Gay, Bisexual, and Other Men Who Have Sex With Men Who Are Not on Oral PrEP may be Less Interested in Available Injectable Products than in Oral PrEP: Examining Individual-Level Determinants of Interest and Barriers Across Products

Abstract

Approval of the first injectable PrEP product (cabotegravir) provides an exciting addition to oral PrEP that could encourage those not currently on PrEP to use it. However, few studies have explored interest in injectable cabotegravir among those at increased risk who are not currently on PrEP. We conducted an online survey with 327 gay, bisexual, and other men who have sex with men (GBM) with limited PrEP history to explore their interest and intentions to use oral and injectable PrEP (cabotegravir), and examine barriers and individual-level predictors of both product types. Results showed that 17% of participants who reported being uninterested/neutral about oral PrEP expressed interest in injectable PrEP. Slightly more racial/ethnic minority GBM who were uninterested/neutral for oral PrEP expressed interest in injectable PrEP than White GBM (23% vs. 14%). Determinants were similar across PrEP types. Findings can directly inform interventions encouraging use of both PrEP products.

Introduction

Improving access to and use of HIV pre-exposure prophylaxis (PrEP) in groups at increased risk for HIV is a key part of the 2019 United States Health and Human Services’ Ending the Epidemic (EHE) plan because of its considerable promise for helping achieve sustained declines in new infections [1, 2]. Since the first antiviral drug was approved for use as PrEP in 2012, use among those at increased risk has risen from just 3% in 2015 to 28% in 2020 [3, 4]. However, current use varies dramatically across racial/ethnic groups, with more than 66% of White men at high risk having been prescribed PrEP in 2020 compared to only 16% of Hispanic/Latino (H/L) men and 9% of Black/African American (B/AA) men at high risk [4]. Continuing to increase PrEP use, particularly among B/AA and H/L gay, bisexual, and other men who have sex with men (GBM), is critical to achieving EHE goals.

Prior to December 2021, the only drugs approved for use as PrEP were oral medications, tenofovir disoproxil fumarate and emtricitabine (TDF/FTC) and tenofovir alafenamide and emtricitabine (TAF/FTC), and guidelines recommended that patients take one pill daily [5]. Follow-up studies of initial oral PrEP efficacy trials showed that taking ≥ 4 doses per week likely achieved more than > 95% protection [6, 7]. However, a number of studies have shown that many in groups who are at increased risk for HIV do not achieve these protective levels of adherence [6–9]. Among those who are at increased risk who have not yet been prescribed PrEP, taking a pill every day may be a significant barrier, particularly among B/AA and H/L GBM [10, 11]. Having to store pill containers is also a key concern among adolescents and other groups at high risk because of the danger of inadvertent disclosure and stigma [12–14]. Together, these findings suggest that some attributes of oral PrEP could limit its effectiveness among those prescribed PrEP and its appeal to those not yet prescribed PrEP.

In December 2021, the US FDA approved Apretude^® (cabotegravir), the first extended release, injectable PrEP product [15]. It is given as two injections one month apart to assess hypersensitivity reactions, and then once every two months thereafter. A sizable literature has shown that a majority of GBM and other groups at increased risk show interest in injectable PrEP and that nearly half may prefer it to oral PrEP [16–18], but it does not appear that many of these studies provided information about specific potential PrEP products before asking participants about their interest. However, these findings were echoed in HTPN 077, a safety trial of long-acting cabotegravir, though participants were asked after receiving injections [19]. Few studies have explored interest in this specific injectable PrEP product, given its known efficacy, protocol, and side effects. Importantly, while injectable PrEP may provide a preferable alternative for many who are currently on oral PrEP, much of its public health promise lies in its potential to encourage more who are not currently on PrEP to use it. Yet, few studies have examined interest in injectable PrEP specifically among those who are not currently using oral PrEP.

Finally, although cost has been consistently among most prohibitive barriers to starting PrEP, many private health insurance plans began covering oral PrEP in 2019 as part of the Affordable Care Act [20] and the “Ready, Set, PrEP” program has begun providing free oral PrEP for at-risk patients without insurance [21]. Significant access and cost barriers still exist, though, including inconsistent compliance among some insurance providers and burdensome documentation requirements. However, individual-level factors also likely play an important role in adoption, and their influence may increase as access improves. The Information-Motivation-Behavioral Skills (IMB) model [22, 23] provides a useful framework for organizing potential determinants, many of which have been shown to influence oral PrEP uptake in past studies. For example, past research suggests that knowledge about PrEP (information; 24, 25) and several motivation-related factors, including low risk perceptions [24, 25], low social norms [26, 27], high PrEP stigma [14, 28], and high medical mistrust [29, 30], affect PrEP uptake. The IMB model suggests that these factors would primarily affect PrEP adoption through behavioral skills, like PrEP self-efficacy, which refers to beliefs that one can execute the tasks needed to attain and use PrEP, if they want to [28, 31]. Understanding associations between these factors and intentions to seek oral and injectable PrEP can contribute directly to the design of interventions that aim to change those factors with the goal of promoting PrEP adoption. Testing whether they are similar to or different across oral and injectable PrEP can similarly inform interventions that address the unique individual barriers that affect each product.

In this study, we presented GBM who were not currently on PrEP with basic information (e.g., efficacy, guidelines for use, side effects) about specific oral PrEP (emtricitabine/tenofovir) and injectable PrEP products (cabotegravir) and asked them to rate their interest in each product if they were able to access it for free. We also asked them to identify barriers to interest in either type of product, and tested whether several individual-level determinants suggested by the IMB model and past research were associated with interest. Specifically, our aims were to: [1] summarize and compare interest in using oral PrEP versus cabotegravir, specifically examining whether those who were less interested in using oral PrEP were more interested in cabotegravir, [2] summarize and compare barriers to use of oral versus injectable PrEP, and [3] to test whether several individual-level determinants identified in the IMB model [22, 23] and past research on oral PrEP [14, 24–30, 32, 33] were associated with interest in injectable PrEP and compare identified determinants with those associated with interest in oral PrEP. Based on past research, we anticipated that GBM would express stronger interest in injectable PrEP versus oral PrEP, but would also identify some unique barriers across the two products. We also anticipated that key determinants, particularly HIV risk perceptions, PrEP stigma, and self-efficacy, would be associated with intentions to use injectable PrEP, and would be similar across both types of products.

Methods

Recruitment

Participants (N = 327) were recruited from a participant registry for a larger study on HIV self-testing [34] that was collected by conducting outreach on several gay-oriented dating platforms (e.g., Grindr, Jack’d), social media (e.g., Facebook, Instagram, Tumblr), and other websites (e.g., Google Search, TrafficJunky) from December 2021 to January 2021. Participants were eligible for this study if they [1] were 18 years or older, [2] were assigned male sex at birth, [3] reported having had anal sex with a man in the past year, [4] were HIV-negative by self-report, [5] were able to speak and read either English or Spanish fluently, and they [6] either had (a) never been on PrEP, or (b) had been on PrEP, but not for at least 6 months. Given the goals of the larger self-testing study, participants were also required to: [7] have access to a location where packages can be securely delivered, [8] currently own a smartphone, and [3] have not tested for HIV in the past 6 months. We invited participants to complete this survey if they had either (a) participated in the eTest study but did not report any PrEP use during the 12-month follow-up period of that study, or (b) were eligible for and interested in the larger study, but chose not to enroll, and reported never having been prescribed or used PrEP. As such, all participants included in this study had not used PrEP for at least the last 18 months. This was confirmed by self-report at the outset of the survey in this study. We recruited a target sample of N ≥ 309 based on a priori power analysis suggesting that, in a multiple regression model with 13 predictors, α = 0.05, and power of 0.80, this sample size would be powered to detect a medium-to-small effect size (f² = 0.06).

Procedures

Participants who had previously completed the HIV self-testing study (53%, N = 160) were asked to verify the contact information (name, phone, email, address, date of birth) they submitted at registration in a phone call with staff before being considered enrolled in that study. All participants were invited to complete this study via email by clicking on a personal link to the survey. After collecting basic demographics, the survey first asked about participants’ awareness of oral PrEP and assessed basic knowledge about oral PrEP. For those who reported ever being prescribed or using PrEP, we then collected a more detailed history of their experience with PrEP. In participants who were not aware of oral PrEP, we provided basic information about it (e.g., common drugs, efficacy, typical adherence recommendations, safety, and side effects) in bullet-point and table formats. A simple, three-question quiz then assessed their understanding of the information presented. The survey session ended for those who missed any question more than twice. We included this to ensure that respondents had a basic understanding of oral PrEP in order to rate their interest. Participants were asked to rate their interest and intentions as though cost were not a factor, and they could access the medications for free. We did so in order to ensure that the focus was specifically on the influence of potential individual-level determinants. After all questions about oral PrEP had been answered, we then asked about participants’ awareness of cabotegravir injectable PrEP. Given the novelty of the product, all participants were then presented basic information about cabotegravir injections and asked to complete a three-question quiz in a format similar to oral PrEP. This ensured that participants had read and understood basic information about injectable PrEP before responding to questions about interest and intentions. Finally, participants responded to measures assessing key determinants of PrEP interest/use. Participants were compensated for their time with a $10 gift card sent via email, and were also entered for a chance to win one of three $100 prizes drawn via lottery. All procedures were approved by the Brown University Institutional Review Board (IRB).

Measures

PrEP (oral/injectable) awareness, interest, and intentions to use.

Awareness of oral and injectable PrEP was assessed using two, single-item questions with the stem, “Have you ever heard of [taking a pill or getting injections] to keep you from getting HIV?” (yes/no). These questions also provided common terms used and both trade and generic drug names for each. Participants rated their interest in and intentions to start using each type of PrEP product on a 1 (not at all) to 5 (very much) scale.

PrEP (oral/injectable) barriers.

Questionnaires assessing barriers to oral and injectable PrEP were assessed using a scale adapted from Golub et al. [35]. In the oral PrEP measure, participants rated 17 items on a scale from 0 (not a concern) to 3 (major concern, meaning I would not take PrEP pills because of this issue). Example items include “having to take a pill every day” and “having to get the medication and pick it up.” For the injectable PrEP measure, participants rated 19 items on the same scale as the oral PrEP measure. Example items include “going to the clinic every 2 months for an injection,” and “having to get an injection once a month for two months beforehand.” Some items were common to both versions but a few in each version were unique to each type of PrEP.

PrEP knowledge was assessed using a 4-item scale [36] that asked about the general effectiveness of PrEP products, the impact of adherence on effectiveness, and PrEP’s impact on other sexually transmitted infections (STIs). The sum of participants’ total correct responses were used in final models.

Attitudes about PrEP were assessed using a 19-item scale described in Golub et al. [31]. Items were rated in a 1 (strongly disagree) to 4 (strongly agree) scale. Example items include “PrEP is an excuse for gay and bisexual men to avoid using condoms” and “gay and bisexual men who take PrEP are being responsible.” The measure showed excellent reliability in this study (α = 0.83).

PrEP social norms were assessed via items we adapted from work in other areas [37]. Two items assessed descriptive social norms, asking “what percentage of [your friends and acquaintances/all gay, bisexual, or other men who have sex with men (MSM)] do you think are using PrEP?” Participants responded by selecting a percentage (0–100%) on a slider bar. Injunctive norms for friends and acquaintances, most gay, bisexual, and other MSM, and family were assessed by asking how much each of these groups “would approve of using PrEP?” Participants rated these items on a 1 (strongly disagree) to 6 (strongly agree) scale. These items had adequate reliability in this sample (α = 0.63).

PrEP Stigma Scale [38] is a 13-item scale used to assess common disapproving beliefs about people who use PrEP. Example items include “someone taking PrEP should keep it hidden” and “someone using PrEP should receive praise for being responsible.” Items are rated on a 1 (strongly disagree) to 5 (strongly agree) scale. The scale showed high reliability in this study (α = 0.86).

PrEP use self-efficacy was assessed using an 8-item scale we created for this study. The scale assesses participants’ degree of confidence that they could carry out the various behaviors needed to obtain and take either oral or injectable PrEP over time, if they wanted to. An example item was “I could find a doctor or medical provider to find out more about PrEP medication if I wanted to.” Items were rated on a 1 (strongly disagree) to 4 (strongly agree) scale, and the scale showed excellent reliability in this sample (α = 0.89).

We also collected an abbreviated, 4-item version of the Perceived Risk for HIV Scale [39]. Example items included “I worry about getting infected with HIV” and “I feel vulnerable to HIV.” The rating scale differed for each item, but was either a 1 to 4 or 1 to 5 scale, with higher ratings reflecting more perceived risk. These items had good reliability in this study (α = 0.79).

Perceptions of HIV severity was assessed using a 5-item scale adapted from Zimmerman et al. [40]. Participants rated each item using the stem, “If you got HIV, how serious or difficult would it be for you to...” and example items included “take pills every day to treat your HIV?” and “Have to tell family, friends, or partners that you are HIV-positive?” Items were rated on a 1 (not at all serious/difficult) to 5 (extremely serious/difficult) scale. These items showed good reliability in this study (α = 0.77).

Medical Mistrust Index is a 7-item scale that assesses respondents’ level of trust in healthcare organizations [41]. Example items include “mistakes are common in healthcare organizations” and “you’d better be cautious when dealing with healthcare organizations.” Items are rated on a 1 (strongly disagree) to 4 (strongly agree) scale. This scale showed very good reliability in this study (α = 0.82).

Data Analysis Plan

We first summarized demographic, risk, and PrEP history data. Although participants were not excluded from either the larger study or this one based on their gender, we restricted these analyses to only cisgender men because only 11 participants who identified as transgender, non-binary, or non-conforming completed the survey, which would not have been sufficient to allow meaningful subgroup analyses. Next, we computed basic summary statistics for oral and injectable PrEP interest, intentions and motivation, comparing differences across PrEP types descriptively and using t-tests. Then, we summarized key barriers identified as ‘mild, moderate’ or ‘severe’ and compared the most commonly identified barriers across PrEP types descriptively. Next, we calculated bivariate correlations between individual-level determinants that would be included in regression models. Finally, we estimated linear regression models of oral PrEP interest and intentions, and cabotegravir interest and intentions, with each of the constructs specified above as predictors. Histograms with overlaid normal distributions confirmed that each outcome variable was approximately normally distributed. All continuous predictors were centered at their means prior to analysis. We included dummy-coded variables for ever having had a PrEP prescription, meeting 2017 CDC criteria for PrEP candidacy, and minority status in the final models in order to explore differences in interest across these groups and account for these variables when interpreting the effects of other predictors. Given previous research suggesting that medical mistrust may be especially pronounced among B/AA and H/L individuals, we specified a two-way interaction between medical mistrust and identification with one of these racial/ethnic minority statuses. Partial ω² was used as a metric of effect size. Although we originally included PrEP attitudes in our full models, there was evidence of multicollinearity between PrEP attitudes and stigma (VIF = 2.60) so only PrEP stigma was retained in the final models. Analyses were conducted in Stata 14 (Stata Corp., College Station, TX), and statistical significance was set at α < 0.05.

Results

Demographic and behavioral characteristics of the sample are reported in Table 1. A third of participants were racial/ethnic minorities (33.6%), with about half living in the northeastern US, and about a quarter each living in the West and South. 70% met the 2017 CDC criteria for PrEP eligibility and 17% reported having had a prescription for PrEP at some point in their lifetimes. We tested whether there were differences in awareness, interest, or intentions across those who participated in the larger study on self-testing versus those who did not, given that there could have been systematic differences across these two groups. We found no statistically significant differences.

Table 1.

Demographic characteristics of participants (N = 327)

Characteristic	Mean (SD) or N (%)
Age (Range: 19–80 years)	36.2 (12.3)
Hispanic/Latino	82 (25.2)
Race
Black/African American	34 (10.4)
American Native	1 (0.3)
Asian	41 (12.4)
Multiracial	20 (6.1)
White	203 (62.1)
Other	16 (4.9)
Choose not to respond	12 (3.7)
Spanish primary language	9 (2.8)
Region of primary residence
Northeast	148 (45.7)
South	79 (24.4)
Midwest	15 (4.6)
West	82 (25.3)
College degree	239 (73.1)
Low income1	78 (23.9)
Unemployed	35 (10.7)
Sexual identity
Gay	285 (87.2)
Bisexual	37 (11.3)
Other	2 (0.6)
Not sure	3 (0.9)
Met CDC PrEP eligibility criteria2	185 (56.5)
Ever prescribed PrEP in lifetime	55 (16.8)

Note.

¹Defined as annual income <$30,000.

²PrEP eligibility criteria reflected 2017 guidelines, and included participants who reported any of the following in the past 6 months: [1] anal sex with a man outside of a sexually-exclusive relationship, [2] having been diagnosed with a sexually transmitted disease (e.g., Chlamydia, Gonorrhea, or Syphilis), or [3] currently having anal sex with a man who is HIV-positive

Overall, nearly all participants reported having heard of oral PrEP, while just under a third were aware of injectable PrEP (cabotegravir; see Table 2). After providing basic information about both types, participants expressed a higher average level of interest in oral PrEP compared to injectable PrEP (t = 6.00, p < .001). Only 16.8% of participants who reported being uninterested (very or somewhat) or neutral about oral PrEP reported being interested (somewhat or very) in injectable PrEP (see Fig. 1). 47% of participants expressed lower interest in injectable relative to oral PrEP, 38% of participants’ ratings stayed the same, and only 15% increased. This same pattern also emerged for intentions to start oral and injectable PrEP. Participants also reported higher intentions to use oral PrEP compared to injectable PrEP (t = 4.50, p < .001). 38% of participants said they somewhat/very much intended to start oral PrEP in the near future, versus only 24% for injectable PrEP. Participants expressed about 16% higher interest in injectable cabotegravir if they could use this product at home, relative to being required to receive injections at a clinic.

Table 2.

Summary statistics of awareness, interest, and intentions to use oral vs. injectable PrEP

Variable	Oral PrEP	Injectable PrEP
	Mean (SD) or N (%)	Mean (SD) or N (%)
Awareness (Yes/No)	318 (97.3)	102 (31.2)
Interest (overall mean)	3.4 (1.3)	2.7 (1.5)
Not at all	39 (11.9)	101 (30.9)
Not very	53 (16.2)	64 (19.6)
Neutral	69 (21.1)	43 (13.2)
Somewhat	80 (24.5)	68 (20.8)
Very	86 (26.3)	51 (15.6)
Intentions (overall mean)	2.9 (1.4)	2.4 (1.4)
Not at all	67 (20.5)	116 (35.5)
Not very	69 (21.1)	71 (21.7)
Neutral	68 (20.8)	59 (18.0)
Somewhat	69 (21.1)	44 (13.5)
Very	54 (16.5)	37 (11.3)

Fig. 1.

Percentage of participants reporting interest in oral PrEP and injectable PrEP (cabotegravir) in each category

The barriers identified by participants were similar for both oral and injectable PrEP (see Table 3). Almost all participants reported at least some concern about both side effects and long-term health effects of both products, but surprisingly, slightly fewer reported these concerns for injectable compared to oral PrEP. Likewise, a similar percentage of participants also noted concerns that both products would not achieve high levels of protection against HIV, despite having presented information about efficacy in the survey. However, a sizable majority endorsed having a unique set of concerns about injectable PrEP. Namely, many were concerned about HIV medications not working if they used injectable PrEP, and about 17% more participants reported being concerned about the frequency of clinic visits required for injections versus oral PrEP. A majority also expressed concern about being able to stick to the required frequency of clinic visits for injections as strictly as they would be encouraged to.

Table 3.

Barriers to oral and injectable PrEP

Oral PrEP			Injectable PrEP
Concern	%	Avg. rating1	Concern	%	Avg. rating1
Potentially experiencing side effects	90.0	1.9 (0.8)	Potentially experiencing side effects	83.2	1.8 (0.8)
Potentially having long-term health effects	82.3	2.1 (0.8)	Potentially having long-term health effects	79.9	2.0 (0.8)
Not protecting me against HIV infection completely	65.5	1.7 (0.8)	If I do become HIV+, medicines not working because of PrEP injections	74.4	2.0 (0.8)
If I do become HIV+, medicines not working because I was taking PrEP	61.0	1.8 (0.8)	Having to go in for a clinic visit every 2 months	72.3	1.8 (0.8)
Having to go in for a clinic visit every 3 months	55.8	1.6 (0.7)	Making it to a clinic visit as strictly as I’d need to	72.0	1.8 (0.8)
Not being at high enough risk for HIV to warrant taking PrEP	54.0	1.9 (0.8)	Not protecting me against HIV infection completely	66.1	1.7 (0.8)
Making me more willing to have anal sex without a condom	50.4	1.6 (0.7)	Not being at high enough risk for HIV to warrant PrEP injections	55.7	1.9 (0.8)
Having to take a pill every day	49.0	1.6 (0.7)	Potentially experiencing pain from injections	54.0	1.5 (0.7)
Having to find a new doctor & make an appointment to get PrEP	46.9	1.5 (0.7)	I don’t like needles	48.4	1.7 (0.8)
Making my partner(s) expect me to have anal sex without a condom	43.1	1.7 (0.7)	Having to find a new doctor & make an appointment to get PrEP injections	46.3	1.5 (0.7)
Having to get the medication filled & pick it up	39.2	1.4 (0.6)	Making me more willing to have anal sex without a condom	45.4	1.5 (0.7)
Having to talk to my doctor about my sex life	33.9	1.5 (0.6)	Having to take pills every day for 30 days or get an injection for two straight months	41.6	1.5 (0.7)
People seeing my pills & wanting to know why I’m taking it	22.7	1.5 (0.7)	Making my partner(s) expect me to have anal sex without a condom	36.9	1.6 (0.7)
Having to take PrEP means I’m putting myself at risk for HIV	26.6	1.5 (0.7)	Having to talk to my doctor about my sex life	31.9	1.4 (0.6)
People seeing my pills & thinking I have HIV	22.7	1.5 (0.7)	Having to take PrEP means I’m putting myself at risk for HIV	25.7	1.5 (0.7)
My friends/acquaintances didn’t have good experiences with PrEP	21.2	1.4 (0.7)	People seeing me going to the doctor for PrEP injections and thinking I have HIV	18.0	1.5 (0.7)
Being ashamed to take PrEP	19.2	1.4 (0.7)	Being ashamed to get PrEP injections	18.0	1.4 (0.7)
			People seeing me going to the doctor to get PrEP injections & wanting to know why	16.8	1.4 (0.7)

Note.

¹Reflects average rating among participants who reported each barrier being at least a minor [1] concern, on a scale of 0 (not a concern) to 3 (major concern, meaning I would not take PrEP pills because of this issue)

Bivariate correlations in individual-level predictors are presented in Table 4, and results of the full regression models are presented in Table 5. Unsurprisingly, interest and intentions for both oral PrEP and cabotegravir were significantly lower among those who did not meet one of the three PrEP eligibility criteria noted in the 2017 CDC PrEP guidelines [5]. In models focusing on oral PrEP, HIV risk perception and PrEP descriptive norms were positively associated with both interest in and intentions to start oral PrEP, with risk perceptions accounting for about 12% of the variance in interest and 15% of intentions, and norms accounting for about 2% of interest and 3% of intentions. PrEP stigma was negatively associated with both interest and intentions for oral PrEP, accounting for about 7% of the variance in interest and 4% of intentions. Many of the same predictors were also significantly associated with injectable cabotegravir interest and intentions. In models focusing on injectable PrEP, HIV risk perceptions and descriptive norms were positively associated with both interest and intentions, with risk perceptions accounting for about 10% of the variance in interest and 11% of intentions, and descriptive norms accounting for about 3% of interest and 4% of intentions. PrEP stigma was also similarly negatively associated with injectable PrEP interest and intentions, accounting for about 3% of these outcomes, respectively. However, two predictors were uniquely associated with injectable PrEP interest/intentions relative to oral PrEP: PrEP self-efficacy and racial/ethnic minority status. PrEP self-efficacy was positively associated with only interest, accounting for about 1% of the overall variance. However, minority status was positively associated with both injectable PrEP outcomes, suggesting that some racial/ethnic minority GBM may be more interested in injectable cabotegravir than White GBM.

Table 4.

Correlation matrix of variables included in regression models

Variable	1	2	3	4	5	6	7	8	9
1. HIV risk perception
2. HIV severity	0.30*
3. PrEP knowledge	0.01	−0.04*
4. PrEP self-efficacy	−0.24*	−0.16*	0.20*
5. PrEP descriptive norms	0.12*	−0.10*	−0.01	0.17*
6. PrEP injunctive norms	−0.11*	−0.15*	0.21*	0.41*	0.21*
7. PrEP attitudes	−0.15*	−0.21*	0.26*	0.39*	0.08*	0.52*
8. PrEP stigma	0.06*	0.17*	−0.21*	−0.43*	−0.19*	−0.63*	−0.69*
9. Importance of HIV	0.14*	0.35*	0.04*	0.12*	0.11*	0.14*	0.15*	−0.16*
10. Medical mistrust	0.12*	0.13*	−0.05*	−0.16*	−0.01	−0.16*	−0.22*	0.32*	−0.03

Note.

¹All variables dummy-coded, comparing identified group versus all others.

^*p < .05

Table 5.

Linear regression of interest and intentions to seek oral and injectable PrEP (N = 327)

	Interest in PrEP
	Oral					Injectable
Variable	β	SE	p	95% CI	ω 2	β	SE	p	95% CI	ω 2
Minority status1	0.03	0.14	0.807	−0.24,0.30	0.00	0.30	0.15	0.043	0.99–1.11	0.01
Medical mistrust	0.10	0.08	0.256	−0.07,0.26	0.00	−0.15	0.09	0.103	−0.34,0.03	0.00
Medical mistrust x minority status	−0.06	0.13	0.657	−0.32,0.20	0.00	0.13	0.15	0.393	−0.16,0.42	0.00
History of PrEP Rx	0.03	0.18	0.859	−0.33,0.40	0.00	0.12	0.20	0.542	−0.28,0.53	0.00
CDC-ineligible2	−0.30	0.15	0.039	−0.59,−0.01	0.01	−0.46	0.16	0.005	−0.78,−0.14	0.02
HIV risk perception	0.16	0.02	< 0.001	0.11,0.20	0.12	0.16	0.03	< 0.001	0.11,0.21	0.10
HIV severity perceptions	0.01	0.07	0.895	−0.14,0.16	0.00	−0.07	0.08	0.398	−0.22–0.09	0.00
Perceived importance of HIV	0.10	0.09	0.286	−0.08,0.29	0.00	0.06	0.10	0.547	−0.14,0.27	0.00
PrEP knowledge	0.11	0.07	0.118	−0.03,0.24	0.00	0.03	0.07	0.666	−0.11,0.18	0.00
PrEP self-efficacy	−0.08	0.08	0.278	−0.23,0.07	0.00	−0.17	0.08	0.044	−0.33,−0.01	0.01
PrEP descriptive norms	0.20	0.07	0.004	0.07,0.33	0.02	0.24	0.08	< 0.001	0.09,0.39	0.03
PrEP injunctive norms	−0.08	0.08	0.332	−0.25,0.09	0.00	0.09	0.10	0.323	−0.09,0.28	0.00
PrEP stigma	−0.45	0.09	< 0.001	−0.62,−0.27	0.07	−0.33	0.10	0.001	−0.52,−0.13	0.03
	Intentions to start PrEP
Minority status1	0.21	0.14	0.131	−0.06,0.49	0.00	0.32	0.14	0.022	0.05–0.60	0.01
Medical mistrust	0.05	0.09	0.557	−0.12,0.22	0.00	−0.17	0.09	0.051	−0.34,0.01	0.00
Medical mistrust x minority status	−0.03	0.14	0.853	−0.29,0.24	0.00	0.02	0.14	0.864	−0.24,0.29	0.00
History of PrEP Rx	−0.13	0.19	0.480	−0.50,0.24	0.00	0.13	0.19	0.481	−0.24,0.50	0.00
CDC-ineligible2	−0.31	0.15	0.040	−0.60,−0.01	0.01	−0.45	0.15	0.003	−0.74,−0.15	0.02
HIV risk perception	0.18	0.02	< 0.001	0.13,0.23	0.15	0.15	0.02	< 0.001	0.10,0.20	0.11
HIV severity perceptions	−0.01	0.08	0.968	−0.15,0.15	0.00	−0.09	0.08	0.214	−0.25,0.06	0.00
Perceived importance of HIV	−0.06	0.10	0.539	−0.24,0.13	0.00	0.13	0.10	0.188	−0.06,0.32	0.00
PrEP knowledge	0.09	0.07	0.199	−0.05,0.22	0.00	−0.03	0.07	0.706	−0.16,0.11	0.00
PrEP self-efficacy	−0.03	0.08	0.684	−0.18,0.12	0.00	−0.13	0.08	0.087	−0.28,0.02	0.00
PrEP descriptive norms	0.21	0.07	0.003	0.07,0.35	0.03	0.25	0.07	< 0.001	0.11,0.39	0.04
PrEP injunctive norms	0.04	0.09	0.662	−0.13,0.21	0.00	0.03	0.09	0.704	−0.14,0.21	0.00
PrEP stigma	−0.36	0.09	< 0.001	−0.54,0.18	0.04	−0.29	0.09	0.002	−0.47,−0.11	0.03

Note.

¹Participants who reported non-White race and/or Hispanic/Latino ethnicity.

²Reflects participants who did not meet one or more of the 2017 CDC PrEP eligibility criteria

Discussion

In this study, we found that almost all GBM who were not on oral PrEP were aware of oral PrEP, but less than a third were aware of cabotegravir, one of the first injectable products approved by the US FDA for use as HIV PrEP. These results are not surprising, given that cabotegravir was approved only a few weeks prior to conducting this study. However, unexpectedly, our findings also showed that interest in injectable cabotegravir was considerably lower on average than oral PrEP, and very few GBM who reported low interest in oral PrEP expressed high interest in injectable PrEP. This contrasts with previous research on hypothetical injectable PrEP products, which have shown that about half of GBM may prefer injectable PrEP [16, 17]. No obvious sample characteristics explain differences between our results and these studies, except that this study provided participants with key details about a specific injectable product, whereas these other studies asked about participants’ preferences about a hypothetical injectable PrEP product. Overall, our results suggest that only about 17% of GBM who expressed low/neutral interest in oral PrEP reported high interest in cabotegravir. If these results reflect the preferences of the broader population of GBM, with overall PrEP uptake currently at 25% in the US [4], our findings would suggest only a further 1 in 5 of GBM who are not currently on PrEP might consider starting PrEP with this specific injectable product. Together, these results suggest that the availability of injectable cabotegravir could lead to modest increases in PrEP uptake. However, offering GBM an option of administering injectable cabotegravir themselves at home may be one way to boost interest.

Like the national data on PrEP uptake in US GBM so far, our results suggest that interest in cabotegravir also varies across racial/ethnic minority groups, but these differences appear to mostly be driven by smaller decreases in interest/intentions across oral versus injectable PrEP among racial/ethnic minority GBM compared to White GBM: Overall average interest in cabotegravir was about 23% lower than oral PrEP among White GBM, but only 15% lower among racial/ethnic minority GBM. However, about 23% of racial/ethnic minority participants who said they were somewhat/very uninterested or neutral about oral PrEP said they were somewhat/very interested in cabotegravir, versus 14% in White GBM. Although the absolute number of participants in specific racial/ethnic groups was low in this study, this similar pattern appeared to apply to both B/AA (21% vs. 14%) and H/L participants (24% vs. 14%). Although these findings suggest that, all else equal, racial/ethnic minority GBM may be more receptive to injectable PrEP compared to White GBM, interest in injectable PrEP was still generally lower than oral PrEP among racial/ethnic minority GBM, as well.

Our findings also suggest that many of the most common barriers GBM report are similar across oral and injectable PrEP. Almost all expressed at least some concern about the potential side effects and long-term health effects for both types of product, and a majority noted concerns about efficacy, which is consistent with past studies of oral PrEP among GBM and transgender women [35]. However, concerns about HIV treatment drugs being less effective because of PrEP were 13% higher for cabotegravir, which could be a reasonable concern given that two patients in the primary cabotegravir clinical trial (HTPN 083) developed resistance [42]. 17% more participants reported at least some concern about having to visit a clinic every two months with cabotegravir versus every three months for oral PrEP, 72% also expressed concern about being able to visit a clinic consistently, needles were a concern for nearly half, and the lead-in requirements were also a concern for about 40%. These are new barriers to PrEP that are unique to injectable cabotegravir and were relevant for large shares of GBM in our sample, supporting concerns that injectable cabotegravir may not be a silver bullet for facilitating more PrEP adoption, especially in those who have been less interested in PrEP generally so far.

Finally, the results of our regression models suggested that, if cost were not a factor, a number of important individual-level determinants were associated with interest and intentions to use both oral and injectable PrEP. As expected, those who did not meet any of the criteria for PrEP candidacy outlined in the CDC’s 2017 PrEP guidelines were less interested in both oral and injectable PrEP, likely because some of these participants were truly not at risk. However, we believed it was important to include all GBM regardless of risk because CDC’s 2021 PrEP guidelines, published the same month we began this study, include updated recommendations that encourage providers to talk to all sexually active, HIV-negative GBM or transgender person about PrEP [43]. Including this term in the model also allowed us to estimate the effects of the other predictors while holding actual risk constant. Beyond behavioral risk, individual-level determinants were remarkably consistent across interest and intentions to use both oral and injectable PrEP. Consistent with many past studies on oral PrEP [24, 25], low HIV risk perception was among the strongest predictors of interest/intentions for both oral and injectable PrEP, suggesting that GBM who believe they are at low risk for HIV express less interest in PrEP overall, regardless of actual risk. PrEP stigma was another strong and consistent predictor across all outcomes, which is also consistent with a growing body of research on oral PrEP [14, 28, 44–46]. Our results add that stigma still may be a significant barrier limiting the use of injectable PrEP, even though presumably it may be easier to conceal one’s use of injectable PrEP and fewer cues indicating one’s use would be available (e.g., keeping pill bottles in the home). Other negative beliefs, like perceptions that others look down on or judge PrEP users, likely limit interest in injectable PrEP in much the same way as oral PrEP. Finally, our findings also showed that descriptive norms were positively associated with all outcomes as well, which is consistent with past studies of oral PrEP [26, 27, 47, 48]. This suggests that GBM who believe that few of their friends/acquaintances and other GBM use PrEP are less interested in any type of PrEP themselves, but that the extent to which these groups and others approve of PrEP use may be less important. Lastly, one factor that was uniquely associated with interest in cabotegravir was PrEP self-efficacy, which are participants’ beliefs that they can carry out tasks that are needed to access and use the product. This, too, is consistent with past studies of oral PrEP [28, 31]. Our findings that self-efficacy was uniquely associated with injectable PrEP interest, however, suggests that one reason GBM may be less interested in injectable PrEP is that they may be less confident about their ability to find a provider who offers it, attend clinic visits as frequently as needed, and continue doing so for as long as they are at risk. We did not find evidence that participants’ perceptions about HIV’s severity, the relative importance of HIV relative to other priorities, PrEP knowledge, or PrEP injunctive norms were associated with interest or intentions to use any PrEP product. Together, these results can directly inform the design of individual-level interventions by identifying specific changeable factors that may be highest priority for these interventions to address. That nearly all of the same factors were significant predictors of interest/intentions for both oral and injectable PrEP also suggests that many interventions addressing these determinants could be similarly effective in promoting use of both products.

Limitations

Although this study has a number of important strengths, several limitations are also important to note. First, participants in this study could only be provided with a limited set of information about cabotegravir, a product that was entirely new to over two-thirds of participants. Since this study was conducted via online survey, it was only possible to concisely present basic information about the product, including its recommended dosing approach and schedule, risks, benefits, and side effects. As such, it could be that participants would have expressed more interest in cabotegravir if they had been given more detailed information or were able to have a two-way discussion that involved answering questions and specifically highlighting its personal benefits for them. Second, although over a third of this sample were racial/ethnic minority participants, the sample was still majority White. Recruiting a sample with a larger percentage of racial/ethnic minority participants would have helped add confidence in some results, especially those around differences in PrEP interest across racial/ethnic groups. Third, this study focused on GBM who were not currently using PrEP, and while a portion had some experience on PrEP, the vast majority had been encouraged to use PrEP at least once and had so far elected not to use it. It is possible that interest may be higher among GBM who currently use PrEP, and that doing so may be key to ensuring optimal coverage of risk over time. It is also possible that interest may be higher among GBM who have not yet been encouraged to consider PrEP. More research is needed to clarify these questions. Finally, participants in this study were primarily recruited online, and so these findings may only generalize to GBM who use the internet.

Conclusions

In summary, this study found that, among GBM who were not currently on PrEP, interest in using cabotegravir as injectable PrEP was considerably lower than oral PrEP. However, slightly more racial/ethnic minority GBM who expressed low interest in oral PrEP reported high interest in injectable cabotegravir when compared to White GBM, so the availability of injectable cabotegravir may have some promise for improving overall PrEP uptake somewhat among racial/ethnic minority GBM. Top barriers to PrEP use were similar across both oral and injectable PrEP, but some unique barriers for injectable PrEP, including having to attend a clinic visit every 2 months, needing adhere to these visits strictly, and fear/pain related to injections, were a concern among significant numbers of participants. Finally, a number of individual-level factors were associated with interest and intentions to use both oral and injectable PrEP, including HIV risk perceptions, PrEP descriptive norms, and PrEP stigma. Although significant predictors were very similar across both types of PrEP products, PrEP self-efficacy was a unique predictor of interest in injectable cabotegravir. These results can directly inform interventions designed to encourage use of PrEP products.

Data Availability

Data are available upon request from the authors.