Boutet Clément, Di Ciaccio Marion, Spire B., Velter A., Sagaon Teyssier Luis. (2025). A competing risks duration model to study PrEP discontinuation among MSM in France : The ERAS 2023 study. AIDS and Behavior, 29, 2711-2727. ISSN 1090-7165.
Titre du document
A competing risks duration model to study PrEP discontinuation among MSM in France : The ERAS 2023 study
Boutet Clément, Di Ciaccio Marion, Spire B., Velter A., Sagaon Teyssier Luis
Source
AIDS and Behavior, 2025,
29, 2711-2727 ISSN 1090-7165
Although PrEP prevents HIV among at-risk populations, high discontinuation rates hinder its effectiveness. We investigated three reasons for PrEP discontinuation in real-life settings in France and associated factors, in order to assess weariness taking PrEP. We used data from the French online survey Enqu & ecirc;te Rapport au Sexe 2023 which targeted gay and bisexual men having sex with men (GBMSM). First, we constructed the outcome 'PrEP duration'. Second, we created a three-category 'reason for PrEP discontinuation' variable as follows: changes in sexual behaviors, a desire to discontinue PrEP, and clinical reasons. Third, we calculated PrEP survival probability. We then performed a Weibull accelerated failure time model with competing risks to assess how the duration of time on PrEP influenced discontinuation, and whether specific factors were associated with different discontinuation reasons. The analysis included 4819 GBMSM. Median PrEP duration was 19 months. Discontinuation because of changes in sexual behaviors was less likely among GBMSM who initiated PrEP prescription in a sexual health structure (AF = 0.604 [0.424-0.858]). A desire to stop PrEP was more likely among those living in regions with the lowest reported seropositivity rates (AF = 2.320 [1.054-5.107]). Discontinuing PrEP for clinical reasons was more likely among participants coming from smaller municipalities (AF = 2.312 [1.099-4.868]). PrEP duration dependence was negative for all three reasons, implying that the longer the time taking PrEP, the lower the probability of discontinuation. We found no weariness with PrEP use among GBMSM; this highlights the need to detail specific factors associated with PrEP discontinuation.
