MOLBPE28 - Poster Exhibition
The association between risk perception and adherence in the FEM-PrEP clinical trial
A. Corneli1, M. Wang2, K. Agot3, K. Ahmed4, L. Olang'o3, S. Makatu4, J. Lombaard5, M. Malahleha4, L. Van Damme6, FEM-PrEP Study Group
1FHI 360, Social and Behavioral Health Sciences, Durham, United States, 2FHI 360, Quantitative Sciences, Durham, United States, 3Impact Research and Development Organization, Kisumu, Kenya, 4Setshaba Research Centre, Soshanguve, South Africa, 5Josha Research, Bloemfontein, South Africa, 6FHI 360, Clinical Sciences, Washington, United States
Background: FEM-PrEP assessed the safety and effectiveness of once-daily emtricitabine and tenofovir disoproxil fumarate (FTC/TDF) in preventing HIV acquisition in women at higher risk in Bondo, Kenya; Bloemfontein and Pretoria, South Africa; and Arusha, Tanzania. One hypothesis for the overall poor adherence observed in the trial was low HIV risk perceptions among participants. Here we describe the association between risk perception and adherence.
Methods: We assessed HIV risk perceptions among all participants at enrollment and at quarterly follow-up visits. Participants were asked if they believed they had “no chance,” “small chance,” “moderate chance,” or “high chance” of becoming infected in the next four weeks. Drug levels were measured at four-week intervals for up to 52 weeks in 150 randomly-selected participants assigned FTC/TDF (50 from each site with HIV infections: Bloemfontein, Bondo, and Pretoria). Logistic regression with robust variance estimation was used to assess the association between risk perceptions (none versus some) and “good-to-excellent” adherence (having both tenofovir in plasma exceeding 10 ng/ml and tenofovir diphosphate in upper-layer packed cells exceeding 100,000 femtomoles). All available drug-level measurements were used to assess the association between baseline risk perception and adherence. Drug-level measurements from weeks 4, 16, 28, and 40 were used to assess the time-varying effect of risk perception reported at baseline and weeks 12, 24, and 36.
Results: We found no overall association between baseline risk perception and adherence (OR: 1.62; 95% CI: 0.85, 3.06) but did observe a significant association in Bloemfontein (OR: 3.64, 95% CI: 1.21, 10.95) during subgroup analysis. When risk perception was considered a time-dependent variable (i.e. perceptions reported at each visit), we observed an overall significant association with adherence (OR: 2.26, 95% CI: 1.29, 3.98), which remained significant when restricting the analysis to Bondo (OR: 4.72; 95% CI: 1.31, 17.00) and Bloemfontein (OR: 4.55; 95% CI: 1.81, 11.46).
Conclusions: Participants' risk perceptions may have influenced their decisions to adhere to the study pill. Additional research is needed to better understand the context influencing risk perceptions among women at the different sites. Future PrEP research and programs should consider exploring women´s risk perceptions during adherence counseling.
Back to the Programme-at-a-Glance