Skin prick testing a better predictor than blood testing for the diagnosis of peanut allergy in Chinese children.

Abstract

Peanut allergy is common in Chinese children, yet the most predictive diagnostic cut-offs for skin prick test (SPT) and blood testing in this population are unclear. We aimed to determine the optimal cut-off values for whole-peanut SPT, specific IgE (sIgE) and component-resolved diagnostics (CRD) for Chinese children based on outcomes of open oral food challenges (OFC) to peanut. We recruited ethnic-Chinese patients 1-18 years old who were suspected of having peanut allergy based on a history of reactions after exposure or sensitization although peanut naïve. Considering the AUC value of 0.8, 80% power and 5% level of significance with two tails, 26 patients were needed. Sensitivities, specificities, positive and negative predictive values, and receiver operating characteristic curves (ROCs) and their area-under-curves (AUCs) for SPT, peanut sIgE, and CRD were compared. Thirty-one subjects participated. Only SPT reached statistical significance (AUC 0.91, p = 0.0001), but not the other tests. Seven retrospective data were added to optimize the power. SPT remained to be the best predictor, followed by Ara h 2 sIgE (AUC 0.72, p = 0.02). An SPT wheal size of 3 mm and Ara h 2 sIgE of 0.14 kU(A)/L yielded the highest Youden's index. The specificity of SPT and Ara h 2 sIgE reached 94% at 6 mm and 0.74 kU(A)/L, respectively. Comparisons of ROCs revealed that SPT was significantly better than Ara h 2 sIgE (p = 0.03) and whole-peanut sIgE (AUC 0.61, p = 0.26). In Chinese children, SPT appeared to be the best predictor for peanut allergy, followed by Ara h 2 sIgE.