Analysis of the correspondence between endoscopy and histology in 135 pediatric esophagogastroduodenoscopies: a retrospective study

Abstract

Aim. To determine the level of correspondence between endoscopic and histological data when performing esophagogastroduodenoscopie (EGD) in children. A secondary aim was to identify predictors that affect the correspondence between endoscopy and histology. Materials and methods. A retrospective analysis of the EGD (patients aged 2–18 years) was carried out. Descriptive statistics and binary logistic regression were used to determine the level of agreement and potential predictors of agreement between endoscopy and histology. Results. 135 EGDS were analyzed. The overall level of disagreement between endoscopy and histology was 39.3% (53 patients). Taken into account histology as the gold standard, the specificity of white light EGD was 45.3%, sensitivity, accuracy, positive and its negative predictive value was 80, 60,7, 53,9, 73,9% respectively. A comparison of endoscopic and histological conclusions for each upper GI organ showed that the collection of biopsies for histological examination in all children improves the diagnostic value of white light EGD by 33.3% in the esophagus, 41.5% in the stomach and 14% in the duodenum. Logistic regression analysis showed that only heartburn and belching turned out to be a statistically significant (p=0.018) predictor of a low degree of correspondence between endoscopic and histological results (OR 0.302, 95% CI 0.112–0.814). Abdominal pain, nausea and vomiting, diarrhea, an established diagnosis of IBD, the presence of a diagnosis of JRA, age, gender, experience of an endoscopist and endoscope model did not have a statistically significant effect on the agreement level between endoscopy and histology (p0.05). Conclusion. EGD as an instrumental intervention is by definition an invasive procedure, and its diagnostic potential should be used to the maximum through the performance of routine biopsies in all children. This is especially important for pediatric gastroenterology because it allows not only to improve the effectiveness of intraluminal diagnostic method by almost 40%, but also to reduce the number of repeated unproductive white light EGD’s without taking tissue for confirming or refuting the presence of the disease in the child.