Liver dysfunction indices as a predictor for in-hospital mortality in acute heart failure patients

Abstract Aim Acute heart failure (AHF) causes significant short and long-term mortality. Thus, to improve the outcomes of AHF patients, risk stratification is mandatory. Our study evaluates the prognostic significance of five liver dysfunction indices: the Albumin/Bilirubin (ALBI) score, the Model for End-Stage Liver Disease Excluding INR (MELD-XI) score, the Model for End-Stage Liver Disease including Albumin (MELD-Alb) score, the liver fibrosis 4 (FIB-4) index, and the liver fibrosis 5 (FIB-5) index, as predictors of all-cause in-hospital mortality in AHF patients. This is the first study, to our knowledge, to compare these five risk scores for the same endpoint. Methods The five scores, as well as the Get With The Guidelines Heart Failure (GWTG-HF) risk score, were calculated at admission for 107 AHF patients followed for all-cause in-hospital mortality. Each score was statistically analyzed as a continuous and categorical (three tertiles) variable using univariate and multivariate logistic regression analysis. To detect the best cutoff value and compare the prognostic accuracy of the scores, a receiver-operating characteristic (ROC) curve was constructed. Results During the median hospitalization period (9 days), 36 (33.6%) patients died. In the fully adjusted logistic regression model, the ALBI, MELD-Alb, FIB-4, and FIB-5 scores were statistically significant as a continuous variable for all-cause in-hospital mortality ( P < 0.05) with a cutoff value − 1.85, 25.5, 2.6, and − 7.92, respectively. Except for the FIB-4 index, they were also significant as a categorical variable ( P < 0.05). The MELD-XI score was not significant for mortality in the adjusted model. In the ROC analysis, the ALBI score demonstrated the highest prognostic accuracy for predicting all-cause in-hospital mortality (AUC = 0.935), followed by the MELD-Alb score (AUC = 0.825) and the FIB-5 index (AUC = 0.759). A model combining ALBI score and GWTG-HF risk score showed a better prognostic accuracy (AUC = 0.946). Conclusion The ALBI score, MELD-Alb score, and FIB-5 index could be useful predictors of all-cause in-hospital mortality in AHF patients, with the ALBI score exhibiting the highest predictive accuracy, and a model that combines the GWTG-HF risk score and ALBI score is of better prognostic value. Further research is needed for validation.