肝脏 ›› 2026, Vol. 31 ›› Issue (9): 1253-1257.

• 肝纤维化及肝硬化 • 上一篇    下一篇

构建预测肝硬化患者隐性肝性脑病首次发作风险的列线图

顾征太, 刘炜炜, 顾凌云, 马陈斌   

  1. 226500 南通 如皋市人民医院感染性疾病科(顾征太,刘炜炜,马陈斌),结核门诊(顾凌云)
  • 收稿日期:2025-09-07 出版日期:2026-09-30 发布日期:2026-11-16
  • 通讯作者: 马陈斌,Email: mcb321@163.com
  • 基金资助:
    南通市科技计划项目(MSZ19122)

A construction of a nomogram model to predict the risk of the first episode of occult hepatic encephalopathy in cirrhotic patients

Gu Zhengtai1, Liu Weiwei1, Gu Lingyun2, Ma Chenbin1   

  1. 1. Department of Infectious Diseases, Rugao Hospital Affiliated to Nantong University, Nantong 226500, China;
    2. Tuberculosis Clinic, Rugao Hospital Affiliated to Nantong University,Nantong 226500, China
  • Received:2025-09-07 Online:2026-09-30 Published:2026-11-16
  • Contact: Ma Chenbin, Email: mcb321@163.com

摘要: 目的 构建预测肝硬化患者隐性肝性脑病(CHE)首次发作风险的列线图。方法 选择如皋市人民医院2021年5月—2023年5月间诊疗的208例肝硬化患者。采用多因素logistic回归分析评估肝硬化患者CHE首次发作的独立危险因素。将独立危险因素作为构建列线图的预测指标,并采用内部数据集进行校正曲线、决策曲线和临床影响曲线分析。结果 208例患者中,CHE首次发作患病率为41.3%。CHE患者糖尿病、腹水、Child-Pugh分级(B级)比例高于无CHE患者(58.1%比35.0%;37.0%比22.0%;39.5%比22.1%),差异有统计学意义(P<0.05)。CHE患者年龄(68.5±5.9)岁、握力(21.89±3.36)kg、终末期肝病模型评分(9.57±5.38)分、白蛋白-胆红素评分(-2.07±0.48)分和PT(16.95±1.98)s,无CHE患者则分别为:(64.2±5.3)岁、(23.29±3.99)kg、(7.12±4.36)分、(-2.40±0.52)分、(15.77±1.37)s。CHE患者血清尿素氮(7.11±1.93)mmol/L、血小板计数(100.56±40.54)×109/L和白蛋白(30.54±3.76)g/L,无CHE患者则分别为:(6.40±2.48)mmol/L、(128.54±46.53)×109/L、(33.32±3.72)g/L,差异有统计学意义(P<0.05)。多因素logistic回归分析显示:年龄、糖尿病、握力、白蛋白、血清尿素氮与肝硬化患者CHE首次发作风险独立相关(OR=1.628、2.072、3.326、0.711、1.426,P<0.05)。列线图校正曲线显示C-index为0.906,具有较高预测准确性;决策曲线分析结果显示列线图的预测风险阈值区间为0.153~0.876,能提供意义显著的临床标准化净收益;临床影响曲线分析显示预测模型阈值概率大于40%时,列线图预测准确性接近实际观察结果。结论 本研究基于肝硬化患者年龄、糖尿病、握力、血清尿素氮和白蛋白构建的列线图能有效定量肝硬化患者CHE首次发作风险,为医护人员评估患者状况和预后情况提供方法。

关键词: 肝硬化, 隐性, 肝性脑病, 列线图

Abstract: Objective To construct a nomogram model in order to predict the risk of the first episode of occult hepatic encephalopathy (CHE) in cirrhotic patients. Methods A total of 208 patients with cirrhosis who were admitted in Rugao Hospital between May 2021 and May 2023 were selected. The independent risk factors for first episode of CHE in cirrhotic patients were analyzed using multifactorial logistic regression. A predictive nomogram were further established based on these risk factors, and the internal data set was used for the calibration curve, decision curve and clinical impact curve analyses. Results Among the 208 patients, the prevalence rate of first-episode of covert hepatic encephalopathy (CHE) was 41.3%. Compared to non-CHE patients, CHE patients showed significantly higher proportions of diabetes (58.1% vs. 35.0%), ascites (37.0% vs.22.0%), and Child-Pugh grade B (39.5% vs. 22.1%), with statistically significant differences (P<0.05). The average age,handgrip strength, model for end-stage liver disease score, albumin-bilirubin score, and prothrombin time were (68.5±5.9) years, (21.89±3.36) kg, (9.57±5.38), (-2.07±0.48), (16.95±1.98) s for CHE patients, and (64.2±5.3) years, (23.29±3.99) kg, 7.12±4.36, -2.40±0.52, and (15.77±1.37) s for non-CHE patients, respectively. The blood urea nitrogen, platelet count, and albumin level were (7.11±1.93) mmol/L, (100.56±40.54) × 109/L, and (30.54±3.76) g/L in CHE patients, respectively, which were significantly higher than those of (6.40±2.48) mmol/L, (128.54±46.53) × 109/L, and (33.32±3.72) g/L in non-CHE patients (all P<0.05). Multivariate logistic regression analysis identified age, diabetes, handgrip strength, albumin, and blood urea nitrogen as independent risk factors for first-episode of CHE in cirrhotic patients (OR=1.628, 2.072, 3.326, 0.711, and 1.426, respectively; P<0.05). The calibration curve of the nomogram demonstrated excellent predictive accuracy with a C-index of 0.906. Decision curve analysis revealed that the nomogram provided clinically meaningful standardized net benefits within a risk threshold range of 0.153~0.876. The clinical impact curve analysis indicated that when the predicted probability threshold exceeded 40%, the nomogram's predictive accuracy closely approximated the actual observed outcomes. Conclusion In this study, a nomogram model constructed based on age, diabetes, handgrip strength, albumin, and blood urea nitrogen in cirrhotic patients was effective in quantifying the risk of a first episode of CHE in cirrhotic patients, which provided a method for healthcare professionals to assess the patient's condition and prognosis.

Key words: Liver cirrhosis, Cryptic, Hepatic encephalopathy, Nomogram