实用肝脏病杂志 ›› 2015, Vol. 18 ›› Issue (4): 387-390.doi: 10.3969/j.issn.1672-5069.2015.04.014

• 肝硬化 • 上一篇    下一篇

Logistic回归分析食管静脉曲张破裂出血的危险因素*

戴光荣,谭玉娥,刘文娜,冯义朝,宁涛   

  1. 710061 西安市 西安交通大学医学院 /716000 陕西省延安市 延安大学附属医院消化内科(戴光荣);消化科(谭玉娥,冯义朝);消化学专业研究生(刘文娜);外科学专业研究生(宁涛)
  • 收稿日期:2014-12-24 出版日期:2015-07-10 发布日期:2016-02-19
  • 作者简介:戴光荣,女,西安交通大学医学院博士研究生,主任医师,延安大学附属医院消化科二病区主任。主要从事肝脏疾病的诊治研究。E-mail:daiguangrong6810@sina.cn
  • 基金资助:
    延安大学附属医院研究生创新基金项目(10YJ11)

Logistic regression analysis on risk factors of esophageal variceal bleeding in patients with uncompensated liver cirrhosis

Dai Guangrong,Tan Yu’e,Liu Wenna,et al.   

  1. Department of Gastroenterology,Affiliated Hospital, Yan’an University,Yan’an 716000,Shaanxi Province,China
  • Received:2014-12-24 Online:2015-07-10 Published:2016-02-19

摘要: 目的 探讨与肝硬化食管静脉曲张破裂出血(EVB)相关的危险因素,寻找一种有效预测EVB发生的无创检测指标。方法 回顾性分析168例肝硬化患者的临床资料,其中出血85例,非出血83例,对可能影响EVB发生的临床指标进行分析,将有显著性差异的因素进行单因素Logistic回归分析,然后对两组中有统计学意义的因素进行多因素非条件Logistic回归分析。结果 两组患者性别、年龄、病因分布差异均无统计学意义(P>0.05);出血患者门静脉直径(DPV)、脾静脉直径(DSV)和脾脏指数(SI)分别为(1.48±0.28) cm、(1.08±0.19) cm和(72.54±16.69)cm2,显著高于非出血组[分别为(1.34±0.29) cm、(1.18±0.25) cm和(66.98±19.21) cm2,P<0.05];出血患者血Na+和血小板/脾长径比值分别为(139.45±3.41) mmol/L和(5.05±2.07)×109/L?cm-1,显著低于非出血组[分别为(140.66±4.41) mmol/L和(8.99±4.36)×109/L?cm-1,P<0.05];出血组Child-Pugh分级与非出血组比较,差异有统计学意义(P<0.05);单因素Logistic回归分析显示两组间DPV、DSV、血小板计数与脾脏长径比值和Child-Pugh分级差异均有统计学意义(OR值分别为6.832、7.283、0.655、2.129,P<0.05),多因素非条件Logistic回归分析显示血小板计数与脾脏长径比值(OR=0.685,P=0.000)是肝硬化患者发生EVB的独立危险因素。结论 血小板计数与脾脏长径比值可作为肝硬化患者发生EVB的无创预测指标。

关键词: 肝硬化, 食管静脉曲张破裂出血, 危险因素, Logistic回归分析

Abstract: Objective To discuss the related risk factors of esophageal variceal bleeding(EVB)in patients with liver cirrhosis and to find out an effective noninvasive index for predicting EVB. Methods A retrospective analysis of clinical data of 168 patients with liver cirrhosis were carried out,and 85 patients with and 83 without EVB were included. The clinical data that may affect the EVB were analyzed by Logistic regression analysis. Results The age,gender and causses of liver cirrhosis in the two groups had no statistically significant difference (P>0.05);The diameter of portal vein (DPV),diameter of splenic vein (DSV) and splenic index (SI) in patients with EVB were (1.48±0.28)cm,(1.08±0.19)cm and (72.54±16.69) cm2,respectively,significantly higher than those in without [(1.34±0.29)cm,(1.18±0.25)cm and (66.98±19.21) cm2,respectively,P<0.05];Serum Na+ and ratio of platelet counts/splenic length in patients with EVB were (139.45±3.41)mmol/L and(5.05±2.07)×109/L?cm-1,respectively,significantly lower than in without [(140.66±4.41) mmol/L and(8.99±4.36)×109/L?cm-1,respectively,P<0.05];The Child-Pugh class between the two groups was significantly different(P<0.05);Logistic regression analysis showed that DPV,DSV,ratio of platelet counts/splenic length and Child-Pugh classification between the two groups were statistically significant(OR=6.832,7.283,0.655,2.129,respectively,P<0.05),and multivariate and unconditioned Logistic regression analysis showed that only ratio of platelet counts/splenic length (OR=0.685,P=0.000)was an independent risk factor for EVB in patients with cirrhosis. Conclusion Ratio of platelet counts/splenic length might be used as an predictive index for EVB in patients with cirrhosis.

Key words: Liver cirrhosis, Esophageal variceal bleeding, Risk factors, Logistic regression