ABSTRACT
A serological survey on bovine brucellosis was carried out 3 times between 2007 and 2009 in 3 districts (Kiboga, Mpigi and Kiruhura) in western Uganda and 2 (Kumi and Mbale) in the east employing the rose bengal test (RBT) for infected-herd screening and an indirect ELISA (iELISA) for testing the serostatus of individual animals. The animal prevalence was significantly higher in the 3 districts of the west (mean 21.5% in 2009) compared with the 2 districts (mean 3.4% in 2008) in the east (P<0.0001), though a significant difference was not observed between Kumi and Mpigi in 2008. In the west, it was the lowest in Mpigi, but a significant increase was observed between 2008 (5.3%) and 2009 (30.0%), as in Kiruhura, in which the prevalence increased from 8.1% in 2007 to 16.8% in 2009. A similar trend was also observed in Kumi, namely, the seropositivity significantly increased from 2.3% in 2007 to 6.2% in 2008 and became remarkably higher than in Mbale (0.64%). As a result, the farm prevalence was also higher in the west, especially in Kiboga in 2007 (77.8%) and 2008 (65.6%), and Mpigi in 2009 (70.8%). The linear predictor of the fitted generalized linear model proved that the logit of RBT positivity increased linearly over the increase in percent positivity values. This study demonstrated an example of an unaided self-help survey as one of the control measures in Uganda.
