[Screening for celiac disease in children attending secondary school around Lucca]

A growing number of clinical and epidemiological data point to the fact that coeliac disease (CD) is a pauci- or asymptomatic occurrence, relatively more frequent than it was supposed in the past, manifested in atypical, silent and latent forms which are often undiagnosed or diagnosed only at a later age. The fact that resolutive treatment is now available means that CD is an ideal field for the application of a screening method. In this context the study by Catassi et al. (1994) is particularly important since it reported a prevalence of 0.38% of CD in healthy children in the Pesaro-Urbino area.

The aim of this study was to verify the incidence of CD in a nonselected pediatric population in the province of Lucca, using the aforesaid screening method.

The eligible population consisted of 1585 students from 5 secondary schools around Lucca, aged between 10 and 15 years old, none of whom were known to be affected by CD. In the first phase of the study anti-gliadin-IgA (AGA-IgA) and IgG antibodies were assayed in capillary blood (collected at school) using Alfa-Gliatest (Eurospital); children with AGA-IgA, AGA-IgG over 7 and 15 U/ml respectively were considered positive. In the second phase children with positive results underwent a further assay of AGA-IgA, AGA-IgG, anti-endomysium antibodies (EMA) and total IgA in venous blood. Lastly, children positive for AGA-IgA and/or EMA, or those positive for AGA-IgG with IgA deficit underwent duodenal jejunal biopsy.

41 children were positive on screening (2.6% of the eligible population, 3.8% of subjects effectively tested). Of these, 39 were assayed for AGA (IgA and IgG), EMA and total IgA in peripheral blood, identifying 4 subjects positive for AGA-IgA and EMA. Of the 4 children selected in this way, only 2 underwent jejunal biopsy and both presented "duodenal mucosa with chronic phlogosis and subtotal villous atrophy". Two cases of CD were formally ascertained with a prevalence of 1 out of 793 (0.13%) of the eligible population and an estimated prevalence of 1 out of 546.5 (0.18%) of the subjects undergoing screening. The cost was approximately Lit. 23,000 per child screened and approximately Lit. 6,100,000 for each coeliac child diagnosed.

The diagnostic iter proved efficacious and enabled 4 "high-risk" children to be selected. If the two subjects who did not undergo biopsy are also formally considered as coeliacs, the prevalence would be 1 out of 396 (0.25%) of the eligible subjects, namely 1 out of 273 (0.37%) of effectively tested subjects. This is a figure which is very similar to that reported by other studies. The 4 children identified here as strongly suspected of CD did not possess any anamnestic and/or objective elements which might have suggested "ex ante" a diagnosis of CD. If confirmed, these data provide concrete evidence of the need to perform mass screenings to identify CD. The economic convenience of this procedure depends on a careful analysis of the costs of the failure to diagnose CD.