肝脏 ›› 2026, Vol. 31 ›› Issue (9): 1277-1282.

• 肝功能衰竭 • 上一篇    下一篇

HBV相关慢加急性肝衰竭患者发生多重耐药菌感染的列线图预测模型构建

陈泠忻, 李娜, 马从鲜, 唐晓红, 刘翠华   

  1. 621000 绵阳 电子科技大学医学院附属绵阳医院·绵阳市中心医院
  • 收稿日期:2025-11-07 出版日期:2026-09-30 发布日期:2026-11-16
  • 通讯作者: 刘翠华,Email: 1803298604@qq.com
  • 基金资助:
    北京肝胆相照公益基金会人工肝专项基金(iGandanF-1082024-RGG086)

Development of a nomogram prediction model for multidrug-resistant infection in patients with HBV-related acute-on-chronic liver failure

Chen Lingxin, Li Na, Ma Congxian, Tang Xiaohong, Liu Cuihua   

  1. Mianyang Hospital Affiliated to Medical College of University of Electronic Science and Technology·Mianyang Central Hospital , Mianyang 621000,China
  • Received:2025-11-07 Online:2026-09-30 Published:2026-11-16
  • Contact: Liu Cuihua,Email: 1803298604@qq.com

摘要: 目的 构建乙型肝炎病毒相关慢加急性肝衰竭(HBV-ACLF)患者发生多重耐药菌(MDR)感染的列线图预测模型,为临床精准预防 MDR 感染提供循证依据。方法 回顾性选取2021年1月至2025年1月电子科技大学医学院附属绵阳医院、绵阳市中心医院收治的符合欧洲肝病学会及亚太肝病学会(APASL)2019年拟定的HBV-ACLF诊断标准的患者为研究对象,统计住院期间MDR感染发生情况。通过电子病历系统收集患者年龄、性别、体质指数、吸烟史、饮酒史等临床资料,采用多因素logistic回归模型分析HBV-ACLF患者发生MDR感染的危险因素并建立列线图;采用校准曲线、受试者工作特征(ROC)曲线、决策曲线分析评价预测效果。结果 基线纳入295例HBV-ACLF患者,男性182例(61.69%),女性113例(38.31%),发生MDR感染56例,发病率为18.98%。多因素logistic回归分析结果显示:高中国重型乙型肝炎研究组-慢加急性肝衰竭(COSSH-ACLF)评分(OR=2.781,95%CI:1.539~5.027)、广谱抗生素使用 (≥7 d ,OR=2.221,95%CI:1.256~3.929)、碳青霉烯类抗生素暴露(是,OR=1.827,95%CI:1.104~3.024) 、糖皮质激素治疗(是,OR=1.505,95%CI:1.039~2.180)、多器官衰竭(是,OR=1.424,95%CI:1.033~1.965)是HBV-ACLF患者发生MDR感染的危险因素(P<0.05)。建立的列线图预测模型经验证:校准曲线显示预测概率与实际感染率高度吻合,平均绝对误差为0.009,C-index 指数为0.885;ROC曲线下面积为0.885(95%CI:0.698~0.783),灵敏度为0.857,特异度为0.899;决策曲线分析结果显示,风险阈值概率为0.05~0.40时,净收益较高。结论 本研究构建的MDR列线图预测模型有较好的校准度、区分度和临床实用性,可用于HBV-ACLF患者MDR风险预测。

关键词: 乙型肝炎病毒相关慢加急性肝衰竭, 多重耐药菌感染, 列线图预测模型

Abstract: Objective This study aimed to develop and validate a nomogram for predicting multidrug-resistant organism (MDR) infection in patients with hepatitis B virus-related acute-on-chronic liver failure (HBV-ACLF), so as to provide an evidence-based tool for the targeted prevention of MDR infections. Methods A retrospective study was conducted on patients with HBV-ACLF (diagnosed according to the EASL and APASL 2019 criteria), admitted to Mianyang Hospital and Mianyang Central Hospital between January 2021 and January 2025. MDR infections occurring during hospitalization were assessed. Clinical data, including age, gender, BMI, smoking, and alcohol history, were collected from electronic medical records. Multivariable logistic regression was used to identify independent risk factors for MDR infection and to construct a nomogram. The model's performance was assessed using calibration curves, the receiver operating characteristic (ROC) curve, and decision curve analysis (DCA). Results The study included 295 HBV-ACLF patients; 182 (61.69%) were male and 113 (38.31%) were female. MDR infection occurred in 56 patients, with an incidence of 18.98%. Multivariable analysis identified the following independent risk factors (all P<0.05): high COSSH-ACLF score (OR=2.781, 95% CI: 1.539~5.027), use of broad-spectrum antibiotics for ≥7 days (OR=2.221, 95% CI: 1.256~3.929), exposure to carbapenems (OR=1.827, 95% CI: 1.104~3.024), glucocorticoid therapy (OR=1.505, 95% CI: 1.039~2.180), and multiple organ failure (OR=1.424, 95% CI: 1.033~1.965). The developed nomogram demonstrated good performance: the calibration curve showed high agreement between predicted and actual probabilities (mean absolute error = 0.009), the C-index was 0.885, and the area under the ROC curve was 0.885 (95% CI: 0.698~0.783) with a sensitivity of 0.857 and specificity of 0.899. Decision curve analysis indicated a superior net benefit within a risk threshold probability range of 0.05 to 0.40. Conclusion The constructed nomogram provides a reliable tool with good calibration, discrimination, and potential clinical utility for predicting the risk of MDR infection in patients with HBV-ACLF.

Key words: HBV-related acute-on-chronic liver failure, Multidrug-resistant infection, Nomogram prediction model