The performance of different prognostic scores in cirrhotic patients admitted to intensive care unit

Background and aim Critically ill cirrhotic patients exhibit high in-hospital mortality and require substantial healthcare resources, including multiorgan support. Egypt has the highest age-standardized death rate from cirrhosis worldwide. This study aimed to compare the performance and accuracy of different prognostic scores for mortality prediction, including Child-Turcotte-Pugh (CTP), Model for End-Stage Liver Disease (MELD), MELD with serum sodium incorporation (MELD-Na), Royal Free Hospital (RFH), as well as general intensive care unit (ICU) prognostic models, Acute Physiology and Chronic Health Evaluation II (APACHE II), and Chronic liver failure -sequential organ failure assessment (CLIF-SOFA), in cirrhotic patients admitted to an ICU in an Egyptian tertiary facility. Patients and methods This prospective observational study included ninety patients admitted to the ICU between July 2021 and June 2022. All patients underwent clinical assessment with the estimation of all included scores within the first 24 h of admission, alongside basic laboratory and radiological investigations. Results Mortality occurred in 56.7% (n = 51) of the patients. Non-survivors exhibited statistically significantly higher APACHE II, CLIF-SOFA, RFH, and CTP scores but lower Glasgow Coma Scale (GCS) than survivors. However, no statistically significant difference was found regarding the performance of the scoring systems. At cut-off levels > 20 for APACHE II and > 46 for CLIF-SOFA, these scores were the most independent predictors of in-hospital mortality. Conclusions While CTP, MELD, and MELD-Na scores demonstrated satisfactory performance in critically ill cirrhotic patients, APACHE II and CLIF-SOFA scores were superior in predicting short-term mortality in this patient population.