Preferences for Starting Daily, On-Demand, and Long-Acting Injectable HIV Preexposure Prophylaxis Among Men Who Have Sex With Men in the United States (2021-2022): Nationwide Online Cross-Sectional Study

Introduction

Daily oral (DO) HIV preexposure prophylaxis (PrEP) effectively prevents HIV acquisition when taken as directed [1]; however, PrEP initiation and adherence are still low among men who have sex with men (MSM) in the United States [2]. Also, many studies report that MSM who start PrEP have low adherence to DO pills [3,4] which results in less protection from HIV acquisition, as adherence is a key factor for effective protection of PrEP [5].

Alternative PrEP options, such as long-acting (LA) injectable and oral on-demand PrEP, may improve PrEP uptake. LA PrEP is approved by the US Food and Drug Administration [6] and has proven to be more effective than DO PrEP in preventing HIV acquisition among MSM [7]. LA PrEP only necessitates a single injection every 2 months, which may result in fewer concerns about adherence among MSM who initiate PrEP or switch from DO to LA PrEP, ultimately leading to better effectiveness [8,9]. LA PrEP could also potentially provide more confidentiality than daily oral PrEP and reduce stigma among MSM [10]. The use of on-demand PrEP, in which individuals align their pill-taking schedule with periods of sexual activity, is reported to decrease HIV transmission risk, despite that it is not a US Food and Drug Administration–approved or recommended dosing schedule by the Centers for Disease Control and Prevention [11-13]. On-demand PrEP could potentially enhance PrEP uptake and adherence, particularly for MSM who are worried about side effects or who have difficulty consistently taking daily pills or prefer to take fewer pills overall [14-17]. These PrEP options offer an opportunity to increase PrEP initiation and adherence. However, there is limited information about the PrEP option preferences of MSM in practical scenarios where they may choose or switch between various PrEP options. Understanding MSM’s preferences and ranking of PrEP options could inform health care provider–patient discussions in clinical settings.

Some sociobehavioral factors could be associated with relative preferences of a particular PrEP regimen among MSM. For example, racial differences in willingness to use LA PrEP were previously reported [18,19]. Also, young age [20], geographical residence [21], and health insurance coverage [22] might shape relative preferences for PrEP initiation among MSM. Behavioral factors, such as the number of partners, unprotected sex [23], history of sexually transmitted infections [24], and illicit drug use [25] could affect patients’ preferences for a particular PrEP regimen. However, little is known about how these sociodemographic and behavioral factors impact patient PrEP option preferences. We aimed to examine relative preferences for starting PrEP regimens and their associations with sociodemographic and behavioral characteristics among a US nationwide online convenience sample of MSM.

Methods

Study Design and Analytic Sample

The American Men’s Internet Survey (AMIS) collects data annually from MSM via a self-administered online survey. The AMIS investigation team recruits participants through English and Spanish-language banner advertisements placed on websites and social networking sites used by MSM as well as by email blasts on LGBTQ+ (lesbian, gay, bisexual, transgender, queer/questioning, and others)-specific email listservs. The advertisements are displayed online based on self-reported demographic information related to online profiles. Participants who click an advertisement or use an email survey link are taken to the eligibility screener. Participants who are eligible for the survey are then taken to the online informed consent page. After participants provide consent, they enter responses to survey questions directly into their own computer, tablet, or smartphone, via a web interface. The online survey includes questions on demographics, sexual behaviors, substance use, HIV and STI testing and diagnosis, and use of HIV prevention services [26,27]. For this analysis, we used data from the 2021 AMIS data collection cycle which were collected from September 2021 to February 2022. Participants were eligible to participate in AMIS if they were aged 15 years or older, were assigned male sex at birth and reported current male sex identity, resided in the United States and provided a US ZIP code, and reported having ever had oral or anal sex or both with a male partner at least once or identified as gay or bisexual if they were in the age group of 15-17 years. For this analysis, additional eligibility criteria were having had oral or anal sex with another man in the past 12 months and no self-reported prior HIV diagnosis.

Outcome Measures

Participants were first given a brief description of each PrEP option: DO, on-demand, and LA PrEP (Table S1 in Multimedia Appendix 1). To determine willingness to start each PrEP option, participants were asked “If [PrEP option] were available from your local doctor and you could access it for free, would you go to your doctor in the next month to start [PrEP option]?” Based on the responses, willingness to use the PrEP option was grouped as “willing to start PrEP option” and “not willing to use PrEP option or not sure.” Current DO PrEP users among the analytic sample were determined by asking “Are you currently taking PrEP?” Participants who currently use DO PrEP were asked about their willingness to switch to on-demand or LA PrEP by asking “If [PrEP option] were available from your local doctor and you could access it for free, would you go to your doctor in the next month to start [PrEP option]?” Participants who had discontinued DO PrEP in the past 12 months (ie, those who used DO PrEP in the past 12 months but were not current users, n=261) were not asked willingness questions. If participants were willing to start multiple PrEP options, they were then asked to rank the PrEP options by preference.

Covariate Measures

Sociodemographic characteristics were age (15-24, 25-29, 30-39, and 40 years and older), race or ethnicity (Non-Hispanic or Latino Black, Hispanic or Latino, Non-Hispanic or Latino White, and other or multiple racial groups), health insurance type (private, public, other or multiple insurances, and no insurance), county of residence urbanicity based on the National Center for Health Statistics urban-rural classification scheme for counties (large central metro, large fringe metro, medium metro, small metro, micropolitan, and noncore) [28], and census region (Northeast, Midwest, South, and West).

Behavioral characteristics were self-reported for the past 12 months for condomless anal sex with a male partner (yes or no), number of male sex partners (one, or two or more), sexually transmitted infection (STI) diagnosis (yes or no), marijuana use (yes or no), and illicit drug use other than marijuana (yes or no).

To determine participants’ prior awareness of PrEP options, they were asked “Before today, have you ever heard of [PrEP option]?” If participants were current oral PrEP users, they were asked about their prescription medication brand (Truvada or Descovy), how many doses of oral PrEP they took in the last 30 days (15 or less doses, 16-29 doses, and 30 doses), and how many months in a row they have been taking oral PrEP (less than 2 months, 2-6 months, 7-12 months, and 12 or more months) to describe PrEP use characteristics.

Statistical Analysis

We described the sociodemographic, behavioral, and PrEP use characteristics in the analytic sample. We report the distribution of willingness to start each PrEP option for participants who were or were not using PrEP at the time they completed the survey, overall and by participant characteristics. We examined the distribution of first preference for the PrEP option among participants willing to use multiple PrEP options. We used log-binomial regression models to examine the association of each characteristic with the willingness to start each PrEP option, using unadjusted and adjusted prevalence ratios. To estimate the adjusted prevalence ratios and 95% CI, we included the sociodemographic (age, race or ethnicity, health insurance, urbanicity, and census region), behavioral (condomless anal sex, number of male sex partners, STI diagnosis, marijuana use, and other illicit drug use), and PrEP option awareness variables in multivariable log-binomial regression models. We retained these variables in the multivariable models since they were shown to have associations with PrEP willingness in previous literature, despite the lack of significance in this univariate analysis. We followed the same approach for those who were currently using PrEP; however, we additionally included PrEP use characteristics (current PrEP prescription medication, number of PrEP doses taken in the last 30 days, and PrEP duration) in our multivariable models to estimate the adjusted prevalence ratios and 95% CIs among this group of participants.

LA PrEP was the first preference for PrEP modality among participants who were not currently on PrEP and were willing to use multiple PrEP regimens. To better understand this finding, we examined characteristics associated with ranking LA PrEP as the first preference versus DO or on-demand PrEP. Among participants who were not currently using oral PrEP, we combined those who reported willingness to use (1) DO and LA PrEP; (2) on-demand and LA PrEP; and (3) DO, on-demand, and LA PrEP and created univariate and multivariable log-binomial regression models to estimate unadjusted and adjusted PRs and 95% CIs. We also examined the characteristics associated with ranking LA PrEP as a first preference versus on-demand PrEP among current DO PrEP users. We used univariate log-binomial regression because a multivariable modeling approach was not possible due to small cell numbers and the distribution of covariates in this group. We conducted the data analysis with SAS version 9.4 (SAS Institute).

Ethical Considerations

The core AMIS study was reviewed and approved by Emory University’s human subjects research ethics board (IRB00047676). Informed consent is collected from participants before their participation in the core AMIS study. Eligible participants were shown a consent form to review and asked whether they wished to participate in the AMIS survey. Participants who provided informed consent were taken to a screen where they could complete the online AMIS survey. Those who did not consent were taken to a screen thanking them for their interest and no further information was collected. Emory University has determined that this substudy does not comprise human subjects research because it only consists of secondary analysis. Therefore, no informed consent process is required for the substudy. Participants did not receive any compensation for their participation. The study data are nonidentifiable and do not pose a risk of loss of confidentiality.

Results

In the analytic sample (N=7760), 45.2% (n=3511) of participants were aged 40 years and older, 14% (n=1086) were Hispanic or Latino, 10% (n=784) were non-Hispanic or Latino Black, and 66.6% (n=5114) were non-Hispanic or Latino White individuals (Table S2 in Multimedia Appendix 1). Most participants had private health insurance, had a college degree or postgraduate education, and were employed with full-time wages. Approximately 45% (n=3467) of the participants reported residing in large central metro areas; 38% (n=2931) resided in the South. 74% (n=5774) of the participants had condomless anal sex and 78% (n=6061) had two or more male sex partners in the past 12 months. 34% of the participants used DO PrEP in the past 12 months (n=2652). Among those who used PrEP in the past 12 months, 90% (n=2379) were currently using DO PrEP. Among those who did not use PrEP in the past 12 months (n=5108, 66%, 7% (n=334) reported ever taking daily oral PrEP.

Among participants who did not use any PrEP in the past 12 months (n=5108, 66%), 54% (n=2794) reported willingness to start at least 1 PrEP option. Overall, the highest willingness was reported for on-demand PrEP (n=2445, 44%), followed by DO PrEP (n=2174, 43%) and LA PrEP (n=1482, 29%). However, 40.6% (n=2075) of participants showed interest in starting multiple PrEP options and LA PrEP was ranked first among those interested in multiple options (Figure 1).

In multivariable modeling, willingness to start LA PrEP was significantly higher among Hispanic or Latino participants compared to White participants (Table 1). Participants who had an STI diagnosis, had condomless anal sex, had 2 or more male sex partners, and used illicit drugs other than marijuana in the past 12 months were significantly more willing to start LA PrEP. Prior awareness of LA PrEP was also associated with increased willingness to start LA PrEP.

In multivariable modeling, willingness to start on-demand PrEP was significantly higher among the youngest age group (15-24 years) compared to those who are 40 years and older and Hispanic or Latino participants compared to White participants (Table 2).

Participants who had an STI diagnosis, had condomless anal sex, and had 2 or more male sex partners in the past 12 months were significantly more willing to start on-demand PrEP. Characteristics associated with ranking LA PrEP as a first option to start PrEP versus DO or on-demand PrEP were region of residence (residing in the West vs Northeast), report of STI diagnosis (vs no diagnosis), report of illicit drug use (vs no drug use) and prior awareness of LA PrEP (vs not being aware of LA PrEP; Figure 2; and Table S3 in Multimedia Appendix 1).

Among participants who were currently using DO PrEP (n=2379, 31% of the whole analytic sample), 58% (n=1386) were willing to switch to on-demand or LA PrEP. Willingness to switch to LA PrEP (n=1121, 47.1%) was higher than willingness to switch to on-demand PrEP (n=756, 31.7%). LA PrEP was ranked first among participants who were open to switching to both options (Figure 3).

In multivariate modeling, there were no meaningful associations with sociodemographic characteristics and willingness to switch to LA PrEP among those currently using DO PrEP, except that willingness was slightly increased in Hispanic or Latino participants compared to White participants (Table 3).

Willingness to switch to LA PrEP was higher among those who used illicit drugs other than marijuana in the past 12 months and among those who had previously heard of LA PrEP. There were no meaningful associations between current PrEP prescription medication and PrEP duration with the willingness to switch to LA PrEP. Those who took 15 or less doses of DO PrEP and who took 16-29 doses in the last 30 days were more willing to switch to LA PrEP, compared to those who took all 30 doses.

In multivariable modeling, willingness to switch to on-demand PrEP was significantly higher among those who lived in rural micropolitan or noncore areas compared to those who lived in large central metro areas (Table 4).

‎

‎

‎

Willingness to switch to on-demand PrEP was also significantly higher among those who had been using PrEP for less than 2 months compared to those who had used PrEP for more than 12 months. Despite a strong association between willingness to switch to on-demand PrEP and fewer doses of DO PrEP taken in the last 30 days in bivariate analyses, this association was not stable in multivariable analyses. Among 150 participants who took 15 or less doses in the last 30 days, 42% intended to take PrEP only when they had sex, and 10% intended to take PrEP on some other schedule.

Characteristics associated with ranking LA PrEP as a first option to switch to versus on-demand PrEP were other or multiple races (vs White, non-Hispanic or Latino), illicit drug use, and prior awareness of LA PrEP (Table S4 in Multimedia Appendix 1).

Discussion

Principal Findings

There was substantial interest in starting LA and on-demand PrEP among US MSM in our nationwide study. Among MSM who were not currently using DO PrEP, the highest willingness was reported for on-demand PrEP. However, a substantial proportion of MSM were willing to start multiple PrEP options, and LA PrEP was ranked first when those willing to start multiple types of PrEP were asked to make a single choice. More than half of US MSM in our study who were currently using DO PrEP expressed interest in switching to LA PrEP or on-demand PrEP; when both options were selected, they expressed preference toward LA PrEP versus on-demand PrEP. Multiple demographic factors, risk behaviors, and prior PrEP experiences were associated with these preferences.

Previous research has highlighted the potential benefits in coverage or uptake of adding LA PrEP as an additional PrEP option because many current PrEP users reported a preference toward LA PrEP over DO PrEP [29,30]. Additionally, prior studies have shown the potential for LA PrEP to increase PrEP uptake overall, as many individuals who could benefit from some form of PrEP preferred LA PrEP to DO PrEP [31,32]. Our findings build upon this existing literature demonstrate that LA PrEP is the highest-ranked option when participants are open to using multiple PrEP options or are willing to switch from daily oral PrEP.

Our findings also highlight that PrEP option preferences are not uniform for US MSM and vary by key characteristics. Hispanic or Latino MSM showed a significantly higher willingness to start or switch to LA PrEP and on-demand PrEP compared to White MSM. Younger MSM expressed a significantly higher willingness to start on-demand PrEP versus older MSM. These results align with previous studies that reported high PrEP willingness among Hispanic participants [18,33] and young MSM [34]. However, despite high PrEP willingness, PrEP uptake and adherence remain suboptimal in young and Hispanic MSM [2,35,36]. In particular, DO PrEP adherence was shown to be low among young MSM [37]. One of the difficulties that MSM often face is the requirement to consistently take their daily medication, which can be challenging to remember [38,39]. Challenges with adherence to daily pills might explain the higher willingness toward on-demand PrEP among young MSM in our study. Nevertheless, recent studies report that LA PrEP could expand access to PrEP for young MSM who are wary of adhering to DO or on-demand PrEP [39].

Willingness to start on-demand or LA PrEP was also associated with behavioral risk factors in our study, such as having a recent STI diagnosis, having two or more male sex partners, and having condomless anal sex. Recent studies estimating the benefits of on-demand PrEP suggested that MSM who have poor adherence to oral PrEP can still achieve similar effectiveness in reducing HIV acquisition by using on-demand PrEP [11,17]. Our findings suggest that MSM who have sexual behavioral risks may particularly benefit from on-demand PrEP. Our findings also align with previous studies suggesting that LA PrEP may be a particularly preferable option among MSM at higher risk of HIV acquisition [30]. Additionally, willingness to use LA PrEP was higher among participants who used illicit drugs versus those who did not. In parallel to these findings, acceptance of LA PrEP was previously shown to be high among people who inject drugs [25].

Among current DO PrEP users, a higher willingness to switch to LA PrEP was associated with a lower adherence to the current daily regimen. Participants who did not take the approved number of doses of DO PrEP in the last 30 days (used 15 or less doses or used 16-29 doses) were more willing to switch to LA PrEP compared to those who fully adhered to the DO protocol (30 doses) in the last 30 days. Willingness to switch from DO PrEP may be related to adherence challenges [40,41]. LA PrEP use includes less frequent dosing compared to DO PrEP. DO PrEP users who are aware of their suboptimal adherence might be more willing to switch to LA PrEP [42]. LA PrEP might facilitate increased adherence among MSM and might potentially increase PrEP uptake.

Among current PrEP users, there was also a clear association between willingness to switch to on-demand PrEP and taking fewer doses of DO PrEP in the past 30 days. It is worth noting that out of the 150 individuals who took 15 or less doses within the past 30 days, 42% stated that they only planned to use PrEP during sexual encounters, whereas 10% indicated that they intended to take it on a different schedule. It is possible that some participants might have intended to use PrEP when necessary, even though they were prescribed DO PrEP. Although this strategy is not common among MSM, those who were taking 15 or less doses of DO PrEP might have believed they were already using on-demand PrEP and may have seen a need to switch to an on-demand regimen [16,43]. In our analysis, this could have resulted in a conflation of participants who were DO PrEP users and who were using DO PrEP only when they needed it.

Our findings also suggest that participants residing in rural micropolitan or noncore areas have a higher willingness to switch to on-demand PrEP than those living in large central metro areas. This might be related to challenges in accessing PrEP from rural pharmacies or clinics, making the intermittent use of medication specifically during sexual encounters more appealing than consistently maintaining a supply [44,45]. Also, higher levels of stigma are associated with using HIV prevention medications in rural areas compared to more urban areas [46,47]. The negative perception toward individuals taking PrEP may be higher in rural areas due to the close-knit social dynamics prevalent in small towns. In these areas, people tend to have higher familiarity with each other’s activities, increasing public knowledge of potentially stigmatized behaviors, which can lead to discrimination from various groups, including community members and health care providers [48,49]. HIV and PrEP stigma often act as barriers to PrEP uptake and persistence [10]. Previous research has highlighted that MSM have concerns about the possibility of other individuals, including sexual partners, becoming aware of their PrEP use if they are taking daily regimens [50]. Furthermore, MSM who experienced intimate partner violence, where their behaviors were monitored by their partners, ranked daily oral PrEP lower than other PrEP options [51]. These findings suggest that switching to the LA PrEP option could provide more confidentiality than taking pills daily; this might help to increase PrEP uptake overall.

Prior LA PrEP awareness was another factor associated with a higher willingness to start or switch to LA PrEP options in this study. Willingness to use PrEP was shown to increase with improved PrEP awareness in prior studies including among young MSM [39,52-54]. These results highlight the significance of health education, public education campaigns via social media, and open, informative dialogue between health care providers and patients regarding the newer PrEP options.

We observed regional variations in the PrEP rankings showing that participants living in the West and South rated LA PrEP as their preferred first option for starting PrEP, compared to those in the Northeast. Considering that the highest rates of HIV [55] and the lowest PrEP uptake levels are being reported in the South [2], a high willingness to use LA PrEP in this region may be promising to create opportunities to increase PrEP uptake. However, although these findings offer insights into regional differences in PrEP option preferences, it is important to acknowledge that we were unable to differentiate the disparities between demographically and socioeconomically varied states within these regions which may have provided further insights.

This study has some limitations. Data were collected via online recruitment using a convenience sampling approach; therefore, our results have limited external generalizability. In addition, because we recruited from online sex-seeking apps, our study sample may have a higher proportion of men who may benefit from PrEP [56]. Also, the survey consisted of only self-reported behaviors, so our data are subject to misclassification: it is possible that less socially desirable responses may be underreported by participants even though the survey was anonymous and self-administered [57]. Although we explored urban or rural differences in PrEP preferences, other geographical factors, such as Medicaid expansion status or PrEP provider density could be important determinants in PrEP preferences and warrant investigation in future studies [58,59]. Finally, we asked PrEP preferences of participants in a hypothetical scenario, where PrEP options were available free of charge. However, costs associated with PrEP might create PrEP access barriers in real-world settings.

Conclusions

Our results highlight the substantial interest among US MSM in starting or switching to on-demand and LA PrEP options. Our findings also indicate that those at higher risk for HIV acquisition or who may be struggling with DO PrEP adherence are substantially more interested in these PrEP options. Increased and varied PrEP options are even more likely to increase overall community PrEP uptake if PrEP options offered by clinicians are informed by understanding differences in preferences among US MSM with different demographic backgrounds, risk profiles, and prior PrEP experiences.