慢加急性肝衰竭合并细菌感染患者抗感染治疗72小时的效果及影响因素

Clinical outcome of patients with acute-on-chronic liver failure and bacterial infection after 72 hours of anti-infective therapy

目的探讨入院时合并细菌感染的慢加急性肝衰竭(ACLF)患者72 h抗感染疗效及影响因素。方法运用电子数据库选取解放军总医院第五医学中心2014年1月-2016年3月收治的267例入院时合并细菌感染的ACLF患者,分析患者的临床特征,根据72 h抗感染疗效将患者分为有效组(n=156)和无效组(n=111)。正态分布计量资料2组间比较采用t检验,非正态分布计量资料2组间比较使用Mann-Whitney U秩和检验。分类资料2组间比较使用χ2检验。应用logistic回归分析影响患者72 h抗感染疗效的危险因素并建立预测模型,利用受试者工作特征曲线(ROC曲线)评价模型的诊断价值。结果 HBV感染(161例)、酒精性肝病(60例)是最常见病因;共发生感染性休克29例。发生312例次细菌感染,腹水(199例)、肺部(77例)是最常见感染部位。培养菌株49株,大肠埃希菌(16株)、肺炎克雷伯杆菌(13株)是最常见菌株。合并真菌感染20例。156例(58. 4%)患者抗感染治疗有效。有效组急性肾损伤发生率、肝性脑病发生率、感染性休克发生率、肺部感染率、合并真菌感染率、WBC、中性粒细胞计数、ALT、AST、ALP、血肌酐、国际标准化比值、MELD分值均低于无效组(P值均<0. 05),腹水感染率、凝血酶原活动度均高于无效组(P值均<0. 05)。有效组30 d病死率、90 d病死率均低于无效组(15. 2%vs 77. 5%,χ2=97. 942,P <0. 001; 35. 2%vs 91. 7%,χ2=71. 977,P <0. 001)。多变量logistic回归显示基线INR、存在感染性休克是影响患者抗感染疗效的危险因素,新预测模型为logistic(P)=-5. 907+2. 260×感染性休克(有=1,无=0)+1. 603×INR,ROC曲线下面积为0. 746,cut-off值-2. 3,敏感度62. 2%,特异度78. 2%,新模型的诊断价值高于基线MELD分值(Delong检验,Z=2. 402,P=0. 016)。结论 INR、感染性休克是预测ACLF合并细菌感染患者72 h抗感染疗效的独立影响因素。

Objective To investigate the clinical outcome of patients with acute-on-chronic liver failure(ACLF)and bacterial infection on admission after 72 hours of anti-infective therapy and related influencing factors.Methods An electronic database was used to select 267 patients with ACLF and bacterial infection on admission who were admitted to the fifth Medical center of Chinese PLA General Hospital from January 2014 to March 2016,and their clinical features were analyzed.According to the outcome after 72 hours of anti-infective therapy,these patients were divided into effective group and ineffective group.The t-test was used for comparison of normally distributed continuous data between two groups,and the Mann-Whitney U rank sum test was used for comparison of non-normally distributed continuous data between two groups.The chi-square test was used for comparison of categorical data between two groups.A logistic regression analysis was used to identify the risk factors for outcome after 72 hours of anti-infective therapy and establish a predictive model,and the receiver operating characteristic(ROC)curve was used to evaluate the diagnostic value of the new model.Results Hepatitis B virus infection observed in 161 patients and alcoholic liver disease observed in 60 patients were the most common causes,and 29 patients experienced septic shock.There were 312 case-times of bacterial infection,with ascites(199 patients)and the lung(77 patients)as the most common infection sites.A total of 49 strains were cultured,among which Escherichia coli(16 strains)and Klebsiella pneumoniae(13 strains)were the most common strains.A total of 20 patients also had fungal infection.Of all patients,156(58.4%)had response to anti-infective therapy.Compared with the ineffective group,the effective group had significantly lower incidence rates of acute kidney injury,hepatic encephalopathy,and septic shock,pulmonary infection rate,fungal infection rate,white blood cell count,neutrophil count,alanine aminotransferase,aspartate aminotransferase,alkaline phosphatase,serum creatinine,international normalized ratio(INR),and Model for End-Stage Liver Disease(MELD)score(all P<0.05)and significantly higher ascites infection rate and prothrombin time activity(both P<0.05).Compared with the ineffective group,the effective group had significantly lower 30-day mortality rate(15.2%vs 77.5%,χ^2=97.942,P<0.001)and 90-day mortality rate(35.2%vs 91.7%,χ^2=71.977,P<0.001).The multivariate logistic regression analysis showed that baseline INR and the presence of septic shock were the risk factors for the outcome of anti-infective therapy.The new predictive model was logistic(p)=-5.907+2.260×septic shock(yes=1,no=0)+1.603×INR.At the cut-off value of-2.3,this model had an area under the ROC curve of 0.746,with a sensitivity of 62.2%,and a specificity of 78.2%.This new model had a higher diagnostic value than MELD score(Delong test,Z=2.402,P=0.016).Conclusion INR and septic shock were independent influencing factors for predicting the therapeutic outcome of patients with ACLF and bacterial infection after 72 hours of anti-infective therapy.