Methods: A multicenter prospective observation study was performed and 433 children were enrolled, comprising of groups A (n = 220; 1 year old) and B (n = 213; 2-6 years old). OFC was performed with cooked egg powder for all the participants and with raw egg powder for patients who passed the initial cooked egg challenge. sIgE values to egg white (EW) and ovomucoid (OM) were measured by 3g and CAP. The relationship between the sIgE and OFC outcomes was analyzed using ROC and logistic regression analyses. Fitted predicted probability curves (PCs) were plotted using logistic regression.
Results: Out of 433 children, 243 (56%) passed cooked egg challenge. From the subset of 243 who passed the challenge, 113 children were given the raw egg challenge. Out of these 113 children, 55 (49%) passed the challenge. The two sIgE tests in the cooked egg challenge group showed similar performance, with area under the curve (AUC) of 0.791 and 0.778 in group A, 0.848 and 0.828 in group B, for OM sIgE with 3g and CAP, respectively. However, 3g showed higher cut-off values than CAP, at 21.5 IUA/L and 3.1 UA/L in group A and 45.4 IUA/L and 9.0 UA/L in group B, respectively. The PCs of EW and OM sIgE by 3g showed a wider distribution of data across the x-axis (sIgE values) and y-axis (probability) in comparison to CAP. In the raw egg challenge group, EW sIgE tests showed similar performance with AUCs at 0.678 and 0.671 in group A and 0.771 and 0.751 in group B with 3g and CAP, respectively. However, PCs for raw egg challenge showed significantly larger 95% confident intervals than those for cooked egg challenge.
Conclusion: Overall results from 3g demonstrated similar but slightly better performance than CAP in predicting the OFC outcomes.