ObjectiveTo study the learning curve of minimal invasive coronary artery bypass grafting (MICS CABG) and the influence on the perioperative clinical effects by analyzing operation time.MethodsFrom March 2012 to November 2020, 212 patients underwent MICS CABG by the same surgeon. Among them, 59 patients (52 males and average age of 62.89±8.27 years) with single vessel bypass grafting were as a single-vessel group and 153 patients (138 males, average age of 59.80±9.22 years) with multi-vessel bypass grafting were as a multi-vessel group. Two sets of operation time-operation sequence scatter plots were made and learning curve was analyzed by cumulative summation (CUSUM) and regression method of operation time. The surgical data of each group before and after the inflection point of the learning curve were compared with the main clinical outcome events within 30 days after surgery.ResultsThere was no death, perioperative myocardial infarction and stroke in 212 MICS CABG patients and no transfer to cardiopulmonary bypass or redo thoracotomy. The learning curve conformed to the cubic fitting formula. In the single-vessel group, CUSUM (x operation number)=–1.93+93.45×x–2.33×x2+0.01×x3, P=0.000, R2=0.986, the tipping point was 27 patients. In the multi-vessel group, CUSUM (x)=y=2.87+1.15×x–1.29× x2+3.463×x3, P=0.000, R2=0.993, and the tipping point was 59 patients. The two sets of case data were compared before and after the learning curve and there was no statistical difference in main clinical outcomes within 30 days (mortality, acute myocardial infarction, stroke, perioperative blood transfusion rate), ventilator tube, and intensive care unit retention.ConclusionThe learning curve of MICS CABG conforms to the cubic formula, and the process transitions from single to multiple vessels bypass. To enter the mature stage of the learning phase, a certain number of patients need to be done. Reasonable surgical procedures and quality control measures can ensure the safety during the learning phase.