Factors associated with brucellosis among patients attending selected hospitals in Kajiado County, Kenya

Abstract

Brucellosis is a bacterial zoonotic disease with serious public health implications and high socio-economic burden among afflicted populations. Brucellosis is a priority zoonotic infection in Kenya but there is limited information on the true burden of the disease due to weak surveillance. Livestock rearing and trade is a key source of livelihood for majority of residents of Kajiado County. The close linkage between pastoralists and livestock, characterized by high dependence on livestock products increase the risk of spillover of infection between humans and animals. Identifying and understanding the modifiable risk factors for brucellosis in necessary to inform public health interventions. There are few studies in Kenya which are address brucellosis risk factors through incident cases. The objective of the study was to describe socio-demographic and clinical characteristics of brucellosis cases in Kajiado, determine risk factors for infection among patients attending select health facilities in Kajiado County and to evaluate the performance of a brucellosis rapid diagnostic kit (Brucella agglutination test) through comparative testing with Enzyme-Linked Immunosorbent Assay test (ELISA). The study area was Mashuru sub-county in Kajiado and the study design was an unmatched, hospital-based case-control study. Patients with fever and two or more clinical features descriptive of brucellosis as per the World Health Organization case definition and a positive ELISA test were classified as cases. Data on sociodemographic, clinical and occupational characteristics were collected using a structured questionnaire and analyzed using Epi info version 7 software. Descriptive data was analyzed using means and proportions while risk factor analysis was done using bivariate and multivariate analysis. Unconditional logistic regression was used to study the association between exposure variables and brucellosis. Forty-three cases and 86 controls were recruited for the study. The mean age for the cases was 49 years (SD=20) while that of the controls was 38 years (SD = 18.8). Majority of both cases (62.7%) and controls (58.1%) were female. The most reported symptoms for cases were headache (83.7%) back pains (62.8%) and joint pains (60.6%). Most controls also reported similar clinical presentation and there was no significant difference in reported symptoms between brucellosis cases and controls. Regular consumption of un-boiled milk and assisting animals in delivery were significantly associated with brucellosis by adjusted odds ratio (aOR) 7.7 (95% CI 1.5–40.1) and aOR 3.7 (95% CI 1.3–13.5), respectively. The sensitivity and specificity of the Brucellosis Rapid Diagnostic Kit (RDT) in use in the study health facilities was 20 % and 90 % respectively with a Predictive positive value of 33% and Predictive negative value negative of 84%. The lack of a significant difference in clinical presentation between cases and controls conforms to other reports that identify brucellosis as one of the most misdiagnosed diseases because it causes a systemic infection that presents in a myriad of non-specific signs. This means that clinical symptoms alone cannot be adequately used for presumptive diagnosis and clinicians should consider epidemiologic linkages and laboratory results in management of suspect brucellosis cases. The risk factor analysis shows that animal handlers; primarily farmers and animal health workers and people who consume unpasteurized milk are at the greatest risk. Public health education on brucellosis transmission and prevention, specifically use of protective personal equipment when assisting animals in delivery and boiling of milk should be offered to farmers and the general public. The results from the comparative testing are indicative of the challenge in diagnosis of brucellosis using point of care tests with low validity values. The low sensitivity values mean a big proportion of cases are being misdiagnosed. More specifically, a significant number of cases are being missed in this high burden region due to low-test sensitivity. As such there is need to validate and replace low quality Brucella test kits with better diagnostic tools.