This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
HIV prevalence remains high among men who have sex with men (MSM) in the United States, yet the majority of research has focused on MSM as individuals, not as dyads, and has discussed HIV risks primarily in the context of casual sex. Nexus is an online prevention program that combines home-based HIV testing and couples HIV testing and counseling (CHTC). It allows partners in dyadic MSM relationships to receive HIV testing and care in the comfort of their designated residence, via video-based chat. By using video-based technologies (eg, VSee video chat), male couples receive counseling and support from a remote online counselor, while testing for HIV at home.
This randomized control trial (RCT) aims to examine the effects of video-based counseling combined with home-based HIV testing on couples’ management of HIV risk, formation and adherence to explicit sexual agreements, and sexual risk-taking.
The research implements a prospective RCT of 400 online-recruited male couples: 200 self-reported concordant-negative couples and 200 self-reported discordant couples. Couples in the control arm will receive one or two home-based HIV self-testing kits and will be asked to report their results via the study’s website. Couples in the experimental arm will receive one or two home-based HIV self-testing kits and will conduct these tests together under the facilitation of a remotely located counselor during a prescheduled VSee-based video CHTC session. Study assessments are taken at baseline, as well as at 3- and 6-month follow-up sessions.
Project Nexus was launched in April 2016 and is ongoing. To date, 219 eligible couples have been enrolled and randomized.
Combining home-based HIV testing with video-based counseling creates an opportunity to expand CHTC to male couples who (1) live outside metro areas, (2) live in rural areas without access to testing services or LGBTQ resources, or (3) feel that current clinic-based testing is not for them (eg, due to fears of discrimination associated with HIV and/or sexuality).
ClinicalTrials.gov NCT02335138; https://clinicaltrials.gov/ct2/show/NCT02335138 (Archived by WebCite at http://www.webcitation.org/6qHxtNIdW)
HIV prevalence remains high among men who have sex with men (MSM) in the United States [
CHTC has been used as an HIV prevention intervention for heterosexual couples in Africa for over 20 years [
A critical aspect of CHTC involves discussing a couple’s sexual agreement. Sexual agreements refer to mutually understood rules between two partners that describe the kinds of sexual behavior that are allowed within and outside of their relationship [
In addition to CHTC, another HIV testing option is home-based HIV testing, which was approved by the US Food and Drug Administration in 2012 [
A vital cornerstone to prevention and linkage to care is HIV testing [
The theoretical basis for intervention is the Couple’s Interdependence Theory (CIT) [
The research activities involved a blind prospective RCT of approximately 400 online-recruited male couples—200 self-reported concordant-negative couples and 200 self-reported discordant couples. Couples in the control arm will receive one or two home-based HIV self-testing kits and will be asked to report their results via the study’s website. One kit will be provided for each person who reports to have previously tested HIV-negative or unknown status; partners in serodiscordant couples who report living with HIV will not be retested. Couples in the experimental arm will receive one or two home-based HIV self-testing kits and will conduct these tests together under the facilitation of a remotely located counselor during a prescheduled VSee-based video CHTC session.
Participants for each male couple must meet the following eligibility criteria: (1) two men who have been in a sexual relationship with each other for more than 6 months; (2) >18 years of age; (3) both participants not having been tested for HIV in the last 3 months, or for serodiscordant couples, the negative partner not having tested for HIV in the last 6 months; (4) reporting no IPV or coercion within the last 12 months; (5) willing to have HIV test kits delivered to an address they provide; (6) have access to the Internet within their home, or the home of at least one partner; and (7) be either self-reported concordant HIV negative or self-reported HIV serodiscordant. Participants for the trial are being recruited from across the United States, with recruitment via online advertisements placed on key social media websites (eg, Facebook and Instagram) and social media sites aimed specifically at MSM (eg, Grindr and Scruff). When men click on the advertisement, they are taken to a page containing basic study information, including a short description of study activities. If they express an interest in participation, they are then taken to the study consent form and if they consent, they are directed to a short eligibility screener. Men who (1) do not consent, (2) do not meet the eligibility criteria, or (3) do not provide an email for a main partner—defined as a sexual relationship with “a man who you feel committed to above all others”—are taken to a screen thanking them for their interest. Men who are eligible to participate must provide an email address for their main partner so that they can be enrolled in the study together. Further, eligible men able to participate and that provide their partner’s email address are directed to a registration process. During the registration process, both partners provide their contact information, including an email address, a mobile phone number, and a mailing address; they are also asked to provide a nickname or preferred name of choice. Once both partners have (1) completed the consent forms, (2) completed the screening questionnaire, (3) proven eligible for the study, and (4) registered on the study website, a joint email is sent to both partners asking them to complete the baseline questionnaire, individually.
Upon individual completion of the baseline survey by both partners, couples are randomized to either the home-based HIV testing arm or video-based CHTC with home-based HIV testing arm using a stratified 1:1 treatment allocation. The strata are based on two levels of serostatus: seroconcordant negative and serodiscordant. The treatment assignments are generated with the use of a pseudo-random number generator with permutated blocks that are used to ensure balance within stratum between the numbers of couples assigned to each treatment. The randomization process generates one of two emails to both study participants for the enrolled male couple, indicating whether they will be receiving home-based HIV testing or video-based CHTC with home-based HIV testing.
The proposed intervention is a combination of home-based HIV testing and CHTC offered remotely via VSee video chat. In the control arm, couples receive one or two home-based HIV testing kits based on the couple’s serostatus; the partner living with HIV in a serodiscordant couple does not repeat HIV testing. These kits sent to couples in the control arm contain one or two HIV testing kits as well as instructions on how to use the kits and how to report their test results in the study portal. In the experimental arm, couples receive one or two home-based HIV testing kits and complete the testing while undergoing a remote, VSee video-based CHTC session. Individual participants testing positive in either the experimental or the control condition are linked to their preference of local HIV care-specific resources within 48 hours. They will receive another HIV test at their preferred local HIV care-specific resource to validate the preliminary positive test result. These participants will receive follow-up from the CHTC counselor at 1-week, 1-month, and 3-month postresult of HIV-positive status to assess their engagement in care.
For couples randomized to the experimental arm, an email informs them that they have the opportunity to receive one or two home-based HIV testing kits and to take part in a video-based counseling session. The email provides details on the expected content of a CHTC session, the expectation that both partners will need to conduct their individual HIV tests and receive their results together in the presence of a remote counselor, as well as further logistical information (ie, length of the counseling session). From this email, couples are instructed to log on to the study website to order their HIV testing kits, with the same options as the control arm.
The CHTC session is conducted via video chat using VSee.
Project Nexus website log-in page.
Calendar function for selecting a couples HIV testing and counseling (CHTC) appointment.
VSee session interface.
HIV test results reporting page on Nexus website.
An online survey of 1285 MSM from across the United States was conducted to examine willingness to use the proposed intervention (unpublished data). Respondents were recruited through targeted advertisements on Facebook over 10 days from October to November 2012. Of the 907 men who self-reported a negative HIV serostatus and provided complete data, 72.0% (653/907) reported they would be likely to use a home-based HIV test, 69.0% (626/907) reported they would be likely to use CHTC, and 75.0% (680/907) reported that they would be likely to use a home-based HIV test together with a VSee video-based CHTC session. The willingness to use the proposed intervention did not vary by age (
The study outcomes are based on the conceptual model of CIT [
Initiation and maintenance of health-enhancing behaviors are conceptualized as including two sets of outcomes: sexual agreements and sexual risk-taking as well as linkage and retention to HIV care.
All surveys include questions taken from previous studies of male couples’ sexual agreements [
Behavioral measures adapted from the National HIV Behavioral Surveillance behavioral inventory, as well as from studies using behavioral measures among thousands of MSM [
The following outcomes as indicators of linkage to care, per the recent recommendations of the Institute of Medicine [
The four elements of Lewis’ model [
Several scales are used to measure this element. The Perceived Severity of HIV Scale involves the perception of the personal, psychosocial, and physical consequences of a particular health threat. A total of 13 items were developed that crossed the three pertinent consequences of a particular health threat: personal, psychosocial, and physical. The Preferences for General Lifestyle Outcomes Scale is defined as the degree to which interacting partners agree about the shared or joint outcomes in their relationship and is composed of two subscales: the Preferences for General Lifestyle Scale and the Preferences for Sexual Health Outcomes Scale. The Preferences for General Lifestyle Scale includes six outcomes, including diet, nutrition, and social activities. The Preferences for Sexual Health Outcomes Scale relates to sexual health, for example, reducing one’s risk for HIV. In addition, scales to measure other predisposing factors of couples are proposed for inclusion. Conflict style determines how respondents typically handle conflict in their relationships, so the Conflict Style Inventory will be included [
In a recent RCT of CHTC, two measures were developed: ability of the participants to respond (1) cognitively and (2) emotionally to the health threat [
Several scales are also used to measure this element. The Outcome Efficacy to Reduce HIV Threat Scale discusses how communal coping involves couples working together and making decisions together to reduce the health threat. Three subscales were created to capture the full range of outcome efficacy related to these three processes of communal coping. For the first subscale, Joint Effort, the stem “My partner and I believe that ‘working together’ versus on our own is an effective strategy” is used. For the second subscale, Communication, the stem “Communicating with my partner is an effective strategy for...” is used. For the third subscale, Planning and Decision Making, the stem “My partner and I making decisions together rather than separately is an effective strategy” is used. The items for each of the three subscales were the same as the items used for the Preference for Sexual Health Outcomes Scale. The Couple Efficacy to Reduce HIV Threat Scale defines couple efficacy as a couple’s confidence that together they can engage in communal coping efforts.
In addition to the outcomes tied to Lewis’ framework [
Dyadic characteristics will be analyzed within and between couples over time by HIV status and study arm (control vs experimental). Simple
The definition of at-risk sex will be serostatus specific. For serodiscordant couples, at-risk sex will be CAI with either their main or outside partners. For seroconcordant-negative and seroconcordant-positive couples, at-risk sex will be CAI with outside sex partners. While data on PrEP use and adherence will be collected, the definition of at-risk sex focuses on condom use given the CDC recommendation for continued condom use for those adopting PrEP. The incidence of at-risk sex acts will be calculated as an incidence density, with the numerator being the number of individual at-risk sex acts and the denominator being person-years of follow-up time. Comparisons of the incidence of at-risk sex acts will be made by comparing incidence densities between the two arms. Incidence rates per couple-year of follow-up will be estimated and compared using exact methods based on the Poisson distribution by using the GEE approach. Baseline covariates include race, age, and duration of relationship. Period incidence rates—6-monthly incidence density rates—of at-risk sex will be estimated by performing a GEE Poisson regression analysis of the 6-monthly counts. This will be implemented using the PROC GENMOD procedure by SAS (SAS Institute Inc) [
Separately tabulated data about disclosure of sex outside the relationship and percentage of couples with agreements involving sex outside the primary relationship will be developed. These analyses will be descriptive and include analyses stratified by race/ethnicity, study arm, and couple serostatus. These analyses will also characterize the prevalence of agreements about sex outside relationships, the extent to which those agreements are adhered, and any shift of adopting safer sexual behaviors with outside partners, again stratified by race/ethnicity and study arm. A focus of the analysis will be identifying differences across the study arms in the percentages of couples who report a shift to safer sexual agreements at follow-up. In addition, the Actor-Partner Interdependence Model (APIM), which uses the dyad (eg, couple) as the unit of analysis, will be used to predict how an individual and his partner’s reports of dyadic characteristics—described within Lewis’ model [
The percentage of HIV-positive respondents who receive a timely (<3 months) comprehensive visit—a visit including a CD4 count, viral load count, and the date of their first care visit—will be tested for significance across the two study arms. The percentages of couples who are seroconcordant negative, seroconcordant positive, and serodiscordant, as part of the description of the analysis samples, will be recorded. To assess whether the concordancy of the couple modifies the intervention effects, formal statistical tests—likelihood ratio tests in mixed models and generalized score tests in marginal models—of the null hypothesis will be conducted. The null hypothesis is that the two-way interaction effect for the aforementioned items and scales between intervention and concordancy is zero. If the statistical test does not reject the null hypothesis, the concordancy main effect in the model will be retained, removing the two-way interaction, and conclude that there is not sufficient evidence to suggest that concordancy modifies the intervention effect.
The safety of the intervention at both the individual and couple levels will be evaluated by examining reported IPV within the relationship and relationship dissolution. At the individual level, prevalence of each individual adverse outcome or any adverse outcome will be calculated. Prevalence of outcomes will be compared across the control and intervention arms, by serostatus of the couple, and by relationship duration using chi-square tests or Fisher’s exact tests, as appropriate. At the couple level, the APIM will be used to predict how an individual and his partner’s reports of dyadic characteristics affect the individual’s experience of IPV and reported relationship dissolution.
In addition to the outcomes, the study will assess feasibility by examining (1) time to recruit 400 couples to the intervention and (2) rate of recruitment per 100 men expressing interest in participation. Acceptability of the intervention will be determined by analysis of data from the satisfaction survey on the intervention’s acceptability to couples. In addition, the percentages of male couples who do not complete the home-testing profile and the percentage of those not returning test results via the study website will be analyzed.
To examine the cost of the proposed intervention, cost data are collected by input type and by activity, using an activity-based costing matrix. Input types include study personnel salaries and benefits, supplies, equipment, HIV testing kits, and training materials. Activities include training (ie, counselors and crisis counseling), recruitment of participants, VSee video-based chat, follow-up, and linkage to care. For the experimental arm, counselors record the time spent per session of the counselor because these are the only additional (ie, incremental) costs incurred relative to the control arm.
Individual participants each receive US $50 for completing each of the main three surveys: the baseline, the 3-month follow-up, and the 6-month follow-up. If all surveys are completed by both members of the male couple, the total incentive amount is US $300 per couple (US $150 per individual participant).
The goal is to enroll and maintain a sample of 400 male couples: 200 serodiscordant and 200 seroconcordant couples, 100 of each couple per arm (experimental vs control). To achieve this, the intervention aims to screen approximately 1000 male couples—500 serodiscordant and 500 seroconcordant male couples—and will exclude those with a recent (<12 months) history of IPV. Assuming 15% of male couples have a recent history of IPV, this will produce approximately 850 couples for randomization, 425 per arm. Allowing for 20% loss to follow-up and additional 20% relationship dissolution, this will produce a sample of 400 male couples who are expected to complete the prospective RCT. The sample size is calculated based on the detection of significant changes in each of the main outcomes (ie, changes in sexual risk-taking, such as CAI), formation and adherence to explicit sexual agreements, and relationship functioning for the management of HIV risk. As an example, we will assume that about 25% of male couples change agreement status after couples counseling and we will use a two-sample test of binomial proportions with type I error rate at 5%. Using these assumptions, it is determined that 52%, 67%, and 85% statistical power is necessary to detect a difference of 10%, 12%, and 15%, respectively, in a change-of-agreement status between the two arms within seroconcordant or serodiscordant groups (ie, comparing 150 male couples per arm). In other words, under the same conditions, the probability is .52, .67, and .85 that the lower bound of the confidence interval for the proportion change in agreement status in the intervention arm is at least 3%, 5%, and 8% different than the proportion in the control arm, respectively, when the alternative hypothesis is true. If data is pooled across seroconcordant and serodiscordant couples, assuming that combining data makes sense in terms of the magnitude and direction of the intervention effect, then 81%, 92%, and 99% statistical power is necessary to observe differences of 10%, 12%, and 15%, respectively (ie, comparing 300 male couples per arm). In this case, at least 80% statistical power will be necessary to detect as little as a 5% difference between the lower bound of the confidence interval for the proportion in the intervention versus the control arm. Based on discussions with participant advocates and partnered community-based organizations, it is projected that any difference exceeding 5% would be scientifically meaningful and would have public health impact. The analysis also considers differences in linkage to HIV care among HIV-positive individuals in each arm, but it is not powered to detect significant differences, given the small number of incident positive cases.
The research and ethics presented in this study have been reviewed and approved by the University of Michigan Institutional Review Board (HUM00102906), in addition to the Data Safety Monitoring Board. The study is also registered on ClinicalTrials.gov (NCT02335138).
Project Nexus was launched in April 2016 and is ongoing. To date, 219 couples have been enrolled and randomized, with 219 couples remaining eligible for continued study participation, as they have reportedly not experienced IPV, ended a relationship, falsified information, or any other criteria that would make them ineligible for study participation. Of all eligible couples, 95% have taken the baseline survey; the remaining couples are all within the time between enrollment and taking the survey. The 3- and 6-month follow-up surveys have maintained high retention rates at over 90%. Of couples randomized to the control arm, 88.0% (95/108) have reported their home HIV testing results. Of couples randomized to the intervention arm, 76.6% (85/111) have scheduled and completed the video-chat-based counseling session. In total, 5.9% (26/438) of participants have tested preliminarily as HIV positive, of which 73% (19/26) were actively linked to care.
Online CHTC via video chat provides an opportunity to expand CHTC to male couples who (1) live outside metro areas, (2) live in rural areas without access to testing services or LGBTQ resources, or (3) feel that current clinic-based testing is not for them (eg, due to fears of discrimination associated with HIV and sexuality). Although home-based HIV testing is now a reality, many still question the lack of counseling available to those who are undergoing testing in the individual household. A video-chat-based CHTC session potentially provides an inexpensive way to remedy this problem and provides an opportunity for those receiving an HIV-positive result to receive assistance in linkage to care. The proposed activities not only have the potential to expand CHTC to male couples who currently do not have access, but may provide opportunities to improve the utility of home-based testing for couples by providing them a forum to discuss prevention planning with a counselor. It is true that some individuals use home-based HIV testing because they do not want counseling and the proposed intervention would probably not be adopted by those people. However, for couples who desire and/or need counseling and do not have access due to physical or sociocultural barriers, a low-cost, video-based counseling session provides the opportunity to reach them, create HIV prevention planning, and locate linkage to care in a safe and comfortable environment.
Actor-Partner Interdependence Model
condomless anal intercourse
US Centers for Disease Control and Prevention
couples HIV testing and counseling
Couple’s Interdependence Theory
generalized estimating equations
intimate partner violence
incidence rate ratio
men who have sex with men
preexposure prophylaxis
randomized controlled trial
None declared.