Objective Coeliac disease is a prevalent disorder but frequently remains un
diagnosed because of varied modes of clinical presentation. In this study,
methods for the detection of coeliac disease were evaluated in a clinical p
ractice setting,
Methods Small intestinal histology, IgA anti-endomysial and IgG anti-gliadi
n antibody tests were performed on 441 unselected, consecutive patients und
er investigation for small intestinal disease. Response to treatment and ot
her clinical events were monitored over the ensuing years,
Results Untreated coeliac disease was diagnosed in 97 patients and was excl
uded in 344, At clinical presentation, the endomysial antibody test was pos
itive in 84 of the 97 untreated coeliac patients (sensitivity 87%) and nega
tive in 340 of the 344 non-coeliac patients (specificity 99%). A typical hi
stological lesion was found in 83 of the 97 coeliac patients (sensitivity 8
6%) but was absent in all 344 non-coeliacs (specificity 100%). The sensitiv
ity of the gliadin antibody test was 69% and the specificity was 71%,
Conclusions In unselected patients attending a gastroenterology clinic, sma
ll bowel histology and endomysial antibody serology show similar predictive
value in the diagnosis of coeliac disease. These results emphasize that a
combination of clinical, histological and serological criteria are required
for effective diagnosis of this disorder. Exclusive reliance on histology
or serology will result in failure to make a diagnosis in a significant pro
portion of patients. Eur J Gastroenterol Hepatol 10: 919-925 (C) 1998 Lipin
cott Williams & Wilkins.