Validation and analysis of prognostic scoring systems for critically ill patients with cirrhosis admitted to ICU

Abstract

Abstract Introduction The number of patients admitted to ICU who have liver cirrhosis is rising. Current prognostic scoring tools to predict ICU mortality have performed poorly in this group. In previous research from a single centre, a novel scoring tool which modifies the Child-Turcotte Pugh score by adding Lactate concentration, the CTP + L score, is strongly associated with mortality. This study aims to validate the use of the CTP + L scoring tool for predicting ICU mortality in patients admitted to a general ICU with cirrhosis, and to determine significant predictive factors for mortality with this group of patients. This study will also explore the use of the Royal Free Hospital (RFH) score in this cohort. Methods A total of 84 patients admitted to the Glasgow Royal Infirmary ICU between June 2012 and Dec 2013 with cirrhosis were included. An additional cohort of 115 patients was obtained from two ICUs in London (St George’s and St Thomas’) collected between October 2007 and July 2009. Liver specific and general ICU scoring tools were calculated for both cohorts, and compared using area under the receiver operating characteristic (ROC) curves. Independent predictors of ICU mortality were identified by univariate analysis. Multivariate analysis was utilised to determine the most predictive factors affecting mortality within these patient groups. Results Within the Glasgow cohort, independent predictors of ICU mortality were identified as Lactate (p < 0.001), Bilirubin (p = 0.0048), PaO2/FiO2 Ratio (p = 0.032) and PT ratio (p = 0.012). Within the London cohort, independent predictors of ICU mortality were Lactate (p < 0.001), PT ratio (p < 0.001), Bilirubin (p = 0.027), PaO2/FiO2 Ratio (p = 0.0011) and Ascites (p = 0.023). The CTP + L and RFH scoring tools had the highest ROC value in both cohorts examined. Conclusion The CTP + L and RFH scoring tool are validated prognostic scoring tools for predicting ICU mortality in patients admitted to a general ICU with cirrhosis.