Background/aims: Upper gastrointestinal endoscopy is usually recommended for the evaluation of esophageal varices in patients with liver cirrhosis. However, the prevalence of varices is extremely variable. We performed this study to determine the predictive values for esophageal varices and to select eligible patients for screening endoscopy.
Methods: Fifty-two patients were enrolled in this study. Laboratory tests including liver biochemistry and complete blood count along with ultrasonography with Doppler measurements and endoscopy were performed.
Results: Esophageal varices were present in 25 patients (48%). Variables associated with the presence of esophageal varices on univariate analysis were serum albumin, total bilirubin, prothrombin time and platelet count (p<0.05). Significant variables in ultrasonography with Doppler measurement were diameter of spleen (13.04+/-2.1 cm vs. 10.39+/-1.6 cm, p<0.001), peak velocity of portal vein (30.2+/-7.5 cm/sec vs. 36.1+/-8.0 cm/sec, p<0.01) and portal vein diameter (1.26+/-0.28 cm vs. 1.13+/-0.18 cm, p<0.05). On multivariate analysis, independent variables were platelet count (odds ratio (OR) 0.922; 95% confidence interval (CI), 0.86-0.99), diameter of spleen (OR 5.4; 95% CI, 1.63-17.88) and platelet count/spleen diameter ratio (OR 1.007; 95% CI, 1.01-1.02). The optimal critical value for the diameter of spleen was 11 cm. The sensitivity and specificity with this value were 84% and 63%, respectively.
Conclusions: Doppler measurement was not helpful in distinguishing the presence of varices. However, clinical tests including biochemistry and ultrasonography would be useful in selecting eligible patients for screening endoscopy. Endoscopic screening for esophageal varices is recommended in cirrhotic patients with splenomegaly.