Background and aims: Liver cirrhosis is associated with a very bad prognosis. As knowledge of prognostic markers are clinical relevant, we have in this essay studied mortality of patients with cirrhosis and explored predictors of death.
Patients and methods: This essay is part of a retrospective study of patients with cirrhosis that was performed in 2006 at Aker University hospital in Oslo. 174 patients who suffered from cirrhosis according to our criteria were included in the study. These patients were identified by performing search in the hospital`s computersystem for 1) diagnoses indicating fibrosis or cirrhosis, 2) diagnoses potentially indicating chronic liver disease or 3) complications potentially caused by cirrhosis. By univariate analyses, age, sex, ethiology, different biochemical variables, Child-Pugh score, Model of End-Stage Liver Disease (MELD)-score, complications of cirrhosis and some comorbid conditions were explored as possible predictors of death.
Results: Age, Child-Pugh and MELD-score were significant predictors of death. High ASAT/ALAT-ratio (>1.81) and high ALP (>3.13 U/L) were also identified as predictors of death. However, separate analyses of patients with or without alcoholic liver disease (ALD), revealed these biochemical variables to be predictive of poor survival only among patients without ALD. Sex, ethiology and presence of ascites or varices had no impact on survival in this study.
Conclusion: Classical risk factors of death among patients with cirrhosis were identified in this study. In addition ASAT/ALAT-ratio and ALP seem to be strong predictors of death among patients without ALD. Multivariate analyses is needed to investigate whether these markers offer independent prognostic information.