Mobile App Reduces Detectable HIV Viral Load in Men Using Stimulants

In a remote randomised controlled trial published in eClinicalMedicine, researchers found that a mobile health application designed to support antiretroviral therapy adherence significantly lowered the odds of detectable HIV-1 viral load at six months among sexual minority men with HIV who also screened positive for a stimulant use disorder. The study, led by Adam W. Carrico and colleagues, enrolled 286 participants from the continental United States and randomised them to either the Supporting Treatment Adherence for Resilience and Thriving (START) application or a website control condition providing referrals to care resources. The primary outcome was detectable viral load, defined as more than 300 copies per millilitre, measured using remote blood samples collected at baseline, six months, and 12 months. The findings indicate that the digital intervention primarily prevented virologic rebound, a key factor in maintaining viral suppression for individuals at high risk of treatment disengagement.

The research addresses a significant gap in digital health evidence, where prior studies have shown moderate improvements in self-reported adherence but inconsistent effects on objective biomarkers. While randomised controlled trials in lower- and middle-income countries have demonstrated that bidirectional text messaging can reduce detectable viral load, evidence for key populations facing disproportionate challenges, such as sexual minority men using stimulants, remained limited. The use of methamphetamine and other stimulants is a prominent driver of the HIV epidemic, with up to one in three sexual minority men who acquire HIV in the United States reporting persistent methamphetamine use. These individuals often experience greater difficulties with adherence, increasing the likelihood of detectable viral load and potentiating onward transmission risk. The START application was designed to address these barriers by integrating positive psychological intervention skills to improve adherence irrespective of whether participants were ready to reduce stimulant use.

The START application is informed by the information-motivation-behavioural skills model, which proposes that optimising adherence requires providing relevant information, enhancing motivation, and supporting self-efficacy for implementing behavioural skills. Consistent with Revised Stress and Coping Theory, the app also embraces the role of positive psychological processes, such as gratitude and mindfulness, in supporting health behaviour change during chronic stress. Participants in the intervention arm had access to six modules containing 31 on-demand videos covering 10 positive psychological intervention skills and eight audio files with meditation exercises. The app also included brief check-ins for self-monitoring of adherence, substance use, and mood, with graphical representations to examine trends over time. In contrast, the control group received access to a website with resources for crisis counselling, mental health, substance use treatment, and HIV care, but did not receive the structured behavioural skills training.

Eligible participants were sexual minority men aged 18 or older, residing in the continental United States, diagnosed with HIV for at least three months, and currently taking antiretroviral therapy. They were required to have a history of detectable viral load in the past year, disengagement from care, or less than 90% adherence, and to screen positive for a moderate or severe stimulant use disorder using the Alcohol, Smoking, and Substance Involvement Screening Test. The study was conducted entirely online, with enrolment, intervention delivery, and outcome assessments performed remotely. From June 2022 through May 2024, 2,346 individuals screened as eligible, and 508 were enrolled. After a run-in period involving baseline measures and a remote blood sample, 286 participants were randomised in a 1:1 allocation to either START or the website control. The randomisation was stratified by current receipt of substance use treatment, and the investigative team remained blinded to assignment, although participants were aware of their condition due to the nature of the intervention.

At six months, the primary outcome showed a statistically significant benefit for the START group. Participants randomised to the mobile application had 58% lower odds of a detectable viral load compared to those in the website control condition, with an adjusted odds ratio of 0.42 and a 95% confidence interval of 0.19 to 0.93. The absolute risk difference was -0.21, corresponding to a number needed to treat of approximately 4.75 to prevent one instance of detectable viral load. This effect was driven primarily by the prevention of virologic rebound to lower-level viraemia, defined as 301 to 838 copies per millilitre. Among the subset of participants who had undetectable viral load at enrolment, the odds of virologic rebound at six months were 75% lower in the START group than in the control group. In the control condition, nearly half of these participants experienced rebound, compared to only a quarter in the intervention arm.

The study also examined secondary outcomes, including optimal antiretroviral therapy adherence, defined as reporting more than 90% adherence without a detectable viral load. At six months, START participants had nearly 2.5-fold higher odds of achieving optimal adherence, although this result was only marginally significant with a p-value of 0.051. The intervention did not produce statistically significant effects on self-reported methamphetamine use severity or positive emotion levels over the 12-month follow-up period. Engagement with the application was tracked, with participants completing a median of 11 check-ins within the first six months. However, engagement declined markedly after six months, with the median number of check-ins dropping to zero. Per-protocol analyses indicated that participants who completed at least two modules experienced comparable benefits to the intent-to-treat group, suggesting that even modest engagement may be sufficient to yield short-term clinical benefits.

From an implementation perspective, the cost of delivering the START application was estimated at $41,002 over 12 months, including fixed start-up costs for reprogramming. The average annual cost per participant was $231, or $19 per month, based on the 143 men randomised to the intervention. This low cost suggests strong potential for scalability, particularly if the application were broadly disseminated. The study was designed to reflect real-world implementation conditions, meaning participants were not provided with cell phones, data plans, or financial incentives for module completion. Weekly text message reminders were sent to those who had not logged in during the first six months, a staffing cost that was included in the financial estimates. The median system usability score for the application was 70, which falls within the acceptable range for user satisfaction.

Despite the significant short-term benefits, the study found that the effects of START on detectable viral load, virologic rebound, and optimal adherence were not durable at 12 months. There were no statistically significant differences between the groups at the one-year mark for any of these outcomes. The authors note that the short-term reductions in detectable viral load may be attributable to increases in optimal adherence observed only at six months. The decline in engagement after the first six months, particularly without the support of weekly reminders, likely contributed to the loss of efficacy. This underscores the need for more comprehensive digital health approaches that can sustain participant engagement over longer periods. The findings suggest that while mobile applications can effectively prevent early virologic rebound, they may not be sufficient on their own to achieve durable viral suppression in this population.

The study has several limitations that should be considered when interpreting the results. Approximately half of those who screened eligible were not randomised, primarily because they could not provide a viable remote blood sample for baseline viral load measurement. This highlights the challenges of using remote phlebotomy in clinical trials and suggests the need for better coordination of these procedures in future studies. The assay used for viral load measurement had a lower limit of detection of 300 copies per millilitre, which precluded the use of the standard clinical definition of viral suppression, typically defined as less than 200 copies per millilitre. The enrolled sample was composed of more non-Hispanic White participants who were younger and more economically disadvantaged compared to the broader population of sexual minority men with HIV. This demographic profile reflects the higher prevalence of methamphetamine use among these groups, but it may limit the generalisability of the findings to other racial and ethnic minorities.

The authors suggest that future efforts to boost and extend the benefits of digital health interventions should consider combining them with evidence-based behavioural therapies for reducing stimulant use, such as remote contingency management. This approach provides financial incentives for verified abstinence and has shown promise in prior studies. Another promising strategy may be augmenting the mobile application with individually delivered telehealth counselling sessions targeting substance use or adherence. The absence of detectable changes in positive emotion in this study may reflect the limitations of retrospective self-report measures, and future trials should incorporate ecological momentary assessment to capture real-time changes. Additionally, objective measures of adherence and stimulant use, such as urine or hair toxicology testing, would provide more precise estimates of the intervention’s effects. These steps are essential for developing more effective and durable digital health tools for key populations affected by HIV.

Subject of Research: Efficacy of a mobile health intervention on HIV viral load in sexual minority men with stimulant use disorders

Article Title: Effect of a mHealth intervention on detectable HIV-1 viral load: a remote randomised controlled trial

Article References: Carrico, A. W., Larson, M. E., Davis-Ewart, L., Ghanooni, D., Miller-Perusse, M., Ackley, D., Atkins, L., Diaz, J. E., Balise, R., McCollister, K. E., Fardone, E., Neilands, T. B., Moskowitz, J. T., Kantor, R., Sam, S. S., Spinelli, M. A., Hirshfield, S., & Horvath, K. J. (2026). Effect of a mHealth intervention on detectable HIV-1 viral load: a remote randomised controlled trial. eClinicalMedicine, 101, Article 104327. https://doi.org/10.1016/j.eclinm.2026.104327

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104327

Keywords: HIV, mHealth, viral load, stimulant use, randomised controlled trial, antiretroviral therapy, sexual minority men, digital health, virologic rebound, adherence, Effect, intervention