OBJECTIVES: To assess and summarize the impact surgeon learning curve effects could have on surgical metrics, clinical outcomes, and cost-benefit decisions in robot-assisted surgery across all surgical procedures.

METHODS: A systematic literature search of MEDLINE, MEDLINE In-Process, Embase, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, and NHS Economic Evaluation Database were searched from 2012 to February 5, 2018.

RESULTS: There were 2,316 total publications identified with potentially relevant information. From the systematic search, 68 studies were included in this review (49 in surgical setting; 19 in simulator/training setting). Of the 49 surgical studies, 80% were retrospective, 73% were single-arm, and all were observational in design. The number of surgeons per study was small; 67% of studies included fewer than five. The most frequent specialties were urology (n=20), general surgery (n=13), and gynecology (n=7). Time-based variables were the most commonly reported to assess learning curves (86% of studies), and the most reported individual metric was operative time (67% of studies). Other measures, including length of hospital stay, morbidity/mortality rates, and specialty-specific metrics, were less commonly reported. Of the major surgical, safety, and recovery metrics captured within this review, operative time (67%), complication rate (18%), and length of stay (10%) were reported most frequently, respectively. When evaluating the evidence base, the variability in results between studies and surgeons makes it challenging to synthesize a reliable estimate for length and burden of the learning curve for any one procedure.

CONCLUSIONS: The literature lacks studies with high-quality designs assessing surgeon learning curves in robot-assisted surgery. In addition, the methods used to assess learning curves are poorly described and frequently inadequate. Opportunities remain to establish quantitative, evidence-based approaches to measuring and defining learning curves in robotic surgery, as well as estimating the clinical and economic burdens associated with surgeon learning.