Evaluation of the repeatability of a crude adult indirect Ostertagia ostertagi ELISA and methods of expressing test results

An indirect enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies against Ostertagia ostertagi using a crude adult worm antigen was evaluated using serum and milk samples from adult cows, as well as from bulk tank milk. Within and between plate repeatabilities were determined. In addition, the effects of factors such as antigen batch, freezing, preserving of the samples and somatic cell counts (SCCs) of the samples were evaluated. Raw optical densities (ODs) and normalized values were compared using the concordance correlation coefficient (CCC), the coefficient of variation (CV), Bland-Altman plots (BA). Based on raw OD values, there was a high repeatability within a plate (CCC approximate to 0.96 and CV < 10%). Repeatability between plates was evaluated following normalization of OD values by four methods. Computing normalized values as (OD - Nt)/(Pst - Nt), gave the most repeatable results, with the CCC being approximately 0.95 and the CV approximate to 11%. When the OD values were higher than 1.2 and 0.3 for the positive and the negative controls, respectively, none of the normalization methods evaluated provided highly repeatable results and it was necessary to repeat the test. Two batches of the crude antigen preparation were evaluated for repeatability, and no difference was found (CCC = 0.96). The use of preservative (bronopol) did not affect test results, nor did freezing the samples for up to 8 months. A significant positive relationship between ELISA OD for milk samples and SCC score was found. Therefore, the use of composite milk samples, which have less variable SCC than samples taken from each quarter, would be more suitable when the udder health status is unknown. The analytical methods used to evaluate repeatability provided a practical way to select among normalization procedures. (C) 2002 Elsevier Science B.V. All rights reserved.