Assessing risk of HIV amongst young people requires knowledge of the spatial distribution of the disease and its association with demographic, socioeconomic, behavioural, and biological factors. The objective of this study was to press forward with such knowledge by analyzing the spatial distribution of HIV prevalence in relation to the demographic, social and behavioural factors reported in the MEASURE Demographic and Health Surveys conducted in four Eastern Africa countries from 2007-2011.
Methods: The study was a cross-sectional study design where data were obtained from MEASURE DHS databases of Ethiopia (2011) Kenya (2008-2009), Tanzania (2007-2008) and Uganda (2011). Statistical analysis employed Stata TM 12 software to perform descriptive analysis for overall characteristics of the study sample. Bivariate analytical tests compared statistical significant differences between the covariates and outcome. Univariate and multiple variable logistic regression models were used to explore factors that were significantly associated with HIV prevalence in young people. Spatial logistic regression analysis was aided by the Bayesian Software BayesX version 2.1 to perform spatial random effect modelling which was used to account for any unexplained spatial autocorrelation in the study area results. Spatial analyses was performed to examine the spatial distribution of the disease using geostatistical techniques such as; spatial autocorrelation and spatial. Final outputs were visualised using Geographical Information Systems techniques (GIS).
Results: The results showed variations of HIV prevalence not only within countries but also across the countries. Each country was characterised by different factors that were associated with HIV prevalence among young people. Across the study area, behavioural factors were significantly associated with HIV. Presence of an STI, a proxy for high-risk sexual behaviour, {(Kenya: POR=13.46; 95% BCI; 2.92-64.41, Uganda: POR=6.83; 95% BCI; 4.14-16.34)} and an early coital debut were significantly associated with HIV in the study area. On the other hand, circumcision (Uganda: POR=0.30; 95% BCI; 0.12-0.80) and condom use provided a protective effect on HIV among The young people. Spatial distribution of HIV in Eastern Africa was mapped at a regional level in aspects of crude prevalence estimates, excess risk and spatial risk. The spatial distribution of HIV was non-random and clustered with significant Moran’s I for Kenya (0.189, p<0.001) and Tanzania (0.056, p=0.04). Cluster analysis revealed a number of significant geo-spots of HIV in Ethiopia (n=53, p<0.001), Kenya (n=34, p<0.001), Tanzania (n=4, p<0.001), and Uganda (n=32, p<0.001).
Conclusions: Since majority of the significant associations were observed in the behavioral category, HIV prevention interventions should be aimed at behaviorist change among young people. The use of spatial risks maps can help policy makers target interventions in areas where they are greatly needed. A future study which focuses on the distribution of HIV/AIDS in East Africa over space and time is recommended to understand how behavioral change will affect the spread of the disease.