Cirrhotics Treated In Intensive Care Unit Have High Short Term Survival in the Absence of Extrahepatic Organ Dysfunction

Abstract

AIM: This retrospective study of 85 cirrhotics aimed to identify variables during the first 24 hours of intensive care unit (ICU) admission predicting mortality up to 30 days of hospital discharge, and to analyse the prognostic accuracy of common severity scores in predicting mortality. MATERIALS AND METHODS: Eighty-five patients with liver cirrhosis admitted to ICU at the Royal Hobart Hospital, a regional Australian center, from 2007 to 2013 inclusive were identified using International Classification of Disease coding and data extracted from medical records. Predictors of mortality were determined via logistic regression and the prognostic accuracy of 5 severity scores calculated by their area under the receiver-operating curve. These included 2 scores commonly used in liver disease; Child Pugh and Model for End-Stage Liver Disease (MELD); as well as 3 scores designed in the intensive care setting: Sequential Organ Failure Assessment (SOFA), Simplified Acute Physiology (SAPS II) and Acute Physiology and Chronic Health Evaluation (APACHE II). RESULTS: Significant variables predicting short-term mortality included infection (excluding spontaneous bacterial peritonitis), requirement for inotropes or mechanical ventilation, elevated creatinine, decreased Glasgow Coma Scale, decreased pH and elevated white cell count. However the presence of cirrhosis-specific complications such as hepatic encephalopathy, variceal bleeding and spontaneous bacterial peritonitis did not predict mortality and liver-specific prognostic severity scores (Child Pugh and MELD) performed more poorly than the other severity scores designed for the ICU setting. CONCLUSIONS: ICU admission for cirrhotics should not be deemed futile in the presence of hepatic dysfunction alone; cardiorespiratory, neurological and renal dysfunction should be taken into account. ICU-specific severity scores better prognosticate short-term mortality compared to liver-specific scores.