Predictors of mortality in hepatic encephalopathy in acute and chronic liver disease: a preliminary observation.

ABSTRACT

BACKGROUND: Several scoring systems are available to predict the outcome of liver cell failure. Scarce information is available on predictors in hepatic encephalopathy. OBJECTIVES: To study clinical and biochemical variables that would predict the outcome in hepatic encephalopathy. METHODS: Fifty consecutive patients with hepatic encephalopathy were included in the study. Variables included clinical and biochemical parameters, discriminant function, QTc interval and the need for ventilator support. Child-Pugh's Turcotte score and Mayo Clinic model for end-stage liver disease scores were calculated at admission. Patients were followed up until discharge or death. Logistic regression analysis was computed with the variables that predicted a favorable outcome. RESULTS: Chronic liver disease precipitated hepatic encephalopathy in 39 patients (group 1) and encephalopathy followed acute liver disease in 11 patients (group 2). In group 1, high serum bilirubin (P<0.001), prolonged QTc interval (P<0.05) and requirement for support systems (P<0.003) predicted a poor outcome. In group 2, higher grades of encephalopathy (P<0.04) and native drug therapy (P<0.007), high serum bilirubin (P<0.05), requirement for support systems (P<0.02) predicted a poor outcome. Mayo Clinic model for end-stage liver disease and discriminant function in both groups and Child-Pugh-Turcotte's score in group 1 did not predict the outcome. Logistic regression identified serum bilirubin in group 1 (OR 8.55, P=0.012) and native drug therapy in group 2 (odds ratio 3.85, P=0.05) as independent poor risk factors. CONCLUSIONS: High serum bilirubin values in chronic liver disease and native drug therapy in acute liver cell failure are simple parameters that would predict a poor outcome in patients with hepatic encephalopathy.