Background The learning curve of robotic thyroid surgery has been evaluated solely in terms of operation time, and not with respect to surgical complications. The aim of this study was to evaluate the learning curve of robotic thyroid surgery with regard to both operation time and temporary hypoparathyroidism using quantitative statistical analysis.
Methods A total of 194 patients who underwent total thyroidectomy and concomitant central compartment node dissection for papillary thyroid carcinoma by a single surgeon between December 2008 and September 2017 were enrolled. The learning curve for operation time was assessed using the cumulative sum (CUSUM) technique, and the number of procedures required to reduce the incidence of temporary hypoparathyroidism to less than 30% was determined using the CUSUM and risk-adjusted CUSUM (RA-CUSUM) techniques. Age, gender, primary tumor size, multifocality, lymph node metastasis, and weight of thyroid gland removed were used as potential risk factors in the RA-CUSUM analysis.
Results The learning curve for operation time was divided into three phases: phase 1 (the initial learning period, 1st – 19th cases), phase 2 (the challenging period, 20th – 121st cases), and phase 3 (the competent phase, 122nd – 194th cases). To reduce the incidence of temporary hypoparathyroidism to <30% required 119 cases, and after adjustment for potential risk factors by RA-CUSUM analysis this extended to 173 cases. Parameters associated with surgical completeness, that is, number of lymph nodes retrieved and stimulated thyroglobulin levels, were maintained without variation during the study period.
Conclusions Technical proficiency for robotic thyroid surgery with respect to the avoidance of surgical complications probably requires a longer learning period than that required for operation time.