Assessment of the within- and between-laboratory repeatability of a commercially available Ostertagia ostertagi milk ELISA

Abstract

The objectives of this study were (1) to assess the within- and between-laboratory repeatability of a commercially available antibody-detection Ostertagia ostertagi ELISA (SVANOVIR(O). ostertagi-Ab, Svanova, Uppsala) and (2) to investigate if the assay could be further simplified by reading optical density at a single instead of double wavelength. A total of 80 bulk-tank milk samples were divided into aliquots and tested in duplicate per ELISA plate on 3 different days in 4 different laboratories (Bristol, Ghent, Hannover, Uppsala). The within- and between-laboratory repeatability of the ELISA was assessed by random effect models and the amount of variance attributable to each source of variation (duplicate within plate, day, laboratory and sample) was expressed as a proportion of the total between-sample variance. Overall, the within-laboratory repeatability was good in each laboratory with a proportion of total between-sample variance that could be attributed to assay variability of 5% in Ghent to 27% in Bristol. The between-laboratory repeatability was high: 15% of the total between-sample variance was attributable to the duplicate, 1% to the day, 2% to the laboratory and 82% to the sample. The range of deviations expected to include 95% of the observations when the same sample is tested in different laboratories was -0.23 to 0.23. The mean difference between the ODR values when the optical density was read at a single or double wavelength was -0.002 and the 95% confidence interval included zero. This study demonstrates that the SVANOVIR(O). ostertagi ELISA has a good repeatability and can be further simplified by reading optical density at a single instead of double wavelength. However, the observed variations in the within-laboratory repeatabilities suggest that regular ring testing is necessary when different laboratories cooperate in a same monitoring programme.