Endoscopic Spine Fusion Learning Curve Stabilizes at About 30 Cases

Surgeons adopting a minimally invasive technique for lumbar fusion appear to climb the steepest part of their learning curve within about 30 operations, according to a retrospective cohort study published in Acta Neurochirurgica.

The research team, led by Subin Kweon with colleagues Yong Jin Park and Young San Ko, reviewed 62 consecutive patients who underwent single-level UBE-TLIF. Because the study was retrospective, it examined records of operations already performed rather than assigning patients prospectively to different groups. All cases were analyzed using the same endoscopic technique with a large-footprint interbody cage.

To map the learning curve, the investigators measured total endoscopic operative time, defined as the interval from endoscope-in to endoscope-out, and applied cumulative sum, or CUSUM, analysis. In this dataset, the CUSUM analysis identified a transition point at case 30, dividing the series into an early phase of 30 patients and a late phase of 32 patients.

Operative time fell substantially across that transition. In the early phase, the mean endoscopic operative time was 124.1 minutes with a standard deviation of 19.2 minutes, while in the late phase it dropped to 91.0 minutes with a standard deviation of 3.9 minutes, a difference the authors reported as statistically significant at p less than 0.001. The narrowing of the standard deviation also indicates that operative times became far more consistent once the transition point was passed.

A key contribution of the study is its step-by-step breakdown of where the time savings actually came from. Comparing these step-specific durations across the two learning phases using one-way and two-way mixed analysis of variance revealed a significant Step-by-Phase interaction, meaning the two steps improved at different rates.

Endplate preparation with cage insertion showed the greater reduction, falling from 64.8 plus or minus 16.0 minutes in the early phase to 41.4 plus or minus 3.0 minutes in the late phase. Decompression also improved, but less dramatically, from 57.9 plus or minus 6.7 minutes to 49.7 plus or minus 2.6 minutes. Pearson correlation analysis further showed that endplate preparation with cage insertion had the strongest association with total operative time, with a correlation coefficient of 0.95.

These findings suggest that the technical component involving disc space preparation and implant placement is the dominant driver of efficiency gains during the learning process.

The study carries some limitations worth noting. It reflects the experience of a single surgical series of 62 patients undergoing single-level fusion, and its conclusions concern operative efficiency rather than clinical outcomes such as fusion rates, complication rates, or patient-reported results, which were not reported in the abstracted findings. The published version is an early-release, peer-reviewed accepted article that remains subject to further editorial edits before the final version of record. Even so, the identification of a roughly 30-case stabilization point and a specific, trainable bottleneck provides concrete guidance for surgeons and training programs introducing this endoscopic fusion technique into practice.

Subject of Research: Learning curve analysis of unilateral biportal endoscopic TLIF with a large-footprint cage

Article Title: Unilateral biportal endoscopic TLIF with large-footprint cage: step-specific learning curve analysis

Article References: Kweon, S., Park, Y. J., & Ko, Y. S. (2026). Unilateral biportal endoscopic TLIF with large-footprint cage: step-specific learning curve analysis. Acta Neurochirurgica. https://doi.org/10.1007/s00701-026-07064-2

Image Credits: AI Generated

DOI: 10.1007/s00701-026-07064-2

Keywords: UBE-TLIF, learning curve, CUSUM analysis, endplate preparation, cage insertion, lumbar fusion, minimally invasive spine surgery, operative time, surgical training, retrospective cohort study, Unilateral, biportal