JMIR Publications

ABSTRACT

Background:

Group sex events (GSEs) are common among cisgender men who have sex with men (MSM), pose a unique risk profile for HIV and STD transmission, and may be on the rise, in part due to online networking platforms. Yet collecting data on GSEs can be challenging, and many gaps exist in our knowledge about GSE participation among MSM.

Objective:

To develop survey questions addressing aggregate and partner-specific group sex behaviors in order to measure prevalence of GSEs and associated risks in persons participating in Project DETECT, including MSM seeking HIV/STD testing at a public STD clinic in Seattle, WA.

Methods:

We developed a computer self-assisted survey that included questions about participant demographics, sexual history, and risk behaviors, including group sex, as a part of Project DETECT, a CDC-funded study evaluating point-of-care HIV tests. Aggregate and partner-specific questions asked about participation in all GSEs, threesomes, and four-or-more-somes including questions about number and HIV status of sex partners and condom use during the events. To evaluate the performance of these questions, we assessed the discrepancies in reporting between the aggregate and partner-specific questions, quantified question refusal rates, and calculated the additional time required to answer the GSE questions. Information about network density (number of partnerships of overlapping duration in the past three months) was estimated and compared for MSM who did and did not report GSEs.

Results:

Among 841 visits by 690 MSM who were asked any survey question about group sex, participation in a GSE of any type in the past three months was reported at 293 visits (34.8%). We found that 9.0% of MSM in the sample reported ≥1 four-or-more-some in the partner-specific question set that were not reported in the aggregate. The proportion of refusals on any given aggregate GSE-related question ranged from 0% to 10.6% (median=2.6%). The proportion of refusals in the partner-specific questions ranged from 0% to 21.7% (median=3.0%), with questions about four-or-more-somes having the highest proportions of refusals. Completing the aggregate group sex questions added 1-2 minutes and the partner-specific questions added an additional 2-4 minutes per partner to the total survey length. As expected, the partner-specific GSE questions documented higher density of sexual networks that was not captured by asking about total partner counts and overlap of specific partnerships.

Conclusions:

We found the Project DETECT behavioral survey was able to obtain nuanced information about overall and partner-specific GSEs. The questions seemed to be acceptable, skip patterns and consistency checks were effective, and survey fatigue was minimal. More research is needed on GSEs, and our survey represents a promising data collection tool to help fill gaps in knowledge about the subject.