BACKGROUND: Recent guidelines recommend that all cirrhotics undergo screening upper endoscopy to identifly risk patients for bleeding from varices. The aim of this study was to determine whether clinical and laboratory variables were predictive of the occurrence and progression of esophageal varices. METHODS: This is a retrospective analysis of cirrhotics who had a screening upper endoscopy during 10 years after liver biopsy. Fifty-eight patients were recruited. Univariate/multivariate analysis was used to evaluate associations between the development and progression of esophageal varices and patients characteristics including platelet count, liver chemistry value, prothrombin time, shunt index(heart to liver uptake ratio) through transrectal TI-201 liver scan, probability(P)-value (Y=3.3431-0.8160 x ALT/AST ratio-0.0343 x prothrombin time+2.6963 x shunt index, P=e(y)/(e(y)+1)), ascites, splenomegaly, hepatic encephalopathy. RESULTS: There were 36 patients without esophageal varices or with stable varices during 10 years and 22 patients with new developed esophageal varices or progressive varices during 10 years as determined by upper endoscopy. On multivariate analysis, patients with a probability(P)-value of > or = 0.647 and a platelet count of < or = 100,500/mm3 had a risk of the development and progression of esophageal varices. CONCLUSIONS: Non-endoscopic predictors (probability(P)-value and platelet count) could be used to stratify patients with cirrhosis for the risk of the development and progression of esophageal varices and such stratification could be used to improve the effectiveness of screening upper endoscopy for esophageal varices.
Ascites ; Biopsy ; Blood Platelets ; Chemistry ; Endoscopy ; Esophageal and Gastric Varices* ; Fibrosis ; Hemorrhage ; Hepatic Encephalopathy ; Humans ; Liver Cirrhosis* ; Liver* ; Mass Screening ; Multivariate Analysis ; Platelet Count ; Prothrombin ; Prothrombin Time ; Retrospective Studies ; Splenomegaly ; Varicose Veins