Utility and Comparative Efficacy of Lok Score, Splenic Stiffness, EVendo Score, and Liaoning Score in Predicting...

Background Esophageal varices (EV) are a critical complication of portal hypertension in cirrhotic patients, with variceal bleeding contributing significantly to morbidity and mortality. Endoscopy remains the gold standard for diagnosis, but its invasive nature and limited availability in low-resource settings necessitate the validation of non-invasive predictive tools. This study aims to compare the diagnostic performance of four non-invasive models, Lok score, splenic stiffness (SS), EVendo score, and Liaoning score, for predicting EV in Pakistani patients with chronic liver disease (CLD). Methods A cross-sectional observational study was conducted at the Department of Hepatogastroenterology, Sindh Institute of Urology and Transplantation (SIUT), Karachi, from January to December 2024. A total of 272 cirrhotic patients undergoing screening endoscopy were enrolled. Clinical, laboratory, and radiological data were collected to compute the Lok, EVendo, and Liaoning scores. Splenic stiffness was measured using transient elastography. The presence of EV was confirmed via upper gastrointestinal (GI) endoscopy. Receiver operating characteristic (ROC) curve analysis was used to assess diagnostic accuracy (area under the receiver operating characteristic {AUROC}), and multivariate regression identified independent predictors of high-risk varices (HRV). Results EV were present in 72.8% of patients, with high-risk varices in 31.3%. All four scores were significantly elevated in patients with EV. Splenic stiffness demonstrated the highest AUROC (0.858), followed by Lok score (0.804), EVendo score (0.76), and Liaoning score (0.755). EVendo score showed the highest diagnostic accuracy (81.56%) and, along with SS, was an independent predictor of HRV. Conclusion Splenic stiffness and EVendo score are reliable non-invasive predictors of esophageal varices and high-risk varices in the Pakistani cirrhotic population. However, further multicentered studies are required to validate our results.