Introduction Eosinophilic oesophagitis (EO) is the underlying diagnosis in at least 10% of those with dysphagia. To make the diagnosis, oesophageal biopsies showing an eosinophil count >15 per hpf are required. It is most frequent in males under aged 50 years.

Methods All patients having a gastroscopy for dysphagia were identified retrospectively for 6 consecutive years from our endoscopy reporting system. Patient demographics, endoscopic findings and whether biopsies were taken were recorded together with histology results.

Results 3068 patients had a gastroscopy with an indication of dysphagia (1489 female, age 15–100 years, average 67.7 y). The number of patients varied little between years (486–550 patients/year). Common endoscopic diagnoses were normal (20.4%), benign stricture (12.6%), oesophagitis (18.1%), Barrett’s (4.8%), dysmotility (3.7%) and hiatus hernia (10%). 1620 (52.8%) had oesophageal biopsies.

44 patients (1.5% of all patients) were diagnosed with EO, 32 of who were males. This equates to 2.8% of those who were biopsied and 4.7% of those biopsied without cancer, stricture or Barrett’s. Although only 13.3% of those with dysphagia were aged 50 years or under, they equated to 45.4% of those diagnosed with EO. Of those with EO, 6 had food bolus, 6 “typical” EO changes e.g., feline oesophagus, ridges etc, 4 an irritable oesophagus and 3 Schatzki rings.

Conclusion EO is a relatively common cause of dysphagia but is almost certainly under-recognised due to lack of oesophageal biopsies at endoscopy. Reliance on endoscopic changes of EO at endoscopy will miss the majority of cases. Although biopsying only those under 50 years would be more cost effective than biopsying all, it would also miss the majority of cases. It may be appropriate for the BSG to use frequency of oesophageal biopsies in dysphagic patients as a quality assurance measure for upper GI endoscopy.

Disclosure of Interest None Declared.