The learning curve on robot-assisted transaxillary thyroidectomy performed by a single endocrine surgeon in a third-level institution in Europe: a cumulative sum (CUSUM) analysis.

Robot-assisted transaxillary thyroidectomy is widely performed in Asian countries, although it is still under discussion in the Western World. However, there have been few studies reporting on the learning curve of robot-assisted transaxillary thyroidectomy. We used the cumulative sum (CUSUM) analysis to assess the learning curve of gasless robot-assisted transaxillary thyroidectomy at a third-level institution in Europe. We included all consecutive patients operated by a single surgeon without previous experience of robotic surgery from February 2012 to January 2023. The primary endpoint of the study was the learning curve extracted from the median operative time using the CUSUM method for the quantitative assessment. Overall, 583 patients were enrolled. The median operative time for thyroid lobectomy and total thyroidectomy was 70 and 90 min, respectively. The CUSUM analysis showed that the learning curve for thyroid lobectomy and total thyroidectomy is 66 and 56 cases, respectively. Moreover, the presence of thyroiditis resulted associated with shorter operative time for total thyroidectomy (p = 0.044), whereas no factors resulted associated with surgical complications. The learning curve for performing robotic transaxillary thyroid lobectomy for a surgeon without previous robotic experience is 66 cases. After that, 56 cases must be performed to acquire proficiency in robotic transaxillary total thyroidectomy. Training programs may reduce the slope of the learning curve.