MODELING THE EFFECT OF MEDIATION ON HIV PREVALENCE IN KENYA USING A LOGISTIC REGRESSION MODEL

Abstract

Over the last decades, major global efforts mounted to address the HIV epidemic has realised notable successes in combating the pandemic. Sub Saharan Africa still remains a global epicenter of the disease, accounting for more than 70% of the global burden of infections. Despite the widespread use of HIV mass media national campaigns as an intervention in HIV prevention due to its numerous advantages since the mid-1980s, HIV prevalence still remains a challenge in especially in some geographic areas and populations. Therefore how HIV mass media interacts with the prevailing HIV risk factors to cause an impact on HIV prevalence remains a question not answered. This study considered Exposure to HIV related media as a mediator variable believed to mediate the relationship between HIV risk factors and HIV prevalence. Two logistic regression models were formulated and used to compare the model with mediation and that without mediation in order to establish the effect of mediation on HIV prevalence. Models were fitted to real data from 2018 Kenya Population-based HIV Impact Assessment survey and estimation of the model parameters was done using Maximum Likelihood Estimation in R. Results of R analysis based on both Akaike’s Information Criterion and the McFadden’s R2 value for model with mediation revealed that the model formulated in presence of mediation was better compared to that without mediation since the effects of mediation variable were found to be more significant on HIV prevalence.