Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
WhatsApp-based AI companion engagement and documented HIV testing and PrEP uptake among young people in South Africa: a real-world longitudinal cohort study
ABSTRACT
Background:
Conversational artificial intelligence (AI) may provide scalable approaches to support engagement with HIV prevention services, particularly where barriers to facility-based care persist. However, evidence linking real-world engagement with conversational AI to documented HIV prevention outcomes remains limited.
Objective:
To evaluate engagement with Aimee, a WhatsApp-based AI health companion integrated into community-based HIV prevention services in South Africa, and examine associations between engagement and documented HIV testing and pre-exposure prophylaxis (PrEP) uptake.
Methods:
We conducted a retrospective longitudinal cohort study of individuals who engaged with Aimee in Gauteng, South Africa, between March 17 and November 30, 2025. Users who accepted the platform terms and conditions and sent at least one message were included. Engagement was characterised using active days, engagement span, message volume, and AI-driven conversation sessions. HIV testing was assessed using a composite of clinic records, provider-reviewed HIV self-test uploads, and documented self-reported testing information. PrEP uptake was assessed using verified dispensing records and documented self-reported current use, initiation, or refill. Associations between engagement and outcomes were examined using regression models adjusted for registration month. To reduce exposure-timing bias, a 14-day landmark analysis classified engagement during the first 14 days after initial Aimee contact, excluded participants with the relevant outcome during that period, and assessed subsequent outcomes. Exploratory time-to-event analyses were also conducted.
Results:
Among 9,310 users, 5,452 (58.6%) returned to Aimee on two or more distinct calendar days. Documented evidence of HIV testing was identified for 3,280 users (35.2%), and documented PrEP uptake for 1,140 (12.2%). In the 14-day landmark analysis, subsequent documented HIV testing was more frequent among returning than one-day users (20.5% vs 7.2%; RR 2.85, 95% CI 2.52–3.24; P<.001). Documented PrEP uptake showed a similar pattern (9.3% vs 3.6%; RR 2.61, 95% CI 2.17–3.14; P<.001). Exploratory time-to-event analyses showed a consistent direction of association. HIV testing and PrEP uptake were also highest among users who engaged in two-way provider conversations.
Conclusions:
Return engagement with a WhatsApp-based AI health companion was associated with higher subsequent documented HIV testing and PrEP uptake within routine HIV prevention services. Because engagement was self-selected, residual confounding is likely, and some outcomes relied on conversational disclosure, these observational findings do not establish causality. Prospective controlled studies using independently verified outcomes are needed to determine whether AI-supported engagement increases HIV prevention service uptake.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.