Establishment Of Early Diagnosis Model Of Liver Failure With Infection

Objectives:Explore to establish and improve the early diagnosis system of liver failure patients with infections,build a good and simple diagnostic model,and evaluate the predictive effect of the model on the possibility of concurrent infection in patients with liver failure.Methods:1.Retrospective analysis of clinical data of 315 patients with liver failure admitted to the First Affiliated Hospital of Soochow University from January 2015 to June 2018.After screening,179 patients were finally included,and they were divided into infection group and non-infection according to clinical diagnosis.The clinical characteristics,Procalcitonin(PCT)and common laboratory indicators,Infection Probability Score(IPS)score,and Sequential Organ Failure Assessment(SOFA)of the two groups were compared.The t-test or rank sum test was used to compare the measurement data between the two groups;The Chi-square test was used to compare the count data between the two groups.Multivariate logistic regression was used to analyze the influencing factors of infection,and a diagnostic model of PCT combined with IPS score was established.Using receiver operating characteristic curve(ROC curve)to analyze the predictive effect of PCT combined with IPS score on infection.2.From January 2015 to June 2018,liver failure patients admitted to the First Affiliated Hospital of Soochow University were screened and included 179 cases as a model group;A total of 247 patients with liver failure were admitted from the First Affiliated Hospital of Soochow University from July 2018 to October 2019 and the Fifth People's Hospital of Suzhou from January 2018 to October 2019,and 122 patients were finally included as the validation group.According to the clinical diagnosis,the cases in the group were divided into infected group and non-infected group.The model group,select 9 variables related to infection,including age,temperature,neutrophil ratio(NE%),PCT,C-reactive protein(CRP),arterial blood lactic acid(Lac),serum albumin(Alb),and end-stage liver disease model(MELD)scores and SOFA scores were modeled by multivariate logistic regression analysis.The diagnostic model was further validated in the validation group,and its prediction performance and accuracy were analyzed using ROC curves.Result:1.In the model group,179 patients with liver failure,123 cases(68.72%)with concurrent infection,including 99 cases of lung infection(80.49%),49 cases of abdominal infection(39.84%),and 40 cases of infection at two or more sites(32.52%).In the verification group,there were 84 cases(68.85%)of 122 patients with liver failure,mainly 63 cases(75%)of lung infections and 36 cases(42.86%)of abdominal infections,and 50 cases(59.52%)were combined with two or more infections.2.The overall recent hospital mortality of patients with liver failure in the model group was 68.16%,in which the mortality of the infected group(75.61%)was significantly higher than that of the non-infected group.The overall short-term hospital mortality in the verification group was 55.74%,and the mortality in the infected group(67.86%)was also significantly higher than that in the non-infected group.3.In the study to evaluate the predictive value of the procalcitonin combined with infection probability score for patients with liver failure,multivariate logistic regression analysis showed that PCT[Odds Ratio(OR)=3.822,95%confidence interval(95%(CI):1.714?8.523,P=0.001]and IPS score(OR=1.125,95%CI:1.030?1.230,P=0.009)were independent predictors of liver failure combined infection.ROC curve analysis showed that PCT combined with IPS scores had the best predictive ability for liver failure complicated infections[Area under the ROC curve(AUC)=0.857,95%CI:0.801 to 0.913],significantly higher than PCT(AUC=0.803)and IPS The scores(AUC=0.802)were statistically significant(P<0.05).The positive likelihood of PCT combined with IPS score was higher(3.40).and the negative likelihood ratio was the lowest(0.28).4.In exploring the establishment of an early diagnosis scoring model of liver failure complicated with infection,multivariate logistic regression analysis showed that age,body temperature,PCT,CRP,Lac,SOFA scores were independent predictors of liver failure complicated with infection(P<0.05).The ROC curve analysis showed that the established early diagnosis model of infection has good predictive power(AUC=0.899,95%CI:0.846?0.939),and the sensitivity and specificity were 86.2%and 80.4%,respectively.The verification results show that the model has good repeatability(AUC=0.953,95%CI:0.899?0.983),and the sensitivity and specificity are 91.7%and 84.2%,respectively.Conclusion:The mortality of patients with liver failure complicated by infection is extremely high.The main types of infection are lung infection and abdominal infection.Studies have found that both the PCT and IPS scores can predict concurrent infections in patients with liver failure,and the predictive power of the two is equivalent,while the PCT combined with IPS scores have better predictive value than the two.Based on this research,an early diagnosis scoring model for concurrent infection of liver failure was successfully constructed,which has a better predictive power for the early stage of concurrent infection of liver failure,has higher overall accuracy,and shows good repeatability and reliability in patients in different centers in the same area.