Prognostic significance of clinical and laboratory parameters in liver cirrhosis. A multivariate statistical approach.

Successive clinical and laboratory data of 73 cirrhotics, who died during a follow-up period, were evaluated. The relationships between the actual survival time and 22 potentially prognostic variables were studied by means of the log-rank test and the likelihood ratio test. Seven variables (presence of ascites, hepatic encephalopathy and oesophageal varices; serum albumin, total bilirubin, gamma-globulin percentage and cholesterol) showed a strict correlation with survival time, and were included in a multiple regression model in order to compute the specific "prognostic weight" of each parameter. Results show that the parameters directly related to liver cell biosynthesis or indicating a structural liver damage should be considered more reliable for predicting theoretical survival in patients with liver cirrhosis.