Assessment Of The Predictive Value Of The Diagnostic Model Based On The Indocyanine Green Clearance Test For Long-term Prognosis Of Liver Failure

Objective: To analyze the correlation between the indocyanine green retention rate at 15 minutes(ICG R15)and two-year mortality rate of patients with liver failure.And to investigate the long-term prognostic value of the ICG R15-MELD model established on ICGR15 and the model for end-stage liver disease(MELD)in patients with liver failure.Methods: The clinical data of 126 patients diagnosed with liver failure were retrospectively analyzed.The ICG R15 of the 126 patients were determined within 48 hours of admission.At the same time,the first clinical and laboratory indexes of these patients,Child-Turcotte-Pugh(CTP)classification and MELD score were collected for analysis.All statistical analyses were performed using SPSS version 18 or Stata statistical software version 8.0.Measurement data were assessed by t-test and count data by the chi-square test.Fisher's linear discriminant function was used to establish the ICG R15-MELD model.Long-term predictive accuracy for patients with liver failure was compared between different models using the area under the receiver operating characteristic(ROC)curve(AUC).Cut-off value of diagnosis of fatality was determined with ROC curve.A p-value of <0.05 was used to indicate statistical significance.Results: The two-year mortality rate of all patients was 62.70%(79/126).Comparison of the survivors versus the non-survivors showed that age,CTP classification,PT,MELD score and ICG R15 were significantly different between the groups(all,P<0.05).The death group had significantly higher PT,MELD score and ICG R15 than the survival group(PT: 33.46±13.07 s vs 25.27±6.32;MELD score: 28.69±6.52 vs 22.73±4.60;ICG R15: 61.03±11.18% vs 46.19±8.55%;all,P<0.05).The ICG R15-MELD model constructed by Fisher's linear discriminant function was: F=0.526×ICG R15+0.716×MELD-27.005.The area under ROC curve(AUC)of the ICG R15-MELD model was 0.909(95%CI: 0.860,0.959)and the critical value was 19.39,with the sensitivity of 79.7% and the specificity of 91.5%.The ICG R15-MELD model had the biggest AUC,which was significantly better than CTP classification(AUC: 0.757)and PT(AUC:0.706),and was also superior to MELD score(AUC:0.786)and ICG R15(AUC:0.857)(all,P<0.05).Conclusion: ICG R15 and MELD score had good predictive values for liver failure,better than CTP classification or PT.Furthermore,the ICG R15-MELD model was superior to any single parameter model for predicting the long-term prognosis of patients with liver failure.