List randomization for prevalence estimation of sensitive behavioral data among women with HIV of reproductive age in Lilongwe, Malawi.
Am J EpidemiolSep 3, 2026
Clinical ResearchInfectious DiseasesObstetrics & GynecologyPreventive Medicine / Public HealthOpen access
Karen Diepstra, Daniel Westreich, Paul N Zivich et al.
✦ AI-curated · Sources linked
In 30 seconds
This randomized controlled trial evaluated the effectiveness of list randomization in estimating the prevalence of sensitive behaviors among women of reproductive age living with HIV in Lilongwe, Malawi. The study found that self-reported condom use was significantly higher using list response (53.1%) compared to direct response (34.7%), with a prevalence difference of 18.5% (95% CI, 6.2%-30.7%).
Key findings
- Self-reported condom use was 53.1% with list response vs 34.7% with direct response.
- The prevalence difference for condom use was 18.5% (95% CI, 6.2%-30.7%).
- List randomization did not significantly change prevalence estimates for most sensitive questions.
- The difference in reporting was less pronounced among participants with higher socioeconomic status.
Why it matters
Understanding the impact of reporting biases is crucial for accurate data collection in sensitive health behaviors. This study highlights the potential of list randomization to improve self-reported data accuracy, which can inform public health strategies.