BACKGROUND: Geographic location represents an ecological measure of HIV status and is a strong predictor of HIV prevalence. Given the complex nature of location effects, there is limited understanding of their impact on policies to reduce HIV prevalence. METHODS: Participants were 3949 and 10 874 respondents from two consecutive Zambia Demographic and Health Surveys from 2001/2007 (mean age for men and women: 30.3 and 27.7 years, HIV prevalence 14.3% in 2001/2002; 30.3 and 28.0 years, HIV prevalence of 14.7% in 2007). A Bayesian geo-additive mixed model based on Markov Chain Monte Carlo techniques was used to map the change in the spatial distribution of HIV/AIDS prevalence at the provincial level during the 6-year period, accounting for important risk factors. RESULTS: Overall HIV/AIDS prevalence changed little over the 6-year period, but the mapping of residual spatial effects at the provincial level suggested different regional patterns. A pronounced change in odds ratios in Lusaka and Copperbelt provinces in 2001/2002 and in Lusaka and Central provinces in 2007 was observed following adjustment for spatial autocorrelation. Western province went from a lower prevalence area in 2001 (13.4%) to a higher prevalence area in 2007 (17.3%). Southern province went from the highest prevalence area in 2001 (17.3%) to a lower prevalence area in 2007 (15.9%). CONCLUSION: Findings from two consecutive surveys corroborate the Zambian government's effort to achieve Millennium Developing Goal (MDG) 6. The novel finding of increased prevalence in Western province warrants further investigation. Spatially adjusted provincial-level HIV/AIDS prevalence maps are a useful tool for informing policies to achieve MDG 6 in Zambia.
BACKGROUND: Geographic location represents an ecological measure of HIV status and is a strong predictor of HIV prevalence. Given the complex nature of location effects, there is limited understanding of their impact on policies to reduce HIV prevalence. METHODS:Participants were 3949 and 10 874 respondents from two consecutive Zambia Demographic and Health Surveys from 2001/2007 (mean age for men and women: 30.3 and 27.7 years, HIV prevalence 14.3% in 2001/2002; 30.3 and 28.0 years, HIV prevalence of 14.7% in 2007). A Bayesian geo-additive mixed model based on Markov Chain Monte Carlo techniques was used to map the change in the spatial distribution of HIV/AIDS prevalence at the provincial level during the 6-year period, accounting for important risk factors. RESULTS: Overall HIV/AIDS prevalence changed little over the 6-year period, but the mapping of residual spatial effects at the provincial level suggested different regional patterns. A pronounced change in odds ratios in Lusaka and Copperbelt provinces in 2001/2002 and in Lusaka and Central provinces in 2007 was observed following adjustment for spatial autocorrelation. Western province went from a lower prevalence area in 2001 (13.4%) to a higher prevalence area in 2007 (17.3%). Southern province went from the highest prevalence area in 2001 (17.3%) to a lower prevalence area in 2007 (15.9%). CONCLUSION: Findings from two consecutive surveys corroborate the Zambian government's effort to achieve Millennium Developing Goal (MDG) 6. The novel finding of increased prevalence in Western province warrants further investigation. Spatially adjusted provincial-level HIV/AIDS prevalence maps are a useful tool for informing policies to achieve MDG 6 in Zambia.
Authors: Ngianga-Bakwin Kandala; Eugene K Campbell; Serai Dan Rakgoasi; Banyana C Madi-Segwagwe; Thabo T Fako Journal: HIV AIDS (Auckl) Date: 2012-07-18
Authors: V A Alegana; P M Atkinson; C Pezzulo; A Sorichetta; D Weiss; T Bird; E Erbach-Schoenberg; A J Tatem Journal: J R Soc Interface Date: 2015-04-06 Impact factor: 4.118