Objectives: To review the recent literature on the diagnostic accuracy of skin prick tests (SPTs) in pediatric food allergy, focusing on adverse reactions to milk and beef. To present data about the test performance characteristics of beef extracts used in SPTs among children with atopic dermatitis (AD) reporting immediate hypersensitivity to beef.
Data sources: MEDLINE search using the following algorithm ["skin prick test" AND "food allergy" OR allergen; 1997-2002; English; all children]. Prospective sensitivity study of SPTs in 34 patients.
Study selection: Thirty-four children with AD (median age 2.29 years) were consecutively recruited between 1992 and 2000 because of immediate reactions to beef. On double-blind, placebo-controlled food challenges (entry criterion), 20 of the patients reacted to beef and 14 did not. Cut-off points for skin prick test wheal positivity was selected by receiving-operator characteristic analysis for fresh and commercial beef allergens. Sensitivity and specificity of skin tests and indices of reproducibility were calculated.
Results: In the literature, the positive predictive accuracies of skin prick tests vary between 69 and 100% and the negative predictive accuracies between 20 to 86% for cow's milk. In our series, SPTs with commercial beef extracts were highly diagnostic (100% sensitivity; 10% false positive rate) and SPTs with fresh beef were highly specific (100%), albeit with a false-positive rate of 21.42%.
Conclusions: From the literature, we conclude that the diagnostic accuracy of SPTs with milk should be reappraised in the workup of cow's milk allergy. Carrying out commercial and fresh food SPTs at the same time substantially reduces costs and diagnostic work. Oral provocation is necessary in 20.68% of children with AD who have immediate symptoms to beef. Greater allergen standardization and streamlining of the workup of cow's milk allergy are desirable future goals.