Development and validation of a nomogram for predicting 28-day mortality in critically ill liver cirrhotic patients with coagulopathy

Although scoring systems exist to predict outcomes in patients with cirrhosis, specific systems for critically ill cirrhotic patients with coagulopathy are lacking. The aim of the study was to establish a reliable predictive model for 28-day mortality in cirrhotic patients with coagulopathy admitted to the intensive care unit. The 880 cirrhotic patients with coagulopathy admitted to the intensive care unit from 2008 to 2022 in the Medical Information Mart for Intensive Care-IV database were randomly allocated to training ( n = 616) and validation ( n = 264) cohorts. Independent predictors of 28-day mortality were identified by logistic regression analysis and used to construct a nomogram. Age, white blood cell count, activated partial thromboplastin time, prothrombin time, total bilirubin level, blood lactate level, and blood bicarbonate level were identified as independent predictors of 28-day mortality. The nomogram showed good discrimination in the training (concordance index, 0.885; 95% confidence interval, 0.855–0.915) and validation (concordance index, 0.909; 95% confidence interval, 0.871–0.948) cohorts. Calibration, receiver operating characteristic, and clinical decision curves showed that the nomogram had good predictive ability. The nomogram exhibited superior discernment and accuracy to SOFA, MELD3.0, ABIC scores and may help identify high-risk patients to allow early intervention and potentially improve clinical outcomes.