JMIR Publications

ABSTRACT

Background:

Disparities in HIV care persist and there is a need to evaluate scalable approaches to improving HIV care engagement and viral suppression in real-world settings.

Objective:

To compare standard of care HIV linkage, adherence, and retention (LAR) clinic-based support services to a mobile health (mHealth)-enhanced linkage, adherence, and retention (mLAR) intervention and assess mLAR implementation.

Methods:

Participants were enrolled from six Baltimore HIV clinics in an individual-level randomized controlled trial. Eligibility criteria included adults living with HIV who had a new HIV diagnosis, were out of care, or were deemed by clinic staff as someone who could benefit from LAR services. Participants in the control group received the referring clinic’s standard LAR services. Participants randomized to the intervention additionally received mHealth-supported patient navigation including an app platform that integrated HIV surveillance data from throughout Maryland. The primary outcome was HIV virologic suppression at 12 months. Participants with non-suppressed viral load or missing viral loads at 12 months (inclusive of a +/- 3-month window) were considered virologic failures. Secondary outcomes included loss to follow-up and mortality at 12 months. Implementation outcomes were assessed via in-depth interviews with participants and patient support staff. This study was implemented in collaboration with patients and stakeholders, including members of the Baltimore City Health Department, the Maryland Department of Health, patient navigators, HIV healthcare providers, and people living with HIV. Members of the Steering Committee provided guidance on the study design and implementation, app development, interpretation of results, and dissemination of findings.

Results:

A total of 451 participants (226 LAR control group and 225 mLAR intervention group) were randomized and 3 (1 LAR and 2 mLAR) withdrew consent before the study endpoint. Virologic suppression was 56.9% (95% CI 50.4%-63.4%, LAR) and 50.3% (95% CI 44.%-57.7%, mLAR) in the two arms [difference 5.8% (-3.4%-14.9%), p-value=0.22]. Sixty-six participants (29.3%) in the LAR arm and 71 (31.8%) in the mLAR arm were missing a viral load test at 12 months and were assumed to be not virologically suppressed. These results were similar when stratified by race, ethnicity and clinic site. In the LAR and mLAR arms, respectively, 28.9% (95% CI 22.9%-34.8%) and 31.8% (95% CI 25.7%-37.9%) of participants were lost to follow-up [difference -2.9% (-11.5%-5.6%), p-value=0.49] while all-cause mortality was 4.9% versus 4.0% [Pearson's chi-squared p<0.662]. In-depth interviews revealed mixed perceptions of the app; some participants and patient navigators valued its usability, while others found it redundant with existing tools like electronic medical record portals and phone communication.

Conclusions:

An app-based mobile health intervention for patients and patient supporters had no impact on virologic suppression, engagement in care, or mortality. In populations with complex socio-structural determinants of health, additional strategies are needed to improve HIV care outcomes. Clinical Trial: ClinicalTrials.gov NCT03934437