Predictors, and Trends of Hepatitis B Virus in Selected Regions of Kenya

Hepatitis B virus (HBV) infection is a silent epidemic; many infected people are asymptomatic and not aware of the infection. In 2022, it was reported that approximately 254 million people were living with chronic HBV infection globally, majority being in sub-Saharan Africa and Asia. In Kenya, the national HBV prevalence is estimated to be 3.5%. Our study aimed at identifying key predictors and transmission trends that could inform development of sustainable prevention models needed to address existing gaps in national framework towards HBV elimination. We targeted participants seeking health services in Baringo and Kisumu county health facilities and conducted community mass testing in the two counties. Participants were interviewed using a study questionnaire and were tested for hepatitis B surface antigen (HBsAg) using a HBsAg rapid test. Venous blood was collected from participants who tested HBsAg+ for further infection confirmation and linkage to care. Logistic regression was performed to assess factors correlated with HBV infection. Out of 3,034 participants, 192 tested positives for HBsAg and the prevalence of HBV infection was 6.3% (95% CI=0.055-0.072). Intrafamilial infections in Baringo were 15.0%. HBV infection prevalence exceeded 10% among those aged 25–49 years, peaking at 13.1% in the 45–49 years age group and lowest at 1.8% in the 16–19-year age group. Overall, males had higher prevalence in younger ages, while females were more affected above 60 years. In multivariable logistic regression, individuals residing in Baringo (aPR = 8.1; 95% CI = 2.2 - 29.4), users of other injectable drugs (aPR = 6.7; 95% CI = 1.3 - 204.0), those traditionally circumcised (aPR 1.02; 95%CI=0.56,1.88) and staying >5 kilometers from a healthcare facility (aPR = 10.4; 95% CI = 2.2 - 49.4), had significantly higher prevalence ratios of being infected with HBV. These different infection predictors underscore the need for different care and prevention approaches models.