Multidimensional Analyses of the Learning Curve of Endoscopic Thyroidectomy.

Endoscopic thyroidectomy has comparable surgical outcomes and superior cosmetic satisfaction to open thyroidectomy. However, steep learning curve is a concern. This study evaluated the learning curve of endoscopic thyroidectomy using various parameters and statistical methods. A total of 90 consecutive patients who underwent endoscopic thyroidectomy using bilateral axillo-breast approach (BABA) between March 2016 and April 2020 were enrolled. Operative time, postoperative drainage amount, and blood loss were assessed by cumulative sum (CUSUM) analysis and moving average to evaluate the learning curve. Using the CUSUM analysis, the peak point of both operative time and drainage amount occurred at the 30th case. No clear single peak was identified in the CUSUM plot for blood loss. The moving average also showed significant reduction in operative time and drainage amount after, approximately, the first 30 cases. The blood loss decreased after the 25th case. We therefore divided the patients into 2 phases: phase 1 (1–30 cases) and phase 2 (31–90 cases). The operative time, drainage amount, and blood loss decreased significantly in the phase 2 compared with phase 1. Lower pain score in first postoperative day and shorter hospital stay were also observed in the phase 2. Although the reduction in transient hypoparathyroidism did not reach statistical significance, no permanent hypoparathyroidism was noted in the phase 2. The learning curve for endoscopic thyroidectomy is approximately 30 cases. Aside from the operative time, drainage amount may also serve as a surrogate for the learning curve evaluation.