The Yield of Routine Tissue Sampling in Pediatric Gastrointestinal Endoscopy

Objectives: Societies' guidelines suggest routine tissue sampling in all children undergoing esophagogastroduodenoscopy and ileocolonoscopy, even in the absence of visible endoscopy abnormalities. We aimed to determine the agreement between endoscopic and histopathological findings in pediatric endoscopy and to assess the yield of routine biopsies from all sites. Methods: Since January 2019, our endoscopy institute protocol has included routine biopsies sampling from the esophagus, stomach, duodenum, ileum, and colon in all diagnostic procedures. Agreement between tests was done using the kappa coefficient (κ). The study included all endoscopies performed during 2019. Results: In total, 541 diagnostic endoscopies were done during the study period with 434 (80%) esophagogastroduodenoscopy and 107 (20%) were ileocolonoscopy. Compared to histology, endoscopic findings performance were: esophagus - sensitivity 33%, specificity 98%; stomach - sensitivity 60%, specificity 89%; duodenum - sensitivity 50%, specificity 97%; duodenal bulb - sensitivity 47%, specificity 89%; terminal ileum - sensitivity 82%, specificity 100%; colon - sensitivity 84%, specificity 96%. Assessment of concordance between endoscopic and histopathologic findings reveals an overall low level of agreement in esophagogastroduodenoscopy (κ of 0.39, 0.51, 0.53, and 0.24 for the esophagus, stomach, duodenal second part, and bulb, respectively), and good agreement in ileocolonoscopy (κ of 0.88 and 0.81 for the ileum and colon, respectively). Conclusions: Endoscopy findings are highly specific for histologic pathology, whereas the absence of findings correlates poorly with histologic findings. Ileocolonoscopy shows better agreement than esophagogastroduodenoscopy. Our data support routine tissue sampling in pediatric endoscopy.