Transmission dynamic Model to predict the impact of behavioral interventions on HIV/AIDS prevalence in Uganda

Introduction:The HIV sero-prevalence among pregnant women is used as a surrogate measure of overall prevalence in Uganda and in most low income countries. Using surveillance data, the objective of this study was to replicate prevalence trends observed since 1985 through to 2005 and explore potential impacts of various behavioral prevention strategies to identify the strategies with the largest effectiveness in reducing HIV prevalence in both the short and long term. Methods:We developed a simple deterministic mathematical model of HIV transmission dynamics including two risk groups. Data from the HIV sero-prevalence surveys among pregnant women from 1985-2005 and the Sero-Behavioral Survey conducted in 2004-2005 among selected sites in Uganda were used to parameterize the model. The impact of condom use, rate of sexual partner changes and reduction of number of sexual partners assuming different mixing patterns, levels of antiretroviral therapy coverage, and varying parameter estimates on prevalence levels were investigated over time. Results:Our model simulation estimates closely replicate observed sero-prevalence trends reported in the literature for the period 1985-2005. Decreasing partner change rate especially in the high activity groups had a greater impact on lowering the overall prevalence estimates in both the short and long term than condom use assuming other model parameters remained constant. Irrespective of the year of initiation of either intervention, partner change rate had a more substantial overall impact on HIV prevalence projections. Conclusion:Our simulations show that without efforts targeted at decreasing partner change rate, the effects of condom use and ART on prevalence projections are negligible.