Application of repeated bulk milk testing for identification of infection dynamics of Neospora caninum in Thai dairy herds

In this prospective study, repeated bulk milk tests were evaluated to establish the Neospora caninum antibody status and to describe infection dynamics and changes in prevalence of N. caninum in Thai dairy herds. Bulk milk from 418 herds in northeast Thailand were collected at three consecutive samplings, in December 2002 (sampling 1), April (sampling 2) and December 2003 (sampling 3). All samples were analysed for presence of N. caninum antibodies by iscom ELISA, and the optical density values were dichotomized at three different cut-offs. Herd status at either sampling 1 or sampling 2 was used to predict herd status at sampling 3. Changes in both sensitivity and specificity at the first samplings at all cut-offs were consistent with commonly seen patterns when evaluating performance of diagnostic tests. In addition, the predictive positive values (PPV) of herd status at each of the two samplings were more affected by time of sampling than choice of cut-off, whereas the predictive negative values (PNV) increased with increasing cut-off. Herd status at the first samplings was also interpreted in combination, i.e. herds that were negative at both samplings were considered negative and positive otherwise in Combination A, while herds positive at both samplings were considered positive in Combination B. When using these combinations, Combination A gave higher PNV but lower PPV than Combination B at all cut-offs. Using the optical density value 0.20 as cut-off to establish the herd N. caninum antibody status, 270 herds (65%) kept their status throughout the study period, but 148 herds converted their status at least once. Interestingly, 136 (86%) of the 158 herds that were considered negative at sampling 1 and sampling 2 remained negative at sampling 3. In addition of the 159 herds considered positive at the first two samplings, 134 (84%) were still positive at sampling 3. We concluded that repeated bulk milk testing at regular intervals provided better information about herd N. caninum status than a single test. Further, our results showed that the infection is prevalent in northeast Thailand, but that a herd can keep a negative infection status despite the frequent presence of dogs.