Coeliac disease is caused by the development of antibodies to a protein, gluten, in cereals in the diet. The antibodies attack the lining of the patient’s gut and impair the absorption of food. Diagnosis may be difficult, but there is a high risk in patients who have weight loss, anaemia or diarrhoea with loose foul-smelling stools. The ‘gold standard’ for diagnosis is to pass a tube into the duodenum to take a small piece of its lining (biopsy) for microscopic examination. Most patients with the disease have positive blood tests for both IgA antibodies to tissue transglutaminase and IgA endomysial antibodies, but not all do. A biopsy is therefore sometimes considered necessary in those with a negative lab test to exclude the diagnosis and in those with a positive lab test because some positives do not have coeliac disease

Dr Andrew Hopper and eight colleagues from the Royal Hallamshire Hospital Sheffield Department of Gastroenterology, Sheffield University and general practice report in the British Medical Journal of 6 April a plan of investigation that detected all cases of coeliac disease in patients referred to the Department for examination of the upper intestinal tract (gastroscopy) without having to perform duodenal biopsy in everyone. They first looked back at the records of 5979 patients who had been examined in 2003. A combination of a positive test for antibodies to tissue transglutaminase with the ‘high risk’ symptoms of weight loss, anaemia or diarrhoea as defined by the British Society of Gastroenterology picked up all the patients who had a positive biopsy for coeliac disease.

The research workers then recruited 2000 volunteers from patients referred to them for examination of the upper gut during the next 26 months. The patients were divided into ‘high risk’ and ‘low risk’ of coeliac disease on the basis of their symptoms. All 2000 had tissue transglutaminase antibodies measured and had biopsies taken. Of the 77 patients who were found to have coeliac disease on biopsy all were either classified as ‘high risk’ or had a positive antibody test or, in 83% of them, both. None of the 1170 ‘low risk’ patients with a negative antibody test, 59% of the total, had coeliac disease. In future such patients should not need routine invasive biopsy.

In a commentary that accompanied the research paper, Drs Graber and Kumar from the VA Medical Center, Northport, NY agreed with the authors that biopsy still has an important role in high risk patients with positive serology because 40% of them did not have coeliac disease when biopsied. They suggested that decision making might be improved by incorporating further lab tests (HLA) that are positive in nearly all patients with coeliac disease. They concluded that at present the Sheffield decision rule is the most cost effective and efficient way to assess coeliac disease.