Reproducibility of a commercial in vitro allergen-specific assay for immunoglobulin E in dogs

Samples of serum taken from 42 dogs with clinical signs and histories indicating pruritic skin disease and/or diseases of the ear were tested in a commercial allergen-specific assay for immunoglobulin E. Dogs which had been treated with glucocorticoids and/or antihistamines were not excluded. The samples were separated into two equal aliquots, given different randomised numbers, and analysed in two batches on two separate days. The laboratory was blinded to the identification numbers and the history of each dog, but knew the purpose of the study. The results for 48 allergens were expressed in modified absorbance units (MAU). The overall median MAU was 29. For each allergen the mean difference between the MAU values of the paired duplicates was determined and the difference was compared to zero by a paired t test. The number of means that were not 0 (P < 0.05) in each allergen group were: seven of 10 grasses, seven of nine weeds, two of 13 trees, six of 10 fungi, and three of six environmental allergens. A single 2 x 2 table for the 48 allergens was created with MAU >= 60 defined as 'positive' and < 60 as 'negative'. There were 116 of 188 (62 per cent) pairs that were reproducibly 'positive' and 1756 of 1828 (96 per cent) pairs that were reproducibly 'negative'.