Multiple organ system failure in critically ill cirrhotic patients. A comparison of two multiple organ dysfunction/failure scoring systems

Abstract

Objectives: The prognosis for critically ill cirrhotic patients depends on the extent of hepatic and extrahepatic organ dysfunction/failure. We hypothesize that a graded multiple organ dysfunction score, sequential organ failure assessment (SOFA), would provide more descriptive and discriminative power for predicting the hospital mortality for critically ill cirrhotic patients than the classical organ system failure (OSF) score, which defines organ failure as an all-or-none phenomenon.

Methods: 160 patients diagnosed with liver cirrhosis were admitted to the medical intensive care unit (ICU) from January 2002 to June 2003. Information considered necessary for calculating the Child-Pugh, OSF and SOFA scores on ICU admission was collected prospectively.

Results: Hepatitis B infection was the most common cause of liver cirrhosis. A significantly progressive increase in mortality rate was associated with OSF and SOFA scores (p < 0.001). Close correlation between OSF and SOFA scores (p < 0.001) suggested that both systems evaluated the same event. In patients with similar organ dysfunction, the number of failed organ system(s) was significantly higher among non-survivors. However, no correlation existed between the SOFA scores and mortality rate in patients with the same OSF number. Meanwhile, both OSF and SOFA scores displayed excellent discriminative power (areas under receiver-operating characteristic (AUROC) were 0.906 and 0.892, respectively), while Child-Pugh scores clearly performed more poorly (AUROC 0.712). Both OSF and SOFA demonstrate a good fit using the Hosmer and Lemeshow goodness-of-fit test.

Conclusions: Both OSF and SOFA scores are excellent tools for predicting prognosis for cirrhotic patients admitted to ICU. Both of them are superior to Child-Pugh score. Hospital mortality for critically ill cirrhotic patients occurs owing to severe failure of a relatively few organs, rather than because of an accumulation of mild dysfunction in many organ systems. Graded organ dysfunction scales provide no further benefit for predicting hospital mortality for critically ill cirrhotic patients.