Utility of EVendo score as a screening tool for the detection of high risk esophageal varices in Pakistani population

Abstract

Introduction: Endoscopy for the screening of esophageal varices (EVs) is costly and cannot be performed in remote areas with limited resources. Recently, certain non-invasive cost effective models have been proposed for the prediction of EVs but have failed recommendation on a larger scale. EVendo score is a recently developed bedside score for the detection of EVs. Therefore, our aim was to determine the utility of EVendo score as a screening tool for the detection of high risk esophageal varices in Pakistani population.

Methods: It was a cross sectional study which was conducted in the department of Hepatogastroenterology from January 2021-June 2022.All patients with newly diagnosed cirrhosis were included in the study while those patients fulfilling the exclusion criteria were excluded. Area under the receiver operating curve (AUROC) was obtained for EVendo score, Aspartate Transaminase to platelet ratio(APRI) and Platelet count to Splenic Diameter and diagnostic accuracy was obtained for these scores in predicting EVs and also in identifying HRVs. Should be showed the location of study in this section Results: A total of 272 patients were enrolled in the study. Among them, 167(61.4%) were males. On screening endoscopy, EVs were noted in 118(43.4%) patients while high risk EVs (HRV) were noted in 47(17.3%) patients respectively. AUROC was obtained for EVendo score, APRI and Platelet count to Splenic Diameter in predicting EVs and also for identifying HRVs and it was 0.93 (p-value <0.001), 0.821(<0.001) and 0.842(p<0.001) respectively for the prediction of EVs with diagnostic accuracy of 86.76% and 0.852 (p-value <0.001), 0.835(<0.001) and 0.814(p<0.001) respectively for identifying HRVs with a diagnostic accuracy of 84.19%. Conclusion: The performance of EVendo score was reliable and better than the other non-invasive scores in predicting EVs in our population with an excellent sensitivity and diagnostic accuracy in predicting the EVs and also in identifying HRVs. However, studies comprising larger sample sizes are required in this regard.