单侧双通道脊柱内镜下手术治疗腰椎间盘突出症的学习曲线分析

Analysis of the learning curve of unilateral biportal spinal endoscopic surgery for lumbar disc herniation

中文关键词:  腰椎间盘突出症  单侧双通道内镜  学习曲线  微创脊柱外科

英文关键词:Lumbar disc herniation  Unilateral biportal endoscopy  Learning curve  Minimally invasive spinal surgery

中文摘要:

  【摘要】 目的：分析单侧双通道脊柱内镜（unilateral biportal endoscopy，UBE）下手术治疗单节段腰椎间盘突出症的学习曲线特征，量化掌握该技术所需熟练手术例数，为UBE技术的规范化培训提供客观依据。方法：回顾分析2024年7月~2025年1月我院由同一主刀医师早期连续完成的60例UBE下手术治疗单节段腰椎间盘突出症患者，其中男性36例，女性24例，年龄23~77岁（52.8±13.9岁）。通过累积和（cumulative summation，CUSUM）分析手术时间，当CUSUM曲线出现明显拐点（累积偏差峰值）且后续曲线趋势发生系统性转折时，将该拐点作为分期临界点，确定学习曲线的各阶段。比较各阶段的手术时间、透视次数、术后住院天数、并发症发生率、术前和术后4周的Oswestry功能障碍指数评分（Oswestry disability index，ODI）及出院当天MacNab疗效，采用莱特学习曲线数学模型建立曲线函数对手术时间进行统计分析，计算到达手术时间能力提升阶段需要的手术台数。结果：CUSUM曲线的两个累积偏差点为1513.92和1593.8，分别对应手术的第16例和第40例，据此确定学习曲线的3个阶段：1~15例为初期学习阶段，16~39例为能力提升阶段，40例后为熟练稳定阶段，并据此分组。3组患者性别、年龄、体质指数（body mass index，BMI）、手术节段均无统计学差异（P＞0.05）。初期学习阶段组手术时间为193.13±54.68min，能力提升阶段组为90.08±22.25min，熟练稳定阶段组为72.76±11.03min。术后住院天数从初期学习阶段的6.4±1.5d，熟练稳定阶段下降到4.57±0.93d，3组患者术后并发症发生率及出院当天MacNab疗效无统计学差异（P＞0.05），3组患者手术时间有统计学差异（P＞0.05）。初期学习阶段并发症的发生率为20%（3/15），能力提升阶段为4.2%（1/24）、熟练稳定阶段未发生并发症（0/21）。初期学习阶段与熟练稳定阶段相比，手术时间及并发症发生率显著性降低（P＜0.05）。莱特学习曲线数学模型建立的曲线函数分析，手术时间到达能力提升阶段需要15例，到达熟练稳定阶段需要39例手术的积累。结论：UBE下手术治疗单节段腰椎间盘突出症安全有效，学习曲线具有明确的阶段性，学习过程划分为3个阶段：初期学习阶段、能力提升阶段和熟练稳定阶段，手术效率随经验积累而稳步提升，术者需积累约40例手术操作方可达到熟练稳定水平。

英文摘要:

  【Abstract】 Objectives: To analyze the learning curve characteristics of surgery for single-segment lumbar disc herniation(LDH) under unilateral biportal endoscopy(UBE), and to quantify the number of cases required to achieve procedural proficiency, therefore providing objective evidence for the standardized training of UBE techniques. Methods: A retrospective analysis was performed on 60 consecutive patients with single-segment LDH who underwent UBE surgery performed by a single surgeon at our institution between July 2024 and January 2025. The cohort comprised 36 males and 24 females, with ages ranging from 23 to 77 years(52.8±13.9 years). Cumulative summation(CUSUM) analysis was applied to analyze operative time. When the CUSUM curve exhibited a distinct inflection point(cumulative deviation peak) and followed by a systematic shift in trend, then this point was defined as the staging threshold to delineate the phases of the learning curve. Operative time, number of fluoroscopic exposures, postoperative length of stay(PLOS), complication rate, preoperative and 4-week postoperative Oswestry disability index(ODI), and MacNab clinical outcomes were compared across the phases. Wright′s learning curve model was used to establish a mathematical curve function for statistical analysis of operative time, and the number of cases required to reach the competence improvement phase was calculated. Results: CUSUM analysis identified two cumulative deviation peaks of 1513.92 and 1593.8, corresponding to the 16th and 40th cases, respectively. Based on these findings, the learning curve was divided into three phases(groups): the learning phase(cases 1-15), the competence improvement phase(cases 16-39), and the proficient and stable phase(cases≥40). No significant differences were found between the three groups in terms of gender, age, body mass index(BMI), or surgical segments(P>0.05). The operative time was 193.13±54.68min in the learning phase, 90.08±22.25min in the competence improvement phase, and 72.76±11.03min in the proficient and stable phase. PLOS decreased from 6.4±1.5d in the learning phase to 4.57±0.93d in the proficient and stable phase. No significant differences were observed between the three groups in postoperative complication rates or MacNab outcomes at discharge(P>0.05), whereas the difference in operative time was statistically significant(P<0.05). The complication rate was 20%(3/15) in the learning phase, 4.2%(1/24) in the competence improvement phase, and no complications occurred in the proficient and stable phase(0/21). Compared with the proficient and stable phase, the operative time and complication rate in the learning phase were significantly reduced(P<0.05). 15 cases were required to reach the competence improvement phase and 39 cases to achieve the proficient and stable phase calculated by the mathematical curve function based on Wright′s learning curve model. Conclusions: UBE surgery for single-segment LDH is safe and effective. The learning curve demonstrates distinct phased characteristics, and can be divided into three phases of learning phase, competence improvement phase, and proficient and stable phase. Surgical efficiency improves steadily with accumulated experience, and approximately 40 cases are required for a surgeon to achieve proficient and stable performance.