Non-endoscopic predictors of esophageal varices in children with chronic liver disease and their utility in resource-constrained countries.

Background: Although endoscopy is the standard diagnostic screening test to identify esophageal varices in patients with chronic liver disease (CLD), selective endoscopy in patients who are at higher risk of having varices may be cost-effective in a resource-constrained country. The aim of this prospective study was to identify non-endoscopic parameters that may predict the presence of varices, especially high-risk esophageal varices in children with CLD.

Methods: From January 2016 through March 2018, consecutive children with CLD without a history of variceal bleeding were prospectively included. Esophagogastroduodenoscopy was done in all the children to detect and to grade esophageal varices. Both univariate and multivariate logistic regression analyses were done using SPSS version 22 to identify factors associated with esophageal varices.

Results: The mean age of 84 children was 9.7 ± 3.2 years (male 44). Esophageal varices were present in 71.4% of children and 55% of them had large varices. On univariate analysis, low platelet count (< 100,000/mm) and splenomegaly were found to be associated with the presence of esophageal varices (p = 0.006 and 0.001, respectively) and large varices (p = 0.03 and 0.01, respectively). On multivariate analysis, both low platelet count and splenomegaly were independent predictors for the presence of esophageal varices (respectively, OR 11.21, 95% CI 1.2-96.9; and OR 11.39, 95% CI 3.19-40.59).

Conclusions: Splenomegaly and low platelet count independently predict the presence of any grade of esophageal varices and can be used as screening tests to select children for endoscopy. This strategy may help in relieving medical, social, and economic costs in resource-constrained countries.