Learning curve insights in Unilateral Biportal Endoscopic (UBE) spinal procedures: proficiency cutoffs and the impact on efficiency and complications

Objective This study systematically assesses the learning curve of Unilateral Biportal Endoscopic (UBE) techniques across various spinal surgeries, focusing on its influence on operative efficiency and complication rates to guide optimized training and practice. Methods Systematic searches in PubMed, Web of Science, Embase, Scopus, and Cochrane Library identified studies on UBE learning curves for patients aged 18 or older, comparing early and mastery phases. Two reviewers independently extracted data on surgery type, operative time, and complications. Study quality was assessed using the Newcastle-Ottawa Scale. We performed subgroup analyses based on different UBE surgery types, examining variations in operative time and complication rates across each procedure. Results Thirteen studies, including 1217 patients, were included, focusing on lumbar spine surgeries. The average learning curve cutoff for UBE procedures was 32.18 cases, with the mastery phase reducing operative time by an average of 48.14 min (95% CI 35.80, 60.47; p < 0.001), although high heterogeneity observed. In the mastery phase, single-group analysis of 707 patients showed a mean operative time of 92.38 min (95% CI 77.35, 107.41). Complication analysis across 12 studies (1182 patients) revealed a significantly lower rate in the mastery phase (odds ratio, OR = 0.29; p < 0.001). Subgroup analyses revealed significant reductions in both operative time and complications for UBE-LIF and UBE-LD/ULBD during the mastery phase, while no significant changes were observed for UBE-PCF. Additionally, specific complications, such as dural tears and incomplete decompression, were significantly reduced in the mastery phase. Conclusion This systematic review confirms a learning curve in UBE surgery, with improvements in operative time and complications. Procedure-specific cutoff points provide guidance for training and future research.