Comparing Trajectories of Change in PrEP Readiness Across an Intervention Pilot Trial to Increase Uptake Among Young South African Women: A Markov Modeling Approach.

Abstract

Background: In South Africa, uptake of HIV oral pre-exposure prophylaxis (PrEP) has been slow among women, who account for 60% of new HIV diagnoses. A gender-enhanced (GE) online group workshop, Masibambane ("Let Us Work Together") (N = 50), demonstrated greater increases in PrEP stage of change versus providing participants with individual access (IA) to online PrEP educational and motivational materials (N = 50). However, participants began the study at different stages, and it was unclear which women may have benefited most from the intervention. We explored how an underused analytic method could offer greater insight into women's change trajectories.

Methods: Participants completed questionnaires at baseline ("pre-intervention"), post-intervention, and three months after baseline ("follow-up") and documented readiness through a 5-point stage-of-change (SOC) score based on the transtheoretical model. We used Markov transition matrices to calculate probabilities of different trajectories.

Results: Participants who reported 3 ("would like to speak to a counselor") were most likely to transition to 5 ("ready to take PrEP") at post-intervention in the GE arm (Ppost, pre(5|3) = 0.60) but not the IA arm (Ppost, pre(5|3) = 0.17). Those who scored 2 ("uncertain") at pre-intervention also exhibited greater probabilities of upward movement in the GE arm.

Discussion: Compared to IA, GE women who expressed uncertainty or a desire for counseling at baseline were more likely to shift into a state of readiness, driving the observed difference across study arms. This analysis demonstrates how Markov modeling can provide greater understanding of population change dynamics than mean values alone.