The study focused on patients with acute to chronic liver failure caused by the hepatitis B virus (HBV-ACLF) and aimed to predict the 28-day risk of death. They included 159 patients in the research, of which 113 in the training and 46 in the validation cohort. Using statistical methods (LASSO and multivariate logistic regression), they selected three independent risk factors: the value of INR, creatinine (Cr) and the inflammatory cytokine interleukin-6 (IL-6), all in logarithmic form. Based on these variables, they created a new prognostic model – a nomogram labeled ICI. The ICI model had good ability to discriminate between survivors and non-survivors, with an AUC of 0.826 in the training and 0.814 in the validation cohort. Calibration graphs showed very good agreement between predicted and actual 28-day mortality. Decision curve analysis indicated a higher net clinical benefit of the ICI model over a wide range of threshold probabilities. When compared with commonly used scoring systems (COSSH-ACLF II, CLIF-C ACLF, AARC, CLIF-OFs, MELD, CTP), the new model showed better discrimination and higher clinical benefit.