Article Text

Download PDFPDF
Barriers to implementing the revised ESPGHAN guidelines for coeliac disease in children: a cross-sectional survey of coeliac screen reporting in laboratories in England
  1. Siba Prosad Paul1,
  2. Sophie Louise Harries2,
  3. Dharamveer Basude1,2
  1. 1 Department of Paediatric Gastroenterology, Bristol Royal Hospital for Children, Bristol, UK
  2. 2 Medical School, University of Bristol, Bristol, UK
  1. Correspondence to Dr Siba Prosad Paul, Department of Paediatric Gastroenterology, Bristol Royal Hospital for Children, Bristol BS2 8BJ, UK; siba.paul{at}


Background European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) guidelines for diagnosing paediatric coeliac disease (CD) were revised in 2012. This enabled serological diagnosis in a selective group of symptomatic children using anti-tissue transglutaminase (anti-tTG) titre, antiendomysial antibodies (EMA) and HLA DQ2/DQ8 status. However, observing variations in the availability of serological tests for CD within our region, we conducted a countrywide survey to explore the diversity of these tests for all paediatric centres.

Methods A nationwide telephone survey among biomedical scientists based in 139 National Health Service hospital trusts providing paediatric services in England was conducted by a single interviewer over a defined 3-week period. Respondents were asked about type of anti-tTG assay, the upper limit of normal (ULN) for anti-tTG titres, availability of EMA and reporting of IgA concentration.

Results Responses were available from 134 (96.4%) laboratories. Anti-tTG titres are performed by 83/134 (62.6%) laboratories and 68/83 (81.4%) of those also offered EMA testing. Four different anti-tTG assays are available in England, but there are 10 different ULN values. The range for ULN varies widely from 4 to 30 IU/mL. Automatic reporting of total IgA concentration for a coeliac serology request occurs in only 24/83 laboratories.

Conclusions Significant heterogeneity exists for serological tests for CD in particular anti-tTG titre reporting even within the same regions. This potentially affects the interpretation of the results by clinicians diagnosing CD and hence harbouring diagnostic inconsistencies in their practice. Standardisation especially of the anti-tTG assays and routine reporting of IgA concentration nationally should be strongly considered to support the current diagnostic process for CD.

  • Gastroenterology
  • General Paediatrics
  • Laboratory medicine
  • Coeliac disease
  • ESPGHAN guidelines
  • Anti-tissue transglutaminase

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic?

  • Serological diagnosis of coeliac disease is highly reliable and beneficial for a number of reasons: it is highly sensitive and specific, less invasive and economical.

  • There is a wide range of anti-tTG assays available. Due to the complexity in measuring immunoglobulins, each method provides a unique upper limit of normal value.

What this study adds?

  • There is a wide range of upper limit of normal for anti-tTG titres employed throughout England even within regions covered by a tertiary paediatric gastroenterology centre.

  • Many laboratories rely on assay technology to identify low IgA values in the samples tested. Majority of laboratories require specific request made by the clinician to report the IgA concentration.

  • Small minority of laboratories still perform qualitative tests preventing clinicians to use the 2012 European Society for Paediatric Gastroenterology, Hepatology and Nutrition coeliac disease guidelines.


Coeliac disease (CD) is defined as ‘an immune mediated systemic disorder elicited by ingestion of gluten and related prolamines in genetically susceptible individuals. It is characterised by the presence of a variable combination of gluten dependent clinical manifestations, CD specific antibodies, human leucocyte antigen HLA-DQ2 or HLA-DQ8 haplotypes and enteropathy’.1 Epidemiological studies from Europe, including the confidential Avon Longitudinal Study of Parents and Children birth cohort study from Bristol, UK, have highlighted a prevalence of 1% in the population.2 3

Small-bowel biopsies and histological changes as specified by the Marsh classification were considered as the gold standard for diagnosing CD in all children until 2012.4 In response to dietary gluten, the intestinal mucosa of a child with CD undergoes characteristic pathological changes: intraepithelial lymphocytosis, villous atrophy and crypt hyperplasia.1 5

Revised ESPGHAN guidelines

In 2012, European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) guidelines for diagnosing CD in children were revised allowing a selective group of symptomatic children to be diagnosed without small-bowel biopsies,1 provided following criteria are satisfied:

  • Anti-tTG titre more than 10 times the upper limit of normal (ULN)

  • Positive EMA

  • Positive HLA-DQ2/DQ8 genotype.

The British Society of Paediatric Gastroenterology, Hepatology and Nutrition and Coeliac UK have adopted the ESPGHAN guidelines with minor modifications. It suggests that in centres where EMA testing is unavailable, a second strongly positive anti-tTG can be used as alternative to diagnose CD serologically.5

All other children with suspicion of CD will still need to undergo an endoscopic assessment and small-bowel biopsies while remaining on a gluten containing diet. This group includes1 5

  • Symptomatic but anti-tTG titre less than 10 times of ULN

  • Asymptomatic but from high risk group (type 1 diabetes mellitus, first-degree relatives of CD cases, Down syndrome, etc) irrespective of anti-tTG titre

  • Symptomatic with anti-tTG titre greater than 10 times ULN but negative HLA DQ2 and/or DQ8 genotype.

These guidelines have been widely implemented in Europe and many countries outside Europe.6 7 In a prospective Italian cohort study with 468 children, diagnosis of CD was made serologically in 11% cases.6 During a follow-up of 1.9 years, this approach had no negative consequences relating to clinical remission, adherence to gluten free diet and quality of life.6 A similar encouraging outcome was reported in another prospective study from Canada where 40/88 children were diagnosed serologically.7

For implementation of the ESPGHAN guidelines, all laboratories need to report their anti-tTG titres quantitatively with clear ULN values. EMA and anti-tTG titres are both IgA-based and information on total IgA concentration becomes essential to identify false-negative anti-tTG titre results.8

Study background

In the South West of England, Bristol Royal Hospital for Children is the only tertiary referral centre for paediatric gastroenterology serving the 12 secondary care district general hospitals providing paediatric care. It was observed that the ULN for anti-tTG titres from different referring centres varied often requiring further enquiry for the ULN for each referral.

We anticipated that these variations would be a wider issue for the whole of the country. A national survey was therefore conducted to understand the diversity of the serological screening available for CD available from centres providing paediatric services in England.

Aims and objectives

The aim of this national survey was to describe the diversity of the serology screening tests for CD in England. The objectives were to determine

  • the number of laboratories providing anti-tTG titre testing in England

  • the type of anti-tTG titre reporting—quantitative (numeric value) or qualitative (positive/negative only)

  • the ULN for anti-tTG titre reported by individual laboratories

  • the type of assay used for the anti-tTG testing

  • the laboratories measuring both anti-tTG and EMA tests

  • the need for additional request for total IgA concentration when CD serology is ordered.


An electronic search of the National Health Service (NHS) database ( was conducted to identify the NHS Trusts providing acute care services in England. Out of 154 acute NHS Trusts identified, 139 were providing either inpatient and/or outpatient paediatric services and were included in the study. This study was conducted and approved as an external Student Selected Component project by the University of Bristol in 2015. As there were no confidential patient data collected, ethical approval was not considered necessary.

In July 2015, a telephone survey was conducted wherein the lead biomedical scientist in each laboratory was briefed by a single interviewer and responded to the structured telephone interview. These responses were then recorded in an electronic database. Figure 1 highlights the number of laboratories providing the various services relating to screening for CD.

Figure 1

Flow chart of the recruitment and results of the survey. BMS, biomedical scientist; EMA, antiendomysial antibodies; labs, laboratories; NHS, National Health Service; tTG, tissue transglutaminase.


Responses were available from 134 (96.4%) of 139 laboratories. Of 134 laboratories, 83 (62%) analysed anti-tTG samples ‘on-site’, and 40 (30%) sent their samples to another NHS laboratory within their region. Eleven laboratories (8%) sent their anti-tTG samples for analysis outside of their region and 2/11 were sending their samples to a private pathology laboratory. Three laboratories reported that they operate as a ‘screening laboratory’ and would send equivocal or positive anti-tTG sample results to another specialist laboratory for confirmation and additional testing. Figure 2 shows a map depicting NHS strategic health authorities in England and the number of laboratories in each region performing anti-tTG titre ‘on-site’ along with the range of anti-tTG titre values available from the laboratories based in that region. For example, 13 laboratories in London were reported as performing ‘on-site’ anti-tTG out of a total of 19 laboratories who participated in the survey and range of anti-tTG titres values ranged from 5 to 15 IU/mL.

Figure 2

Number of laboratories in England performing on-site anti-tTG titre testing. NHS, National Health Service; NI, number of labs responding to the interview in that region; NP, number of labs performing anti-tTG titres in that region; RT, range of anti-tTG titres offered by individual labs located in that region; tTG, tissue transglutaminase.

We identified four specific types of anti-tTG assays used by the laboratories across England, namely Phadia, Elisa, Bioflash and Multiplex. Phadia (n=42) and Elisa (n=33) are the most common among these. The use of these assays appeared to be randomly distributed among laboratories even within the same region. The ULN values are usually determined by the manufacturer’s recommendations, but it was important to note that there were significant variations in the ULN by different laboratories using the same assay. Table 1 highlights the different anti-tTG assays with their corresponding ULN values. The values for the ULN for anti-tTG titres ranged from 4 to 30 IU/mL. The most common value used nationally for ULN of anti-tTG titre was 10 IU/mL. Figure 3 represents the proportion of laboratories using different values for ULN.

Table 1

Anti-tissue transglutaminase assays used by different laboratories with their corresponding upper limit of normal values

Figure 3

Number of laboratories using each anti-tTG value. tTG, tissue transglutaminase.

Out of 83 laboratories providing ‘on-site’ anti-tTG testing, 80 (96.4%) reported the results quantitatively. However, the remaining three laboratories were still reporting the anti-tTG titre result qualitatively as either positive or negative. Sixty-eight (81.5%) of the 83 laboratories performed EMA testing ‘on-site’ and the remaining 15 sent their samples to another laboratory for analysis if requested by the clinician. One laboratory reportedly performs IgA-EMA as an initial screen and follows this with anti-tTG testing only on the positive samples.

Automatic reporting of total IgA concentration was available only from 24/83 laboratories (29.6%). Majority of the laboratories (56, 67.7%) require a separate request by the clinicians. Two of the 83 laboratories send their samples to another laboratory for total IgA only when a positive anti-tTG titre is detected.


This is the first national survey exploring the commonly performed serological tests for diagnosis of CD in England. This provides an interesting insight into the issues which significantly influence the screening and more importantly the correct diagnosis of CD in children based on the revised ESPGHAN guidelines. We explore these factors separately as follows:

Non-availability of on-site tests

We would anticipate that requests for coeliac serology would be quite frequent in all paediatric services. It was noted that 51 laboratories (38%) did not provide this ‘on-site’. Sending samples off site is likely to significantly delay the decision on the appropriate diagnostic pathway for CD.

Wide variation in anti-tTG titres and assays

The ULN varies widely between laboratories even when same assay methods (table 1) are used. Although the serological pathway accommodates for these variations, the range of 4–30 IU/mL needs interpretation for 10 times the ULN value which will therefore vary from 40 to 300 IU/mL. Critically, these variations occur within a region served by a single tertiary centre.

For example, the tertiary paediatric gastroenterology service in Bristol receives reports from 12 paediatric centres of the South West of England which are catered for by seven laboratories with ULN range of 4–20 IU/ml. A similar situation prevails in other regions in England (figure 2). It was interesting to note that the ULN varied even if the same assay method was used. Further to this, each laboratory applies sensitivity checks on the test, which can result in a change of the ULN.

Qualitative anti-tTG titre reporting by laboratories

For the paediatricians with access to qualitative results, the serological diagnostic pathway cannot be applied. The tertiary centre receiving this referral may have to either consider arranging an outpatient review and repeat quantitative anti-tTG titres or use the pathway of endoscopic duodenal biopsies. Delay in reaching a diagnosis will occur in both situations and the latter would have significant impact on endoscopy workload and cost to the NHS. For example, the approximate cost for biopsy-based diagnosis in our centre is £1300 whereas a serological diagnosis costs only £65.

Availability of EMA test

It was noted that 81.5% of laboratories conducting anti-tTG testing have access to EMA test on-site. Although this is a qualitative test, it has maintained its importance in the revised diagnostic pathway due to its high sensitivity. For the serological diagnosis in the absence of EMA result, a repeat anti-tTG testing is required which adds to the time delay in reaching the diagnosis.

Total IgA reporting

Majority of laboratories (67.7%) do not include the total IgA analysis automatically in their report for coeliac serology. If the clinicians do not request this specifically, they may potentially miss a diagnosis of CD in children with IgA deficiency. Some of the laboratories report performing a qualitative test to exclude IgA deficiency. However, the clinicians may not be made aware of this information. This may often lead to repeat anti-tTG tests and delay in diagnosis of CD.


Our survey performed 3 years after the introduction of the ESPGHAN guidelines on CD provides a unique perspective to the paediatrician and paediatric gastroenterologists responsible for diagnosis of CD. This highlights a number of factors as discussed above which can delay the clinicians receiving adequate serological results. The wide variations of the anti-tTG titre results are likely to impede making consistent decisions for appropriate diagnostic pathway to confirm the diagnosis of CD. Furthermore, qualitative reporting of anti-tTG titres by few laboratories will prevent implementation of the serological pathway. Non-reporting of IgA concentration by nearly two-thirds of laboratories poses the risk of missing a number of children with CD who also have coexistent IgA deficiency. All these factors are likely to affect the standards of care provided for children with CD.


We conclude that there is marked heterogeneity in serological tests available for confirmation of the diagnosis of CD across England. This creates significant barriers for the implementation of serological pathway for diagnosis of CD in children which is being widely used in the country. Standardisation of laboratory assay and ULN values across the NHS hospitals at least within the regional networks is necessary for safe implementation of the ESPGHAN guidelines for diagnosing CD in children.



  • Contributors All the authors have contributed equally to the study and in writing and revising the paper.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.