This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
While the human immunodeficiency virus (HIV) incidence rate has remained steady in most groups, the overall incidence of HIV among men who have sex with men (MSM) has been steadily increasing in the United States. eHealth is a platform for health behavior change interventions and provides new opportunities for the delivery of HIV prevention messages.
The purpose of this systematic review was to examine the use of eHealth interventions for HIV prevention in high-risk MSM.
We systematically searched PubMed, OVID, ISI Web of Knowledge, Google Scholar, and Google for articles and grey literature reporting the original results of any studies related to HIV prevention in MSM and developed a standard data collection form to extract information on study characteristics and outcome data.
In total, 13 articles met the inclusion criteria, of which five articles targeted HIV testing behaviors and eight focused on decreasing HIV risk behaviors. Interventions included Web-based education modules, text messaging (SMS, short message service), chat rooms, and social networking. The methodological quality of articles ranged from 49.4-94.6%. Wide variation in the interventions meant synthesis of the results using meta-analysis would not be appropriate.
This review shows evidence that eHealth for HIV prevention in high-risk MSM has the potential to be effective in the short term for reducing HIV risk behaviors and increasing testing rates. Given that many of these studies were short term and had other limitations, but showed strong preliminary evidence of improving outcomes, additional work needs to rigorously assess the use of eHealth strategies for HIV prevention in high-risk MSM.
Men who have sex with men (MSM) are the population most heavily affected by infection with human immunodeficiency virus (HIV) [
The number of new HIV infections among MSM increased 12% from 2008-2010, with a 22% increase among MSM aged 13-24 years. Notably, young African American MSM account for a disproportionate number of new HIV cases in the United States. There were more new HIV infections (54%) among young African American MSM (aged 13-29 years) than any other racial or ethnic age group of MSM [
While many HIV prevention interventions have been delivered face to face, the emergence of eHealth as a platform for health behavior change provides new opportunities for developing HIV prevention strategies [
The Internet is an important delivery method for eHealth tools. As access to the Internet increases, Americans’ willingness to use the Internet as a source of health information has proliferated, suggesting Web-based interventions are an important modality for health behavior change interventions [
A growing number of eHealth HIV prevention interventions have been developed for MSM [
We searched articles published from January 2000 to April 2014 in the following electronic databases: PubMed, PsycINFO, Embase, ISI Web of Knowledge, Google Scholar, and Google for grey literature of US and international studies. We visually scanned the reference lists of retrieved documents to identify additional relevant manuscripts. Our search terms included HIV, online, mobile technology, AIDS, technology, electronic health, eHealth, chat room, social networking, mobile applications, mobile health applications, mobile phone, mHealth, text messaging, telemedicine, HIV treatment, PLWH, reminder systems, information systems, Computers, Handheld/ or Cellular Phone/ or mobile applications, HIV/ or HIV.mp or HIV Infections/; Cellular Phone/ or mobile application.mp; HIV/, HIV Infections/ or PLWH.mp, HIV infection, intervention, mobile applications, and mobile HIV applications.
Included studies had to (1) focus on an eHealth intervention only and could not use eHealth solely as a recruitment or data collection tool, (2) focus on HIV prevention or testing and not on HIV care, (3) be published in English, (4) be published between January 2000 and April 2014, (5) be quasi-experimental or a randomized controlled trial (RCT), (6) have a behavioral outcome measure, and (7) focus on adult MSM. We did not include adolescent studies in our review since a recent systematic review was published [
A quality assessment tool (
HIV prevention intervention quality assessment tool.
|
Completely adequate (%) | Partially adequate (%) | Inadequate, not stated, or impossible to tell (%) | |
Representativeness | All key characteristics of study population described (50) | Some key characteristics described (25) | Minimal to no description of key characteristics and inclusion/exclusion criteria (0) | |
Detailed inclusion/exclusion criteria described (50) | Some description of inclusion/exclusion criteria (25) |
|
||
Bias and confounding | Study population corresponded to larger population in all key factors (25) | Sample population differed in some minor factors to larger population (12.5) | Sample population differed in several key factors to larger population (0) | |
Equivalent outcome assessment (25) | Minor differences in outcome assessment (12.5) | Major differences in outcome assessment (0) | ||
Study accounted for confounding interventions with respect to effectiveness of intervention (25) | Study only partially accounted for confounding interventions with respect to effectiveness of intervention (12.5) | Study did not account for confounding interventions with respect to effectiveness of intervention (0) | ||
Compliance rate >80% (25) | Compliance rate between 80-50% (16.7) | Compliance rate <50% (8.3) | ||
Description of intervention | Protocol could be replicated given description of intervention and /or monitoring (100) | Some minor details excluded from explanation of intervention and/or monitoring (66.7) | No details given in description of intervention and monitoring (0) | |
Some major details excluded from explanation of intervention and/or monitoring (33.3) | ||||
Outcomes and follow-up | Outcome assessment procedure clearly defined (50) | Outcome assessment procedure somewhat defined (25) | Outcome assessment procedure not defined (0) | |
Groups equivalent in attrition (50) | Some difference in attrition (25) | Major difference in attrition (0) | ||
Statistical analysis | Statistical methods fully described and appropriate (50) | Statistical methods partially described and appropriate (25) | Statistical methods not described or absent (0) | |
Tests addressed differences between groups and variability (50) | Tests addressed some differences between groups and variability (25) | Did not address differences between groups and variability (0) | ||
Strength of evidence | Significant positive intervention effects (100) | Significant effect but not in the stated relevant outcome measure (50) | No significant intervention effect (0) | |
Positive and statistically significant ( |
|
|
||
Group equivalence | Meets all 4 criteria (100) | Meets 3 criteria (75) | Meets no criteria (0) | |
1. Include one or more separate control or comparison study groups. | Meets 2 criteria (50) |
|
||
2. Include clear description of study group comparability. | Meets 1 criteria (25) |
|
||
3. Include clear description of randomization method used or rationale for not using randomization technique in instances when it is not feasible |
|
|
||
4. Include appropriate statistical controls when equivalence is not achieved |
|
|
Data were extracted based on objectives, study design, sample size, type and duration of interventions, outcome measures reported, and findings. To further characterize the intervention, we abstracted the theoretical framework used to guide the intervention design, if reported. Data were also abstracted according to country.
Screening process flowchart.
A total of 13 articles met the inclusion criteria. The articles were published between 2008 and 2013.
Each of the studies had different interventions. Interventions were not clearly described in approximately 46% of the studies. The length of the intervention period ranged from about 15 minutes [
Existing studies of eHealth HIV prevention interventions for adult MSM.
Study | Study design | eHealth Strategy | Length of study | Study population | Results | Mean quality score (range) |
Blas, 2010 [ |
RCT | Web-based Intervention | Mean of 125.5 days of observation | Intervention (N=239); |
Increased HIV testing rates | 76.79% |
Bourne, 2011 [ |
Pre-post test design | SMS reminders | SMS reminders every month for 3-6 months | Intervention (N=714); |
Increased HIV re-testing rates | 51.14 |
Bowen, 2008 [ |
Pre-Post study | Web-based education modules | Mean 19.39 days (SD 7.33 days) | Rural MSM (N=475) | Decrease high-risk sexual risk behaviors; | 89.89 |
Carpenter, 2010 [ |
RCT | Web-based skills training and motivational intervention | Intervention 1.5-2 h; |
MSM (N=112) | Reduction in high-risk HIV behavior | 64.89 |
Christensen, 2013 [ |
RCT | Virtual Simulation Intervention | 3-month follow-up questionnaire | Intervention (N=437) |
Shame reduction; shame reduction as a predictor of UAIa | 75 |
Hirshfield, 2012 [ |
RCT | Web-based media intervention (prevention videos & webpage) | Baseline survey, Intervention 60 day follow up | Intervention (N=2483) |
More likely to disclose HIV status to partners; less likely to report UAI | 89.89 |
Ko, 2013 [ |
Quasi-Experimental, Non-Equivalent control | Web-based peer leader intervention | Baseline survey, 6-month intervention, follow-up survey | Intervention (N=499); Comparison (N=538) | Increased HIV testing, reduced UAI | 60.11 |
Lau, 2008 [ |
RCT | Web-based educational tool | 6-month study period | Intervention (N=140); |
Efficacy of the intervention was not supported | 65.49 |
Mustanski, 2013 [ |
RCT | Web-based media intervention | 12-wk study period | Intervention (N=50); |
Decrease sexual risk behavior | 94.64 |
Reback, 2012 [ |
Pre-post test design | Text Messaging | 2-wk intervention | Meth-using MSM (N=52) | Decreased frequency of methamphetamine use; Decrease high-risk sexual behaviors. | 83.93 |
Rhodes, 2011 [ |
Single-group pretest-post-test design | Chat Rooms | 6-month implementation phase; 1-month follow-up | MSM (N=346 [pretest], 315 [posttest]) | Increased HIV testing rates | 64.89 |
Rosser, 2010 [ |
RCT | Interactive Website | 3-wk intervention | MSM (N=650) | Reduction in risk behavior | 49.41 |
Young, 2013 [ |
RCT | Web Based, Peer leader led groups | 12-wk intervention; 12-wk follow-up | 112 MSM |
Increased requests for an HIV home test | 80.36 |
aUAI: unprotected anal intercourse
Two studies used videos for educating high-risk MSM. In a study conducted in Peru, 5-minute videos were created using the Health Belief Model and Stages of Change Theory to encourage MSM to get tested for HIV. The videos incorporated ways to overcome eight reasons why MSM do not get tested for HIV (eg, fear or lack of confidentiality) [
One study developed and tested multicomponent Internet sites that targeted high-risk sexual behaviors. The intervention, Sexpulse, was a multifaceted Internet intervention that targeted men who use the Internet to seek sex with men and was informed by the Sexual Health Model. Sexpulse was designed by a multidisciplinary team of health professionals, computer scientists, and e-learning specialists and had the following components: a risk assessment tool, an online chat simulation, and virtual peers. Use of the system successfully reduced high-risk sexual behavior in study participants [
Keep It Up! (KIU!) was an online, interactive HIV prevention program. The IMB model of HIV risk behavior change was used to guide the development of the KIU! intervention. It has 7 modules completed across 3 sessions that were done at least 24 hours apart and takes about 2 hours to complete. Keep It Up! was designed to be delivered to young MSM upon receiving an HIV negative test result. In an RCT, the participants in the intervention arm had a significantly lower rate of unprotected anal sex acts at the 12-week follow-up [
Socially Optimized Learning in Virtual Environments (SOLVE) is a downloadable simulation video game designed to simulate and immerse high-risk young adult MSM in affectively charged risky situations. This intervention was informed by the Theory of Planned Behavior, and Social Cognitive Theory. Christensen et al tested this intervention compared to a wait list control condition in an online RCT. After 3 months, participants in the SOLVE treatment condition reported greater reductions in shame. The direct effect on risky sexual behavior at follow-up was not significant [
Finally one study developed and tested a Web-based education module tailored to the information needs of MSM residing in rural areas. There were two 20-minute education sessions that participants watched 6 months apart. Each session consisted of three modules focused on the concepts in the IMB model. Post-intervention behavior change included reduced anal sex and significant increases in condom use [
Three studies in our review used text messaging or short messaging service (SMS) as an intervention; two studies used it to increase HIV testing rates and one study to reduce high risk behaviors. The two studies that targeted increasing HIV testing rates were conducted outside of the United States. In one of the SMS studies set in Australia, clinicians sent reminders to patients who had previously come to a sexual health clinic to come back for follow-up testing. SMS reminders increased HIV re-testing rates after 9 months [
One study used a chat room intervention named CyBER/testing, informed by the Natural Helping Theory, in which an interventionist entered the chat room from 9 a.m.-5 p.m., Monday to Friday [
In the HOPE study, social network sites were used for the delivery of HIV prevention information; 16 peer leaders were randomly assigned to deliver information about HIV (intervention) or general health (control) via Facebook groups for over 12 weeks. Participants randomized to the HIV prevention information group were significantly more likely to request an HIV testing kit than control group participants [
In another social networking intervention study, Internet popular opinion leaders (iPOL) were used to disseminate HIV prevention information via popular social networking sites [
This review of eHealth interventions for HIV prevention among adult MSM has drawn together the evidence base specific to behavioral interventions for MSM and found evidence for eHealth interventions being associated with reductions in high risk behaviors and increases in HIV testing rates. Nonetheless, the studies that demonstrated a decrease in sexual risk behavior had different study designs and outcome measures that make it difficult to synthesize the evidence.
Only one US study in our review solely focused on HIV testing as an outcome measure. Given that the US National HIV/AIDS Strategy has established a goal of increasing the awareness of HIV status in the US population from 79% to 90% by 2015, HIV testing is an important HIV prevention measure. In fact, current recommendations are to repeat HIV testing every 3-6 months for high-risk MSM [
Bourne et al found that SMS can be used to increase HIV testing rates in high-risk MSM [
From the results presented above, we can infer that eHealth interventions reduce risky sexual behaviors and increase HIV testing. This review has provided evidence that eHealth interventions have the potential for promoting HIV prevention behaviors in adult MSM. Even so, there are a number of limitations in many of the studies we reviewed. For example, in the study conducted by Reback et al (2012) [
In another study, Young et al (2013) [
Moreover, there is a need for long-term (12 months) follow-up data after the completion of eHealth HIV prevention interventions. In our review, only 1 study assessed the long-term effects (12 months) of the eHealth intervention and found that it did not have a long-term effect on reducing sexual risk behaviors [
Several limitations of this review should be considered when interpreting the findings. The potential heterogeneity of interventions and outcomes are important to note and make the synthesis of the evidence from these studies challenging. Notably, even though we attempted to be as inclusive as possible, our searches may have excluded relevant studies from this systematic review that did not meet our search word criteria, and/or we excluded conference abstracts that met this review’s criteria but were not peer-reviewed articles.
Our results have important implications for the use of eHealth interventions for HIV prevention in MSM. This review demonstrates eHealth interventions appear potentially useful for reducing HIV risk behavior and increasing HIV testing rates. The detailed data across the studies allows us to comprehensively identify and describe elements that are essential to the effectiveness of eHealth interventions for promoting HIV prevention among adult MSM. Given the limitations of many of these studies as well as the potential for eHealth to transform health behaviors, additional work needs to rigorously assess the use of eHealth strategies for HIV prevention in high-risk MSM. Future work is needed that employs these interventions in longer and larger trials and to assess their efficacy in improving outcomes.
acquired immunodeficiency syndrome
Centers for Disease Control and Prevention
human immunodeficiency virus
Information-Motivation-Behavioral Skills Model
men who have sex with men
randomized controlled trial
short message service
unprotected anal intercourse
This publication was supported by a cooperative agreement between Columbia University School of Nursing and the Centers for Disease Control and Prevention (CDC; 1U01PS00371501; PI: R Schnall). RS is supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant No. KL2 TR000081, formerly the National Center for Research Resources, Grant No. KL2 RR024157. The findings and conclusions in this report do not necessarily represent the views of the National Institute of Health or the CDC.
None declared.