Back

Impact of HIV Self-Testing on Recent HIV Testing Among Women in Uganda: A Propensity Score Matched Analysis Using the 2022 UDHS

Emesu, G. K.; Najjuma, S.; Tiikabulamu, P.; Mukose, A. D.; Kagaayi, J.

2026-08-06 hiv aids
10.64898/2026.08.04.26359666 medRxiv
Show abstract

Abstract Background: HIV self-testing (HIVST) has been promoted as a strategy to reach individuals who do not access facility-based testing. However, evidence on whether HIVST leads to more frequent testing among women of reproductive age in Uganda remains limited. This study evaluated the impact of HIV self-testing on recent HIV testing among women using nationally representative data. Methods: Data were drawn from the 2022 Uganda Demographic and Health Survey (UDHS), including 6,438 women aged 15-49 years. The primary outcome was recent HIV testing, defined as having tested for HIV within the 12 months preceding the survey. The treatment variable was ever having used HIV self-testing. Propensity score matching (PSM) with 1:1 nearest neighbour matching (caliper = 0.05) was used to balance observed covariates including parity, media exposure, education, residence, wealth quintile, health insurance, and age group. The average treatment effect on the treated (ATT) was estimated. Results: Among 6,438 women, 23.87% (1,537) reported ever using HIV self-testing. Recent HIV testing was observed in 67.4% of HIVST users compared to 47.4% of non-users (unmatched difference = 20%). After matching, HIVST use increased the likelihood of recent testing by 15.6 percent (ATT = 15.6%; SE = 0.052; t = 2.99). Covariate balance was achieved post-matching, with mean bias reduced from 20.5% to 0.7%, and the B statistic falling from 50.4% to 2.5% (below the 25% threshold). All standardized differences were substantially reduced, with education showing perfect balance (100% reduction) and wealth showing 97.6% reduction. Conclusion: HIV self-testing significantly increases recent HIV testing among women of reproductive age in Uganda. Expanding access to HIVST, particularly for women with lower education, those in poorer wealth quintiles, and those without media exposure, could improve testing frequency and support progress toward the UNAIDS 95-95-95 targets.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.