Learning Curve for Aortic Root Replacement.

Background: The objective of this study was to assess the number of root procedures performed for a newly graduated surgeon to overcome the learning curve. Methods: An institutional database was used to obtain details for patients undergoing root surgery during a 10-year period for a newly graduated surgeon. To assess learning curve, strucchange R package (engineering method to objectively identify structural changes in data) was used to analyze operative efficiency. A cumulative sum failure analysis curve was plotted to indicate the magnitude and direction of changes in efficiency for this single surgeon. A composite of adverse outcome was assessed across learning curve phases. Results: Strucchange identified 3 segments to the learning curve. These occurred at 38 (95% CI, 25-54) and 148 (95% CI, 135-152) cases. Based on this, there were 3 segments to the learning curve: after the first 25 to 50 cases, 50 to 150 cases, and then 150 cases onward. The composite of adverse outcome was compared across the phases of the learning curve. In phase 1, 44.0% experienced the outcome, which was reduced to 20.0% in phase 2 and 21.1% in phase 3 ( Conclusions: There appear to be 3 segments of the learning curve for the new surgeon when performing root replacement. The first occurs after around 25 to 50 cases, after which there is a rapid decline in operative time and complications. The second phase occurs between cases 50 and 150, during which advanced proficiency is developed. Aortic fellowship programs should aim to graduate fellows with at least 50 root procedures as first operator.