Abstract / Summary
HIV testing uptake among pregnant women remains a public health priority in Burkina Faso. This study aimed to identify and rank key predictors of HIV testing uptake using machine learning (ML) approaches and to estimate potential gains under scenario-based intervention simulations. We conducted a cross-sectional secondary analysis of the 2021 Burkina Faso Demographic and Health Survey, including 4,703 women aged 15–49 years who had given birth within the two years preceding the survey. HIV testing uptake was defined as having been tested for HIV and having received the test result during antenatal care (ANC) or before delivery. Four ML algorithms (penalized logistic regression, Random Forest (RF), Extreme Gradient Boosting, and Support Vector Machine with a radial basis function kernel) were trained and evaluated. Model performance was assessed using weighted AUC-ROC and other standard classification metrics. Scenario-based intervention simulations targeting modifiable predictors were then conducted using the best-performing model. Overall, 55.0% (95% CI: 52.7–57.3%) of women reported HIV testing uptake during ANC or before delivery, with substantial regional disparities, ranging from 17.0% (95% CI: 11.2–22.9%) in the Est region to 81.0% (95% CI: 76.5–85.4%) in the Centre region. RF achieved the best cross-validated discrimination (AUC-ROC = 0.785) and showed comparable performance on the independent test dataset (AUC-ROC = 0.792). The five most important predictors of HIV testing uptake were region (relative importance = 100%), knowledge of HIV and/or prevention of vertical transmission (PVT) (95.6%), wealth index (43.1%), mother’s occupation (37.4%), and area of residence (32.3%). Scenario-based simulations suggested that increasing the coverage of good HIV/PVT knowledge to 80% could increase predicted HIV testing by 8.6% (95% CI: 6.9–10.6%). Compared with knowledge-only interventions at corresponding knowledge targets, the combined scenario incorporating media exposure and education yielded modest additional relative gains across regions, ranging from 0.6% to 13.8%. HIV testing uptake among pregnant women in Burkina Faso is characterized by substantial regional inequalities, and knowledge of HIV and/or PVT emerged as the dominant modifiable predictor. Combining national efforts to improve knowledge with region-specific strategies targeting media exposure and educational attainment may provide a promising approach for developing context-specific interventions to increase HIV testing uptake among pregnant women and reduce regional disparities.