Technical performance assessment of robotic surgery: a systematic scoping review

Defining performance errors in robotic surgery is critical for the assessment of robotic surgery skill. Our goal was to identify and categorize explicitly defined intraoperative technical errors in robotic surgery, how skill assessment was performed, and how ratings were conducted either manually by experts or via automated ratings. This scoping review included studies involving general, urologic, obstetrics/gynecologic, and thoracic surgery, and general skills as practiced in inanimate, virtual reality, in vivo/ex vivo animal, cadaver, and human operations. Primary empirical and consensus-building studies were included if they addressed intraoperative performance assessment or error definition and identification. MEDLINE (Ovid), Embase (Ovid), and Compendex were queried for results from 2012 to May 19, 2022. Of 2642 studies screened, 185 were included. The majority (n = 109, 60%) were US-based and involved either simulated surgical procedures using inanimate models (n = 88), virtual reality (n = 72), or intraoperative performance assessments of robotic surgeries in humans (n = 44); 36 studies combined two or more of these settings. Performance errors were explicitly defined in 104 articles (56%), and 64 used previously defined performance rating scales. The method of rating was split between manual (n = 137) and automated ratings (n = 85). Measures of performance vary considerably. More conceptual work is warranted to explicitly define errors that can inform robotic skill assessment. This is important given the growing interest in developing efficient and reliable objective measures of performance which are likely to rely on automated assessment methods.