Reproducibility of food atopy patch tests over time in the general child population

Background  The atopy patch test (APT) is no longer an experimental method; it is increasingly being used as a standard diagnostic tool for the characterization of patients with aeroallergen- and food-triggered disorders. Some technical aspects of this test still remain to be answered. We aimed to study the reproducibility of this test over time in the general child population.
Methods  In a general population of 118 children, we investigated the reproducibility of duplicate APTs with four food allergens in their native form, which were repeated at set intervals from the first test: 7 days (group 1), 14 days (group 2), and 21 days (group 3).
Results  We observed very poor reproducibility on both sides of the back in all three studied subgroups. The reproducibility rates and Cohen's κ values did not improve when we did not consider the side of the back. There were no differences in the prevalence of atopy between the subjects with reproducible and nonreproducible APT results. All three groups studied showed no difference in the prevalence rates of atopy. There was no relationship between APT and skin prick test positivity for the same allergen. Questionnaire-derived data about previous food-related reactions did not help in the evaluation of the doubtful nonreproducible APT results with food allergens.
Conclusions  Our results show that the reproducibility of food APTs is poor and unsatisfactory over time, and there is an urgent need for the development of optimal, stable, and good-quality APT testing substances.