Like in other Sub-Sahara African countries, HIV has had significant devastating effects in
Zimbabwe. Understanding the underlying and proximate determinants of infections in the
population is important for knowing who needs intervention, where interventions are needed and
for designing specific and relevant interventions.
No studies appear to have assessed distributional patterns and risk factors of HIV prevalence
simultaneously across the ten provinces in Zimbabwe using a combination of the proximatedeterminant
framework approach and modern geostatistical techniques to identify spatial HIV
hotspots. The overall aim of this study was to investigate the prevalence, spatial patterns and
factors associated with HIV infection in Zimbabwe using the proximate-determinant framework
approach and modern geostatistical techniques.
Methods
This study used the Zimbabwe demographic and health survey of 2010/2011 which included a
representative sample of 9171 women and 7104 men- aged 15 to 49. Following the proximate
determinants framework, multilevel models were fitted separately for men and women. Global
and local spatial autocorrelations were assessed. Spatial regression models were also fitted to
adjust for spatial random effects and non-random effects. Significant difference between hotspots
and cold spots was examined at community level.
Results
The overall prevalence estimate was 15.4% (95%CI: 14.8% - 16.0%) [17.7% among women and
12.2% among men]. While prevalence was highest in women in the middle age categories
v
(29.3%), a corresponding dose response relationship was observed among men. Highest
prevalence was obtained in urban dwellers, widowed men (60%) and women (56%), and
Matabeleland South. A dose response association was found between HIV prevalence and
duration of cohabitation, total lifetime partners in both gender populations, and age of most
recent partner in men.
Evidence of global and local spatial autocorrelation was found. Spatial scan techniques identified
three hotspots including Matabeleland South province. Prominent significant underlying risk
factors of HIV infection in both gender populations were age group [men: AOR40-44: 5.19;
(95%CI:1.92 – 14.02); AOR45-49 : 4.76 (95%CI: 1.70 – 13.30)] [women: AOR30-34: 2.17;
(95%CI:1.32 – 3.54); AOR35-39: 1.91; (95%CI:1.14 – 3.19)] and widowed marital status[men:
AOR: 7.37; (95%CI: 3.10 – 17.52)][women: AOR: 4.13; (95%CI: 2.49 – 6.85)], likewise
Matabeleland South region in men [AOR: 2.32; (95%CI: 1.49-3.62)]. On the other hand, total
lifetime partners [men: AOR5-9: 3.15; (95%CI: 2.08 – 4.77)][women: AOR5-9: 3.74; (95%CI:
2.52 – 5.53)], and symptoms of sexually transmitted infections emerged as proximate predictors
of the epidemic.
No evidence of significant heterogeneity in geographical distribution of HIV prevalence was
found after adjusting for significant underlying and proximate risk factors. However, persons
living within hotspots are on the average 30% more likely to be within the richest wealth quintile
(OR: 1.30; p: <0.001), more likely to have higher total lifetime partners and symptoms of
sexually transmitted infections.