Objectives EMERGENCi is a prospective, national, cohort study of emergency endoscopy and severe upper GI bleeds in children. Objectives were to produce national data of the clinical presentations, patient co-morbidities, indications, waiting times and endoscopic treatments for emergency endoscopy.

Methods Units were identified through the UK Paediatric Gastroenterology and Paediatric Surgical Societies (BSPGHAN & BAPS). Once registered, fortnightly emails were sent over a 6 month period asking for reports in <16 years of severe upper GI bleeds requiring endoscopy (UGIB) and/or other emergency upper endoscopies (OEE). Cases were reported in REDCap.

Results 28 centres provided denominator data with regard to the services they provide (covering 90% of the UK population). 22 provided prospective data for UGIB and 18 for OEE covering 70 and 60% of the UK population respectively.

98 cases were reported over a 6 month period: 34 UGIB, 55 OEE, (38 foreign body and 17 others); 9 less severe UGI bleeds not fitting the definitions were excluded from further analysis.

Of 25 centres reporting, 14(56%) had 0 UGIB and 20/25(80%) had ≤2 over the 6 months. Endoscopic interventions for GI bleed were undertaken in only 6/25 centres.

The mean age of the UGIB group was 6.7 years, 29% were ≤1 year. 19(56%) had significant co-morbidities. Presenting symptoms were one or both of melaena and haematemesis. Of the 20 providing sufficient data for a Sheffield score, 25%(4/20) were high (≥8) at presentation (median score 2.5, range 1–24, interquartile range 3.25). Main findings at endoscopy; 8(24%) had no abnormalities, 14(41%) had UGI ulcers (6 duodenal, 6 gastric and 2 oesophageal), 9(26%) oesophagitis and gastritis, 8(24%) varices.

13(38%) required endoscopic treatment, 6 for varices, 4 for GU, 2 DU, 1 for blood in upper GI tract. 3 required surgery. Two patients died, one within 48 hours of the bleed in PICU in the context of sepsis and multi-organ failure. 14 patients required inter-hospital transfer, median time from hospital presentation to endoscopy was 97 hours for patients needing transfer and 24 hours for those not.

For the OEE (N=55), mean age was 6.3 years, 26% ≤1 year. 21(38%) had significant co-morbidities. Main indications were foreign bodies (25, 45%) - coins (15), battery (2), button battery (5), magnets and a toy. 13(24%) food bolus obstruction, 11(20%) caustic substance ingestion, 5 oesophageal stricture. 9/55(16%) endoscopies revealed no significant findings, 37(65%) required treatment at endoscopy. 50% (27 patients) had required inter-hospital transfer. Median time from first hospital presentation to endoscopy was 21 hours in those requiring transfer and 14 hours in those not.

Conclusions This is the first national prospective study of its kind examining the most urgent and severe endoscopy cases in under 16s. These data indicate that very small numbers of centres are performing endoscopic treatments for severe UGI bleeds. Inter-hospital transfers appears to be much quicker for surgical indications than UGIB although we did not find evidence of poor outcomes in the UGIB due to delayed transfer. The planning, location and skill mix of national emergency endoscopy services require careful consideration.