Introduction The use of anti-gliadin antibodies (AGA) and anti-endomysial antibodies (EMA) is well established as a means of screening for coeliac disease (CD). Since the presence of EMA is highly specific for CD, a negative EMA result may deter some doctors from referring patients with positive AGA for duodenal (D2) biopsy or gastroenterological assessment. In CD there is a prevalence of IgA deficiency of around 3%, with corresponding negative EMA. In addition, isolated positive AGA are found in other gastrointestinal (GI) conditions, such as Crohn's disease and chronic liver disease.

Methods Aims: (1) To determine the proportion of patients with positive IgG AGA and/or IgA AGA but negative EMA who had undergone endoscopy and D2 biopsy. (2) To invite those who had not been investigated to attend either the gastroenterology clinic or for direct endoscopy and duodenal biopsy. (3) To determine the prevalence of CD and other gastrointestinal pathology.

We identified patients with positive IgG and/or IgA AGA but negative EMA from the pathology database from February 2007 to August 2008. The medical records were reviewed to determine if there were features of CD on duodenal biopsy if performed, or alternative GI diagnoses. We wrote to the referring doctor and the general practitioner, requesting referral to gastroenterology if D2 biopsy had not been undertaken. If no reply, we contacted the patient directly.

Results 1597 requests for coeliac serology were made during the 18 months: 186 were both AGA and EMA positive. Forty-six patients were AGA positive with negative EMA: 24 had undergone D2 biopsy; 2 had died prior to further investigation (86 and 89 years old); 1 was ineligible for NHS care; 1 patient had no contact details; and 1 patient had repeat EMA and AGA, which were negative. Of the remaining 17 patients, 8 agreed to undergo OGD and D2 biopsy.

Conclusion CD is not uncommonly present in patients with isolated positive AGA and all such patients should be referred for OGD and D2 biopsy. Further investigation should also be considered as other gastrointestinal disorders are prevalent in these patients, probably in part reflecting the presence of symptoms or abnormal blood tests that prompted the request for AGA. The majority of requests for coeliac serology originate from primary care or non-GI physicians, many of whom fail to appropriately follow-up positive AGA. Laboratories should consider providing advice on AGA interpretation with test results.