腹腔镜胆囊切除术后胆汁漏的危险因素分析及预测模型构建

Risk factors for bile leakage after laparoscopic cholecystectomy and construction of a predictive model

摘要:

背景与目的 胆汁漏是腹腔镜胆囊切除术（LC）后常见且严重的并发症之一，影响患者恢复并增加再干预风险。本研究旨在分析LC术后胆汁漏的相关危险因素，并构建预测模型评价其临床应用价值。方法 回顾性分析2020年9月—2025年5月河南省南阳市第一人民医院收治的1 630例LC患者临床资料。采用单因素及多因素Logistic回归分析胆汁漏发生的独立危险因素，并建立回归方程。通过ROC曲线评价模型预测效能。结果 1 630例患者中，胆汁漏发生81例（4.97%）。多因素分析显示，胆总管直径≤3 mm、胆囊颈部结石、解剖变异、胆囊三角粘连、与周围脏器粘连、胆囊壁厚度≥5 mm及胆囊三角电灼为独立危险因素（均P<0.05）。构建回归方程：Logit（P）=-9.126+1.362×胆囊颈部结石+0.784×胆总管直径+1.695×与周围脏器粘连+1.108×胆囊三角电灼+0.895×胆囊三角粘连+0.679×解剖变异+0.559×胆囊壁厚度。模型AUC为0.903，敏感度82.72%，特异度84.44%。结论 LC术后胆汁漏发生与多种解剖及术中操作因素密切相关，所构建的回归模型具有良好的预测效能，可用于术前及术中风险评估。

Abstract:

Background and Aims Bile leakage is a common and clinically significant complication after laparoscopic cholecystectomy (LC), which may delay recovery and increase the need for reintervention. This study aimed to identify risk factors for bile leakage after LC and to develop a predictive regression model.Methods A retrospective analysis was conducted on 1 630 patients who underwent LC between September 2020 and May 2025. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors, and a predictive model was established. The performance of the model was evaluated using receiver operating characteristic (ROC) analysis.Results Bile leakage occurred in 81 patients (4.97%). Multivariate analysis identified common bile duct diameter ≤3 mm, gallbladder neck stones, anatomical variation, Calot's triangle adhesion, adhesion to surrounding organs, gallbladder wall thickness ≥5 mm, and electrocautery in Calot's triangle as independent risk factors (all P<0.05). The regression equation was: Logit (P)=-9.126+1.362×gallbladder neck stones +0.784×CBD diameter+1.695×adhesion to surrounding organs +1.108×electrocautery+0.895×Calot adhesion +0.679×anatomical variation +0.559×wall thickness. The model showed good discrimination with an AUC of 0.903, sensitivity of 82.72%, and specificity of 84.44%.Conclusion Multiple anatomical and intraoperative factors are associated with bile leakage after LC. The proposed regression model demonstrates good predictive performance and may assist in perioperative risk assessment.