Pre-exposure prophylaxis (PrEP) is an effective HIV prevention tool that relies on good adherence in high-risk scenarios. To understand the factors that predict adherence, technology such as mobile applications like UPrEPU—allowing for logging users’ daily behaviours at close to the time they have sex or PrEP intake—can be used as a person-centred, self-care intervention. This study aims to develop a machine learning model using logs of sexual activities and user attributes recorded in the UPrEPU mobile application in Taiwan to predict whether a sexual event was protected by oral PrEP among gay, bisexual and other men who have sex with men (GBMSM).
We used data from the UPrEPU app collected between January 2022 and May 2023 in Taiwan. The dataset included information on users’ sex events, such as the timing and users’ sex roles (e.g. versatile, receptive or insertive partner), and the dynamic user-based attributes related to sexual behaviours and PrEP use. Various subsets of these features were employed in CatBoost models to predict whether the sex events were associated with correct PrEP use. We evaluated the models’ performance using five-fold cross-validation. The influential features were identified through feature importance analysis and Shapley Additive Explanations (SHAP) values to explain the models.
A total of 198 users recorded 2356 anal sex events on UPrEPU. The model with dynamic user-based attributes outperformed those without them. The most parsimonious model had a good prediction performance (accuracy = 75%, precision = 78%, recall = 90%, F1-score = 83%) and identified the key features of PrEP protection. The model with five dynamic user-based attributes—age, cumulative PrEP use, condom use and the proportion of anal sex events with HIV-negative partners not on PrEP—significantly outperformed the model based on event-level attributes alone.
Behavioural patterns significantly influence PrEP adherence among GBMSM. Person-centred mobile applications such as UPrEPU provide valuable data for tailored, just-in-time interventions, enhancing adherence. Recognizing these patterns can guide person-centred interventions. Incorporating these insights into clinical care or digital tools may improve consultations and support timely, informed HIV prevention decisions.
Low bone mineral density (BMD) has been described in children and young people with perinatally acquired HIV (PHIV), which may be related to both traditional (e.g. low body mass index and malnutrition) and HIV-related risk factors (e.g. longstanding exposure to HIV and antiretroviral therapy [ART], with immune suppression, chronic immune activation and inflammation). Here, we evaluate BMD in a U.K. cohort of young people with PHIV by age and ART.
This longitudinal, observational study was conducted at a U.K. tertiary PHIV service between November 2018 and March 2022. Bone health was assessed in 130 individuals aged 15–19 (n = 50), 20−24 (n = 50) and 25 years and older (n = 30) by dual-energy X-ray absorptiometry, bone mineralization and turnover markers. Low BMD was defined as lumbar spine (LS) and/or femur-BMD z-score below −2, relative to age, sex and ethnicity-matched U.K. population-based normative controls. Two-year follow-up evaluation was performed in those aged 15−19 (n = 42) and 20−24 years (n = 43) at enrolment, which included a group who switched from tenofovir disoproxil fumarate (TDF) to tenofovir alafenamide (TAF) ART at baseline. Bayesian logistic regression models examined predictors of low BMD and the effect of ART-backbone on BMD accrual.
At baseline, 57% were female and 82% of black ethnicity, with 31 (24%) on TDF-ART. Sixteen (12%) had low baseline BMD. Over a median follow-up duration of 26 (interquartile range [IQR] 25–29) months, BMD accrual was lower-than-expected in those aged 15−19 years (mean change LS-BMD z-score −0.15 (standard deviation [SD] 0.44)), when compared to normative controls. No associations were seen with HIV parameters or the ART regimen. Participants who switched to TAF-ART had similar BMD accrual 26 (IQR 24–32) months post switch, when compared to those on non-TAF/TDF-ART (mean change LS-BMD z-score TAF −0.01 [SD 0.41] vs. non-TAF/TDF −0.03 [SD 0.54]).
While rates of low BMD were reassuringly low in this cohort, lower-than-expected BMD accrual was observed in younger individuals, relative to normative controls. Overall, BMD accrual on TAF-ART was non-inferior to non-TAF/TDF-ART.
Key populations (KP), including men who have sex with men, people who inject drugs, sex workers, transgender people and people in closed settings, are disproportionately affected by HIV and face structural and legal barriers to care. While community-led responses are central to reaching KP, services are often disease-specific and disconnected from national primary healthcare (PHC) systems. PHC, defined by WHO as a whole-of-society approach to delivering integrated and person-centred services, is rarely designed to meet the broader health needs of KP, who also experience high burdens of non-communicable diseases, mental health conditions and violence. This paper describes three service delivery models, supported by PEPFAR, that integrate HIV and PHC services for KP in Vietnam, Nigeria and Eswatini.
The three models are community-led, client-centred, and tailored to KP health and social needs. Each integrates HIV services—including testing, antiretroviral therapy, viral load monitoring, pre-exposure prophylaxis (PrEP) and advanced HIV disease management—alongside broader PHC services such as mental healthcare, sexual and reproductive health, non-communicable disease screening and tuberculosis services. All models include structural and community-based interventions such as gender-based violence support, stigma reduction, peer navigation and economic empowerment. These services are delivered in safe, trusted spaces by multidisciplinary teams including peer and clinical providers. While the models demonstrate alignment with PHC principles (accessibility, cultural competence, continuity and community empowerment), challenges remain related to integration within national health systems, financing and provider training. Recent U.S. global health policy shifts, including reductions in funding for KP-specific programming and limited PrEP access, pose additional threats to programme sustainability and client trust.
Integrated models of HIV and PHC for KP can improve access, engagement and health outcomes across a range of services. They represent promising approaches for addressing intersecting health and structural needs, particularly in settings where stigma and criminalization persist. Sustained progress will require inclusion of KP in PHC policies and planning, protection of community-led services and domestic financing strategies that ensure continuity in the face of shifting donor priorities.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：