Text Messaging, Behavioral Therapy Improve Adherence in Global Study

Text messaging, cognitive behavioral therapy and supporter interventions improved antiretroviral adherence or viral suppression in a worldwide 85-trial meta-analysis. But these interventions appeared to have only a small impact that waned over time.

Good adherence is essential to the clinical, public health and social goals of antiretroviral therapy, noted Canadian and U.S. researchers who conducted this analysis. But individual trials of adherence strategies yield varying results, so overall interpretation is difficult. To get a better understanding of which adherence interventions work best in people with HIV, the researchers turned to network meta-analysis, which allows comparison of interventions not directly compared in trials. This network meta-analysis was the first to consider all antiretroviral adherence interventions (nine of them) in a single analysis.

The investigators searched electronic databases (up July 16, 2015) and HIV meeting abstracts (from 2013 through July 16, 2015) to identify randomized controlled trials comparing an adherence intervention with standard of care. Trial outcomes could be antiretroviral adherence or viral suppression. From selected trials, the investigators extracted data on study characteristics, interventions, baseline patient characteristics and outcomes. They used logistic regression to analyze trial outcomes and meta-regression to account for populations at risk for poor adherence (drug injectors, other substance users, people with mental health disorders and people with poor adherence at study entry).

The analysis included 85 trials, most of which defined adherence as higher than 95% or 100% adherence. Seventy-one trials enrolling 11,582 patients reported adherence outcomes; 21 of these trials were in low- or middle-income countries (LMIC). Forty-seven trials enrolling 7355 patients reported viral suppression; 12 of these trials were in LMIC.

In a global analysis including all trials assessing adherence, text messaging proved superior to standard of care (odds ratio [OR] 1.48, 95% confidence interval [CI] 1.00 to 2.19). In the global analysis of viral suppression, two interventions did better than standard of care: cognitive behavioral therapy (OR 1.46, 95% CI 1.05 to 2.12) and supporter interventions (OR 1.28, 95% CI 1.01 to 1.71). (Supporter interventions involve an adherence helper.) In the LMIC adherence analysis, text messaging proved superior to standard of care (OR 1.49, 95% CI 1.04 to 2.09), but no individual intervention improved viral suppression in LMIC.

For the adherence outcome, combined interventions generally proved superior to single interventions, particularly (1) supporter plus telephone interventions, (2) cognitive behavioral therapy plus incentives and (3) cognitive behavioral therapy plus a device reminder. For both adherence and viral suppression, analysis of whether trials assessed adherence during or after the intervention suggested that intervention effects wane over time.

The Canadian-U.S. team concludes that supportive strategies (peer support, text messaging, counseling) and behavioral strategies improve adherence. Although measured effects were small and seemed to decline after the intervention stopped, combining effective interventions appeared to increase adherence. The investigators suggest that certain interventions may be more feasible in certain regions -- such as cognitive behavioral therapy in high-income countries and text messaging in LMIC.

The findings of this analysis led the World Health Organization to expand its adherence recommendations from an emphasis on text messaging to also include other options: peer counseling, reminder devices, cognitive behavioral therapy, behavioral skills training and medication adherence training.