Analysis Of Risk Factors Of Death In Patients With Cirrhosis And Hepatorenal Syndrome And Establishment Of Prediction Model Based On Histogram

Objective:This topic collects clinical data of patients with HRS and liver cirrhosis diagnosed in Shanxi Provincial People’s Hospital,analyzing general data,laboratory indicators,diagnosis,Child-pugh score,and complications of liver cirrhosis,further explores risk factors for death.Establishing a nomograph model,we validate it to provide a basis for prediction.Methods:Clinical data of patients with liver cirrhosis and HRS who were definitely diagnosed in our hospital during the period from January 1,2015 to January 1,2022 were collected,and they were screened to meet the inclusion criteria.After being excluded by the exclusion criteria,a total of 286 subjects were finally studied.They were divided into a modeling group and a validation group using a 7:3 ratio,with 200 cases in the modeling group and 86 cases in the validation group.They were divided into a survival group(166 cases)and a death group(34 cases)based on the outcome at discharge.The data of patients in the modeling group were compared between the death group and the survival group,and univariate and multivariate logistic regression analysis was performed.On this basis,a nomogram model was established and validated using a validation group.Results:1.General data characteristics:The etiology of liver cirrhosis is complex,with hepatitis B cirrhosis accounting for the largest proportion.2.Comparison between the death group and the survival group:There was no statistical difference in gender and age between the two groups;In terms of laboratory indicators,a total of 12 items showed statistical differences(P<0.05).WBC,MCV,PT,INR,TT,AST,TBIL,UA,SCR,BUN,K~+in the death group were higher than those in the survivors,while ALB was the opposite.There were significant differences in diagnosis and Child-pugh score between the two groups(P<0.05).In terms of complications,there were statistically significant differences between the two groups in ascites,HE,splenomegaly,esophageal and gastric varices,and SBP(P<0.05).There was no statistically significant difference in gastrointestinal bleeding.3.Univariate and multivariate logistic regression analysis:Relevant factors were included in the univariate logistic regression analysis.WBC,RBC,MCV,MCH,PT,INR,APTT,TT,ALT,AST,ALB,TBIL,UA,BUN,SCR,K+,ascites,HE,splenomegaly,esophageal and gastric varices,SBP,HRS-AKI,and HRS-NAKI were all related factors for mortality in patients(P<0.05).A single factor with statistical differences was included in the multivariate logistic regression model,and the results showed that there were 6 factors that ultimately entered the model.Patients with HRS-AKI and HRS-NAKI are risk factors for death,and the risk of death will increase 6.333 times and 5.946 times,respectively.Among the complications of liver cirrhosis,hepatic encephalopathy and spontaneous peritonitis are independent risk factors for death.The risk of death from hepatic encephalopathy is 1.501times higher,and the risk of death from spontaneous peritonitis is 4.291 times higher.In terms of clinical indicators,every 1 increase in urea nitrogenμMol/L,the risk of death will increase 1.117 times;Each unit increase in INR increases the risk of death by 3.277 times(P<0.05).4.Based on the logistic regression results,a nomograph prediction model was established using R language software,and a calibration curve of the model was drawn using R language based on the validation group data.There was good consistency between the actual value of the death of the enrolled patients and the predicted value of the model.The ROC curve of the nomogram was drawn,with an area under the curve(AUC)of 0.864,a95%CI of(0.777-0.950),a sensitivity of 0.801,and a specificity of 0.805.This indicates that the nomogram model has good discrimination,providing an intuitive and convenient tool for clinicians to assess the risk of death.Conclusion:1.The proportion of hepatitis B virus infection in patients with liver cirrhosis is the highest.2.Elevated HRS-AKI,HRS-NAKI,HE,SBP,INR,and BUN are risk factors for death in patients with liver cirrhosis.The nomograph model established on this basis has high accuracy and sensitivity,providing an intuitive and convenient tool for predicting patient mortality.