Perceptions of HIV Research Participation among Gay, Bisexual, and Other Men who have Sex with Men and Transgender and Non-Binary Adults

Abstract

Human immunodeficiency virus (HIV) continues to disproportionately affect gay, bisexual, and other men who have sex with men (GBM) and transgender and nonbinary (trans/NB) individuals. This study investigated attitudes toward participation in HIV survey research, guided by Emanuel’s framework for ethical clinical research (e.g., risk–benefit ratio, fair participant selection, respect for participants, social value, and collaborative partnership). GBM (n=294) and trans/NB (n=86) persons recruited at a Pride event in Milwaukee completed a survey assessing risks and benefits of participation in, and comfort responding to, sexual health surveys. Participants reported few ethical concerns (e.g., privacy and confidentiality), with notable differences by race, sexual orientation and gender identity, and prior research experiences. Implications for HIV research with GBM and trans/NB individuals are discussed.

Human immunodeficiency virus (HIV) is a significant public health concern. In the United States, HIV disproportionately affects gay, bisexual, and other men who have sex with men (GBM) and individuals who identify as transgender and nonbinary (trans/NB; Becasen et al., 2019; Crepaz et al., 2019). Approximately 1.2 million individuals in the United States are living with HIV, and 62% of these individuals identify as GBM; within the United States, 69% of new HIV diagnoses are among GBM (Centers for Disease Control & Prevention [CDC], 2020). Risk for HIV among GBM varies by race and ethnicity, with Black GBM at greatest risk when compared to other GBM racial groups (Crepaz et al., 2019). Recent estimates indicate that half of Black GBM will be diagnosed with HIV over their lifetime, compared to one in four Hispanic GBM and one in 11 White GBM (National Center for HIV/AIDS, Viral Hepatitis, STD, & TB Prevention, 2016). Additionally, HIV prevalence is greater among transgender people (9.2%), compared to the general U.S. population (0.5%). Transgender women are at elevated risk of HIV, compared to both the general population and GBM (11.1%; Rosenberg et al., 2016), with an estimated HIV prevalence of 14% (Becasen et al., 2019). Further, Black and Hispanic transgender women are at the greatest risk for HIV among those identifying as trans/NB (Becasen et al., 2019).

Research examining engagement with HIV prevention behaviors, such as sexual health and risk behaviors, HIV testing and treatment, and use of preexposure prophylaxis (PrEP; a pill that prevents HIV infection) indicates the critical need to identify strategies to increase uptake of existing prevention strategies and ultimately reduce HIV infections among GBM and trans/NB persons (Franks et al., 2018; Lui et al., 2018; Noble et al., 2017; Zarwell et al., 2021). Survey research is an essential component to understanding engagement with HIV prevention and treatment among GBM and trans/NB persons, given how widely survey research is used due to its practicality and ease of implementation (i.e., reaching large groups of people in a short span of time at a low cost; Safdar et al., 2016). However, little published research has examined perceptions of risks and benefits to participation in HIV survey research broadly, or among GBM and trans/NB persons specifically. The majority of existing research exploring research participation ethics among sexual and gender minority persons has focused on youth (e.g., Macapagal et al., 2017; Mustanski, 2011) due to concerns related to recruitment and confidentiality among youth that may not have widely disclosed their sexual orientation; these concerns may not be applicable to adult populations, who may have differing research participation attitudes and concerns. Attitudes toward research participation may also be impacted by demographic factors (e.g., race; George et al., 2014) and prior research participation, explored in greater detail below.

The current study sought to examine two aims: (1) ethical concerns (e.g., privacy confidentiality) related to participation in HIV survey research and (2) variations in ethical concerns by race, sexual orientation and gender identity, and prior research participation. The current study is largely informed by Emanuel’s (2008) framework for ethical clinical research, considering relevant ethical principles including (1) favorable risk–benefit ratio (i.e., maximizing benefits and minimizing risks), (2) fair participant selection (e.g., research considerations with vulnerable populations), (3) respect for participants (e.g., confidentiality, dissemination of results), (4) social value (i.e., participant and community benefits from the research), and (5) collaborative partnership (e.g., cultural competence and dissemination of research findings).

Attitudes Toward Participation in HIV Survey Research

Most existing research on attitudes toward participation in HIV research among United States adults has either (1) focused on nonsurvey research such as HIV vaccine research (e.g., Chin et al., 2016; Cottingham et al., 2018; Dubé et al., 2017) or vignettes (Slomka et al., 2008) or (2) focused on other vulnerable populations (e.g., Black women, people who inject drugs; Abadie et al., 2018; Overstreet et al., 2018). Further, little research has examined attitudes toward HIV survey research among GBM and trans/NB persons, despite the potential for survey research to inform strategies for reducing HIV. Important ethical research considerations here are (1) a favorable risk–benefit ratio, highlighting the importance of minimizing risks and maximizing benefits for participants, and (2) respect for persons, considering ways to minimize harms and ensure confidentiality (Emanuel, 2008). In general, concerns regarding privacy and confidentiality represent barriers to participation in survey research, whereas compensation and altruism facilitate participation in survey research (Couper et al., 2008).

The primary ethical concerns in HIV-related research surround respect for persons, including stigma, privacy and confidentiality, and the protection of vulnerable populations in research, such as minoritized racial/ethnic groups (Muthuswamy, 2005). For example, people who inject drugs have reported confidentiality and privacy concerns as a barrier to participation in HIV-related research and cited trust in the researcher as an ongoing concern for research participation (Abadie et al., 2018). Other research indicates nonfinancial perceived benefits are a driving factor in research participation among women who use drugs, GBM, and women living with HIV. For example, participants report community (e.g., believing research benefits the community), political (e.g., seeing research participation as activism), and personal motivations (e.g., emotional gratification and new personal insights) for participating in studies measuring HIV risk (Bell & Salmon, 2011; Chin et al., 2016; Overstreet et al., 2018). The current study sought to examine differences between participants with and without prior research experience to identify factors that may facilitate research participation, such as seeing the potential benefits of research participation for yourself or the community.

GBM and trans/NB Participants as Distinct Populations

Understanding participants’ perceptions of the risks and benefits to survey research and comfort responding to sensitive survey items is needed to tailor future research endeavors and aid in the recruitment of populations at elevated risk for HIV, specifically GBM and trans/NB persons; further, gaining insight into participants’ attitudes toward HIV survey research may aid in developing strategies for recruiting more representative samples. Historically, trans/NB persons have either been excluded from HIV research or not considered as a distinct population from GBM (Poteat et al., 2016). As per the principle of fair participant selection, vulnerable populations should be explicitly considered for research in order that the benefits of research are distributed fairly in society (Emanuel, 2008); further, the social value of HIV survey research is bolstered when the populations most affected by HIV are included in the research process. However, the majority of HIV prevention research has been conducted with GBM, with transgender women reporting a lack of HIV-related research opportunities (Logie et al., 2012; Reisner et al., 2020). This disparate rate of participation in HIV research may lead to research fatigue for GBM as an “overresearched” population (e.g., Ashley, 2021; Clark, 2008), or disengagement from research within the trans/NB community, resulting in underrepresentation and a lack of needed information on affected populations. Notably, one study identified barriers and facilitators to transgender women’s participation in HIV 2 Journal of Empirical Research on Human Research Ethics 0(0) research, including mistrust, safety and confidentiality concerns, and the importance of incentives (Reisner et al., 2020); these findings echo trans/NB persons’ concerns about participation in health research more broadly (Owen-Smith et al., 2016). Another study with transgender women indicated the need for transgender cultural competency, gender-affirming environments, and established community partnerships with trans organizations to bolster research participation (Andrasik et al., 2014). Research regarding transgender men’s attitudes toward HIV research participation is limited, perhaps due to increased HIV risk among transgender women, in particular. However, stigma and lack of cultural competence have been found to decrease sexual health services utilization among transgender men (Scheim et al., 2016), which may similarly reduce willingness to participate in HIV research. The lack of trans/ NB participation in research, as well as oversampling of GBM in existing HIV research, illustrates the need to understand the perspectives of these two distinct populations at increased risk for HIV.

Minoritized Race and HIV Prevention Research Participation

Germane to racial minority populations is the risk–benefit ratio, as minoritized racial/ethnic groups may perceive greater risk and fewer benefits of research participation given long-standing distrust in the health care system and research community due to historical (e.g., Tuskegee syphilis study) and contemporary (e.g., health care discrimination) factors (e.g., George et al., 2014; Kennedy et al., 2007; Scharff et al., 2010). Multiple marginalization due to one’s sexual orientation and/or gender identity and minoritized race (i.e., intersectional stigma) may introduce additional concerns related to participation in research (Quinn et al., 2019). As an example, focus group data indicates Black GBM advocates for collaborative partnerships and striving to disseminate research findings to the Black gay community (e.g., ethical principle of social value) as factors informing attitudes toward participation in HIV vaccine research (Andrasik et al., 2014).

Present Study

The present study addressed two aims: (1) to assess ethical concerns (e.g., privacy, confidentiality) related to participation in HIV prevention research and (2) to examine variation by prior research participation, sexual orientation and gender identity, and race and ethnicity.

Method

Participants and Procedure

Participants were recruited at a community Pride festival in Milwaukee in 2018 at which there were over 45,000 attendees (Rytlewski, 2019). Research staff advertised the study at a table in a Health and Wellness area of the event and recruited passersby. Eligibility criteria included: (1) being 18 years of age or older; (2) self-identifying as a cisgender GBM or identifying as transgender, NB, or another gender identity that does not align with sex assigned at birth; (3) not being visibly intoxicated or impaired; and (4) being able to complete the paper survey in English. Participants received an informational letter explaining the risks and benefits to participating in the anonymous study. Due to the anonymous, low-risk nature of the study, the survey was approved for a waiver of informed consent; thus, after reading the informational letter, participating in the survey constituted consent. Participants completed a brief pen-and-paper survey assessing sexual health, mental and behavioral health, and housing and financial instability, followed by a series of questions regarding the risk and benefits to research participation. Completion of the questionnaire took approximately 5–10 min. Participants were offered a choice of incentive for study participation, either five dollars in cash or a small gift (boa) of equal value. Participants were also offered a list of resources regarding mental health, housing, local sexually transmitted infections testing resources, and PrEP services. All procedures were approved by the Institutional Review Board the Medical College of Wisconsin.

The sample (N=380) primarily identified as cisgender men (n=294, 77.4%), gay (n=256, 67.4%), and non- Hispanic White (n=219, 57.6%), with an average age of 31.72 years (SD=13.38). Most participants were employed full-time (n=227, 59.7%) and were persons living without HIV (n=281, 73.9%). A slight majority of the sample (n= 206, 54.2%) had prior experience participating in a study about gender identity, sexual behavior, or drug and alcohol use, the same topics assessed via the survey battery of the current study. Table 1 presents all demographic information.

Table 1.

Descriptive Statistics of Participant (N = 380) Demographic Characteristics

Variable	n (%)
Gender Identity
Man	294 (77.4%)
Transgender Man	36 (9.5%)
Transgender Woman	23 (6.1%)
Non-Binary/Another Identity	27 (7.1%)
Sexual Orientation
Gay	256 (67.4%)
Bisexual	72 (18.9%)
Pansexual	18 (4.7%)
Heterosexual	15 (3.9%)
Another Orientation	19 (5.0%)
Race/Ethnicity
Non-Hispanic White	219 (57.6%)
Non-Hispanic Black	89 (23.4%)
Hispanic	42 (11.1%)
Another Race/Ethnicity	30 (7.9%)
Employment Status
Employed Full-Time	227 (59.7%)
Employed Part-Time	88 (23.2%)
Unemployed	65 (17.1%)
Relationship Status
Not in a Relationship	169 (44.5%)
In a Committed Relationship	156 (41.1%)
In an Open Relationship	55 (14.5%)
HIV Status
Person Living with HIV	27 (7.1%)
Person Living without HIV	281 (73.9%)
Don’t Know	72 (18.9%)
Prior Participation in Research
Yes	206 (54.2%)
No	174 (45.8%)
Age*	31.72 (13.38)

Note.

^*Continuous variables = mean (standard deviation)

Measures

Demographics

Participants provided demographic information including age, race and ethnicity, gender identity, sexual orientation, employment status, and relationship status. Race (multiple-choice, select all that apply) and ethnicity (“are you Hispanic or Latino/Latinx?”, yes/no) were subsequently recoded and collapsed for analyses into (1) non-Hispanic White and (2) racially minoritized groups (i.e., non-Hispanic Black, Hispanic, and other races). Gender identity was assessed using a two-step process (Reisner et al., 2016) asking (1) sex assigned at birth and (2) current gender identity (male, female, transgender man, transgender woman, NB, and another identity). Sexual orientation response options included: gay, lesbian, same-gender-loving, bisexual, straight/heterosexual, and another identity, with the option to specify. From these items, gender identity and sexual orientation were recoded and collapsed for analyses, consistent with the published literature (Poteat et al., 2016), into (1) cisgender GBM or (2) trans/NB (e.g., transgender man, transgender woman, and NB or another identity) to examine differences between two groups both disproportionately affected by HIV.

Risk and Benefits to Research Participation

Participants completed a 10-item questionnaire used to assess privacy and confidentiality concerns, benefits to research participation, opinions on research compensation, and comfort with responding to questions about sensitive topics (e.g., HIV, sexual behaviors, sexual orientation, and gender identity). Questions were adapted from Macapagal et al. (2017) survey assessing sexual and gender minority adolescents’ perceptions of the risk and benefits of participation in sexuality surveys. See Tables 2 and 3 for survey questions and response options.

Table 2.

Perception of Risk and Benefits among Total Sample (N = 380)

Question & Response Options	n (%)
In terms of your worries about privacy, how did filling out this health survey compare to filling out a form at the doctor’s office?
Worry more at the doctor’s office	91 (23.9%)
About the same	271 (71.3%)
Worry more about filling out this survey	18 (4.7%)
How confident are you that the researchers who are conducting this study will keep what you told them private?
Very confident	248 (65.3%)
Somewhat confident	118 (31.1%)
Not at all confident	14 (3.7%)
Do you think that people with your gender and/or sexual identity are…
Researched too often	26 (6.8%)
Researched enough	120 (31.6%)
Not researched enough	234 (61.6%)
Do you think that people with your gender and/or sexual identity will benefit from participating in research like this?
Yes	238 (62.6%)
Maybe	125 (32.9%)
No	17 (4.5%)
Do you feel that the incentive provided for this health survey is appropriate for your participation in this study?*
Yes	325 (85.5%)
No	38 (10.0%)

Note.

^*Only 363 participants provided a response to this item

Table 3.

Comfort with Health Survey Topics among Total Sample (N = 380)

Question & Response Options	n (%)
How comfortable were you answering questions about gender identity?
Comfortable	295 (77.7%)
Not comfortable	85 (22.4%)
How comfortable were you answering questions about sexual behavior?
Comfortable	287 (75.5%)
Not comfortable	93 (24.5%)
How comfortable were you answering questions about drug use behavior?
Comfortable	287 (75.5%)
Not comfortable	93 (24.5%)
How comfortable were you answering questions about silicone injection outside of a medical setting?
Comfortable	282 (74.2%)
Not comfortable	98 (25.8%)
How comfortable were you answering questions about hormone use outside of a medical setting?
Comfortable	278 (73.2%)
Not comfortable	102 (26.8%)

Analysis Plan

Missing data were addressed viamultiple imputation (Enders, 2017). Missingness ranged from 0.50% to 5.04% on items of interest. Analyses were conducted using SPSS 26.0. Aim 1 was addressed via descriptive statistics, used to assess the perceived risk and benefits to research participation among the total sample. Aim 2 was addressed using chi-square tests with Bonferroni posthoc tests and Cramer’s V effect size to examine differences by (1) prior research participation (yes vs. no), (2) sexual orientation/gender identity (GBM vs. trans/NB), and (3) race/ethnicity (non-Hispanic White vs. racially minoritized groups).

Results

Aim 1: Perception of Risk and Benefits among Total Sample

Table 2 contains descriptive statistics of questions and response options among the total sample. The majority of participants (n=362, 95.3%) did not report greater privacy concerns completing the health survey in comparison to filling out a form at the doctor’s office. Similarly, more than half of the participants were very confident (n=248, 65.3%) that their responses would remain confidential. Further, participants tended to report that gender and/or sexual identity is not researched enough (n=234, 61.6%). The majority of the sample believed that the research results would benefit sexual/gender minority individuals (n =238, 62.6%) and that study incentives were adequate (n= 325, 85.5%). Those who reported the incentive amount was inappropriate were offered the opportunity via an open-ended question to respond to what they felt would be a more appropriate incentive amount. Among the participants who responded, the most frequent alternative incentive amount was ten dollars (n=12), followed by 20 dollars (n=4). Two participants noted that the small gift and five dollars would be more appropriate. Individual participants provided the following alternative incentives: (1) gift cards, (2) 100 dollars, and (3) incentives not necessary. Participants also reported their comfort with answering questions about gender identity, sexual behavior, drug use behavior, silicone injection, and hormone use outside of a medical setting. Across items, approximately three-quarters of the total sample reported that they were comfortable responding to these survey topics (see Table 3).

Aim 2: Variation by Prior Research Participation, Sexual Orientation and Gender Identity, and Race and Ethnicity

Perceptions of Risks and Benefits by Prior Research Participation

Tables 4 and 5 contain bivariate associations of perception of risk and benefits by prior research participation. Minimal differences emerged between participants who had prior research participation and those who did not. Participants who previously participated in research on gender identity, sexual behavior, or drug/alcohol use were significantly more likely to respond that people with their gender and/or sexual identity would benefit from participating in research such as the current study, as compared to those without prior research experience. The effect size was small (Cramer’s V ≤ 0.2). No other significant differences emerged.

Table 4.

Bivariate Associations of Perception of Risks and Benefits by Prior Research Participation

Question & Response Options	Prior Research (n = 206)	No Prior Research (n = 174)	χ2 (df)	Cramer’s V
In terms of your worries about privacy, how did filling out this health survey compare to filling out a form at the doctor’s office?			4.14 (2)	0.10
Worry more at the doctor’s office	54 (26.2%)	37 (21.3%)
About the same	146 (70.9%)	125 (71.8%)
Worry more about filling out this survey	6 (2.9%)	12 (6.9%)
How confident are you that the researchers who are conducting this study will keep what you told them private?			4.25 (2)	0.11
Very confident	142 (68.9%)	106 (60.9%)
Somewhat confident	55 (26.7%)	63 (36.2%)
Not at all confident	9 (4.4%)	5 (2.9%)
Do you think that people with your gender and/or sexual identity are…			4.96 (2)	0.11
Researched too often	15 (7.3%)	11 (6.3%)
Researched enough	44 (36.7%)	65 (37.4%)
Not researched enough	136 (66.0%)	98 (56.3%)
Do you think that people with your gender and/or sexual identity will benefit from participating in research like this?			13.24 (2)	0.19**
Yes	146 (70.9%)a	92 (52.9%)b
Maybe	52 (25.2%)a	73 (42.0%)b
No	8 (3.9%)a	9 (5.2%)a
Do you feel that the incentive provided for this health survey is appropriate for your participation in this study?+			0.69 (1)	0.04
Yes	177 (90.8%)	148 (88.1%)
No	18 (9.2%)	20 (11.9%)

Note. Prior research = Participant answered “yes” to “Have you ever participated in study asking about your gender identity, sexual behavior, or drug/alcohol use before?”; Non-matching letter subscripts in same row denotes significant differences in column proportions (p < .05);

⁺Only 363 participants provided a response to this item;

^**denotes p < .01

Table 5.

Comfort with Health Survey Topics by Prior Research Participation

Question & Response Options	Prior Research (n = 206)	No Prior Research (n = 174)	χ2 (df)	Cramer’s V
How comfortable were you answering questions about gender identity?			1.48 (1)	0.06
Comfortable	155 (75.2%)	140 (80.5%)
Not comfortable	51 (24.8%)	34 (19.5%)
How comfortable were you answering questions about sexual behavior?			0.14 (1)	0.02
Comfortable	154 (74.8%)	133 (76.4%)
Not comfortable	52 (25.2%)	41 (23.6%)
How comfortable were you answering questions about drug use behavior?			2.49 (1)	0.08
Comfortable	149 (72.3%)	138 (79.3%)
Not comfortable	57 (27.7%)	36 (20.7%)
How comfortable were you answering questions about silicone injection outside of a medical setting?			2.62 (1)	0.08
Comfortable	146 (70.9%)	136 (78.2%)
Not comfortable	60 (29.1%)	38 (21.8%)
How comfortable were you answering questions about hormone use outside of a medical setting?			2.43 (1)	0.08
Comfortable	144 (69.9%)	134 (77.0%)
Not comfortable	62 (30.1%)	40 (23.0%)

Note. Prior research = Participant answered “yes” to “Have you ever participated in study asking about your gender identity, sexual behavior, or drug/alcohol use before?”; No associations were significant.

Perceptions of Risks and Benefits by Sexual Orientation/Gender Identity

Sexual orientation and gender identity were collapsed and dichotomized into GBM (n=294, 77.4%) versus trans/NB (n=86, 22.6%) to ensure subsamples of sufficient size for analyses. Differences in perception of risk and benefits across GBM and trans/NB participants were identified (Tables 6 and 7). Compared to GBM participants, trans/NB participants (1) reported having more privacy concerns when completing a health survey at the doctor’s office versus completing our survey and (2) were significantly more likely to report that people with their gender and/or sexual identity are not researched enough. Moderate effect size (≥0.6 Cramer’s V ≥ 0.2) was noted on the GBM and trans/NB difference in perceptions of being researched enough. No significant differences were noted in terms of confidentiality concerns, research benefits, adequacy of compensation, or comfort with specific survey topics.

Table 6.

Bivariate Associations of Perception of Risks and Benefits by Gender Identity

Question & Response Options	GBM+ (n = 294)	Trans/NB (n = 86)	χ2 (df)	Cramer’s V
In terms of your worries about privacy, how did filling out this health survey compare to filling out a form at the doctor’s office?			9.05 (2)	0.15*
Worry more at the doctor’s office	60 (20.4%)a	31 (36.0%)b
About the same	220 (74.8%)a	51 (59.3%)b
Worry more about filling out this survey	14 (4.8%)a	4 (4.7%)a
How confident are you that the researchers who are conducting this study will keep what you told them private?			1.68 (2)	0.07
Very confident	195 (66.3%)	53 (61.6%)
Somewhat confident	90 (30.6%)	28 (32.6%)
Not at all confident	9 (3.1%)	5 (5.8%)
Do you think that people with your gender and/or sexual identity are…			23.35 (2)	0.25***
Researched too often	17 (5.8%)a	9 (10.5%)a
Researched enough	111 (37.8%)a	9 (10.5%)b
Not researched enough	166 (56.5%)a	68 (79.1%)b
Do you think that people with your gender and/or sexual identity will benefit from participating in research like this?			2.51 (2)	0.08
Yes	182 (61.9%)	56 (65.1%)
Maybe	101 (34.4%)	24 (27.9%)
No	11 (3.7%)	6 (7.0%)
Do you feel that the incentive provided for this health survey is appropriate for your participation in this study?+			1.85 (1)	0.07
Yes	251 (86.8%)	74 (93.7%)
No	33 (11.6%)	5 (6.3%)

Note. GBM+ = non-heterosexual cisgender men; trans/NB = transgender and non-binary; Non-matching letter subscripts in same row denotes significant differences in column proportions (p < .05);

⁺Only 363 participants provided a response to this item;

^*denotes p < .05,

^***denotes p < .001

Table 7.

Comfort with Health Survey Topics by Gender Identity

Question & Response Options	GBM+ (n = 294)	Trans/NB (n = 86)	χ2 (df)	Cramer’s V
How comfortable were you answering questions about gender identity?			0.01 (1)	0.004
Comfortable	228 (77.6%)	67 (77.9%)
Not comfortable	66 (22.4%)	19 (22.1%)
How comfortable were you answering questions about sexual behavior?			0.31 (1)	0.03
Comfortable	224 (76.2%)	63 (73.3%)
Not comfortable	70 (23.8%)	23 (26.7%)
How comfortable were you answering questions about drug use behavior?			2.88 (1)	0.09
Comfortable	228 (77.6%)	59 (68.6%)
Not comfortable	66 (22.4%)	27 (31.4%)
How comfortable were you answering questions about silicone injection outside of a medical setting?			2.66 (1)	0.08
Comfortable	224 (76.2%)	58 (67.4%)
Not comfortable	70 (23.8%)	28 (32.6%)
How comfortable were you answering questions about hormone use outside of a medical setting?			1.85 (1)	0.07
Comfortable	220 (74.8%)	58 (67.4%)
Not comfortable	74 (25.2%)	28 (32.6%)

Note. GBM+ = non-heterosexual cisgender men; trans/NB = transgender and non-binary; No associations were significant.

Perceptions of Risks and Benefits by Race

Race/ethnicity was collapsed and dichotomized into non-Hispanic White (n= 219, 57.6%) versus racially minoritized groups (n=161, 42.4%) to ensure subsamples of sufficient size for analyses. Differences in perceptions of risk and benefits, as well as comfort with survey question topics, between non-Hispanic White and racially minoritized participants, were identified (Tables 8 and 9). Compared to non-Hispanic White participants, racially minoritized participants were: (1) significantly less confident that researchers would keep study results confidential, (2) significantly more likely to report that people with their gender and/or sexual identity are researched too often, and (3) significantly less likely to indicate that the study incentives were appropriate. Moderate effect size (≥0.6 Cramer’s V ≥ 0.2) was noted on the non-Hispanic White and racially minoritized participant difference in perception of incentive adequacy; other effect sizes were small (Cramer’s V ≤ 0.2). No significant differences were noted in terms of privacy concerns compared to the doctor’s office or perceived research benefits. In terms of comfort with survey topics, racially minoritized participants were significantly less comfortable with all survey topics, including gender identity, sexual behavior, drug use behavior, silicone injection, and hormone use outside of a medical setting. Effects sizes were small (Cramer’s V ≤ 0.2).

Table 8.

Bivariate Associations of Perception of Risks and Benefits by Race

Question & Response Options	Non-Hispanic White (n = 219)	Minoritized Race (n = 161)	χ2 (df)	Cramer’s V
In terms of your worries about privacy, how did filling out this health survey compare to filling out a form at the doctor’s office?			2.89 (2)	0.09
Worry more at the doctor’s office	55 (25.1%)	36 (22.4%)
About the same	157 (71.7%)	114 (70.8%)
Worry more about filling out this survey	7 (3.2%)	11 (6.8%)
How confident are you that the researchers who are conducting this study will keep what you told them private?			6.23 (2)	0.13*
Very confident	154 (70.3%)a	94 (58.4%)b
Somewhat confident	57 (26.0%)a	61 (37.9%)b
Not at all confident	8 (3.7%)a	6 (3.7%)a
Do you think that people with your gender and/or sexual identity are…			8.72 (2)	0.15*
Researched too often	9 (4.1%)a	17 (10.6%)b
Researched enough	64 (29.2%)a	56 (34.8%)a
Not researched enough	146 (66.7%)a	88 (54.7%)b
Do you think that people with your gender and/or sexual identity will benefit from participating in research like this?			3.66 (2)	0.10
Yes	139 (63.5%)	99 (61.5%)
Maybe	74 (33.8%)	51 (31.7%)
No	6 (2.7%)	11 (6.8%)
Do you feel that the incentive provided for this health survey is appropriate for your participation in this study?+			15.79 (1)	0.21***
Yes	203 (94.9%)a	122 (81.9%)b
No	11 (5.1%)a	27 (18.1%)b

Note. Minoritized Race = non-Hispanic Black, Hispanic, and other races; Non-matching letter subscripts in same row denotes significant differences in column proportions (p < .05);

⁺Only 363 participants provided a response to this item;

^*denotes p < .05,

^***denotes p < .001

Table 9.

Comfort with Health Survey Topics by Race

Question & Response Options	Non-Hispanic White (n = 219)	Minoritized Race (n = 161)	χ2 (df)	Cramer’s V
How comfortable were you answering questions about gender identity?			12.14 (1)	0.18***
Comfortable	184 (84.0%)a	111 (68.9%)b
Not comfortable	35 (16.0%)a	50 (31.1%)b
How comfortable were you answering questions about sexual behavior?			6.55 (1)	0.13*
Comfortable	176 (80.4%)a	111 (68.9%)b
Not comfortable	43 (19.6%)a	50 (31.1%)b
How comfortable were you answering questions about drug use behavior?			9.25 (1)	0.16**
Comfortable	178 (81.3%)a	109 (67.7%)b
Not comfortable	41 (18.7%)a	52 (32.3%)b
How comfortable were you answering questions about silicone injection outside of a medical setting?			10.23 (1)	0.16**
Comfortable	176 (80.4%)a	106 (65.8%)b
Not comfortable	43 (19.6%)a	55 (34.2%)b
How comfortable were you answering questions about hormone use outside of a medical setting?			7.62 (1)	0.14**
Comfortable	172 (78.5%)a	106 (65.8%)b
Not comfortable	47 (21.5%)a	55 (34.2%)b

Note. Minoritized Race = non-Hispanic Black, Hispanic, and other races; Non-matching letter subscripts in same row denotes significant differences in column proportions (p < .05);

^*denotes p < .05,

^**denotes p < .01,

^***denotes p < .001

Discussion

Prevalence of HIV is heightened among GBM and trans/NB persons (Becasen et al., 2019; Crepaz et al., 2019). As such, the inclusion of persons from these populations in HIV prevention research is imperative and ethically necessary as per the concepts of fair participant selection and social value (Emanuel, 2008). However, little is known about attitudes toward participation in HIV survey research, or potential demographic (e.g., sexual orientation/gender identity and race/ethnicity) differences in attitudes. As such, the current study sought to examine among a U.S. adult sample: (1) overall perceptions of ethical considerations of participation in HIV survey research and (2) potential differences between GBM and trans/NB persons, racial groups, and those with and without prior research participation in perceptions of HIV survey research. Key findings and implications for the conduct of HIV prevention research with GBM and trans/NB persons are discussed in detail below.

Regarding aim 1, collectively, participants had a positive perception of participation in the survey and endorsed a few ethical concerns. Participants overwhelmingly reported that they did not perceive any problems with privacy and confidentiality when completing the current survey. Further, most participants felt a need for more research among GBM and trans/NB persons, but fewer were sure of the benefits of research participation; taken together, these findings inform our understanding of the favorable risk/benefit ratio, with participants reporting few risks but also few benefits of HIV survey research participation. Regarding aim 2, only one difference was noted between respondents with and without prior experience participating in sexual health research: those with previous experience were more likely to see the benefits of research participation. This finding may be due, in part, to positive experiences participating in research or motivation to participate due to perceived individual or community benefits (i.e., benefits and social value). Indeed, these findings are consistent with prior research which indicates that personal or community benefits derived from research influences motivation to participate (Bell & Salmon, 2011; Chin et al., 2016; Overstreet et al., 2018). Regarding differences in perceptions by sexual orientation/gender identity, GBM and trans/NB persons did not differ significantly in their comfort with research questions. Further, GBM and trans/NB persons endorsed similar perceptions of benefits and positive attitudes toward incentives. Two differences were noted between GBM and trans/NB participants. The first was in relation to research inclusion, with trans/NB persons significantly more likely to endorse that people with their gender and/or sexual identity are not researched enough. This difference may stem from the exclusion of trans/NB persons from much HIV prevention and treatment research (Logie et al., 2012; Reisner et al., 2020), or trans/NB persons considered as one homogenous group along with GBM (Poteat et al., 2016). Despite the differences among GBM-trans/NB identified in the current study, the majority of GBM also reported that people with their gender and/or sexual identity are not researched enough. Second, trans/NB persons were more likely to have greater privacy concerns at the doctor’s office as compared to completing the survey, which is likely a reflection of the pervasive healthcare discrimination faced by trans/ NB persons (e.g., lack of health insurance coverage for needed care and lack of affirming/knowledgeable clinicians and clinics, Cicero et al., 2019) than survey participation itself.

We found significant differences were noted by race and ethnicity, with racially minoritized participants overwhelmingly endorsing more negative perceptions of research participation, including greater confidentiality concerns, less comfort responding to survey questions, and feeling overresearched and undercompensated compared to non-Hispanic White participants. These findings are consistent with the extant literature, which indicates mistreatment of minoritized racial groups by health care, government, and other structural entities that may lead to distrust of the research community (George et al., 2014; Kennedy et al., 2007; Scharff et al., 2010).

Limitations

Findings should be considered in the context of study limitations. Our participants were recruited from a Pride event in the Midwest. Recruitment of sexual and/or gender minority populations often employs events or venue-based sampling to reach the population (Meyer & Wilson, 2009); however, sampling at venues or events, such as festivals, may limit the generalizability of the sample, as Pride attendees may differ from other sexual orientation and gender identity diverse persons (e.g., more connected to the community or more out with identity; Meyer and Wilson, 2009). Differences may exist in in-person versus online recruitment as well. For example, in-person respondents to the U.S. Transgender Survey were younger and more likely to be people of color, transgender women, and have lower education levels and income compared to participants recruited online (Reisner et al., 2014). As well, participants had to be English-speaking to participate, resulting in a lack of generalizability to the greater U.S. population; future research is needed with non-English-speaking participants. Finally, we dichotomized demographic characteristics for comparison due to the small sample size, limiting the ability to detect differences within marginalized groups, such as trans men and trans women, or different racial subgroups.

Best Practices

Despite these limitations, our findings have direct implications for best practices in future HIV prevention and treatment research. Recommendations for future research endeavors include: (1) demonstrating potential individual and community benefits to HIV research participation, (2) incorporating strategies to recruit more trans/NB persons in HIV research, and (3) addressing racially minoritized participant concerns via community-driven research and researcher cultural competence. These implications should be considered in the context of study limitations, namely a geographically limited sample recruited via one venue which may impact the generalizability of results.

Demonstrating Potential Individual and Community Benefits to Research Participation and Findings

Taken together, these findings highlight the need for improved communication to participants regarding the benefits of research participation, addressing Emanuel’s (2008) principle of a favorable risk–benefit ratio. Communicating clearly in recruitment and consent documents the ways in which the findings may be used to enact meaningful change is needed to encourage participation. Beyond indicating potential benefits via recruitment and consent, disseminating research findings to study participants is one strategy to clearly demonstrate the positive outcomes of study participation; this is consistent with the principles of respect for participants and collaborative partnership, which encourages community-based dissemination of research findings (Emanuel, 2008). Consulting with existing, published email templates for disseminating study findings to participants, which contain the study question, study results, a thank you to participants, and information for future studies, may be of utility (e.g., Melvin et al., 2020). Alternatively, nontechnical executive summaries and abstracts may be distributed to participants; Nixon et al. (2013) published examples of these materials from an HIV survey with high school students in South Africa.

Further, consistent with the principle of a collaborative partnership (Emanuel, 2008), it is imperative for research teams to demonstrate a commitment to the population whom they are researching, a key component to research methodologies including community-based participatory research (CBPR) and integrated knowledge translation (IKT; Jull et al., 2017). Central to both CBPR and IKT is the involvement of community members and stakeholders throughout the research process, including research design, data collection, and dissemination. To demonstrate the social value (Emanuel, 2008) of the research to the community at large, study findings should be disseminated broadly to the affected community (e.g., Knerr et al., 2016). Study findings can be published to a website, provided to community partners in the form of a technical report, distributed via newsletter to community health clinics and organizations providing HIV-related services, presented at community meetings, and communicated via other media outreach approaches (e.g., television, radio, and magazine; McDavitt et al., 2016; Knerr et al., 2016). Indeed, research findings from the current Pride survey are often disseminated in presentations to the community in Milwaukee. As an additional resource, Brown et al. (2021) provide keen insights into community-based HIV research via a case study, with a good participatory practices checklist that can be used to guide collaborative HIV research endeavors.

Targeted Recruitment of and Collaboration with GBM and Trans/NB Participants

The vast majority of HIV research has been conducted with GBM (Poteat et al., 2016); however, we found that GBM did not endorse a sense of research fatigue, with most GBM reporting the need for more research among their population. It is important to note that the use of convenience sampling may have impacted responses to this question, as individuals willing to participate in the study may be experiencing less research fatigue than those who chose not to participate. Efforts to recruit and engage GBM in HIV research should be continued, including venue-based sampling (as used in the current study), among other targeted recruitment strategies (e.g., snowball and respondent-driven sampling; Gama et al., 2017) to adhere to the ethical principle of fair participant selection (Emanuel, 2008). Although findings illustrate that, overall, both GBM and trans/NB participants feel underresearched, trans/NB persons report a greater need for research with persons of their gender identity; the current study echoes findings in the extant literature, with transgender women reporting a lack of opportunity to participate in HIV prevention research (Reisner et al., 2020). In fact, Singh (2016) asserts the need for HIV prevention research with exclusively trans/NB samples, given that they are underrepresented in existing research and have unique needs related to HIV prevention.

In conducting research with trans/NB populations that adheres to the principle of a collaborative partnership (Emanuel, 2008), Vincent (2018) outlined myriad ethical research strategies for recruitment of and collaboration with trans/NB persons including (1) researchers being aware of trans/NB history, (2) transparency with participants, (3) intentional use of vocabulary (e.g., gender identity), (4) reliance on feminist methodology (e.g., considering the power dynamic between researcher and participant), (5) inclusion of intersectionality, and (6) respect of trans/NB spaces (e.g., heavy reliance on trans/NB community partners for recruitment). Such ethical considerations offer a user-friendly framework to train researchers interested in trans/NB health. Additionally, for survey research, in particular, careful consideration of the wording of items assessing sexual orientation and gender identity is imperative, including clearly separating the constructs of sexual orientation and gender identity (Suen et al., 2020) and the use of standardized methods for collecting gender identity information, such as the two-step method (Reisner et al., 2016).

Building Trust and Collaboration with Racially Minoritized Participants

It is imperative to include racially minoritized GBM and trans/NB persons in HIV survey research given disparate rates of HIV infection among this population (Crepaz et al., 2019; National Center for HIV/AIDS, Viral Hepatitis, STD, & TB Prevention, 2016); however, racially minoritized participants reported less trust in the research process within our study reflecting the broader literature. Thus, strategies to build trust and collaboration with racially minoritized participants are needed to aid in the recruitment of underserved groups. These findings also indicate the importance of building collaborations with the community, including employing community members as research staff (True et al., 2017). Indeed, Vermund et al. (2018) highlight the need for more research staff who are of the same racial and ethnic groups as those most affected by HIV (i.e., Black and Hispanic), detailing a research scholar program aimed at increasing the number of persons from minoritized racial groups conducting HIV prevention research.

Wallerstein et al. (2019) suggest a culture-centered approach (Dutta, 2007) to CBPR that may be of utility when working with racially minoritized groups. Central to this approach is (1) listening to and valuing community knowledge, (2) questioning the power dynamic between researcher and community, and (3) using resources to enact change within the structural factors leading to health inequities, consistent with the ethical principles of a collaborative partnership with fair participant selection (Emanuel, 2008). Other suggestions for improving recruitment include (1) in-person direct interaction including an initial in-person meeting even for online studies; (2) diverse methods for participant contact (e.g., cell phone, email, and in-person); (3) resources that explain the research process (e.g., brochures, flyers); (4) speaking directly with participants about how information is shared, explaining confidentiality, and allowing for questions; (5) using online and social media recruitment; and (6) partnering with community organizations and churches to distribute recruitment materials (Qualls, 2002; Watson et al., 2016). Such guidelines again provide a training-friendly structure in the education of future sexual health and other researchers.

Research Agenda

Future studies investigating attitudes toward participation in HIV prevention and treatment research with larger, more geographically diverse samples, including non-English-speaking participants, recruited through multiple approaches (e.g., in-person, online, and through peer referral) are needed. Larger samples would also permit the examination of the impact of multiple, intersecting identities on attitudes toward research participation. Finally, future HIV prevention and treatment research should strive to recruit more trans/NB persons, treating trans/NB participants as a research population distinct from other populations at increased risk of HIV.

Educational Implications

Training students and investigators who are conducting HIV treatment and prevention research in community-driven research methodologies (i.e., CBPR and IKT) is imperative, either at their home institution or through other readily available materials. As but one example, the Clinical and Translational Science Institute at the University of Rochester Medical Center offers a free course in CBPR that is available to researchers of all levels (CTSI, n.d.). Further, funded training programs aimed at HIV prevention and treatment research such as the NIMH AIDS Research Training Program (National Institute of Mental Health, n.d.) or NIH-Funded STD/AIDS Research Training Fellowship Program (University of Washington Center for AIDS and STD, n.d.) should strive to recruit applicants with lived experience in communities disproportionately affected by HIV (e.g., trans/ NB individuals, GBM, and individuals with minoritized races). As well, findings from the current study may inform research training programming, highlighting the need for greater research opportunities for trans/NB persons and increased efforts building connections to minoritized racial communities.