Diagnostic yield of gastroscopy in a general surgical unit

Gastroscopy is the gold standard for investigating upper gastrointestinal c omplaints. Open access gastroscopy has advantages, but may not always be fe asible in a small unit(1). Our 2-consultant general surgical unit provides gastroscopy on demand for medical and surgical patients. We audited the 499 gastroscopies carried out in our unit in 1997 to assess yield and diagnost ic trends. The overall diagnostic yield of 69 per cent was comparable with published data from centres in Ireland(1,2) and overseas(3,4). Diagnostic y ield in our series was significantly lower for medical patients (59 per cen t) than for surgical patients (72 per cent) (p<0.05). Yield also varied con siderably according to indication for referral. Patients referred with anae mia had a low yield (41 per cent). The indications associated with high yie lds mere haematemesis (95 per cent), dysphagia (91 per cent), and persisten t vomiting (85 per cent).