Letters Coeliac disease serology

Targeting coeliac disease serology

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e8120 (Published 10 December 2012) Cite this as: BMJ 2012;345:e8120
  1. David J Unsworth, consultant immunologist1,
  2. Francis J Smith, biomedical scientist2,
  3. Robert J Lock, consultant clinical scientist1
  1. 1Department of Immunology and Immunogenetics, Southmead Hospital, North Bristol NHS Trust, Bristol BS10 5NB, UK
  2. 2Department of Immunology, Gloucester Royal Hospital, Gloucestershire Hospitals NHS Foundation Trust, Gloucester, UK
  1. joe.unsworth{at}nbt.nhs.uk

We agree that testing first degree relatives of patients with coeliac disease (CD) for gluten enteropathy is worthwhile but caution against less targeted testing, especially in patients with non-specific symptoms.1

The prevalence of CD in England is estimated at 1.0-1.2%.2 3 Prevalence is higher in “at risk” groups: 5-6% in first degree relatives and 6% in people with type 1 diabetes.4 5

We reviewed requests for coeliac serology from two large hospitals and their primary care providers. All sera were tested for anti-tissue transglutaminase antibodies (tTG). Confirming positivity by anti-endomysium antibody testing improves positive predictive values. Berrill and colleagues used IgA anti-tTG alone when reviewing their case,1 but this test can give false positive results. For primary testing we advise confirmation using the IgA anti-endomysium test.

The figure shows that increasing awareness of CD before the 2009 National Institute for Health and Clinical Excellence (NICE) guidelines resulted in an explosion of requests for serological testing, particularly from primary care. NICE guidance has neither focused testing nor improved yield. Disappointingly, our data suggest that requesting is becoming less effective. Requesting has more than doubled in five years, with the primary care component increasing from 69% to 80% of total requests. The proportion of patients with dual positive serology in primary care has fallen to 1.2% of the tested population.


Numbers of requests for anti-tissue transglutaminase in two NHS trusts

In this time of austerity patient selection for CD serology needs to be better targeted. In primary care, CD detection is no better than would be achieved by random screening. In the hospital setting (many cases post-GP referral, no doubt) CD detection was highest (around 5%) in children (with failure to thrive, family history, and type 1 diabetes) and in adult diabetes and gastroenterology clinics.


Cite this as: BMJ 2012;345:e8120


  • Competing interests: None declared.