Objectives: Paediatric endoscopy is an important diagnostic tool; however, there is little published data to guide clinicians in selecting patients for endoscopy. This study aimed to evaluate a single centre's experience of newly presenting children focusing on presenting symptoms, investigations, and diagnostic yield.
Methods: Clinical factors and endoscopic plus histological findings over a 6-month period were assessed. Only first diagnostic endoscopies were included. All biopsies were reviewed in a weekly histopathology multidisciplinary team meeting with a final agreed outcome. Abnormal histology was used as the criterion standard for reporting abnormality.
Results: A total of 218 endoscopies were reviewed in 164 children. Approximately 65% were histologically normal (49% of children had macroscopically and histologically normal findings). Macroscopic and histological abnormalities (respectively) were 44% and 28% of oesophagogastroduodenoscopy (OGD) patients, 25% and 25% of colonoscopy alone, and 53% and 53% of those undergoing both OGD and colonoscopy (OGD&Col). For OGD-only patients, excluding those with raised anti-tissue transglutaminase antibodies, vomiting led to the highest rate of abnormal histology (22%). For colonoscopy-only and OGD&Col patients, per rectum bleeding led to the highest rates of abnormal histology (14% and 29%, respectively), after excluding those with laboratory abnormalities (anaemia and raised erythrocyte sedimentation rate) suggestive of inflammatory bowel disease.
Conclusions: This study showed that half of all first diagnostic endoscopies in our unit had neither macroscopic nor histological abnormalities. There was discrepancy between macroscopic abnormalities and histological findings in OGD. Prospective studies are needed to develop guidelines in appropriately predicting abnormality and selecting patients for endoscopy.