Gold standard is double blind challenge, but who has time for that? And dangerous!
Mostly based on history – combination of characteristic features without other, more likely, explanation.
EAACI guidance 2023 says where type 1 allergy suspected (signs/symptoms but also timing and consistency of reaction):
- Do skin prick testing and/or specific IgE testing as first line
- After that, if still doubt then for peanut, hazelnut or cashew, if in doubt do component tests Ara h 2, Cor a 14, Ana o 3 respectively (if available) – otherwise do skin prick or IgE if not done already.
- Where peanut or sesame allergy still in doubt, do basophil activation test (BAT – if available – nowhere in Scotland, as far as I know)
- “Reassessment of food allergic children, at regular intervals, depending on age, food and patient’s history, is suggested for possible development of spontaneous tolerance”
Ara h 2 (cut off 0.44) has 82% sensitivity and 92% specificity cf 84 and 86% for SPT of 4mm, so equivalent. Cor a 14 (cut off 0.64) has 73 and 95%, so not great sensitivity. Ana o 3 (cut off 0.4) pretty good – 96 and 94%.
If random reactions, then consider hidden allergens: celery, mustard, cochineal, lupin, soy, fenugreek, other legumes such as pea/bean/lentil protein, insects/mealworm, pink peppercorns.
Panel tests
=multiple specific IgE tests done at the same time (the ultimate being the ALEX3 test, where 280+ different antigens are tested simultaneously) – likely reduced sensitivity, compared with individual test, but more importantly, potential for false positives, with attendant harms (including iatrogenic food allergy, if that food then avoided unnecessarily).
Advances of Alex3 over Alex2 include new molecular allergens including alpha-Gal, chicken (Gal d 7), celery (Api g 7, predicts anaphylaxis), wheat (Tri a 36, and Tri a 37), oak pollen, and honeybee venom; inclusion of lentil and pea.