Validation of the EVendo score for assessing high-risk varices in Western European patients with cirrhosis
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Background: Gastroesophageal varices (GEV) and associated bleeding are serious complications of cirrhosis, leading to considerable morbidity and mortality. While esophagogastroduodenoscopy (EGD) remains the gold standard for detecting GEV, it is an invasive procedure. The EVendo score, a non-invasive predictive tool, has demonstrated promise in identifying GEV and varices needing treatment (VNT). However, its effectiveness has not been extensively studied in Western European populations. This study aims to evaluate the utility of the EVendo score in predicting GEV and VNT among patients with cirrhosis. Methods: This retrospective study, conducted at Klinikum Stuttgart in Germany, screened over 700 patients with cirrhosis who underwent EGD from 2014 to 2021. After applying strict inclusion and exclusion criteria, 86 patients were included. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were assessed for EVendo score cutoffs of ≤ 3.9, ≤ 4.5, and a cohort-optimized cutoff of ≤ 4.3. Results: GEV was present in 94.2% of patients, with 41.9% requiring treatment (VNT). At ≤ 3.9, the EVendo score showed high sensitivity for GEV (79.8%) and VNT (87.2%), but specificity remained 0% for GEV and 24% for VNT. Increasing the cutoff to ≤ 4.5 slightly reduced the sensitivity to 77% for GEV and 85% for VNT, with specificity improving to 26% for VNT. This cutoff avoided 22.1% of EGDs, missing three VNT cases. The cohort-specific cutoff of ≤ 4.3 provided the highest area under the receiver operating characteristic curve (0.57), avoiding 22.1% of EGDs and missing two VNT cases. Conclusion: The EVendo score shows promise for reducing unnecessary EGDs, with high sensitivity across different cutoffs. However, its limited specificity for GEV suggests the need for further refinement, particularly in advanced liver disease stages, to optimize clinical utility. Further validation in diverse, larger cohorts is recommended.
Abstract
Background: Gastroesophageal varices (GEV) and associated bleeding are serious complications of cirrhosis, leading to considerable morbidity and mortality. While esophagogastroduodenoscopy (EGD) remains the gold standard for detecting GEV, it is an invasive procedure. The EVendo score, a non-invasive predictive tool, has demonstrated promise in identifying GEV and varices needing treatment (VNT). However, its effectiveness has not been extensively studied in Western European populations. This study aims to evaluate the utility of the EVendo score in predicting GEV and VNT among patients with cirrhosis. Methods: This retrospective study, conducted at Klinikum Stuttgart in Germany, screened over 700 patients with cirrhosis who underwent EGD from 2014 to 2021. After applying strict inclusion and exclusion criteria, 86 patients were included. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were assessed for EVendo score cutoffs of ≤ 3.9, ≤ 4.5, and a cohort-optimized cutoff of ≤ 4.3. Results: GEV was present in 94.2% of patients, with 41.9% requiring treatment (VNT). At ≤ 3.9, the EVendo score showed high sensitivity for GEV (79.8%) and VNT (87.2%), but specificity remained 0% for GEV and 24% for VNT. Increasing the cutoff to ≤ 4.5 slightly reduced the sensitivity to 77% for GEV and 85% for VNT, with specificity improving to 26% for VNT. This cutoff avoided 22.1% of EGDs, missing three VNT cases. The cohort-specific cutoff of ≤ 4.3 provided the highest area under the receiver operating characteristic curve (0.57), avoiding 22.1% of EGDs and missing two VNT cases. Conclusion: The EVendo score shows promise for reducing unnecessary EGDs, with high sensitivity across different cutoffs. However, its limited specificity for GEV suggests the need for further refinement, particularly in advanced liver disease stages, to optimize clinical utility. Further validation in diverse, larger cohorts is recommended.