Multiple patch-test reactions: a pilot evaluation of a combination approach to visualize patterns of multiple sensitivity in patch-test databases and a proposal for a multiple sensitivity index.

Background: The variety of patterns of multiple sensitivity in patch-test data remains poorly defined. Studies addressing this topic have primarily concerned the occurrence of pairs of allergens, and some reports hypothesize a predisposing factor that influences the individual sensitivity of patients to multiple reactions.

Objective: The aim of this study was to address this topic by reanalyzing a matched data set from two patch-test units in Cleveland, OH, and one unit in Cologne, Germany, focusing on multiple reactions to identical allergens of the standard screening trays over the same 4-year period.

Methods: Based on the statistical FREQ procedure (SAS Institute Inc., Cary, NC), we propose a program for addressing the difficulties in computing and visualizing patterns of multiple sensitivity. Additionally, we propose the "multiple sensitivity index" (MSI) as an absolute measure for characterizing the occurrence of an allergen with others on a selected panel.

Results: For the seventeen allergens examined, 131072 possible combinations were evaluated in a total of 2881 patients. Of patients tested, 12.4% had multiple positive patch-test reactions to two to seven allergens. However, because of the small number of patients with the crucial number of possible combinations, no cluster patterns were evident in the three- to seven-allergen combinations. Pairs of allergens most frequently observed were nickel sulfate and potassium dichromate (n = 23), formaldehyde and quaternium-15 (n = 18), and nickel sulfate and formaldehyde (n = 13). We found that nickel sulfate, once again the most frequent sensitizer, occurred in the majority of noncombined cases (MSI = -0.280).

Conclusion: Larger patch-test databases require evaluation to obtain further evidence of cluster patterns of multiple sensitivity and to validate the MSI.