eHealth interventions to address HIV and other sexually transmitted infections, sexual risk behavior, substance use, and mental ill-health in men who have sex with men: Systematic review and meta-analysis

Background: Men who have sex with men experience disproportionately high levels of HIV and other sexually transmitted infections (STIs), sexual risk behavior, substance use, and mental ill-health. These experiences are interrelated, and these interrelations are potentiated by structural conditions of discrimination, stigma, and unequal access to appropriate health services, and they magnify each other and have intersecting causal pathways, worsening both risk for each condition and risk for the negative sequelae of each condition. eHealth interventions could address these issues simultaneously and thus have wide-ranging and greater effects than would be for any 1 outcome alone.

Objective: We systematically reviewed the evidence for the effectiveness of eHealth interventions in addressing these outcomes separately or together.

Methods: We searched 19 databases for randomized trials of interactive or noninteractive eHealth interventions delivered via mobile phone apps, internet, or other electronic media to populations consisting entirely or principally of men who have sex with men to prevent HIV, STIs, sexual risk behavior, alcohol and drug use, or common mental illnesses. We extracted data and appraised each study, estimated meta-analyses where possible by using random effects and robust variance estimation, and assessed the certainty of our findings (closeness of the estimated effect to the true effect) by using GRADE (Grading of Recommendations, Assessment, Development and Evaluations).

Results: We included 14 trials, of which 13 included active versus control comparisons; none reported mental health outcomes, and all drew from 12 months or less of follow-up postintervention. Findings for STIs drew on low numbers of studies and did not suggest consistent short-term (<3 months postintervention; d=0.17, 95% CI -0.18 to 0.52; I²=0%; 2 studies) or midterm (3-12 months postintervention, no meta-analysis, 1 study) evidence of effectiveness. Eight studies considering sexual risk behavior outcomes suggested a short-term, nonsignificant reduction (d=-0.14, 95% CI -0.30 to 0.03) with very low certainty, but 6 studies reporting midterm follow-ups suggested a significant impact on reducing sexual risk behavior (d=-0.12, 95% CI -0.19 to -0.05) with low certainty. Meta-analyses could not be undertaken for alcohol and drug use (2 heterogeneous studies) or for HIV infections (1 study for each of short-term or midterm follow-up), and alcohol outcomes alone were not captured in the included studies. Certainty was graded as low to very low for most outcomes, including all meta-analyses.

Conclusions: To create a comprehensive eHealth intervention that targets multiple outcomes, intervention evaluations should seek to generalize both mechanisms and components that are successfully used to achieve change in 1 outcome over multiple outcomes. However, additional evaluations of interventions seeking to address outcomes other than sexual risk behavior are needed before development and evaluation of a joined-up intervention.