肝脏 ›› 2020, Vol. 25 ›› Issue (10): 1030-1033.

• 肝纤维化及肝硬化 • 上一篇    下一篇

酒精性肝硬化合并自发性细菌性腹膜炎患者长期预后预测模型的建立与验证

王瑞瑞, 江宇泳, 王宪波, 杨晋翔   

  1. 100029 北京中医药大学(王瑞瑞); 北京中医药大学第三附属医院消化科(王瑞瑞,杨晋翔); 首都医科大学附属北京地坛医院中西医结合科(江宇泳,王宪波)
  • 收稿日期:2020-04-20 出版日期:2020-10-31 发布日期:2020-12-18
  • 通讯作者: 杨晋翔,Email:yjx571102@163.com
  • 基金资助:
    国家科技重大专项(2018ZX10725505-001); 北京中医药大学校级自主项目(2019-JYB-XS-247)

A prognostic model for long-term survival in alcoholic liver cirrhosis patients complicated with spontaneous bacterial peritonitis

WANG Rui-rui1,2, JIANG Yu-yong3, WANG Xian-bo3, YANG Jin-xiang2   

  1. 1. Beijing University of Chinese Medicine, Beijing 100029, China;
    2. Gastroenterology Department in The Third Affiliated Hospital of Beijing University of Chinese Medicine, Beijing 100029, China;
    3. Department of Integrated Traditional Chinese and Western Medicine in Beijing Ditan Hospital Affiliated to Capital Medical University, Beijing 100015, China
  • Received:2020-04-20 Online:2020-10-31 Published:2020-12-18
  • Contact: YANG Jin-xiang,Email:yjx571102@163.com

摘要: 目的 探讨酒精性肝硬化(ALC)合并自发性细菌性腹膜炎(SBP)长期预后的影响因素。方法 采集于2016年1月至2019年2月确诊的酒精性肝硬化合并自发性细菌性腹膜炎患者的临床信息,以2018年10月31日为时间节点分为建模组(123例)和验证组(36例),建模组依据随访1年时的结局分为生存组(76例)和死亡组(47例),运用Logistic回归分析得出影响因素并建立预测模型,验证组进行模型验证。结果 年龄(OR:0.935,95% CI:0.883~0.990)、自发性细菌性腹膜炎病史(OR:0.195,95% CI:0.082~0.462)、合并高血压(OR:0.364,95% CI:0.134~0.990)、并发上消化道出血(OR:0.364,95% CI:0.140~0.942)、并发肝肾综合征(OR:0.271,95% CI:0.094~0.777)、Maddery判别函数(OR:0.981,95% CI:0.969~0.994)和白蛋白水平(OR:1.061,95% CI:0.999~1.127)为ALC合并SBP患者预后的独立影响因素(P<0.05);预测模型的ROC曲线下面积为0.805(95% CI:0.727~0.883);内部验证的灵敏度为87.50%,特异度为95.00%,总正确率为91.67%。结论 年龄、既往SBP病史、合并高血压、并发上消化道出血、并发肝肾综合征、Maddery判别函数和白蛋白水平是导致ALC并发SBP患者长期预后差的高危因素;本研究建立的预测模型可较准确地预测患者的长期预后。

关键词: 酒精性肝硬化, 自发性细菌性腹膜炎, 长期预后, 危险因素, 预测模型

Abstract: Objective To investigate the factors affecting the long-term prognosis of patients with alcoholic cirrhosis (ALC) complicated with spontaneous bacterial peritonitis (SBP) and to develop a prognostic model. Methods The clinical information of 159 ALC patients complicated with SBP were collected. They were divided into modeling group (123 cases) and verifying group (36 cases). Patients in the modeling group were subdivided into death group (47 cases) and survival group (76 cases) according to their outcomes after 1-year. Clinical information of all patients was collected and compared. Significant single risk variants were selected for further Logistic regression analysis to obtain the risk factors for building a prognostic model. The resulting model was further validated in the verifying group of patients. Results Logistic regression analysis showed that age(OR=0.935,95% CI:0.883~0.990), history of SBP(OR=0.195,95% CI:0.082~0.462), history of hypertension(OR=0.364,95% CI:0.134~0.990), complicated upper gastrointestinal bleeding(OR=0.364,95% CI:0.140~0.942), complicated with hepatorenal syndrome(OR=0.271,95% CI:0.094~0.777), Maddery discriminant function(OR=0.981,95% CI:0.969~0.994)and albumin level(OR=1.061,95% CI:0.999~1.127) were independent risk factors associated with survival (P<0.05). The area under the ROC curve of the prediction model was 0.805 (95% CI:0.727-0.883).The sensitivity, specificity, and the total accuracy of the model was 87.50%, 95.00%, and 91.67% respectively. Conclusion Age, history of SBP, history of hypertension, complicated with upper gastrointestinal bleeding, complicated with hepatorenal syndrome, Maddery discriminant function and albumin level are the high risk factors for the poor long-term prognosis in ALC patients with SBP. The prediction model established on these factors is accurate for predicting long-term prognosis of ALC patients with SBP.

Key words: Alcoholic liver cirrhosis, Spontaneous bacterial peritonitis, Long-term prognosis, Risk factors, Prognostic model