Negative predictive value of skin prick tests for predicting IgE-mediated food allergy is greater than 95%. This means that there is still a small percentage (~5%) of patients with negative skin tests who may still have IgE-mediated food allergy.
Generally, try to avoid using the term "negative" IgE test, and use the term "undetectable". It is important to note that in 10–25% of food allergy reactions, sIgE can be undetectable by blood test (http://buff.ly/1e7HAmY).
8 foods cause 90% of food allergies (click to enlarge the image). The likelihood of a negative oral food challenge is shown in relation to the respective values of skin prick test (SPT) and serum IgE (sIgE):
Annual testing is often the practice for determining whether allergy to milk, egg, soy, and wheat have been outgrown, and the testing interval is extended to 2 to 3 years for allergy to peanut, tree nuts, fish, and crustacean shellfish. However, the 2010 Food Allergy Guidelines noted that these testing schedules are not supported by objective evidence.
Skin testing has approximately 30% greater sensitivity than serum IgE (Cleveland Clinic J of Med, 2011, 78-9, 585-592), and a negative skin test typically rules out clinical environmental or food allergy in approximately 90% of patients.
Comparison of diagnostic methods for peanut, egg, and milk allergy - skin prick test (SPT) vs. specific IgE (sIgE) (click to see the spreadsheet). Sensitivity of blood allergy testing is 25-30% lower than that of skin testing, based on comparative studies (CCJM 2011).
Inability to eat peanuts in spite of negative skin tests and serum-specific IgE testing. What to do? AAAAI Ask the Expert, 2012.
Comments from Twitter:
Stuart Carr @allergydoc4kidz: NPV of SPT (good extract or fresh food) very high, but less so for sIgE, where undetectable result much less meaningful. Also need to avoid suggesting positive SPT alone confirms allergy, or wheal size predicts severity of reaction.