IDENTIFICATION OF CELIAC-DISEASE IN PRIMARY-CARE

Background: Coeliac disease is common yet often undiagnosed because sy mptoms may be trivial, nonspecific, or non-gastrointestinal, or becaus e of lack of clinician awareness. Serum IgA-class endomysial antibodie s (EmA) have high specificity for coeliac disease and may facilitate c ase-finding by clinicians other than gastroenterologists. We assessed the appropriateness and diagnostic yield of of requests for EmA by pri mary care general practitioners in a defined geographic area of Northe rn Ireland. Methods: We identified patients who had EmA examination re quests by their general practitioners during 1994-1996. Individual pat ient questionnaires were posted to the general practitioners concerned , seeking information on indications for testing, management after the result, and final diagnosis. We compared new patient diagnosis rates in two catchment areas, one served by a large district general hospita l with, and the other by smaller hospitals without, a medical gastroen terology facility. Results: A total of 239 patients had coeliac profil e testing by 69 of 177 general practitioners in the area Data were ava ilable for 181 patients not previously known to have coeliac disease, of whom 20 (11%) had EmA. All EmA-positive patients were referred to h ospital, where 19 underwent small-bowel biopsy, which confined coeliac disease in all 19. Only 7 (35%) of the 20 had diarrhoea, and there wa s no significant difference in EmA prevalence among patients tested wi th and without diarrhoea. Although the mean number of new patients (pe r 100,000 population annually) diagnosed by biopsy was II at the large hospital compared with 5 elsewhere, the numbers identified by EmA in general practice for the 2 catchment areas were similar (2 and 3, resp ectively). Conclusion: General practitioners have an important role in the identification of patients with coeliac disease, particularly whe re there is no local medical gastroenterology facility, and this is fa cilitated by EmA testing.