Study Area. Baringo County is situated in the Rift Valley Region and shares borders with 8 counties namely, West Pokot to the North West, Turkana to the North, Samburu to the North East, Laikipia to the East, Nakuru to the South, Kericho and Uasin Gishu Counties to the South West, and Elgeyo Marakwet to the West. The County is divided into 6 Sub-Counties, namely Baringo South, Mogotio, Eldama Ravine, Baringo Central, Baringo North and Tiaty. It is predominantly inhabited by the Tugen, Pokot and Ilchamus ethnic groups, who are livestock keepers; with minority groups such as Endorois, Nubians, Ogiek, Kikuyu and Turkana. The Tugens mostly practice agro-pastoralism. This mixture of land use allows for complex human animal interactions and interphase usually compounded by the high population density and diversity [12]. It is these complex dynamics that our study was aiming to unravel with respect to brucellosis. Baringo County is classified as arid and semi-arid regions. Most parts of East Pokot, Baringo Central, Baringo South, Baringo North and Mogotio sub-counties are arid and semi-arid except for Eldama Ravine Sub-County, which is a highland zone. The rainfall varies from 1000 mm to 1500 mm in the highlands to 600 mm per annum in the lowlands. The sub-counties due to their varied altitudes receive different levels of rainfall. Eldama Ravine Sub-County receives the highest amount of rainfall. The lowlands sub-counties of Mogotio, East Pokot and Baringo North receive up to 600 mm of rainfall per year. The region is occupied by nomadic communities that qualify it as at higher risk region for the prevalence of brucellosis [27].
Study Design. In a cross-sectional study involving quantitative approach of data collection in Baringo County, 8 locations in Koibatek and Marigat sub-counties were targeted: Torongo, Koibatek, Ravine, Lembus Kwen, Marigat, Eldume, Kimalel and Loboi. The study targeted household heads and domestic ruminants; cattle, goats, sheep and camels. Livestock keepers who were in close contact with the animals were the key respondents for interview and were enrolled as study participants.
Study Population and Sampling Procedures
Study Population. The study population consisted of farmers, herders and their livestock (sheep, goats, cattle and camels) from Koibatek and Marigat Sub-counties in Baringo County who were in close contact with their livestock. Majority of livestock were predominantly owned by pastoralists who migrate throughout the dry season looking for water and pastures in small groups of families or in large groups of villagers.
Sample Size Determination. The sample size was determined by Cochran formula (1977) which allowed for the calculation of an ideal sample size given a desired level of precision of 5%, 95% confidence interval and 30% estimated proportion of attribute present in the population. Based on these estimates, 640 animal blood samples (322 cattle, 105 camels, 155 goats and 58 sheep) in 640 households with 640 human blood samples were collected.
Sampling Procedure. Probability sampling techniques using cluster and simple random sampling methods were used to practically access households that had domestic ruminants. A random sample of 50 villages was selected using a table of random numbers, which gave 30 pastoral villages from Marigat and 20 agro-pastoral villages from Koibatek. The data collection team comprising of laboratorians and veterinary officers who worked in pairs were recruited for purposes of data collection. The team was trained on the research protocol, data collection instruments and ethical issues pertaining to the study. Pretesting of the data collection instruments was done in one of the nearby villages that were excluded from the study. The data collection tools were then customized wherever necessary prior to actual administration. A pair of two enumerators covered at least one village in a day to administer a minimum of 8 questionnaires at random and these were uploaded in Open Data Kit (ODK) in real-time. Once the team was within the prescribed geocode, the compound to be assessed was identified using the ‘spin bottle method.’ Using a flat surface, the enumerator could spin the bottle until it settled and take the direction facing the mouth of the bottle until he/she reached a household with a domestic ruminant.
The first household in that direction was selected and the team administered their questionnaire following informed consent. Once the enumerator finished administration of the questionnaire, he or she stood at the door of the just completed house and spun the bottle again to pick the direction of the mouth of the next bottle. The enumerator again walked to the next household until all eight eligible households were interviewed. An enumerator that reached the end of the village before completing the numbers required, went back to the center of the village, and span the bottle once again. In case the enumerator double selected the previous household, that household was excluded, and the exercise was repeated until another eligible household with domestic ruminants was selected.
The questionnaire captured data on socio-demographics, animal ownership data such as handling, animal product consumption of animal products practices, health-seeking behavior and awareness of brucellosis. For every sampled animal, individual animal level risk aspects were captured; sex, age, method of production, history of abortion, breed, herd size, animal management system, vaccination history, history of introduction of new animals into the herd and presentation of brucellosis like symptoms. This dataset was reported in our previous publication [12]. The assessed households were sensitized on clinical presentation and on signs and symptoms associated with brucellosis and asked to present themselves to nearby health facilities distributed across the study region for medical assistance should they present with the sensitized signs or symptoms associated with brucellosis.
Sample Collection Procedure for Animals
Methods of Data Collection for animals. Livestock selected for sample collection were individually restrained and 5 ml blood collected in plain vacuum plastic tubes (Vacutainer®). Using the halter, the head was elevated slightly, drawn to the side opposite the jugular vein to be sampled and tied to a stationary surface. The vein was occluded by digital pressure in the jugular groove low in the neck. A vacutainer needle, attached to a vacutainer holder, was placed into the distended jugular vein at approximately a 450 angle (angle varies with depth of vein) cranial to the occluding digit. When positioned in the vein, a vacutainer was inserted onto the needle and a blood sample collected. When the desired volume was collected, the occluding pressure was removed. The tube was detached from the needle and the needle removed from the vein. The samples were then labelled and transported in a cool box and ice packs (~4°C) to the hospital laboratory, where they were centrifuged on the same day of collection at 5000 rpm for 5 minutes to obtain serum. The serum and blood samples were transported in freezers and stored at −20°C till testing at Maseno University Laboratory, Kisumu, Kenya.
Methods of Data Collection for humans. Prior to bleeding, the medial site of the elbow region was disinfected using cotton wool soaked in methylated spirit. Blood was aseptically collected from the brachial vein using a disposable 5ml syringe by laboratory personnel. The blood was immediately transferred into a plain vacutainer (Red top) and assigned a unique identification number. After centrifugation, the serum was kept at -20 0C in a freezer till analysis. Approximately 1 ml extracted serum from each human subject was aliquoted into cryotubes for this study. All human sera were also transported at −20°C until further molecular and serological testing at Maseno University Laboratory, Kisumu, Kenya.
Other Methods of Data Collection. Quantitative data was collected using the Open Data Kit (ODK) software that captured the demographic characteristics, location signs and symptoms of brucellosis, abortions, treatment, perceived socio-economic effects on livestock production and reproduction performance [12]. These were pre-tested and customized accordingly prior to actual administration. The data collection exercise was conducted in Tugen, the local language, Kiswahili or in special cases, where the respondent was knowledgeable, in English.
Questionnaire interview method. A brief structured questionnaire was administered to the head of the household by an enumerator for a period of 35 min and covered specifically animal’s sex, location, history of abortion and retained placenta. For human sampling, information on the age, gender, and location of residence of each sampled human participant was recorded.
Blood Sample Testing Rationale. This study’s overall aim was focused on the serological and molecular epidemiology of Brucella in Baringo County, Kenya. Therefore, the initial laboratory procedure was to establish the serological c-ELISA and thereafter, perform DNA extraction techniques for all positive samples for PCR testing to detect the genus Brucella and to identify Brucella species. Results for the initial serological c-ELISA guided the subsequent molecular tests and analysis for Brucella species in humans and livestock in Koibatek and Marigat sub-counties of Baringo County, Kenya.
Brucella Antibody ELISA. The animal Brucella antibody tests were performed using the PrioCHECK® Brucella Antibody 2.0 ELISA (ThermoFisher Scientific, USA), while human Brucella was performed using the Human Brucella-IgM ELISA kit (MyBiosource, USA), as per manufacturer’s instructions. These were indirect ELISA for the detection of antibodies against B. abortus and B. melitensis in the serum obtained from humans, cattle, camels, sheep and goats. In brief, the serum samples were dispensed in the coated wells of a microtiter plate. Antibodies directed against B. abortus and B. melitensis, present in the test sample, were bound to the antigen during incubation. The bound antibodies were then detected using an anti-Ig monoclonal antibody, conjugated to an enzyme that generated a color signal. The color development occurred when specific antibodies against B. abortus or B. melitensis were present in the test sample, which then indicated the presence of Brucella antibodies in the sample.
Molecular Analysis and speciation
DNA extraction. DNA extraction from the human samples were performed using the DNeasy® Blood &Tissue Kit (Cat #s 69504 and 69506) as recommended by the manufacturer (ThermoFisher Scientific, USA). About 20 µl of proteinase K was pipetted into a 2 ml microcentrifuge tube and 100µl anti-coagulated treated blood added with the volume adjusted to 220 µl with PBS. About 200 µl of buffer was added and mixed thoroughly by vortexing and blood samples incubated at 560c for 10 minutes. Another 200 µl of ethanol was added and mixed thoroughly by vortexing. The mixture was pipetted into a DNeasy® mini-spin column placed in a 2ml collection tube and centrifuged at >6000rpm for 1 minute and the flow through and collection tube discarded. The spin column was placed in a new 2ml microcentrifuge tube. The DNA was eluted by adding 200 µl of buffer AE to the center of the spin column membrane and incubated for 1 minute at room temperature. Finally, this was centrifuged for 1 minute at >6000rpm (Quick Start Protocol). The DNA quality and quantity for each of the extracts were assessed using a NanoDrop 2000c Spectrophotometer (ThermoFisher Scientific, USA) before being stored at − 20°C until PCR was done.
Real-Time PCR. The real-time PCR was done to detect the genus Brucella in the extracted DNA samples using primers and probe nucleotide sequences that target the Bcs31 gene. The PCR procedure was done as previously described [28]. A sample that generated a clear amplification plot and a corresponding threshold value lower than 40 in one or all the duplicate samples was considered as positive for Brucella. A further speciation assay was performed on the entire genus for PCR positive samples using the B. abortus and B. melitensis-specific oligonucleotide probes and primers. Similarly, only samples that had both clear amplification plot and a Ct value < 40 was considered as positive for any of the two Brucella species.
Ethical Considerations. Ethical approval was obtained from Maseno University Ethical Review Committee (REF: MSU/DRPI/MUERC/00600/18) and permission to conduct research in Baringo County was obtained from National Commission for Science, Technology, and Innovation permit No. NACOSTI/P/18/4661/26645. Introductory letter to the Chief Officer Livestock and Fisheries for Kabarnet was obtained from the office of the Dean, Maseno University. Research authorization was obtained from Sub-County Health Coordinators: Eldama Ravine and Baringo South through Chief Officers of Medical Services and Preventive and Promotive Services for Baringo County. Informed written consent was obtained from study participants and both privacy and confidentiality were guaranteed throughout the study period. The right to participate in the study was voluntary and the participant enjoyed the right to withdraw at any time without penalty. For confidentiality purposes, unique codes were used for both humans and domestic ruminants in the households where samples and data were drawn.