ERCP后胰腺炎的危险因素分析及风险预测模型构建

Risk factors for pancreatitis following ERCP and development of its risk prediction model

摘要:

背景与目的 内镜逆行胰胆管造影术（ERCP）后胰腺炎（PEP）是最常见且影响预后的主要并发症，其发生增加住院时间及医疗负担。尽管已有部分危险因素被识别，但临床上仍缺乏基于常规变量、便于应用的风险评估工具。本研究旨在分析PEP的危险因素，并构建实用的风险预测模型。方法 回顾性分析2020年1月—2025年6月于内蒙古医科大学赤峰市医院537例行ERCP患者的临床资料，按是否发生PEP分组。采用单因素及多因素Logistic回归筛选独立危险因素，并据此建立列线图模型。通过受试者工作特征（ROC）曲线及校准曲线评价模型性能。结果 537例患者中93例（17.3%）发生PEP。多因素分析显示，高血压史（OR=2.014）、胰管显影（OR=4.663）、十二指肠乳头球囊扩张术（OR=2.459）及插管时间>57 min（OR=3.530）为PEP的独立危险因素（均P<0.05）。基于上述变量构建的列线图模型具有良好的区分度（AUC=0.750）及校准度（Hosmer-Lemeshow检验P=0.793）。在接受十二指肠乳头球囊扩张术患者中，放置与未放置胰管支架患者的PEP发生率差异无统计学意义（P>0.05）。结论 高血压史、胰管显影、十二指肠乳头球囊扩张术及插管时间延长与PEP的发生密切相关。所建立的预测模型具有较好的预测效能，可用于ERCP患者围手术期风险分层及个体化预防策略制定。

Abstract:

Background and Aims Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) is the most common complication and is associated with prolonged hospitalization and increased healthcare burden. Although several risk factors have been identified, a simple and clinically applicable prediction tool based on routine variables remains lacking. This study aimed to identify risk factors for PEP and to develop a practical prediction model.Methods A retrospective analysis was conducted on 537 patients who underwent ERCP between January 2020 and June 2025. Patients were divided into PEP and non-PEP groups. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors. A nomogram was constructed and evaluated using receiver operating characteristic (ROC) and calibration curves.Results PEP occurred in 93 patients (17.3%). Multivariate analysis identified a history of hypertension (OR=2.014), pancreatic duct opacification (OR=4.663), endoscopic papillary balloon dilation (OR=2.459), and cannulation time >57 min (OR=3.530) as independent risk factors (all P<0.05). The nomogram demonstrated good discrimination (AUC=0.750) and calibration (Hosmer-Lemeshow test, P=0.793). Among patients undergoing endoscopic papillary balloon dilation, there was no significant difference in the incidence of PEP between those with and without pancreatic duct stent placement (P>0.05).Conclusion A history of hypertension, pancreatic duct opacification, endoscopic papillary balloon dilation, and prolonged cannulation time are closely associated with the occurrence of PEP. The proposed nomogram provides a simple and effective tool for individualized risk prediction and may assist in optimizing preventive strategies for ERCP patients.