Accuracy of patient prediction of allergy skin test results

Background: The ability of patients to self-diagnose allergy is unknown. Objective: To estimate the ability of patients to correctly predict the results of allergy skin tests. Methods: We conducted a structured interview of 86 patients with chronic rhinitis or asthma undergoing aeroallergen skin tests. We asked, "Do you expect the skin tests to be positive or negative?" and "What do you expect the allergy tests to be positive for?" Responses to these questions were correlated with the results of aeroallergen skin tests. Skin tests were performed using the prick technique and included cat, grass pollen, tree pollen, weed pollen, dust mites, and molds. Results: Seventy-three participants provided usable responses. Of those with a positive skin test, the number (percentage) of participants who predicted correctly was 10/18 (56%) for cat, 4/14 (29%) for tree, 7/26 (27%) for weeds, 5/23 (22%) for dust mite, 2/12 (17%) for grass, and 1/8 (12%) for mold. Of those with a negative skin test, the number (percentage) of participants who predicted correctly was 47/50 (94%) for dust mite, 51/59 (86%) for trees, 56/65 (86%) for mold, 52/61 (85%) for grass, 45/55 (82%) for cat, and 38/47 (81%) for weeds. Conclusions: (1) Patients have limited ability to correctly predict positive skin tests to aeroallergen. (2) Patients are able to predict negative skin tests with reasonable accuracy. (3) "What do you think you are allergic to?" may be a good screening question for patients with asthma and rhinitis.