Predicting hospital mortality in cirrhotic patients: comparison of Child-Pugh and Acute Physiology, Age and Chronic Health Evaluation (APACHE III) scoring systems

The severity of hepatic abnormalities and extent of dysfunction of other organ systems influences prognosis for cirrhosis. The Child-Pugh system has been used to classify cirrhotic patients into good, intermediate, or poor risk categories in evaluation and therapy. Disregard for cardiorespiratory, renal, electrolyte balance, and acid base status limits its predictive accuracy. We evaluated the accuracy of Acute Physiology and Chronic Health Evaluation (APACHE III) to predict short term hospital mortality in patients with liver cirrhosis.A total of 282 patients were prospectively enrolled. Child-Pugh and APACHE III scores were recorded on day 1 for each patient.Mean age was 51.7+/-11.3 yr, with 65% men and 35% women; 57% presented with upper GI bleeding, 47% encephalopathy, 9% hepatorenal syndrome, and 7% hepatocellular carcinoma. Sixty-three patients (22%) died. Major causes of death were upper GI bleeding 38%, liver failure 21%, hepatorenal syndrome 19%, hepatocellular carcinoma 4%, and spontaneous bacterial peritonitis 6%. Child-Pugh and APACHE III scores for survivors (8.6+/-2.3 and 58.9+/-35.1) were lower than those for nonsurvivors (10.9+/-2.7 and 87.4+/-30.3) (p0.001). Using discriminant analysis, APACHE III correctly identified 75% of cases vs 67% of cases for Child-Pugh (p0.05). When information regarding ascites and prothrombin time was added to APACHE III, 81% of cases were correctly classified.The APACHE III scoring system is superior to Child-Pugh for prognosticating short term survival of cirrhotic patients. Prognostic accuracy of APACHE III can be enhanced by incorporating information regarding ascites and prothrombin time prolongation.