J Brasch1, W Uter, J Geier, A Schnuch. 1. Department of Dermatology, University of Kiel, Kiel, Germany. jbrasch@dermatology.uni-kiel.de
Abstract
BACKGROUND: Patch testing with a standard allergen series often yields positive reactions to more than 1 allergen in a patient. OBJECTIVE: To identify all significantly associated pairs of positive reactions and to assess their relation to the strength of the reactions and to the irritative potential of the allergens. METHODS: Based on the filed data of 57,822 patients, associations between positive reactions to 2 different allergens were quantified with odds ratios. Statistical methods included Fisher's exact test, the Bonferroni adjustment to account for the effect of multiple testing, and the Spearman rank correlation. RESULTS: Out of the 32,779 patients with complete readings of 24 standard allergens, 7,501 had shown more than 1 positive reaction. Statistically significant associations were detected for 166 out of the 276 possible different combinations of 2 distinct positive reactions, including combinations that had not been identified before. Patients with a strong reaction or a positive reaction to an allergen with a high irritative potential tended to have additional positive reactions to further allergens more often than others, but the number of significant associations was not dependent on these parameters. CONCLUSION: There are more significant associations that have to be taken into account for patch testing than has been known so far. Although irritation can favor a higher number of positive reactions, significant associations of positive reactions to distinct allergens are probably caused by other mechanisms that require further analyses. Copyright 2001 by W.B. Saunders Company
BACKGROUND: Patch testing with a standard allergen series often yields positive reactions to more than 1 allergen in a patient. OBJECTIVE: To identify all significantly associated pairs of positive reactions and to assess their relation to the strength of the reactions and to the irritative potential of the allergens. METHODS: Based on the filed data of 57,822 patients, associations between positive reactions to 2 different allergens were quantified with odds ratios. Statistical methods included Fisher's exact test, the Bonferroni adjustment to account for the effect of multiple testing, and the Spearman rank correlation. RESULTS: Out of the 32,779 patients with complete readings of 24 standard allergens, 7,501 had shown more than 1 positive reaction. Statistically significant associations were detected for 166 out of the 276 possible different combinations of 2 distinct positive reactions, including combinations that had not been identified before. Patients with a strong reaction or a positive reaction to an allergen with a high irritative potential tended to have additional positive reactions to further allergens more often than others, but the number of significant associations was not dependent on these parameters. CONCLUSION: There are more significant associations that have to be taken into account for patch testing than has been known so far. Although irritation can favor a higher number of positive reactions, significant associations of positive reactions to distinct allergens are probably caused by other mechanisms that require further analyses. Copyright 2001 by W.B. Saunders Company