences using unadjusted and adjusted metrics. Results There were 1195 cases performed by 114 surgeons. A visual inflection point occurs by a surgeon's 10th procedure for approach conversion, major morbidity or mortality, and overall procedural success after which outcomes stabilize. There was a significant decrease in the rate of approach conversion (7.7% and 2.5%), reoperation (18.9% and 10.8%), and major morbidity or mortality (21.7% and 12.9%), as well as an increase in the rate of procedural success (72.9% and 85.3%) with increasing experience between groups. In a multivariable logistic regression model, case sequences of >10 were an independent predictor of decreased approach conversion (odds ratio [OR]: 0.27; 95% confidence interval [CI]: 0.09–0.84) and increased rate procedural success (OR: 1.96; 95% CI: 1.00–3.84). Conclusions The learning curve for RA‐CABG is initially steep, but stable clinical outcomes are achieved after the 10th procedure.