Abstract
Abstract
Objective
To evaluate and compare learning-curve assessment in laparoscopic and robotic colorectal surgery and to develop a standardized framework for future learning-curve reporting.
Background
Learning curves in minimally invasive colorectal surgery are widely reported, but inconsistent definitions limit interpretability and comparability. This systematic review analyzed learning-curve assessment in laparoscopic and robotic colorectal surgery and proposes a standardized framework.
Methods
MEDLINE, Web of Science, and CENTRAL were systematically searched according to PRISMA guidelines (registered in PROSPERO, CRD42026128860). Included were randomized, prospective, and retrospective studies on learning curves in colorectal surgery. Primary outcome was the number of procedures required to reach the initial learning phase. Secondary outcomes included definitions of learning curves, analytical methods, and classification of learning phases. Analyses were stratified by procedure type (colon vs. rectal resections), and pooled breakpoint regression was used to identify learning phases.
Results
A total of 114 studies comprising 167 learning-curve analyses were included. CUSUM or risk-adjusted CUSUM was used in 66.4% of analyses, whereas 33.3% applied arbitrary definitions; operative time was the most frequently used endpoint (79.6%). Operative-time plateaus were identified after 38 colonic and 56 rectal procedures. Reported median competency thresholds were numerically lower for robotic than laparoscopic surgery in colonic resections (24 vs. 30 cases;
p =
0.659) and rectal resections (28 vs. 38 cases;
p =
0.140), but these differences were not statistically significant.
Conclusions
Learning-curve assessment in minimally invasive colorectal surgery remains heterogeneous. Robotic series report numerically lower competency thresholds than laparoscopic series; however, this likely reflects skill transfer from prior laparoscopic experience and modern training pathways rather than an intrinsically shorter learning curve of the robotic platform. Differences in prior surgeon experience, training, and study era limit direct comparison. Standardized assessment and prospective studies with comparable baseline exposure are needed.