Kim JD, Cho EJ, Ahn C, et al. - Researchers sought to create a model to foretell outcome in patients with acute liver failure (ALF) caused by hepatitis A. They used a derivation set comprising 294 patients diagnosed with hepatitis A–related ALF (ALFA) from Korea, as well as a validation set including 56 patients from Japan, India, and United Kingdom. Incorporating age, international normalized ratio, bilirubin, ammonia, creatinine, and hemoglobin levels obtained on the day of ALF diagnosis, they constructed a risk-prediction model (ALFA score) using a multivariate proportional hazard model. Compared with King’s College criteria, Model for End-Stage Liver Disease (MELD), and MELD including sodium, a significantly better performance was displayed by the ALFA score in the validation set; however, the performance was not statistically significant than that of the Acute Liver Failure Study Group (ALFSG) index specific for hepatitis A virus and new ALFSG indices. A good calibration of the model was observed in both sets. Overall, a refined prediction of outcome in this patient population was provided by this disease-specific score.