AGREEMENT BETWEEN MULTIDETECTOR COMPUTED TOMOGRAPHY AND ENDOSCOPY FOR THE EVALUATION OF GASTRO-ESOPHAGEAL VARICES IN PATIENTS WITH PORTAL HYPERTENSION.

Abstract

Aim: The aim is to study the agreement between MDCT and endoscopy in diagnosing gastroesophageal varices in patients of chronic liver disease with portal hypertension, keeping endoscopy as gold standard. Materials and Methods: The study included 29 cirrhotic patients with portal hypertension assessed between August 2018 to August 2020 who underwent upper gastrointestinal (GI) endoscopy and MDCT within 12 weeks. Two radiologists reviewed the scans, to determine the grade of esophageal varices. To evaluate the accuracy of MDCT findings by Radiologists 1 and 2 with respect to endoscopy findings (GOLD STANDARD), McNemar’s chi square test with validity parameters such as Sensitivity, Specificity, Positive predictive value, Negative predictive value and accuracy was computed. Cohens kappa was used to ascertain the inter-observer agreement for grading of varices between MDCT and endoscopy. Results: The sensitivity of MDCT for radiologist A was 76.7%, specificity 93%, Accuracy 89.7%, PPV 91.6%, NPV 93.9% and for radiologist B, 79%, 93.4%, 89.6%, 81.4% and 93.1% respectively. For the inter-rater agreement the kappa value between Radiologist A v/s upper GI endoscopy was 0.716, Radiologist B v/s upper GI endoscopy was 0.720 and Radiologist A v/s Radiologist B was 0.808. MDCT detected para esophageal varices in 16 cases, gastric fundus varices in 9 cases and splenorenal collaterals in 3 cases, palisade vein dilatation was positive in 9 cases, negative in 20 cases. 3 cases of HCC were incidentally found during examination. Conclusion: MDCT is a good replacement for endoscopy in the detection of gastroesophageal varices, especially those with high grade varices (grade 2 and 3). MDCT helps in detection of palisade vein and other extra-luminal findings like HCC, other portosystemic collaterals.