Abstract: Objective To summarize the experience of robotic thyroidectomy in our hospital and the learning curve of robotic thyroidectomy. Methods The first 200 cases with robotic thyroidectomy using bilateral axillo-breast approach (BABA) in our hospital from May 2015 to August 2017 were retrospectively studied. Clinical data including operation time and pathological data were analyzed. Results The operative time decreased significantly and reached the plateau after first 35 cases. Eighty cases later, further decrease in operative time and second plateau were found. The patients then were divided into three groups including group A (first 35 cases), group B (36 th to 80 th cases) and group C (81 st to 200 th cases). There was no significant difference in gender, age and maximum tumor size among three groups (P>0.05). Operative time reduced sequentially from groups A (145.6±42.6 min) to group B (116.2±29.2 min) to group C (88.2±26.1 min) (P=0.001). The decrease in both hospital stays (P=0.031) and intraoperative blood loss (P=0.001) and the increase in number of lymph node dissected (P=0.001) were present among three groups. Conclusions The learning curve of robotic thyroidectomy with BABA was about 35 cases. When the cases reached to 80 cases, the procedure would be safe and effective.

Key words: Robot, Thyroidectomy, Lymph node dissection, Learning curve