Long-acting pre-exposure prophylaxis preferences among pregnant and postpartum women in Kenya: results from a discrete choice experiment

Abstract

Background

New long-acting HIV pre-exposure prophylaxis (LA-PrEP) methods may address adherence barriers during pregnant and postpartum periods, when HIV risk is elevated. Understanding their preferences for LA-PrEP is essential for person-centered HIV prevention in maternal and child health (MCH) systems, yet evidence on preferred attributes is limited.

Objective

To estimate pregnant and postpartum women’s preferred PrEP attributes using a discrete choice experiment (DCE) at important timepoints in the perinatal period

Study design

From February 2023 to July 2024, we conducted a DCE among 513 HIV-negative pregnant and postpartum women taking daily oral PrEP in Kisumu and Siaya, Kenya, enrolled between 24–32 weeks gestation and a high HIV risk score. Participants completed the DCE with 12 choice sets at their third antepartum (median gestational age: 37.0 weeks) and/or 6-month postpartum visits. Attributes included effectiveness, form and dosing, safety data, side effects, collection place, cost, and multipurpose prevention (postpartum only). We fit effects-coded choice data to a conditional logit model, latent class analysis (LCA) for preference heterogeneity, and univariate multinomial logistic regressions to predict class membership by individual characteristics.

Results

A total of 513 women completed the DCE at least once (151 at third antepartum, 509 at 6-month postpartum). Every 2-month injections were strongly preferred, showing the highest positive preference weight (pregnant: 1.22, 95% CI: 1.12–1.33; postpartum: 1.24, 95% CI: 1.18–1.30). Four latent classes emerged: “Flexible PrEP Adopters” (37.2%), “Safe and Effective Injection Preference” (16.5%), “Strong Injection Preference” (37.7%), and “Oral PrEP Preference” (8.6%). Higher parity was associated with lower odds of membership in “Flexible PrEP Adopters” (OR=0.6, 95% CI: 0.4–0.8, P=.001), “Safe and Effective Injection Preference” (OR=0.6, 95% CI: 0.4–0.8, P=.003), and “Strong Injection Preference” (OR=0.7, 95% CI: 0.5–1, P=.027) compared to “Oral PrEP preference.”

Conclusions

Strong preferences for every 2-month injectables emphasize the need to prioritize LA-PrEP in this population. ANC settings can support diverse PrEP preference profiles with tailored counseling to account for individual preferences, PrEP experience, and obstetric history.