Operative performance evolution in robotic transaxillary thyroidectomy: an 884-case institutional experience.

Remote-access thyroidectomy via the robotic transaxillary approach (RTAT) avoids an anterior cervical scar while providing a stable, magnified operative field. How a high-volume RTAT programme evolves across its full institutional lifespan-from initial case acquisition through deliberate expansion of surgical indications-has not been comprehensively characterised in the Western literature. Retrospective analysis of 884 consecutive RTAT procedures at a single tertiary referral centre. Cases were stratified into three chronological descriptive phases: early (cases 1-295), middle (cases 296-589), and late (cases 590-884). Era boundaries were defined a priori and were not optimised by formal breakpoint or CUSUM analysis. Operative extent, operative time data, conversion rate, postoperative complications, and access-specific morbidity were compared across eras. Independent predictors of operative time were assessed by multivariable linear regression in the 355 cases with complete operative time and covariate data. The cohort was 75.7% female (of 879 patients with documented sex); mean age was 45.1 ± 12.4 years. Among 355 cases with recorded operative time (40.2% of the cohort, predominantly from the early and middle eras), mean time fell from 159.7 ± 39.0 min (early) to 133.4 ± 34.5 min (middle; p < 0.001); late-era data were not systematically available. Total thyroidectomy rose from 42.0% to 50.0% across eras (p = 0.02); malignancy proportion increased from 8.5% to 14.9%; and mean specimen weight rose from 23.8 g to 31.9 g. Conversion to open surgery occurred in 6 patients (0.7%). Transient RLN weakness affected 40 patients (4.52%); no permanent palsy was recorded. Working-space dysaesthesia (2.49%) and brachial plexopathy (0.90%) were confined almost entirely to the early era. Chronological case number, total thyroidectomy, and BMI were independently associated with operative time (R² = 0.38). In 884 consecutive RTAT procedures, operative time declined across successive programme phases, access-specific morbidity fell substantially, and the case mix broadened toward bilateral resection and oncological disease without a corresponding rise in major complications. The absence of permanent RLN injury or hypoparathyroidism across the full series represents a noteworthy safety observation during a period of deliberate indication expansion. These findings contribute descriptive evidence supporting staged RTAT programme development at an experienced centre.