“Knowing your status and knowing your partner’s status is really where it starts”: A Qualitative Exploration of the Process by which a Sexual Partner’s HIV-Status Can Influence Sexual Decision Making

Abstract

Gay and bisexual men are at disproportionate risk for HIV infection. While prevention efforts often emphasize consistent condom use, there is growing evidence that men are using seroadaptive safer sex strategies, such as sero-sorting and sero-positioning. This qualitative analysis of 204 HIV-negative and HIV-positive gay and bisexual men explores the ways that a sexual partners’ HIV-status can influence safer sex strategies and sexual decisions. The majority of the respondents reported that they were influenced by their partner’s HIV-status. Those respondents who reported no influence discussed adhering to safer sex rules that were not dependent on partner status and a lack of concern about HIV. Conversely, respondents who reported influence identified three primary areas of influence: psychological impacts, partner preference and selection, and specific behavioral intentions and strategies. A conceptual model explicating a potential process by which respondents use partner serostatus information in shaping sexual decisions is presented.

The number of individuals infected with HIV in the United States continues to rise with approximately 50,000 people becoming newly infected every year (Centers for Disease Control and Prevention [CDC], 2015b). New HIV infections remain concentrated among men who have sex with men (MSM), including gay and bisexual men, who, despite constituting only 2–4% of the population (Hamel et al., 2014; Purcell et al., 2012), accounted for approximately 63% of all new HIV infections in 2010 (CDC, 2015b). Furthermore, according to the Centers for Disease Control and Prevention, MSM of color and young MSM bear the greatest burden of new infections; for example, while HIV incidence has remained relatively stable, there was a 22% increase in the rate of new infections among young MSM (ages 13–24) between 2008 and 2010 (CDC, 2015b). This persistence in new HIV infections remains despite the implementation of a comprehensive national strategy for addressing the epidemic (Office of National AIDS Policy, 2015) and a heightened focus on HIV-prevention and education in the United States (The Kaiser Family Foundation, 2014). Understanding the conditions and situations that lead to HIV vulnerability is critical to reducing HIV infection rates.

The prevention environment has changed significantly over the last thirty years. New HIV treatments that suppress viral loads and transmissibility have been developed (Attia, Egger, Müller, Zwahlen, & Low, 2009; Cohen, Muessig, Smith, Powers, & Kashuba, 2012), new preventative regimens such as pre-exposure prophylaxis (PrEP) have become more readily available (Golub, Gamarel, Rendina, Surace, & Lelutiu-Weinberger, 2013; Grant et al., 2010) and trends towards risk reduction strategies that do not require universal condom use have increased (Eaton, Kalichman, O’Connell, & Karchner, 2009; Golden, Stekler, Hughes, & Wood, 2008; Grace et al., 2014; Snowden, Wei, McFarland, & Raymond, 2014; Wei et al., 2011). Attending to these changes in gay men’s preventative strategies is becoming an important focus for research and program development.

A growing body of research suggests that many gay and bisexual men are adopting more nuanced approaches to prevention that extend beyond skills building and universal condom use. Many of these men are embracing seroadaptive sexual practices where sexual risk decisions are based on the HIV status of a sexual partner (Eaton et al., 2009; Golden et al., 2008; Grace et al., 2014; Snowden et al., 2014; Wei et al., 2011). Typical seroadaptive strategies include serosorting (i.e., selecting sero-concordant partners), sero-positioning (i.e., choosing less risky sexual positions with sero-discordant partners), and condom sero-sorting (i.e., choosing not to use condoms for anal sex with sero-concordant partners) (Eaton et al., 2009; Golden et al., 2008; Grace et al., 2014; McFarland et al., 2012; Vallabhaneni et al., 2012). In all of these strategies, individuals take into account the HIV-status of their sexual partner, which has some direct or indirect influence on their sexual negotiation and/or sexual behavior.

The success of these alternative strategies depends on a number of key factors, namely, that individuals are aware of their own HIV status and that they are explicitly discussing and sharing HIV-status information with sexual partners. However, research suggests that the effectiveness of these seroadaptive strategies is potentially undermined due to insufficient HIV-testing (Hamel et al., 2014; CDC, 2015a), undiagnosed HIV-infections among gay men (CDC, 2015a; Wilson et al., 2010) and decreased communication about HIV with sexual partners, often due to HIV-related stigma (Bird & Voisin, 2013; Hamel et al., 2014; Wei et al., 2011; Wilson et al., 2010). This decreased communication about HIV can lead individuals to make assumptions about partners’ HIV statuses, a phenomenon referred to as sero-guessing, that can be biased towards believing a partner is sero-concordant (Flowers, Duncan, & Frankis, 2000; Grace et al., 2014; Parsons et al., 2006). Given these challenges, there have been mixed conclusions regarding the effectiveness of seroadaptive strategies (Golden et al., 2008; Philip, Yu, Donnell, Vittinghoff, & Buchbinder, 2010; Vallabhaneni et al., 2012; Wilson et al., 2010).

More information is needed about the beliefs gay and bisexual men hold about their partners’ HIV statuses and the ways these beliefs can influence sexual negotiation and behavior. Furthermore, it is important that this range of influence be explored not only in sero-concordant (i.e., both individuals’ HIV-statuses are the same) or sero-discordant (i.e., the HIV-statuses of the individuals are different) sexual partnerships but also in sero-nonconcordant partnerships, where a partner’s HIV-status is unknown and there can be a broader interpretation about potential risk (e.g., increased precautions when a partner is assumed to be sero-discordant; increased sexual risk when a partner is assumed to be sero-concordant). Towards this end, this paper seeks to fill in some of these gaps in the literature. Using a phenomenological approach, we explore respondents’ beliefs about how a partner’s HIV-status can influence their sexual decisions across sero-concordant, sero-discordant, and sero-nonconcordant sexual partnerships and analyze these beliefs as they relate to their sexual behavior with their last partner.

Methods

The data analyzed in this study come from [information redacted to remove author identifying information]. Data collection was completed in 2011. The main study from which the data for the present analysis were obtained was designed to examine whether Conversational Interviewing, an alternative method for conducting survey interviews where interviewers can assist respondents in interpreting questions uniformly and as intended, enhanced respondents’ comprehension of survey questions when compared to Standardized Interviewing techniques that emphasize little assistance other than repeating the question as written and/or providing definitions. Questions regarding the influence of partner HIV-status on sexual behavior were included as a part of the main study. While the exploration of influence was not the primary objective of the main study, the available data presented an opportunity to take an in-depth look at how HIV status influences risk behavior.

Respondents were recruited actively from a local gay bathhouse and passively through notices posted at area bathhouses or through online advertisements. Men age 18 or older who reported engaging in anal intercourse with another man in the prior 3 months qualified for the study. Eligible men were randomly assigned to either the Conversational (N=100) or the Standardized (N=104) conditions. The same questionnaire was administered in both conditions via face-to-face computer assisted personal interview technology (CAPI). Interviewers were blind to the experiment and conducted all their interviews in only one condition (either Conversational or Standardized). Immediately after the survey interview was completed, respondents participated in a second interview to debrief their interpretations of the interview questions. This debriefing interview was conducted with a separate group of specially trained interviewers. All survey and debriefing interviews were video and/or audio recorded, with audio recordings available to the interviewers during the debriefing interviews to help respondents remember their responses to the survey questions. Each study visit took approximately two hours. All study procedures were approved by the Institutional Review Board (IRB) at the University of California San Francisco. Informed consent was obtained from all individual participants included in the study.

This paper is based on an analysis of a specific cluster of survey questions relating to whether the respondent knew the HIV status of their last sexual partner, the specific sexual activities in which they engaged with their last sexual partner, and whether the HIV-status of their last sexual partner influenced the type of sex in which they engaged. The specific HIV-status related questions included: What was the HIV status of your last sex partner? Was he HIV-positive, HIV-negative, or didn’t you know his HIV status? Did his HIV status influence the kind of sex you had with him or did not knowing his HIV-status influence the kind of sex you had with him? Aware of the potential disparity in responses based on randomization to different arms, this cluster of questions was revisited in the debriefing interview and respondents in both conditions were asked: What did you think was meant by “influenced the kind of sex you had” in this question?

Data Analysis

We utilized a phenomenological qualitative approach in this paper to explore the respondents’ beliefs about influence and the processes by which they believed it affected sexual decision-making. We drew data from survey items that provided basic demographic information about the respondents and their last sexual partners, the audio-video recordings of the survey interviews, and the audio-recorded debriefing interviews. These debriefing interviews allowed all respondents to speak more broadly and in more detail about how partner HIV-status influenced the kind of sex they had. Overall, the combination of these different data sources provided a unique opportunity to learn more about the impact partner HIV status had on the respondents’ sexual decisions.

The digital recordings were transcribed verbatim and we used the software Transana to assist with the management of our data and to identify analytical saturation of themes across different respondent categories (Woods, 2014). Two coders conducted the initial analysis, comparing and contrasting their independent analytical results; the entire research team was involved in analytical discussions about the data. Data analysis was informed by Strauss and Corbin’s framework for analysis and included three primary processes: open-coding, axial coding, and selective coding (Creswell, 2012; Strauss & Corbin, 1990). During the open coding phase, broad codes related to influence and partner HIV-status were identified and the data and codes were used to categorize whether or not participants reported being influenced by their sexual partners’ HIV-statuses. Our analysis and cross comparisons yielded three primary categories (those who were not influenced, those who were influenced, and those who reported mixed responses to influence). Within those three primary categories, seventeen codes related to ‘no influence’ and thirty-eight codes related to ‘influence’ were identified. These codes were then categorized into five distinct thematic sets (two for ‘no influence’ and three for ‘influence’) that reflected similar and/or complimentary experiences of and beliefs about the role of partner status in influencing respondents. The relationships and interconnections between these primary qualitative themes and unique demographic categories (i.e., respondent HIV status and sero-concordance) were explored during the axial coding phase and further developed and explained as a part of the selective coding stage (Creswell, 2012; Strauss & Corbin, 1990). The representative quotes used below and in the tables are identified by unique pseudonyms to protect the identity of the respondents.

Results

Demographics

This analysis is based on 204 respondents. Respondents’ ages ranged between 19 and 72 years of age with a mean age of 44.1 years. The sample was diverse with regards to race/ethnicity, with 59.4% (n=123) self-identifying as White/Caucasian, 19.8% (n=38) self-identifying as Black/African American, 5.2% (n=13) self-identifying as Asian, and 6.8% (n=13) self-identifying as some other race/ethnicity or refusing to answer the question. 20.3% (n=39) of the respondents reported being Latino/Hispanic. 61.8% (n=126) of the respondents reported being HIV-negative and 37.2% (n=76) reported being HIV-positive. Only 1% (n=2) of the respondents indicated that they did not know or refused to disclose their HIV-status. In terms of the HIV status of their last sexual partner, 48% (n=98) of the respondents reported that their last partner was believed to be sero-concordant, 16.2% (n=33) reported the last partner was sero-discordant, and 35.8% (n=73) reported that the last partner was sero-nonconcordant (i.e., they did not know the partner’s HIV serostatus). Only 27.0% (n=55) considered this partner to be their main sexual partner while 47.5% (n=97) said they were a casual sex partner and 25.5% (n=52) characterized them with a designation other than main or casual.

More than half of the respondents (51%) initially reported during the survey interview that the HIV-status of their last partner influenced the type of sex in which they engaged. Respondents discussed the concept of influence in more depth during the debriefing interviews and often offered a broader assessment of this influence on their sexual encounters, in general. Based on these qualitative discussions, the percentage of respondents reporting some level of influence increased to 59.8% (n=131) and the percentage of respondents reporting no influence decreased to 40.2% (n=88). Fifteen respondents discussed reasons why they were not influenced but also ways in which they believed they could potentially be influenced by their sexual partners’ HIV status. These respondents were represented in both categories bringing the total number of cases in the qualitative analysis to 219. Table I describes the demographic characteristics of the sample as well as a breakdown of respondents by whether their behavior was “influenced” by partners HIV status.

Table I.

Demographic Information

				N	%
Age
	19–24 years			17	8.3%
	25–34 years			34	16.7%
	35–44 years			46	22.5%
	45–54 years			65	31.9%
	55+ years			42	20.6%
Race /Ethnicity
	White/Caucasian			123	59.4%
	Black/African American			38	19.8%
	Latino/Hispanic			39	20.3%
	Asian			13	5.2%
	Other/Mixed/Refused			13	6.8%
HIV Status of respondent
	HIV-Negative			126	61.8%
	HIV-Positive			76	37.2%
	HIV-Unknown			2	1%
Dyadic serostatus of last sexual partner
	Sero-Concordant			98	48%
	Sero-Discordant			33	16.2%
	Sero-Nonconcordant			73	35.8%
Type of last sexual partner
	Main Sexual Partner			55	27%
	Casual Sexual Partner			97	47.5%
	Other			52	25.5%
Does a partner’s HIV-status ‘influence’ the respondent’s sexual decisions
	Reported Influence (including mixed influence)			131	59.8%
	Reported No Influence (including mixed influence)			88	40.2%
	No Influence
		HIV-Status of respondents
		HIV-Negative		48	54.5%
		HIV-Positive		38	43.2%
		HIV-Unknown		2	2.3%
		Dyadic serostatus of last sexual partner
		Sero-Concordant		31	35.2%
		Sero-Discordant		16	18.2%
		Sero-Nonconcordant		41	46.6%
	Influence
		HIV-Status of respondents
		HIV-Negative		89	67.9%
		HIV-Positive		42	32.1%
		HIV-Unknown		0	0%
		Dyadic Serostatus of last sexual partner
		Sero-Concordant		75	57.3%
		Sero-Discordant		19	14.5%
		Sero-Nonconcordant		37	28.2%

Qualitative Themes

No Influence of partner HIV-status on sexual decisions

The reasons given by respondents who reported that they were not influenced by their partners’ HIV status were categorized into two basic thematic areas: (1) Strict adherence to an established set of rules and (2) Limited concern about HIV or HIV transmission

Strict adherence to an established set of rules

Many of the respondents reported that they were not influenced by their partner’s HIV status because they did not deviate from their established set of rules and these rules were not dependent on the HIV status of their partner. These safer sex rules varied across respondents depending on their beliefs about what was sexually safe but each set of rules was consistently maintained with all sexual partners. For example, Julian reported that he treats all sexual partners as though they are HIV-positive, stating:

…I treat everyone, regardless of their status, what they are saying or not … that they are (HIV+), so I am very conscious of what I am going to be doing. I always use a condom because that is the best way to protect yourself…I don’t know if they are lying to me. I don’t know if they got it since the last time they got tested…you don’t certainly know. You have to treat everyone like there might be a chance you could get something…(HIV-negative; 41yrs; Latino/Hispanic; Concordant Partner)

Other examples of safer sex rules included always using condoms for anal sex, never engaging in receptive and/or insertive anal sex (depending on the HIV status of the respondent), and never ejaculating inside a sexual partner’s rectum.

Limited concern about HIV or HIV transmission

Several respondents also reported that their sexual partners’ HIV status did not influence their sexual behavior because they were not concerned about HIV. For several of these respondents, their lack of concern was related to not wanting to think about or discuss HIV with their sexual partners with Daniel stating:

Most of the time the people that I’ve met never talk about their HIV status because I feel that it is a responsibility that I’m taking on myself by not asking it because I don’t want the experience to be clouded by this whole thing about status…I made the decision and I live with…it is a risk that I take every time I have sexual activity. (HIV-negative; 53yrs; Other Race/Ethnicity; Nonconcordant partner)

A few respondents reported that they were generally not worried about HIV infection because they believed that HIV was unimportant, such as Al who stated:

personally it really doesn’t matter to me - you’re negative or positive - to me. People will eventually die it doesn’t matter if you don’t get this you get high blood pressure or pneumonia or whatever you know. It is not a perfect world, so what. (HIV-positive; 45yrs; Asian; Nonconcordant partner)

Finally, another group reported that they were not worried about HIV infection because they practiced safer sex and/or had an undetectable viral load, which led them to believe they were less able to transmit the virus. Respondents in this group were different from those in the ‘strict rules’ category because of their emphasis on the decreased concern around HIV, in general, and, specifically, HIV-transmission.

Influence of partner HIV-status on sexual decisions

The majority of the respondents, however, reported that their sexual decisions were influenced by the HIV-status of their sexual partners. The interpretation of “influence” varied across this diverse sample of respondents, encompassing a range of beliefs about HIV and sexual risk. These definitions were dependent on the HIV-status of the respondent, with most themes uniquely clustered by respondents’ HIV status. There were three primary thematic areas regarding how respondents were (or could be) influenced by a partner’s HIV status. Respondents described influence as related to 1) psychological reactions; 2) partner preference and selection; and 3) behavioral intentions and strategies for making sexual risk decisions (See Table II).

Table II.

Themes related to reported influence of sexual partner HIV status

Psychological Reactions
Jacob (HIV-Negative; 46yrs; Black; Concordant)	“It made me feel more at ease knowing that this person was negative. It enhanced the sex…”
Connor (HIV-Negative; 65yrs; White; Discordant)	“I didn’t particularly enjoy, I found him very attractive, but I think the fact that he was positive caused my erection to fade somewhat…There was something about him, he wasn’t very relaxed.”
Tristan (HIV-Positive; 38yrs; Black; Concordant)	“Well, being HIV-positive and dealing with someone else who is HIV- positive I think it makes it a little more comfortable as far as the honesty and letting your guard down. Moreso than with someone who is negative or not knowing their status. There is still a guard kind of thing where you feel like you are lying to that person. There is still not that comfortability thing.”
Partner Preference and Selection
Ethan (HIV-Negative; 30yrs; White; Concordant)	“…if he was HIV-positive we wouldn’t have a relationship unfortunately. I think that people that are HIV-positive have a hard time finding people that are HIV-positive anyway. I guess it’s not worth the risk…”
Nick (HIV-Negative; 59yrs; White; Concordant)	“To me there are two categories. If someone looks like they are HIV-positive and someone who may be positive but looks healthy. I don’t have sex if someone looks like they are HIV-positive. Meaning I have my own method of determining what that look is. So I would just avoid having sex with them…”
Ned (HIV-Positive; 53yrs; White; Concordant )	“there’s a part of me that almost prefers being with someone who is positive…I’ve always got the concern that I might infect somebody who’s HIV-negative and that doesn’t sit well with me at all.”
Behavioral Intentions and Strategies for Making Sexual Risk Decisions
Owen (HIV-Negative; 23yrs; White; Concordant)	“because I knew he was HIV-negative, that would influence my decision to have sex without a condom that one time that we did because there was no condom around. So yeah, it did because I have been seeing him for a while and his status has been negative…and I know that he uses condoms with other people. So both his status and his behavior pattern both influenced my decision.”
Austin (HIV-Negative; 42yrs; Black; Non-Concordant)	“some things I didn’t do for not knowing for sure…We hadn’t discussed it and he hadn’t volunteered that information to me. It didn’t bother me. I just didn’t do certain things because I didn’t know his status”
Justin (HIV-Positive; 59yrs; White; Discordant)	“In that particular case I knew he was negative and he knew I was positive from the get go. So he set clear boundaries. I was fine him fucking me without a condom, but he wasn’t fine with me fucking him without a condom which was fine with me. Those were mutually agreed upon boundaries”

Psychological Reactions

Both HIV-negative and HIV-positive respondents reported that the HIV-status of their sexual partners had, or could have, a psychological impact on them. These psychological reactions could be connected to potential, current, or past sexual partners. Several of the HIV-negative respondents explained that having HIV-positive partners made them feel uncomfortable and one participant reported that having an HIV-negative partner helped to decrease his fears of infection. Moreover, several other HIV-negative respondents reported that having a same status sexual partner enhanced the sex and/or increased their comfort with the sexual acts, making them more relaxed and confident and helping to put them more at ease. Logan, for example, reported that:

It is very, very influential…I do safe sex period, but…I just feel a lot more uninhibited with the casual one because I know his status (which is HIV-negative) and I trust him and he trusts me. I don’t have to worry about it. My (previous) partner, even though I have a lot of trust with him it (his being HIV-positive) just came to weigh on my mind in an unconscious way - I wish it didn’t. (HIV-Negative; 49yrs; White; Concordant Partner)

Among the HIV-positive respondents, several explained that having a same status sexual partner made them more comfortable, relaxed, and less concerned about engaging in sex (regardless of risk). Conversely, another few respondents reported that having HIV-negative sexual partners decreased their comfort and increased their concern and worry about engaging in sex, in general, and, for one respondent, condomless anal sex specifically. Finally, a number of respondents reported that the HIV-statuses of their sexual partners was important to them specifically because they had a fear of transmitting the virus to others and a desire to keep their HIV-negative partners negative. However, there were some inconsistencies with regards to condom use where respondents used a sero-adaptive strategy (i.e., withdrawal) to decrease the risk of infection and thereby reduced their psychological concerns regarding infecting others.

Partner Preference and Selection

The second theme was centered on partner preference and selection. There were two primary reasons that HIV-status influenced the selection of partners: the HIV-negative respondents were chiefly concerned with remaining HIV-negative and the HIV-positive respondents were largely concerned with not infecting others with HIV, although several of the HIV-positive respondents reported that they ultimately left risk decisions up to their HIV-negative partners.

About one-fifth of the respondents explicitly identified partner selection and preference as being influenced by HIV-status in the qualitative discussions. HIV-negative respondents represented in this theme generally reported that they would not have sex, or would be reluctant to have sex, with someone known to be HIV-positive. Often, this was explicitly related to concerns about infection, for example, Mario stated:

…I wouldn’t be having sex with someone I know that is HIV-positive. Not even with a condom - it is too risky. I’m very strict on that. I want to stay healthy. (HIV-Negative; 52yrs; Latino/Hispanic; Concordant Partner)

At other times, preference and selection was related to stigmatizing beliefs about people who were infected with HIV, with Bruce stating:

Negative and clean, I go with; positive, I will not have sex with you because you do not take care of yourself (HIV-negative; 39yrs; Asian; Nonconcordant Partner)

Several of the HIV-positive respondents also reported that they generally looked for, and had, sex with other HIV-positive men, with Luis explaining:

Now I only have sex with other positive men …when you have sex with other positive men, they have an understanding and no worries. If there is a positive man and a negative man…it’s a very different way of thinking like being straight and being gay… (HIV-Positive; 40yrs; Latino/Hispanic; Concordant Partner)

None of the HIV-negative or HIV-positive men who discussed this theme reported having a sero-discordant sexual partner, which supported their stated behavioral intentions. However, several respondents reported that they did not know the HIV-status of their last sexual partner. This engagement with partners whose HIV-status was unknown while intending to engage in sero-concordant partnerships suggests that there may be challenges between the desire to serosort, the actual practice of serosorting, and the degree of sexual communication needed to negotiate this type of partner selection.

Behavioral Intentions and Strategies for Making Sexual Risk Decisions

Finally, the influence of partner HIV status information on specific sexual risk intentions and strategies represented the most prevalent theme discussed by respondents. The majority of respondents who reported influence identified ways that their sexual behavior was, or could be, directly influenced by their sexual partner’s HIV status. Most of the respondents in this category described specific, status-based influences on their willingness to engage in particular sexual behaviors. These influences were often dependent on the respondents’ HIV-status, which is used as a sub-classification below.

HIV-negative respondents

About a third of the HIV-negative respondents discussed ways that their partners’ HIV-status influenced their sexual risk decisions. Many of these respondents discussed influence in terms that were more general rather than referencing specific sexual behavior. In most of these cases, they stated broadly that the sex with HIV-positive partners would be more constrained and/or that they would be more careful to avoid sexual risk with someone who was HIV-positive. Overall, though, the participants in this group were clear that they would engage in sexual activity differently with a partner known to be HIV-positive, with Jackson stating:

…knowing your status and knowing your partner’s status is really where it starts. If one of you happens to be compromised, of course you are going to have a different approach to the way you have sex…(HIV-Negative; 36yrs; Black; Concordant Partner)

There were also a number of respondents who reported that having an HIV-positive partner would specifically influence sexual risk decisions around anal sex. In these cases, the respondents reported that they would not engage in any anal sex at all or that they would be less likely to engage in receptive anal sex with a partner who was HIV-positive. For example, Kevin explained:

…Now I might have more concern if we were to have anal sex…The type of sex I may be having brings up the HIV question. So I think, whether I’m wrong or right, oral sex is a low risk activity…Whereas having anal sex I might be more concerned… (HIV-Negative, 40yrs; White, Nonconcordant Partner)

Many of the HIV-negative respondents also explained that having an HIV-negative partner could influence their specific sexual risk decisions. For example, several of these respondents reported that having an HIV-negative sexual partner, or a partner they assumed to be HIV-negative, could potentially open up options for engaging in sexually riskier activity and made them more comfortable in choosing to take sexual risks as Aaron explained:

Him informing me of his HIV status makes me more receptive to the idea of unprotected sex with him where I am bottoming. If he were to say flat out I am HIV-positive and I would like to top you, I would say ‘not without a condom’ (HIV-Negative, 30yrs; White; Concordant partner)

Having a sexual partner whose HIV-status was unknown resulted in mixed influences for the HIV-negative respondents. For some respondents, having a partner whose status was unknown led them to be more cautious and to limit certain activities that could lead to HIV infection. For others, having a partner whose HIV status was unknown allowed them to feel more comfortable engaging in sex, in general, and taking more sexual risks in some encounters. In one of these cases, Robert reported that he occasionally assumed his partners were also HIV-negative as a means of decreasing his concern about sexual risk, explaining:

…if, for example, and this has happened to me a couple of times, really wanting to have this guy inside of me and he’s rubbing up against me and he’s already lubricating naturally and a little bit of his penis is going inside of me and I am ‘oh fuck, you gotta stop’ and I need to think. But if you know at least if you have the solution in your mind that the person is negative, you say to yourself “whew, that was a risky thing” but it was not as risky. (HIV-Negative; 43yrs; Latino/Hispanic; Nonconcordant Partner)

A number of respondents also indicated that this comfort in engaging in sex and/or risk would have decreased had they known their partner was HIV-positive.

Although most of the men indicated an intention to use condoms with all sexual partners, several respondents also discussed the ways in which a partners’ HIV status could impact their concerns regarding accidental and/or unplanned sexual risk or if they were to ‘slip-up’ from their planned sexual behavior. For example, two respondents explained that they would be less concerned if they engaged in unplanned sexual risk with someone who was believed to be HIV-negative. Likewise, another two reported that not knowing a partner’s HIV status made them more willing, or at least less concerned about engaging in unplanned sexual risk, with Jeremiah stating, “…they don’t ask don’t tell and presume that their partner is negative they may let the person slip up and slide their dick in without a condom real quick” (HIV-negative, 34yrs, Black, Nonconcordant Partner). Conversely, Jeremiah also reported that he would be more careful not to “slip up” or engage in “unplanned” sexual risk if he knew his sexual partner was HIV-positive. Therefore, while these cases were not related to a general or specific plan regarding risk, they were related to the respondent’s concerns about deviations from their planned behavior and the level of this concern was influenced by the HIV-statuses of their sexual partners.

HIV-positive respondents

Many HIV-positive respondents also discussed ways that their partners’ HIV-status influenced, or could influence, their sexual risk decisions. Most of the discussions revolved around their strategies for engaging sero-discordant or nonconcordant partners. However, a number of the respondents did report that having a sexual partner who was also HIV-positive gave them the option of not using condoms for anal sex with Allen explaining:

Yes, it did influence because I hooked up with him primarily because I know I’m gonna have better sex without that. I think there is better sex without protection anyway. That is why I prefer to hook up with positive men. (HIV-Positive; 30yrs; Asian; Concordant Partner)

Nearly half of the HIV-positive respondents reported that they specifically worked to limit sexual risk with HIV-negative partners, with a number of them explicitly stating that they did not want to infect anyone with HIV. While several of the respondents reported that they would leave the decision of using condoms to their HIV-negative sexual partner, most of the respondents discussed strategies that required them to take at least partial responsibility for reducing risk. For example, one respondent reported that he would not allow an HIV-negative sexual partner to be receptive during oral sex. Additionally, two other respondents reported that they would not engage in any anal sex with a partner who was HIV-negative or whose status they did not know, with David stating, “If I am with someone and I don’t know their HIV status then anal sex is off the table” (HIV-Positive; 52yrs; Black; Concordant Partner).

Several of the respondents reported generally that they would engage in sexually safer activities with a partner who was HIV-negative and nearly half of the respondents reported that they would definitely use condoms for anal sex if their sexual partner were HIV-negative. For example, Parker, reported:

I wouldn’t engage in any kind of anal sex with him without a rubber. Which means that we don’t engage in anal sex…there is no reason to expose him to this virus since this is not our main sexual outlet to begin with. (HIV-Positive; 53yrs; Native American; Discordant Partner)

A few respondents reported using non-condom based methods for reducing anal sexual risk with HIV-negative partners, such as not ejaculating into a partner during receptive anal sex or consenting to not using a condom with an HIV-negative, insertive partner. For example, John reported that he used a withdrawal method (no rectal ejaculation) to help decrease the potential risk, stating:

We had great sex the past two days but he let me fuck him bareback and I said ‘[NAME], do you know what you are doing?’ But I promised him that I would not come to orgasm in his ass. (HIV-Positive; 45yrs; White; Nonconcordant Partner)

Noah also reported using withdrawal methods because his viral load was undetectable and he believed he was less infectious, explaining that:

We made a decision ahead of time not to use a condom even though [NAME] is HIV-negative…I was turned on by being with [NAME] but I did not want to come inside of him and thus cause damage to him in that way. It was consensual…I would hate for that to happen ever…and I’m on medication and stuff where my viral load is supposed to be, I mean it’s an educated guess. (HIV-Positive; 46yrs; White; Discordant Partner)

Finally, a couple of the respondents also explained that having a partner whose status was unknown led them to be more cautious and to limit certain activities that could lead to HIV infection. However, one respondent did report that having a partner whose status was unknown allowed him to not have to think about condoms or safer sex.

Discussion

The role that HIV plays in sexual risk decisions and negotiation is multi-faceted and complex. A number of respondents in this study reported that their sexual risk decisions were not impacted by their sexual partners’ HIV status. However, the majority of the respondents described a variety of ways that their sexual desires, intentions, and behaviors were influenced by their sexual partners’ HIV status. For these respondents it was clear that sexual desire, pleasure, and fulfillment were occasionally at odds with consistent condom use as a primary risk reduction strategy. Given this, many of these respondents’ developed safer sex strategies that sought to strike a balance between sexual pleasure and decreased risk of HIV transmission. Towards that end, the HIV-status of their potential partners played a role in influencing their feelings about risk before, during and after their sexual encounters, the types of partners they preferred, and the kind of sex in which they engaged.

Those respondents who reported that they were not influenced by their sexual partners’ HIV status generally explained that they either strictly followed established safer sex strategies that were not dependent on the sexual situation or that they were not worried about becoming infected with HIV or infecting others. A number of these participants also reported that they were uninterested in engaging in discussions about HIV with their sexual partners, which led them to develop strategies that were not dependent on disclosure. In other words, they did not report conflictual or intense psychological or emotional reactions to HIV and did not see HIV as a vital factor in partner selection or their sexual decision-making processes.

Conversely, our analysis of those respondents who reported that they were influenced by their sexual partners’ HIV status revealed three primary and interrelated themes: the psychological reactions and the impact on risk assessments and sexual pleasure; partner preference and selection; and behavioral intentions, strategies for making sexual risk decisions, and deviations from those strategies. The exploration of these three themes helped to better explicate the ways in which knowing, and, in some cases not knowing, a partner’s serostatus could impact sexual risk decisions. Furthermore, these three themes helped us develop a conceptual model outlining one process by which a partners’ serostatus could influence behavioral intentions, outcomes, and deviations (See figure 1).

Figure 1.

Conceptual Model of Influence

The first ‘influence’ theme related to psychological reactions to a partners’ serostatus. The patterns of these psychological reactions generally included increased sexual comfort, relaxation, confidence, intimacy, and enjoyment within perceived to be sero-concordant relationships and increased anxiety, worry, and concern within sero-discordant partnerships. Furthermore, these psychological reactions appeared to influence decisions across the range of sexual negotiation. For example, for some respondents, their concerns about HIV influenced what they wanted in potential partners and had a direct impact on their partner choices (qualitative theme II in the model). For others, the HIV-status of actual sexual partners affected the emotional pleasure of the sexual experience and the behavioral intentions and safer sex strategies used in the sexual encounter (qualitative theme III in the model). Finally, these psychological reactions could influence how the respondents justified or cognitively processed deviations from their strategies after the sexual encounter (qualitative theme III in the model). Therefore, these psychological reactions appeared to represent a critical, underlying factor in shaping partner choice, behavioral intentions and strategies, and the engagement in sexual risk.

The second ‘influence’ theme revolved around partner preference and selection. A number of respondents reported that they desired and/or sought out sero-concordant sexual partnerships. The primary, and explicit, reasons for preferring sero-concordant partnerships were related to concerns about HIV-transmission: HIV-negative men wanted to stay HIV-negative and HIV-positive men wanted to keep from infecting others. There was also a more implicit, secondary reason for preferring sero-concordant relationships related to HIV-based stigma and the ways in which HIV was seen as a dividing characteristic within the gay community, which has been identified in other research (Bird & Voisin, 2013). This theme was evident in the more disparaging language used by some of the HIV-negative men in describing people infected with HIV. For example, their descriptions of HIV-positive men as “compromised” and as not ‘tak(ing) care of (them)selves” and HIV-negative men as “clean” and “healthy” betrayed potentially negative moral implications, assumptions, and stereotypes regarding people who were HIV-infected. On the other side, several of the HIV-positive respondents described difficulty relating to sero-discordant partners. Furthermore, these underlying stereotypes could have an important impact on the assumptions made about partner HIV-status and, thereby, on their potential sexual risk decisions. While only some respondents explicitly stated that they serosorted or that they preferred sero-concordant partners, nearly 60% of the overall sample of respondents who reported influence indicated that their last partner was sero-concordant. Less than 15% reported that their last sexual partner was sero-discordant. Therefore, while not all respondents explicitly spoke about serosorting or preferring sero-concordant partners, the phenomenon was evident in the data and represented a primary influence in the conceptual model.

It is important to note that these data were collected prior to the Food and Drug Administration’s approval of Truvada™ as HIV pre-exposure prophylaxis (PrEP). We anticipate that biomedical prevention interventions (e.g., PrEP and treatment as prevention) will likely influence sexual decision-making. For instance, qualitative research with young gay men in Oakland indicated that PrEP use may diminish fears of contracting HIV and, in some cases, lead to a de-emphasis of HIV serostatus as an influencing factor in partner selection (Myers & Koester, 2016). Therefore, further uptake of PrEP use within this population may lead to greater acceptance of people living with HIV as sexual and romantic partners and reduce HIV-related stigma. This could transform the role a partner’s HIV status plays in influencing sexual behavior over time and should be studied in future research.

The third ‘influence’ theme explored the impact of partners’ serostatus on specific behavioral intentions and safer sex strategies. This theme related to both the respondents’ psychological reactions to a partners’ HIV status (e.g., increased/decreased comfort, concern, anxiety) and to partner preference and selection, where partner HIV-status (as opposed to sexual activity) was explicitly connected to HIV-transmission risk. For the HIV-negative men, there was a general intention to use condoms with sexual partners, especially when anal sex was involved, but that intention was stronger with known to be sero-discordant partners and weaker with partners who were assumed, believed to be, or known to also be HIV-negative. For many of these men, this translated into being more cognizant of and careful to avoid sexual risk activities, particularly when engaging in receptive anal sex with HIV-positive partners. Conversely, there appeared to be more flexibility in safer sex strategies with sero-concordant partners that translated to an increased willingness to engage in sexual risk and decreased concern about “slip-ups” or “accidental” or “unanticipated” risks with known, or believed to be HIV-negative partners. Therefore, they had general intentions to engage in decreased sexual risk but were willing to deviate from their intentions and strategies based on the HIV-status of their sexual partner.

The HIV-positive respondents were more focused on sero-concordant partnerships where they discussed increased sexual options and an explicit intention to not use condoms with other HIV-positive men. However, for most of these respondents, when they were with HIV-negative partners, there was an explicit desire not to transmit HIV and a specific intention to actively avoid risk by using condoms or alternative sexual safety methods such as not ejaculating into the rectum of HIV-negative partners. In this, their intentions and safer sex strategies were explicitly based on the HIV-status of their partner and they were less likely to deviate from those intentions and strategies.

Interestingly, nearly 30% of the respondents who reported that the HIV-status of their sexual partner influenced the type of sex in which they engaged reported that they did not know the HIV-status of their last partner. While many of these themes revolved around what respondents knew about their partners’ serostatus, not knowing a partner’s serostatus presented a more complicated level of influence on partner selection, behavioral intentions, and risk-reduction strategies. In general, sero-nonconcordant relationships led to intentions for increased safety for HIV-positive respondents. However, there were varied reactions for the HIV-negative respondents. For some men, not knowing a partner’s HIV status influenced them to decrease potential risk. For others, not knowing a sexual partner was sero-discordant or believing them to be sero-concordant was sometimes used to justify deviations from behavioral intentions or, at the very least, decrease concern if a deviation occurred. In this, it appeared that they were primarily interested in not having sex, or engaging in sexual risk, with men known to be HIV-positive but were willing to and comfortable with engaging in sex and risk with someone whose HIV-status was unknown. Therefore, sero-nonconcordance appeared to represent a third partner status that had its own unique influence on the respondents’ sexual risk intentions and strategies.

These findings and the conceptual model present important implications for research and practice. Both HIV-negative and HIV-positive respondents described multifaceted sets of beliefs about how HIV influenced their sexual risk decisions. They also identified concerns about underlying psychological and stigmatizing factors and issues related to sexual pleasure, intimacy, and fulfillment that underscore the complexity of influence as a phenomenon. While each of the qualitative themes were distinct the interconnections between them, which was evident in the qualitative analysis and outlined in the conceptual model, suggest a complex process by which partner serostatus information (or lack thereof) was used in making decisions about partners and sexual risk. For example, participants intertwined knowledge of a potential partners’ status with whether they wanted to have sex with him, the anticipation of sexual comfort and/or pleasure, their behavioral intentions, the sexual risks they were willing to take, and how they felt about the sex they had.

The conceptual model can help researchers and practitioners expand their focus when exploring issues of influence and the underlying conscious and unconscious factors that lead to partner selection, the development of behavioral intentions and risk reduction strategies, and the reasons why some men engage in sexual activities that might increase the risk of HIV transmission. Additionally, the findings indicate that questions regarding influence should be broad in order to capture the breadth of the phenomenon. In other words, it is not sufficient to ask only about how a partner’s HIV status influenced specific sexual behaviors but questions should also explore beliefs, attitudes, and feelings related to HIV and the objectives of sexual encounters.

Finally, the model identifies specific points of potential intervention. For example, interventions could explore the underlying psychological beliefs an individual holds about HIV as a means of understanding how they select partners and the conceptions (or misconceptions) that increase (or decrease) the effectiveness of their protective strategies. This may be especially important as more men begin using PrEP. Additionally, exploring the unique psychological factors related to sexual fulfillment or pleasure would shed light on how individuals develop their set of behavioral strategies and why and with whom they adhere or deviate from those strategies. These types of interventions may be more responsive to the individuals’ unique needs and may result in tailored strategies that can both decrease the risk of HIV transmission and better achieve the objectives of the sexual encounter.

Limitations

There were several limitations to the study. The first limitation was that the question of influence was not the primary focus of the study, which was to assess the best way to administer survey questionnaires. Because of this, there was some variability regarding how interviewers followed up on the respondents’ qualitative answers, resulting in more in-depth information from some respondents over others. Regardless, all of the respondents were given the opportunity to discuss their own perceptions of influence and most discussed these topics in some depth.

Also of note was that respondents were recruited online and from bathhouses and that this was not a randomly selected sample. Therefore, a second limitation is that the study results may be unique to a specific sub-set of gay and bisexual men.

Finally, a third limitation was related to the design of the influence questions. In the survey, the influence question was very specific to the respondents’ last sexual partner. In the debriefing interview, the question was asked more broadly, such that a respondent might limit his response to his last partner, expand the discussion to encompass a broad strategy with all partners, or reflect on what he thought the interviewer ‘meant’ by using the word influence. Given this, there is a chance that some of the respondents’ beliefs about influence were missed.

However, these data also highlighted how the complicated, and sometimes ambiguous, nature of influence as a phenomenon among gay and bisexual men makes constructing effective survey questions more difficult. Therefore, while the wording of the question could be seen as a limitation, it also allowed us to explore ideas for constructing more effective quantitative questions. For example, it is important to fully describe the researchers’ definition of “influence” in the question and to then ask respondents about influence both broadly and with specific partners in mind. This technique would likely decrease ambiguity about the meaning of influence, cast a wide net in identifying when and if influence occurred, and also provide some specific examples of influence in the respondents sexual partnerships.

Conclusion

Understanding how and why HIV transmission continues to occur among gay and bisexual men is critical to reducing HIV infection rates. It is clear in these results that HIV played a major role in the sexual lives of both the HIV-negative and HIV-positive respondents, many of whom were taking the HIV-statuses of their sexual partners into consideration when making partnership and sexual risk decisions. Given this, our data helps to confirm trends that have been found in previous studies, namely those that suggest men are using seroadaptive strategies, such as serosorting and seropositioning, as alternatives to universal condom use (Eaton et al., 2009; Golden et al., 2008; Grace et al., 2014; McConnell, Bragg, Shiboski, & Grant, 2010; Snowden et al., 2014; Wei et al., 2011). What these previous studies have not clearly identified are the diverse beliefs that gay men hold about the HIV-statuses of their sexual partners or the process by which these beliefs influence partner selection and the potential risks in which they engage. The results of this study, and the subsequent conceptual model derived from these results, help to fill in this important gap in the literature by providing an in-depth exploration of the psychological responses to anticipated, perceived, and actual partner HIV-status and its influence on partner preference and selection, behavioral intentions, and safer sex strategies.

These results also identify a number of important complications to HIV-prevention efforts, especially for HIV-negative individuals, that should be studied further. First, engaging in explicit discussions about, and accurate assessments of, a sexual partner’s HIV status are critical to the successful implementation of alternative HIV-prevention strategies. However, this study indicates that sero-guessing and assumptions about partner status can have an important impact on risk decisions. Therefore, factors related to sexual communication should be addressed in future HIV-related research and intervention development. This type of information can be used in determining the potential vulnerabilities in sero-adaptive strategies and help prevention specialists assess how these strategies may inadvertently expose gay and bisexual men to HIV.

Second, there is sometimes a tension between wanting to decrease HIV transmission while simultaneously achieving sexual fulfillment and it is clear that many men are engaging in sero-adaptive strategies as a means of managing this tension by increasing sexual pleasure while also mitigating potential risk. It is important that new preventative strategies acknowledge the importance of sexual pleasure as a valid objective and decrease the pathologization of decisions to forgo condom use, which increases conflict and de-legitimizes sexual pleasure as a valid objective. This will help prevention specialists in better responding to emerging trends in the sexual behaviors of gay and bisexual men that may not map onto prevention dogma. Having a better understanding of primary sexual objectives and the role partner HIV-status has on negotiating sexual partnerships can assist HIV-prevention specialists in developing interventions that respond to the evolving needs of this at-risk population and help strengthen a larger preventative toolbox for both HIV-negative and HIV-positive gay men to use in decreasing HIV transmission risk.