The Effectiveness of Electronic Health Interventions for Promoting HIV-Preventive Behaviors among Men Who Have Sex with Men: Meta-Analysis Based on an Integrative Framework of Design and Implementation Features

Background: The disproportionately high prevalence of HIV among men who have sex with men (MSM) is a global concern. Despite the increasing utilization of electronic health (eHealth) technology in the delivery of HIV prevention interventions, few studies have systematically explored its effectiveness and association with various intervention characteristics. Objective: This study aimed to conduct a meta-analysis of the effectiveness of eHealth technology–based interventions for promoting HIV-preventive behaviors among MSM and to determine effectiveness predictors within a framework integrating design and implementation features. Methods: A systematic literature search using terms related to eHealth technology, HIV, the MSM population, and an experimental study design was performed using 5 databases (ie, MEDLINE, PsycINFO, EMBASE, Web of Science, and ProQuest Dissertations & Theses) and other sources (eg, bibliographies of relevant reviews and JMIR Publications). First, primary meta-analyses were conducted to estimate the effectiveness of eHealth interventions (d₊) in changing 3 HIV-preventive behaviors among MSM: unprotected anal intercourse (UAI), HIV testing, and multiple sex partnership (MSP). Moderation analyses were then conducted to examine a priori effectiveness predictors including behavioral treatment components (eg, theory use, tailoring strategy use, navigation style, and treatment duration), eHealth technology components (eg, operation mode and modality type), and intervention adherence. Results: A total of 46 studies were included. The overall effect sizes at end point were small but significant for all outcomes (UAI: d₊=−.21, P<.001; HIV testing: d₊=.38, P<.001; MSP: d₊=−.26, P=.02). The intervention effects on UAI were significantly larger when compared with preintervention groups than with concurrent groups. Greater UAI reductions were associated with the increased use of tailoring strategies, provision of feedback, and tunneling navigation in interventions with a concurrent group, whereas reductions were associated with the use of self-paced navigation in interventions with a preintervention group. Greater uptake of HIV testing was associated with longer treatment duration; computer-mediated communication; and the use of messaging, social media, or a combined technology modality. Higher intervention adherence consistently predicted larger effects on UAI and HIV testing. Conclusions: This study provided empirical evidence for the effectiveness of eHealth interventions in promoting HIV-preventive behaviors among MSM. Features of treatment content and eHealth technology might best predict the intervention effects on UAI and HIV testing, respectively. Most importantly, intervention adherence tended to play an important role in achieving better effectiveness. The findings could help inform the development of efficacious interventions for HIV prevention in the future.