Cumulative sum analysis for the learning curve of minimally invasive mitral valve repair.

Minimally invasive mitral valve repair, recently, has become an alternative procedure to conventional mitral valve surgery, given its clinical benefits. Understanding the learning curve of a new procedure is important prior to its introduction. This study aimed to evaluate the learning curve for minimally invasive mitral valve repair and safety during the start-up period. The first 100 consecutive patients who underwent isolated minimally invasive mitral valve repair for mitral valve regurgitation were evaluated. The procedure was performed by a single surgeon at a single institution. Calculated cumulative sum analysis and cubic spline curve analysis were performed to evaluate the learning curves for the total procedure (TP), extracorporeal circulation (ECC), and aortic cross-clamping (ACC) times. ACC time was affected by the complexity of individual mitral valve repair; therefore, we analyzed the TP minus ACC (TP-ACC) time as a true learning curve by subtracting the ACC time from the TP time to exclude the difference of the complexity. Additionally, the operative outcome was assessed. Overall, the average TP, ECC, ACC, TP-ACC times were 211 ± 41, 133 ± 35, 108 ± 31, and 104 ± 4.9 min, respectively. All cubic spline curves depicted a decreasing trend, and improvements in TP, ECC, and ACC times were observed after 56 cases, while those of the TP-ACC time were observed after 68 cases. None of the patients experienced hospital mortality, reoperation for bleeding, respiratory failure, cerebral infarction with a disability, or recurrence of mitral valve regurgitation. Acute renal failure occurred in one patient. In conclusion, minimally invasive mitral valve repair can be introduced safely and provide a favorable outcome. However, a learning curve exists for the operative time factors. Approximately 60 operations are required to achieve a consistent operative time.