Prediction of Esophageal Varices in Patients with Liver Cirrhosis

ABSTRACT

BACKGROUND: Patients with liver cirrhosis are regularly examined for the evaluation of esophageal varices. Those with large varices should be treated with beta-blockers. The aim of this study was to determine whether clinical variables were predictive of the presence of esophageal varices or high-grade varices. METHODS: The medical records of 257 patients, diagnosed as having liver cirrhosis and underwent esophagogastro-duodenoscopy (EGD), were reviewed. None had a history of malignancy and variceal hemorrhage before EGD. Clinical findings were analyzed in relation to the presence of esophageal varices and high grade esophageal varices. RESULTS: The multiple logistic regression analysis showed an independent association between the presence of esophageal varices and ascites and low platelet count. Only low platelet count showed association with high grade esophageal varices. The cut-off value of predictive model for high grade esophageal varices was 73,766/microliter and its sensitivity was 75%, specificity 65.7%, negative predictive value 96.2%, and positive predictive value 18.6%. CONCLUSION: Our analysis showed that low platelet count and ascites were independent predictive factors for esophageal varices. For high grade varices, the platelet count was the only predictive factor. Endoscopic screening for detection of high grade varix is highly recommended when the platelet count is lower than 74,000/microliter is observed in patients with liver cirrhosis.