Introduction Brucellosis is an important zoonotic disease in Kenya, and identifying the bacteria in milk is important in assessing the risk of exposure in people. Methods A cross-sectional study that involved 175 households was implemented in the pastoral counties of Marsabit and Isiolo in Kenya. Pooled milk samples (n = 164) were collected at the household level, and another 372 were collected from domesticated lactating animals (312 goats, 7 sheep, 50 cattle and 3 camels). Real-time polymerase chain reaction (qPCR) testing of the milk samples was performed to identify Brucella species. Brucella anti-LPS IgG antibodies were also detected in bovine milk samples using an indirect enzyme-linked immunosorbent assay (ELISA). Results Based on the qPCR, the prevalence of the pathogen at the animal level (considering samples from individual animals) was 2.4% (95% confidence interval (CI) 1.1–4.5) and 3.0% (CI: 1.0–7.0) in pooled samples. All 14 samples found positive by qPCR were from goats, with 10 contaminated with B. abortus and 4 with B. melitensis. The Brucella spp. antibody prevalence in bovine milk using the milk ELISA was 26.0% (95% CI: 14.6–40.3) in individual animal samples and 46.3% (95% CI: 30.7–62.6) in pooled samples. Conclusion The study is the first in Kenya to test for Brucella spp. directly from milk using qPCR without culturing for the bacteria. It also detected B. abortus in goats, suggesting transmission of brucellosis between cattle and goats. The high prevalence of Brucella spp. is a significant public health risk, and there is a need for intervention strategies necessary in the study area.
Brucella . Milk . ELISA . PCR . Pastoral