The diagnostic yield of upper gastrointestinal endoscopy in the investigation of anaemia

Computer records of patients endoscoped over a 34-month period were studied to assess the diagnostic yield of gastrointestinal endoscopy in patients with anaemia. Patients with obvious gastrointestinal bleeding and known gastrointestinal pathology were excluded. On hundred and thirty-six patients were endoscoped for anaemia. Eighty-three of them (61%) had iron deficiency anaemia and 53 (39%) had other types of anaemia. The ages of the patients with iron deficiency anaemia (mean 56 years) were significantly lower than those of other anaemias (means 65 years), (p less than 0.003). Patient characteristics were otherwise comparable. There were significant endoscopic findings (ulcers, carcinoma and haemorrhagic or erosive gastritis) in 26 of 83 patients (31%) with iron deficiency anaemia, in 11 of 53 patients with other anaemias (21%) and 37 of 136 patients (27%) combined. Significant endoscopic findings were found in 506 of 2224 patients (23%) endoscoped during this period who were not anaemic, did not have obvious gastrointestinal haemorrhage and were not known to have gastrointestinal diseases. The diagnostic yield for iron deficiency anaemia was significantly higher than for the non anaemic group (p less than 0.05). There was no difference between the diagnostic yields of iron deficiency and other anaemias, other anaemias and the non anaemic group, or total anaemias and the non anaemic group. Gastrointestinal symptoms and history of analgesic or steroid usage did not appear to increase the incidence of gastrointestinal lesions in either iron deficiency anaemia or other anaemias. Twenty-three of 41 patients (56%) who had no cause for anaemia found at the end of all investigations were colonscoped.