CROI 2025 Abstract eBook

Abstract eBook

Poster Abstracts

engagement with a digital family planning and reproductive health (FP/ RH) intervention (“CyberRwanda”) influenced the intervention’s impact for Rwandan adolescents. Methods: A 3-arm, cluster randomized, hybrid effectiveness-implementation trial was conducted with 6,078 students in 60 secondary schools from 2021 2023. Schools were randomized 1:1:1 to CyberRwanda self-service (self-guided access to digital platform), CyberRwanda facilitated (self-guided and group access plus in-person activities), or comparison (no access). Dose-response, treatment on the treated, and complier average causal effects (CACE) were estimated to determine the intervention’s impact among participants with varying levels of school exposure (any, <12 vs. ≥12 months) and platform engagement (high vs. low frequency use, high-frequency plus ≥12 months exposure). The primary outcomes were childbearing, contraceptive use, and HIV testing. Secondary outcomes included FP/RH knowledge, attitudes, self-efficacy, and behavior. Generalized linear mixed models and instrumental variable models were used to estimate prevalence differences (PD) comparing all CyberRwanda participants to the comparison. Results: Overall, 91.5% of CyberRwanda participants were exposed to the intervention at school. Of these, 62.1% had ≥12 months of exposure; 31.7% were high-frequency users. Exposure time and engagement were found to be the primary drivers of impact. No effects were detected among participants exposed for <12 months; however, the ≥12-month exposure subgroup had significant improvements on several secondary outcomes, including condom use among sexually active participants (56.6% CyberRwanda vs. 44.2% comparison, PD: 0.13, 95% confidence interval: 0.06, 0.21). CACE analyses revealed that, if sexually active participants had ≥12 months exposure and were high-frequency users, condom use could increase by 37 percentage points relative to the comparison arm. There were no effects on the primary outcomes for any of the exposure groups. Conclusions: When provided with access for a year or more, CyberRwanda has the potential to increase condom use and reduce risk of critical FP/RH outcomes including HIV and unintended pregnancy. These results underscore the need for sustained access to and engagement with digital platforms to achieve impact. 1311 Digital Strategy to Improve PrEP Adherence Among MSM: A Stepped Wedge Randomized Trial in China Chunyan Li 1 , Zhuoheng Yin 2 , Quanmin Li 3 , Linghua Li 3 , Songjie Wu 4 , Ke Liang 4 , Gifty Marley 2 , Chengxin Fan 2 , Yifan Dai 2 , Haojie Huang 5 , Aniruddha Hazra 6 , Jonathan Lio 7 , Joseph D. Tucker 8 , Renslow Sherer 7 , Weiming Tang 8 1 University of Tokyo, Tokyo, Japan, 2 University of North Carolina Project - China, Beijing, China, 3 Guangzhou Eighth People's Hospital, Guangzhou, China, 4 Wuhan University Zhongnan Hospital, Wuhan, China, 5 Wuhan LGBT Center, Wuhan, China, 6 University of Chicago Medical Center, Chicago, IL, USA, 7 University of Chicago, Chicago, IL, USA, 8 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Methods: This trial is nested in a PrEP demonstration trial in Guangzhou and Wuhan, China (NCT04754139). Participants were provided TDF/FTC for 12 months and followed up quarterly (M3, M6, M9, M12). MSM were randomly allocated to four sequential groups receiving weekly WeChat-based (a popular Chinese chat media) digital messages in a quarterly staggered order. Intervention messages were co-created through open calls and three co creation rounds with 19 PrEP users in the trial, including short videos, images, and infographics. Per-protocol analysis and generalized linear mixed-effects models were used to assess primary outcomes, including self-reported optimal PrEP adherence in past 30 days (daily users: 6-7 pills/week, on-demand users: full compliance with the 2-1-1 dosing scheme per sex event) and study retention (i.e. whether loss to follow-up). Results: From July 2021 to September 2024, 1087 MSM (mean age=28.0, IQR=24.0-31.4) enrolled. About one quarter of participants were college students, and three quarters had 2 or more sex partners in the past 3 months. About half of the participants reported using on-demand regimen (2:1:1) throughout the study. Participants enrolled in each SWT cluster at baseline were 251, 281, 276, and 279, respectively. About 33.8% (n=368) of the participants dropped out or lost to follow-up before M12, with no significant difference across SWT clusters. From M3 to M12, over 75% of the participants reported optimal adherence to PrEP, with a non-significant secular trend of declining Background: Digital Health intervention is promising for encouraging behavior change. We conducted a stepped-wedge trial (SWT) to evaluate the efficacy of co-created digital PrEP adherence intervention to facilitate adherence among Chinese MSM PrEP users.

adherence over time. The digital intervention marginally improved adherence levels (aOR=1.18, 95%C.I.: 0.85-1.60), but not statistically significant. On demand PrEP users versus daily PrEP users were less likely to report optimal adherence (aOR=0.06, 95%C.I.: 0.04-0.09). MSM with higher income or having less sex partners are more likely to report optimal adherence. Conclusions: Digital PrEP adherence intervention messages with Chinese MSM has the potential to improve self-reported adherence. Our findings of varied adherence performance among individuals highlight the need for future research into differentiated PrEP interventions. 1312 Impact Analysis and Machine Learning Forecasts of HIV Viral Load Suppression in 21 African Countries Amobi A. Onovo 1 , Melaku Dessie 2 , Gonza Omoro 3 , Sara Herbst 4 1 Henry M Jackson Foundation, Bethesda, MD, USA, 2 United States Agency for International Development, Washington, DC, USA, 3 Walter Reed Project–Kisumu, Kisumu, Kenya, 4 PEPFAR, Washington, DC, USA Background: Viral load suppression (VLS) is crucial for controlling HIV transmission and improving health outcomes in people living with HIV. While PEPFAR-supported programs have made substantial progress in improving VLS, innovative approaches are needed to sustain and enhance VLS in resource limited settings. This analysis integrates machine learning and the difference in-difference (DiD) method to forecast future VLS and assess the impact of programmatic efforts across 21 PEPFAR-supported countries in sub-Saharan Africa. Methods: Historical aggregate PEPFAR program data from 2017 to 2024 were analyzed to identify key factors associated with VLS improvement. A program indicator matrix, aligned with PEPFAR's Monitoring, Evaluation, and Reporting (MER) guidance, was computed for key HIV metrics. This matrix was integrated with socio-demographic, socioeconomic, and health indicators from the World Bank's World Development Indicators to create a comprehensive analytical dataset. A DiD analysis was conducted to compare changes in VLS between 2017 and 2024. An XGBoost machine learning algorithm was used to develop a time-series model forecasting VLS from 2025 to 2030, tuned and validated using five-fold cross-validation with root mean square error (RMSE) as the primary performance metric. Results: The XGBoost model exhibited strong predictive accuracy when validated against actual data from 2017-2024, achieving an RMSE of 1.61 on the training set and 2.30 on the testing set. This small difference indicates minimal overfitting and reliable generalization to unseen data. The model identified key predictors of VLS, including viral load coverage, new enrollments in PrEP, and age demographics. Notably, the 15-64 age group had a more significant impact on VLS outcomes compared to the pediatric population (ages 0-14). The DiD analysis revealed a statistically significant increase in VLS from 2017 (82.93%) to 2024 (95.59%) across all 21 countries, with an overall estimated 15.27% increase (p=0.0016). The forecast suggests a continued upward trend in VLS through 2030, contingent on sustained programmatic interventions identified. Conclusions: This analysis demonstrates the potential of machine learning to identify targeted HIV interventions that can be modified to improve program performance. It underscores the need for innovative, data-driven strategies to enhance viral load testing, PrEP enrollment, and ART uptake for sustained VLS in sub-Saharan Africa.

Poster Abstracts

CROI 2025 436