Introduction The prevalence of coeliac disease in the UK is 1%, but only approximately 1 in 6 patients have been diagnosed.¹NICE Guidelines give a list of indications for testing for coeliac disease.²12 months of coeliac serology (anti-tTG) requests in our local laboratory were reviewed to assess variation in serology requesting and detection of coeliac disease across 23 local GP practices. This was compared with requests from secondary care.

Method All anti-tTG requests, with clinical indications, between December 2013 and December 2014 were obtained from the York Hospital (YH) Laboratory. All positive D2 biopsies were obtained from the YH Pathology Database and were cross-referenced to obtain the crude rate of referral for D2 biopsy. Only newly diagnosed patients were analysed. Categorical data was analysed by means of a Chi-squared contingency table with Yates’ correction. A two-tailed P value of <0.05 was considered significant.

Results There were a total of 15,183 (male 36%) anti-tTG requests – 11,321 from primary care and 3,862 from secondary care. The most common indications for serological testing were “anaemia” (13%), “no details given” (11.3%), “diarrhoea” (10.5%) and “abdominal pain” (9.2%). “Osteoporosis”, osteomalacia”, “fracture”, or similar was given as a clinical indication for 0% of the requests. 133 (1.2%) of the requests from primary care yielded a positive result compared with 96 (2.5%) of the requests from secondary care (P < 0.0001). Of these 229 patients (38% male), 204 (89%) went on to have a D2 biopsy (mean waiting time 6.6 weeks). Of those biopsied, 191 (94%) had coeliac disease confirmed. Percentage of total practice list size tested for coeliac serology varied from 0.05% to 21.4% with percentage testing positive varying from 0% to 6.51% (see Table 1).

Conclusion There is poor selection of patients for coeliac serology testing in the community. Better education for checking coeliac serology in poor bone health is needed. The crude positive serology rate in the community was no better than random sampling from the population. This data suggests either the wrong patients are being tested or those with undiagnosed coeliac disease do not have regular contact with primary care. The reasons for variation in crude testing rates need to be explored further.

Disclosure of interest None Declared.

References

1. West, et al. Am J Gastroenterol. 2014

2. NICE CG86. May 2009