A Mobile Health Tool to Capture Social Determinants of Health and their Impact on HIV Treatment Outcomes among People Who Use Drugs: Pilot Study (Preprint)

Abstract

Active substance use, food or housing insecurity, and criminal justice system involvement can disrupt HIV care for people living with HIV (PLWH) and opioid use disorder (OUD). These social determinants of health are not routinely captured in clinical settings.

We evaluated whether real-time reports of social and behavioral factors using a smartphone app could predict viral non-suppression and missed HIV care visits to inform future mobile health interventions.

We enrolled 59 participants from the AIDS Linked to the IntraVenous Experience (ALIVE) Study in Baltimore, Maryland into a 12-month sub-study between February 2017 and October 2018. Participants were eligible if they had OUD and had either a measured HIV RNA ≥1000 copies/mL or a ≥1 month lapse in antiretroviral therapy in the preceding 2 years. Participants received a smartphone and reported HIV medication adherence, drug use or injection, and several disruptive life events, including not having a place to sleep at night, skipping a meal due to lack of income, being stopped by police, being arrested, or experiencing violence on a weekly basis through a survey on a mobile health application. We described weekly survey completion and investigated which factors were associated with viral non-suppression (HIV RNA ≥200) or a missed HIV care visit using logistic regression with generalized estimating equations adjusted for age, gender, smartphone comfort, and drug use.

Participants were predominately male (61.0%), Black (89.8%), and a median of 53 years old. At baseline, 21.6% (N=8) were virally unsuppressed. Participants completed an average of 23.3 total surveys (SD: 16.3) and reported missing a dose of ART, using or injecting drugs, or experiencing any disruptive life events on an average of 13.1 (SD: 9.8) surveys over one year. Reporting use of any drugs (adjusted odds ratio [aOR]: 2.3, 95% Confidence Interval [CI]: 1.4-3.7), injecting drugs (aOR: 2.3, 95% CI: 1.3-3.9), and non-completion of all surveys (aOR: 1.6, 95% CI: 1.1-2.2) were associated with missing a scheduled HIV care visit over the subsequent 30 days. Missing ≥2 antiretroviral medication doses within a one-week period was associated with HIV viral non-suppression (aOR: 3.7, 95% CI: 1.2-11.1) on laboratory tests drawn within the subsequent 30 days.

Mobile health applications can capture risk factors that predict viral non-suppression and missed clinic visits among PLWH who have OUD. Using mobile health tools to detect socio-behavioral factors that occur prior to treatment disengagement may facilitate early intervention by healthcare teams.