The Use of Daily and On-Demand Oral Pre-Exposure Prophylaxis Dosing Strategies Among Young Adult Gay, Bisexual and Other Men who have Sex with Men Enrolled in an mHealth Adherence Intervention

Abstract

Pre-exposure prophylaxis (PrEP) is a highly effective HIV prevention option for gay, bisexual and other men who have sex with men (GBMSM). However, with newer PrEP options, a greater understanding of whether and why GBMSM switch dosing strategies is needed to inform clinical practice and research. We assessed the dosing strategies (daily or on-demand) of GBMSM enrolled in an mHealth PrEP adherence pilot intervention at four timepoints over approximately 10 months. Among GBMSM with complete data (n = 66), a consistent daily dosing strategy was used by most (73%) participants across all time points, while on-demand PrEP was used at least once during the study period by 27% of participants. A higher percentage of on-demand PrEP users self-reported as Asian/Pacific Islander and had less positive attitudes toward PrEP, adjusting for key sociodemographic variables and intervention arm. Daily PrEP users reported high numbers of sexual partners, and the primary reason that they would switch to on-demand PrEP is reduced sexual activity. At the final assessment, 75% of participants were taking daily PrEP, of whom 27% reported that they would like to switch to another option, including on-demand and long-acting injectable PrEP. While findings were largely descriptive, they showed that switches in PrEP dosing strategies are relatively common and PrEP strategy choice may vary across racial and ethnic groups.

Introduction

Pre-exposure prophylaxis (PrEP) provides a high level of protection (> 90%) against HIV acquisition for gay, bisexual, and other men who have sex with men (GBMSM) when taken correctly and consistently [1, 2]. GBMSM accounted for 62.9% of population living with HIV, and 70% of new HIV infections, in the United States (US) in 2019, with a disproportionate representation among Black and Latino GBMSM [3]. Despite its effectiveness for HIV prevention, PrEP uptake and adherence remain suboptimal among GBMSM [4, 5] to meet the Ending the HIV Epidemic in the US (EHE) goal of reducing new HIV infections by 90% by 2030 [6]. A study with a large US sample of GBMSM found that most participants (77.4%) had never used PrEP, and an additional 6% of participants had previously used, but were not currently on, PrEP [4]. More recently, data from the Centers for Disease Control and Prevention (CDC) showed that only one-quarter of the 1.2 million people in the US for whom PrEP is recommended in 2020 are taking the medication, and coverage among Black (9%) and Hispanic/Latino (16%) persons is substantially lower compared to their white counterparts (66%) [7]. Clearly, there is substantial room to optimize PrEP uptake among key at-risk communities, including GBMSM.

Contributing to the deficient uptake of PrEP may be concerns about daily oral PrEP, which may motivate some persons on PrEP to seek alternative formulations (i.e., long-acting injectable PrEP) and/or dosing strategies (on-demand oral PrEP). Injectable PrEP (i.e., cabotegravir) was approved in January 2022 after trials demonstrating that it may be more effective than daily oral PrEP for GBMSM and transgender women [8], however important questions about how to implement it remain [9]. On-demand PrEP, the focus of this study, is not currently FDA-authorized in the US, however it is recommended for GBMSM by the International Antiviral Society-USA Panel [10] as a reflection of the strong evidence for its effectiveness as a dosing strategy when two pills are taken 2–24 h before sex, one pill taken 24 after the first dose, and a final pill taken 24 h after the most recent dose [11]. A number of studies examined GBMSM preferences for different PrEP modalities. Among 761 young adult (18–29 years of age) GBMSM residing in California and recruited through geosocial networking applications, 265 (38.6%) were unwilling to try daily oral PrEP [12]. However, most (85.3%) of the sample was willing to try alternatives, including on-demand (74.3%) or injectable PrEP (73.6%). Another study of GBMSM residing in three cities (Atlanta, Detroit, and New York City) asked participants whether they were likely to use different PrEP options in the future other than daily oral PrEP (32% were currently PrEP users), with high endorsement for injection every 1–3 months (74%) and on-demand PrEP (67%) [13]. A recent study 147 GBMSM in New York City who were receiving sexual health services showed that 35% believed on-demand PrEP as a dosing strategy would be a good choice for them [14]. Belief that on-demand PrEP was a good option was predicted by participant report of taking more than 30 min to prepare for anal sex, a preference for taking PrEP only when it is needed, and the belief that using this strategy would decrease side effects [14].

While the studies described above suggest that GBMSM have positive attitudes toward non-daily PrEP options, when given the choice between using a daily or an on-demand PrEP dosing strategy in cohort studies conducted in Europe, most GBMSM initially chose a daily dosing strategy (73–77%) over an on-demand PrEP strategy (23–27%) [15–18]. In one of those studies [17], approximately one-third (cumulative probability = 34.2%) of GBMSM switched their PrEP dosing strategy one or more times, with somewhat more switches from daily to on-demand PrEP (n = 113 participants) than from on-demand to daily PrEP (n = 90 participants). An examination of PrEP users (n = 2338) in a large healthcare system in San Francisco showed that, as of August 2019, nearly 12% of patients on PrEP were prescribed an on-demand PrEP dosing strategy [19]. After three months, just over half (51%) of patients initially prescribed an on-demand strategy reported using that dosing strategy, while 19% switched to a daily dosing strategy, 15% used a combination of daily and on-demand, 4% used a different non-daily strategy, and 3% never started PrEP.

Studies on GBMSM have examined demographic and psychosocial correlates of switching PrEP dosing strategies over time. Switching from an on-demand to a daily dosing strategy has been associated with younger age, not reporting having sex with partners living with HIV, not engaging in chemsex (i.e., sex under the influence of substance use), and not self-reporting as sexually compulsive [17]. In contrast, switching from a daily to an on-demand PrEP dosing strategy was associated with younger age and sexual dissatisfaction [17]. Moreover, GBMSM report different concerns with either a daily or on-demand PrEP dosing strategy. In a study on GBMSM (n = 970) in Australia, those who are dissatisfied with daily PrEP dosing report concerns about drug toxicity, infrequent sex reducing the need for daily pills, dislike of pills and difficulty taking pills daily [20]. In the same study, those who were worried about using on-demand PrEP reported concerns about its effectiveness, taking it 2 h before sex, having spontaneous or unplanned sex, and about having sex too often which would require a daily regimen.

Overall, results of existing studies show that while GBMSM report relatively high interest in using an on-demand PrEP dosing strategy [13, 21, 22], there is limited uptake of on-demand PrEP when given a choice between this option and daily dosing [15–19]. However, most assessments of non-daily PrEP use in cohort studies have been conducted in Europe [15–18], and no studies to our knowledge have examined the use of daily and non-daily PrEP dosing strategies in the context of a PrEP adherence intervention. Because of their longitudinal design, intervention studies may offer potential insights into how GBMSM may switch between different PrEP dosing strategies over time to inform both real-world patterns of PrEP use and highlight potential challenges of measuring and addressing different PrEP dosing strategies in intervention contexts. To fill these gaps, the purpose of the current study was to:

1. Characterize patterns of PrEP dosing strategies among young (18–29 year old) US-based GBMSM enrolled in an mHealth PrEP adherence intervention.

2. Examine associations between sociodemographic and sexual behavior factors with using an exclusive daily PrEP strategy or using an on-demand dosing strategy at least once during the intervention period.

3. Better understand GBMSM’s attitudes toward the use of daily and on-demand PrEP dosing strategies, including intentions to stay on their current dosing pattern (daily or on-demand) or switch to a new formulation or dosing strategy (i.e., daily, on-demand, or injectable PrEP).

Methods

Participants

Data were drawn from four online surveys conducted as part of a longitudinal mHealth intervention, called PrEP iT!, study to improve PrEP adherence among 80 GBMSM between 18 and 29 years of age. The first three surveys (baseline, month 3, month 6) were conducted at 3-month intervals starting in February 2021 and were a part of the original study design. An additional fourth survey (launched in February 2022) was conducted to explore in greater detail participants’ attitudes toward using different PrEP options and was administered approximately three months after the month 6 survey, with the number of months ranging from 2.9 to 6.2 months. Eligibility criteria included: (a) having a prescription for PrEP by a healthcare provider; (b) being 18–29 years old; (c) assigned male at birth; (d) current identification as male; (e) having a gay, bisexual or other non-heterosexual identity, or has had sex with a man in the past year; (f) self-reported HIV-negative status; (g) having started or restarted PrEP in the past 30 days, or started PrEP more than 30 days ago and self-reported not always taking PrEP in the past 30 days; g) being English-speaking (as the intervention was built in English); (h) having the ability to meet with project staff over a 2–3 week period; and (i) having regular access to SMS and the internet, either through a mobile device, tablet computer, and/or laptop computer.

Study Procedures

PrEP iT! is a six month, two-arm pilot RCT grounded in the Information, Motivation, and Behavioral Skills (IMB) model [23–25] to evaluate the feasibility, acceptability, and preliminary impact of an mHealth intervention among 80 GBMSM who are taking PrEP and self-reported imperfect adherence at baseline. The PrEP iT! intervention has the following components: (a) PrEP adherence, appointment, and sexual activity self-monitoring with reinforcement messages; (b) IMB-related content addressing PrEP options, PrEP adherence, PrEP healthcare engagement, PrEP stigma, and sexual risk, as well as other related content; (c) optional text message reminders for PrEP adherence and healthcare appointments; and (d) an “Ask the Expert” feature where participants can send anonymous questions to the investigators, and answers were updated to the app and visible to all users. Importantly, while intervention content described both daily and on-demand PrEP strategies, participants were not encouraged or discouraged to prioritize one dosing strategy over the other.

After clicking on paid social media ads, potential participants were redirected to a brief online eligibility screener. Next, study staff contacted eligible participants to schedule a baseline study (primarily virtual using Zoom videoconferencing software, although two participants were in-person baseline visits), during which we confirmed participants’ eligibility, collected consent, and assisted participants to complete an online baseline survey. Participants were randomized to the intervention (PrEP iT! plus usual PrEP care) or control (usual PrEP care only) arm and assessed quarterly for up to six months (i.e., completed online surveys at baseline, month 3, and month 6). As noted earlier, a fourth survey (referred to from here forward as the “final assessment”) that was not a part of the original study plan was administered to better understand attitudes and behaviors around alternative PrEP dosing strategies compared to daily oral PrEP, including questions about their intentions to switch from their current regimen to a daily, on-demand, or long-acting injectable formulation. All study procedures were approved by San Diego State University Institutional Review Board (HS-2019–0291).

Measures

The sociodemographic and stigma measures described below were taken from the baseline survey, and the sexual behavior and PrEP dosing strategies variables were assessed at all time points.

Sociodemographic Characteristics.

Participants completed survey items assessing their age (in years), race/ethnicity (recategorized into Asian/Pacific Islander; Black/African American; Hispanic/Latino; non-Hispanic White), education level (recategorized as college degree vs. no college degree), current employment status, health insurance status (recategorized as own insurance, on parent or guardian insurance, or none), whether they had skipped meals in the past 3 months, outness to family members (recategorized as more out to more than half vs. half or less), outness to peers (recategorized as more out to more than half vs. half or less), and outness to medical providers (recategorized as talked about sexuality sometimes or openly with medical provider vs. less open).

PrEP Stigma and Positive Attitudes.

A 10-item measure [26] assessed participants’ PrEP stigma (PS; 7 items; e.g., “People who are on PrEP sleep around.”) and positive attitudes toward PrEP (PA; 3 items; e.g., “People who are on PrEP are making a smart decision to protect their health.”). Each item had a 1 to 5 response option, which were subsequently coded so that higher scores reflecting higher levels of stigma and less positive attitudes toward PrEP. The subscales showed acceptable internal validity among a sample of primarily (93%) young GBMSM, with a Cronbach’s alpha of 0.85 and 0.69 for the PS and PA subscales, respectively [26]. The Cronbach’s alpha for the current sample for the full scale was 0.70, and 0.71 and 0.47 for the PS and PA subscales, respectively.

Sexual Behaviors.

Participants were asked to report the number of male sexual partners they had in the past three months. In addition, they were asked to indicate whether they engaged in condomless sex with HIV-negative, HIV serostatus unknown, or HIV-positive male partners in the past three months (i.e., yes or no to each partner type). Reponses were recoded to reflect condomless sex with HIV serostatus unknown or partners living with HIV in the past 3 months (yes or no).

PrEP Dosing Strategy.

During the first three videoconference-based study visits (baseline, month 3, and month 6), participants reported to the study coordinator whether they were taking PrEP daily, taking it only around sex events (i.e., on-demand), or had discontinued taking PrEP. Similar items were asked on the final (i.e., fourth) assessment, however participants reported their response in an online survey.

Pros and Cons of Daily and On-Demand PrEP and Intentions to Switch to a Different PrEP Formulation or Strategy.

The variables described next were only included in the final assessment to explore in greater depth participants’ use of, and attitudes toward, different PrEP dosing options. Survey items focused primarily on comparing daily PrEP to on-demand PrEP dosing strategies as long-acting injectable PrEP had only been FDA-approved shortly before the survey was administered. As with earlier study visits, participants reported whether they were currently taking daily oral PrEP, on-demand oral PrEP, or had discontinued PrEP. Among those who discontinued PrEP, they were asked to report whether they were taking PrEP daily or on-demand just prior to discontinuation. Next, both current PrEP users and recently discontinued PrEP users reported the barriers they faced in taking PrEP using their chosen dosing strategy. Participants taking daily PrEP were asked to report the reasons why they would switch to an on-demand dosing strategy if they were to do so. Participants who are currently on, or last used, an on-demand PrEP dosing strategy were asked why they prefer that strategy to a daily dosing strategy. Finally, participants were asked if they intended to stay on their current dosing strategy and, if not, what they intended to switch to: daily, on-demand, or long-acting injectable PrEP, or “I want to switch but I’m not sure yet to what.”

Analytic Plan

Descriptive statistics (e.g., means and percentages) were calculated for participants’ sociodemographic, outness, attitudes toward PrEP, sexual behaviors, and daily and on-demand PrEP dosing strategy variables. A new variable was created to reflect whether a participant consistently used a daily PrEP dosing strategy throughout each of the four study visits (the “Daily PrEP Strategy Only” group), or whether an on-demand PrEP dosing strategy was used at one or more study visits during the study period (the “On-demand PrEP Strategy at least Once” group). Chi-squared and Fisher’s exact tests examined bivariate differences in sociodemographic, outness, and attitudes toward PrEP variables between PrEP strategy groups (n = 66 total participants). Next, variables that were significantly different between PrEP strategy groups in the bivariate analyses were included in separate multivariable logistic regression models to assess their odds with being in the on-demand group [1] rather than the daily only group (0). The first model adjusted for age and randomization arm, and the second model adjusted for race/ethnicity, age, and intervention arm to better understand the contributions of these factors and because of the relatively small sample size.

All available data from the full sample (n = 80) were used at each timepoint to compare sexual behavior among participants who used a daily or on-demand PrEP strategy, or discontinued PrEP. To normalize data, the number of sexual partners was capped at 20 at each three-month assessment, resulting in seven entries that were modified (two at month 3; two at month 6; and three at the final assessment). Wilcoxon-Mann-Whitney and Kruskal Wallis tests examined bivariate differences in the number of male sex partners in the past three months (continuous), and chi-squared and Fisher’s exact tests compared differences in condomless sex with HIV-unknown or positive partners in the past three months (dichotomous), between PrEP strategy groups (i.e., daily only or on-demand at least once during the study period) or who discontinued PrEP. Variables that were significantly associated with the number of male sex partners or condomless sex in the bivariate analyses were included in separate multivariable regression or logistic regression models, respectively, and adjusted for race/ethnicity, age, and intervention arm as described above.

Frequencies and percentages were used to assess reasons for switching, as well as intention to switch, to a different PrEP formulation or dosing strategy. All analyses were conducted using STATA/MP 17.0, with a pre-determined alpha level 0.10 given the exploratory nature of this study.

Results

Patterns of PrEP Use During Study Period

A majority of participants used a daily dosing strategy at each time point, including 85% (68/80) at baseline (15% on-demand), 83% (62/75) at month 3 (9% on-demand; 8% discontinued), 68% (50/74) at month 6 (17% on-demand; 15% discontinued), and 75% (54/72) at the final assessment (8% on-demand; 17% discontinued). Complete data on PrEP dosing strategy were available for 66 (among 80 total) participants at all timepoints; except for age, no other sociodemographic differences were found between those with complete and those who had incomplete data (n = 14). Participants with complete data were older (25.4 years) than those with incomplete data (23.7 years; t(78) = 2.21, p = .03).

Patterns of PrEP use for study participants (n = 66) with complete data at all four study timepoints are shown in Fig. 1. Overall, 19 different patterns of PrEP dosing were reported by study participants. Approximately 61% (n = 40) of participants used a daily PrEP dosing strategy for the duration of the study. An additional 8 participants (12%) only used a daily PrEP dosing strategy, but discontinued PrEP at one or more time points. The remaining 18 participants (27%) used an on-demand PrEP dosing strategy at one or more study timepoints, with the most common pattern among this group using an on-demand strategy at the beginning of the study period but switching to a daily strategy thereafter (n = 3). Only 1 participant reported using an on-demand PrEP dosing strategy throughout the entire study period.

Fig. 1.

Daily and on-demand PrEP dosing strategy and PrEP discontinuation for participants with complete data at all assessment points in the PrEP iT! intervention trial (n = 66). Note: Final Assessment conducted 3 to 4 months after month 6 assessment.

The Association Between Sociodemographic Characteristics, Outness, Attitudes Toward PrEP and PrEP Dosing Strategy

In the findings reported below, we compared participants with complete data across all study timepoints who consistently reported a current or most recent use of a daily PrEP dosing strategy at each study timepoint (n = 48) to those who reported using an on-demand dosing strategy on at least one study timepoint (n = 18).

Bivariate Analyses.

Sociodemographic characteristics of participants with complete data (n = 66) by PrEP dosing strategy are shown in Table 1. Overall, the mean age of participants was 25.4 years, and most held a college degree (78.5%), were currently employed (83.3%), owned their own health insurance (78.5%), and did not live with their family (81.8%). A small proportion (10.8%) of participants had skipped a meal in the past 3 months, suggesting that most were financially stable. Just over half of participants (51.5%) identified as a racial or ethnic minority, including 19.7% who were Asian/Pacific Islander, 15.2% Black, and 16.7% Hispanic/Latino. There was a statistically significant difference across racial and ethnic groups in the proportion of men who used an on-demand PrEP dosing strategy at least once; 61.5% of Asian/Pacific Islander men, 30.0% of Black men, 21.9% of non-Hispanic white participants used an on-demand PrEP strategy at least once during the study period, while no Hispanic/Latino participants did so (Fisher’s exact test = 0.005).

Table 1.

Sociodemographic variables by PrEP dosing strategy for participants with complete data at all assessment points in the PrEP iT! intervention trial (n = 66)

		Total	Daily PrEP Strategy Only	On-demand PrEP Strategy at least Once
		(n = 66)	(n = 48)	(n = 18)
Age (M; SD)a		25.4 (2.6)	25.5 (2.7)	25.2 (2.4)
		Col % b (n)	Col % b (n)	Col % b (n)
Race/Ethnicity**	Asian/Pacific Islander	19.7 (13)	10.4 (5)	44.4 (8)
	Black/African American	15.2 (10)	14.6 (7)	16.7 (3)
	Hispanic/Latino	16.7 (11)	22.9 (11)	0 (0)
	Non-Hispanic White	48.5 (32)	52.1 (25)	38.9 (7)
College Degree	Yes	78.5 (51)	81.3 (39)	70.6 (12)
	No	21.5 (14)	18.8 (9)	29.4 (5)
Currently Employed	Yes	83.3 (55)	87.5 (42)	72.2 (13)
	No	16.7 (11)	12.5 (6)	27.8 (5)
Health Insurance	Yes, I have my own	78.5 (51)	80.9 (38)	72.2 (13)
	Yes, parent/guardian	20.0 (13)	17.0 (8)	27.8 (5)
	No health insurance	1.5 (1)	2.1 (1)	0 (0)
Skip Meals Past 3 Months	Yes	10.8 (7)	8.5 (4)	16.7 (3)
	No	89.2 (58)	91.5 (43)	83.3 (15)
Live with family	Yes	18.2 (12)	18.8 (9)	16.7 (3)
	No	81.8 (54)	81.3 (39)	83.3 (15)
Out to half or more family*	Yes	78.8 (52)	85.4 (41)	61.1 (11)
	No	21.2 (14)	14.6 (7)	38.9 (7)
Out to half or more peers*	Yes	87.9 (58)	93.8 (45)	72.24 (13)
	No	12.1 (8)	6.3 (3)	27.8 (5)
Participant Discussed Sexuality Sometimes, or Openly Talked about Sexuality with, Provider#	Yes	77.3 (51)	83.3 (40)	61.1 (11)
	No	22.7 (15)	16.7 (8)	38.9 (7)
Intervention Condition	Treatment	48.5 (32)	45.8 (22)	55.6 (10)
	Control	51.5 (34)	54.2 (26)	44.4 (8)

^a.Mean (Standard Deviation)

^b.Column percentage; may not add up to 100% due to rounding

Notes:

^#p < .10;

^*p < .05;

^**p < .01;

^***p < .001

Most of the 66 participants with PrEP dosing strategy data at all timepoints reported being out to a majority of family members (78.8%) and peers (87.9%), and, similarly, reported that their medical provider definitely knew of their sexuality, and they sometimes or openly talked about it with their provider (77.3%; see Table 1). A higher proportion of men who were out to less than half of their family (39% on-demand vs. 15% daily only; (X²(2,n = 66) = 4.63, p = .031) and peers (28% on-demand vs. 6% daily only; Fisher’s exact test = 0.030) reported using an on-demand PrEP dosing strategy at least once during the study period compared to participants who only used a daily strategy. Analyses also demonstrated a similar pattern among men who discussed their sexuality rarely or never with their medical provider (39% on-demand vs. 17% daily only; X²(2, n = 66) = 3.68, p = .055).

The mean PrEP stigma total scale score was 1.97, with no differences found between men who always used a daily PrEP dosing strategy (M = 1.92) and those who used an on-demand strategy at least once (M = 2.09; not shown in tables). Similarly, the PS subscale had an average score of 2.14, with similar scores for men who report only daily dosing (M = 2.11) compared to those who reported using on-demand at least once (M = 2.21). However, men who only used a daily PrEP dosing strategy had more positive attitudes toward PrEP (i.e., the PA subscale), compared to those who used an on-demand PrEP strategy at least once (M = 1.47 vs. 1.81) with higher scores reflecting less positive attitudes; t(64) = −2.55, p = .013).

Multivariable Analyses.

The top half of Table 2 shows the results of the logistic regression model to assess the odds of being in the “On-demand PrEP Strategy 1 + Time” group (vs. Daily PrEP Strategy Only group), adjusted for age and intervention arm. Consistent with the bivariate analyses, Asian/Pacific Islander race, lower outness to family, peers, and medical provider, and less positive attitudes toward PrEP were associated with higher odds of using an on-demand PrEP strategy at least once during the study period. However, when race/ethnicity was included in the model with age and intervention arm (bottom half of Table 2), outness to family, peers, and medical provider was no longer significantly associated with the use of an on-demand PrEP strategy, although less positive attitudes toward PrEP remained significant (OR = 4.86 [1.22, 19.40], p = .03). Moreover, Asian/Pacific Islander race was associated with higher odds of using an on-demand PrEP strategy in all models (ORs between 4.13 and 5.06). In follow-up analyses, a lower proportion of Asian/Pacific Islander participants reported being out to half or more family members (46.2% vs. 70.0% Black/African American, 81.8% Hispanic/Latino, and 93.8% Non-Hispanic White participants; Fisher’s exact test = 0.003) and peers (61.5% vs. 80.0% Black/African American, 100.0% Hispanic/Latino, and 96.9% Non-Hispanic White participants; Fisher’s exact test = 0.005). Lower proportions of both Asian/Pacific Islander (53.9%) and Black/African American (50%) participants discussed their sexuality sometimes or openly talked about their sexuality with their provider than Hispanic/Latino and non-Hispanic White participants (100.0% and 87.5%, respectively; Fisher’s exact test = 0.003).

Table 2.

Adjusted odds of being in the “On-Demand PrEP Strategy at least Once” group for participants with complete data at all assessment points in the PrEP iT! intervention trial (n = 66)

Adjusting for Age & Intervention Arma
	OR (95%CI: L.L., U.L.) b
Race/Ethnicity c,d
Asian/Pacific Islander	5.91 (1.43,24.50)**
Black/African American	1.53 (0.30,7.75)
Out to half or more family	0.53 (0.28, 0.98)*
Out to half or more peers	0.36 (0.16, 0.83)*
Participant Discussed Sexuality Sometimes or Openly Talked about Sexuality with Provider	0.79 (0.62, 1.01)#
Positive Attitudes Toward PrEP scale
Less positive attitudes	4.43 (1.36, 14.45)*
Adjusting for Race/Ethnicity,c,dAge, & Intervention Arma	OR (95%CI: L.L., U.L.) b
Out to half or more family	0.66 (0.31,1.43)
Asian/Pacific Islander	4.13 (0.86,19.88)#
Black/African American	1.25 (0.23,6.81)
Out to half or more peers	0.57 (0.23,1.43)
Asian/Pacific Islander	4.21 (0.91, 19.48)#
Black/African American	1.32 (0.25,6.96)
Participant Discussed Sexuality Sometimes or Openly Talked about Sexuality with Provider	0.91 (0.68, 1.20)
Asian/Pacific Islander	5.06 (1.15, 22.30)*
Black/African American	1.24 (0.22, 7.11)
Positive Attitudes Toward PrEP scale
Less positive attitudes	4.86 (1.22, 19.40)*
Asian/Pacific Islander	6.34 (1.40, 28.75)*
Black/African American	1.37 (0.23, 8.23)

^a.Age and Intervention Arm were not significant in any of the models and, therefore, are not shown in table

^b.Odds Ratio (95% Confidence Interval: Lower Limit, Upper Limit)

^c.100% of Hispanic/Latino participants took daily PrEP, and therefore were omitted from the model

^d.Reference group: Non-Hispanic White

Note:

^#p < .10;

^*p < .05;

^**p < .01;

^***p < .001.

Sexual Behavior

All available data for all participants (n = 80) at each time point were used to assess the association between the number of male sex partners, as well as condomless sex with HIV-unknown status or partners living with HIV, and PrEP dosing strategy (daily, on-demand, or discontinued; see Table 3). In the bivariate analyses, the median number of male sex partners was significantly higher for participants on a daily PrEP dosing strategy than participants who used an on-demand strategy or discontinued PrEP at baseline (3 vs. 1 vs. no discontinued, respectively; z = 2.26, p = .02), month 6 (3 vs. 1 vs. 1, respectively; χ2(2, n = 73) = 6.35, p = .04), and the final assessment (3 vs. 1 vs. 1, respectively; χ2(2, n = 71) = 15.43, p = .0004). However, in the multivariable regression models adjusted for intervention arm, age, and race/ethnicity, participants who reported an on-demand PrEP strategy (β = −3.84, p = .07) or who discontinued using PrEP (β = −2.75, p = .008) had significantly fewer sex partners than participants who took PrEP daily only on the final assessment (see Table 4).

Table 3.

Differences in sexual behavior by PrEP dosing strategy for participants in the PrEP iT! intervention trial (n = 80)

		Total	Daily PrEP Strategy Only	On-demand PrEP Strategy at least Once	Discontinued
Median and mean sexual partners (past 3 months)
		Mdn; M(SD)a	Mdn; M(SD)	Mdn; M(SD)	Mdn; M(SD)
Baseline (n = 77)*		3; 3.7 (3.9)	3; 4.0 (4.1)	1; 1.8 (1.4)	n/a
3 months (n = 73)		2; 3.9 (5.1)	2; 4.4 (5.4)	1; 1.9 (1.6)	1; 2.2 (3.4)
6 months (n = 73)*		2; 4.1 (5.4)	3; 4.6 (5.3)	1; 3.1 (5.3)	1; 3.0 (6.1)
Final assessment (n = 71)***		2; 3.8 (4.9)	3; 4.6 (5.4)	1; 1.7 (1.6)	1; 1.0 (0.4)
Condomless Sex with HIV-unknown/positive partner (past 3 months)
		Col % (n)	Col % (n)	Col % (n)	Col % (n)
Baseline (n = 75)	Yes	32.0 (24)	33.3 (22)	22.2 (2)	n/a
	No	68.0 (51)	66.7 (44)	77.8 (7)	n/a
3 months (n = 74)	Yes	27.0 (20)	29.5 (18)	14.3 (1)	16.7 (1)
	No	73.0 (54)	70.5 (43)	85.7 (6)	83.3 (5)
6 months (n = 74)	Yes	24.3 (18)	28.0 (14)	23.1 (3)	9.1 (1)
	No	75.7 (56)	72.0 (36)	76.9 (10)	90.9 (10)
Final assessment (n = 71)*	Yes	31.0 (22)	39.6 (21)	16.7 (1)	0 (0)
	No	69.0 (49)	60.4 (32)	83.3 (5)	100.0 (12)

^a.Mdn = Median; M(SD) = Mean (Standard Deviation)

Notes:

^#p < .10;

^*p < .05;

^**p < .01;

^***p < .001; frequencies less than 80 are due to attrition at each time point

Table 4.

Regression results for the adjusted association between using an on-demand PrEP strategy at least once during the study (vs. daily strategy) and number of past 3 months sex partners for participants in the PrEP iT! intervention trial (n = 80)

Number of Male Sexual Partners (past 3 months)
Adjusted for Age & Intervention Arma		Coefficient	Standard Error
Baseline (n = 77)
PrEP Strategyb	On-demand	−2.29#	1.22
3 Month (n = 73)
PrEP Strategyb	On-demand	−2.61	2.02
	Discontinued	−1.97	2.17
6 Month (n = 73)
PrEP Strategyb	On-demand	−1.88	1.70
	Discontinued	−1.62	1.86
Final Survey (n = 71)
PrEP Strategyb	On-demand	−3.29	2.03
	Discontinued	−3.70*	1.50
Adjusting for Race/Ethnicity, Age, & Intervention Arma
Baseline (n = 77)
PrEP Strategyb	On-demand	−2.26	1.36
Race/Ethnicityc	Asian/Pacific Islander	− 0.38	1.29
	Black/African American	− 0.79	1.29
	Hispanic/Latino	− 0.89	1.30
3 Month (n = 73)
PrEP Strategyb	On-demand	−2.55	2.06
	Discontinued	−1.34	2.40
Race/Ethnicityc	Asian/Pacific Islander	−1.13	1.74
	Black/African American	−3.28#	1.83
	Hispanic/Latino	− 0.56	1.69
6 Month (n = 73)
PrEP Strategyb	On-demand	−2.90	1.85
	Discontinued	−2.33	1.92
Race/Ethnicityc	Asian/Pacific Islander	1.43	1.85
	Black/African American	−1.29	1.81
	Hispanic/Latino	−1.78	1.85
Final Survey (n = 71)
PrEP Strategyb	On-demand	−3.84#	2.08
	Discontinued	−4.22***	1.53
Race/Ethnicityc	Asian/Pacific Islander	0.15	1.48
	Black/African American	−2.56	1.60
	Hispanic/Latino	−1.50	1.64

^a.Age and Intervention Arm were not significant in any of the models and, therefore, are not shown in table

^b.Reference group: Daily PrEP

^c.Reference group: Non-Hispanic White

Notes:

^#p < .10;

^*p < .05;

^**p < .01;

^***p < .001; frequencies less than 80 are due to attrition at each time point

Between 24.3% (at month 6) and 32.0% (at baseline) of men reported having condomless sex with unknown or positive HIV status partners across the study time period (see Table 3). The highest percentage of condomless sex with HIV status unknown or positive partners at each time point was reported by those who used a daily PrEP dosing strategy compared to participants who used an on-demand PrEP strategy at least once during the study or those who discontinued PrEP, although this was only statistically different in the bivariate analyses at the final assessment (39.6% daily dosing vs. 16.7% on-demand dosing vs. 0% discontinued; Fisher’s exact test = 0.009). However, in the multivariable logistic regression analyses adjusted for intervention arm and age, or for intervention arm, age, and race/ethnicity (not shown in table), PrEP strategy and discontinuation of PrEP was not associated with condomless sex with HIV-unknown status or sexual partners living with HIV.

Use of On-Demand PrEP Dosing Strategy, and Reasons for, and Intentions to Switch, Dosing Strategies among Participants Completing the Final Assessment

The remainder of the results are taken from the final assessment, in which we asked participants (n = 72) in-depth about their use of, and attitudes toward, daily and on-demand PrEP dosing strategies. At the final assessment, three-quarters (n = 54) of men reported currently using a daily PrEP dosing strategy, 6 (8.3%) an on-demand strategy, and 12 (16.7%) had discontinued PrEP. Of those that had discontinued, most (75%, n = 9) used a daily dosing strategy before they quit, and the remaining (n = 3) used an on-demand strategy. Below, daily PrEP users include participants who were using a daily strategy at the time of the final assessment, or who had used a daily PrEP strategy just before discontinuing (n = 63); on-demand PrEP users include participants who were using an on-demand strategy at the time of the assessment or who used an on-demand strategy just prior to discontinuing (n = 9).

When asked about the barriers men faced when taking daily PrEP (n = 63), the most common were forgetfulness (n = 33), having problems making appointments (n = 15), difficulty taking PrEP while traveling (n = 14), and lack of, or difficulties with, insurance (n = 13), the cost of the prescription (n = 6), and mental health (n = 6). In contrast, on-demand PrEP users (n = 9) reported that the cost of the prescription was the most important barrier (n = 3), followed by forgetfulness (n = 2), difficulty making appointments (n = 2), and lack of, or difficulty with, insurance (n = 2).

The most endorsed reason that daily PrEP users reported why they would switch to an on-demand PrEP strategy if they were to do so was having infrequent sex (n = 30). Other common reasons included concerns about the long-term effects of daily PrEP (n = 23), the convenience of on-demand dosing (n = 22), concerns about side-effects (n = 14), the ability to accurately predict when sex will occur (n = 13), problems with taking a daily pill (n = 12), and to save money (n = 9). Similarly, participants using PrEP on-demand (n = 9) reported that they are doing so because they don’t have sex often (n = 4), concerns about side-effects (n = 4), convenience (n = 3) and being able to accurately predict when sex will occur (n = 3).

Finally, we asked participants if they intended to switch to a different PrEP formulation or dosing strategy. Most daily PrEP users (73%; n = 44 of 60 responses) reported that they wanted to remain on daily PrEP, while four men wanted to switch to an on-demand strategy, four would like to switch to injectable PrEP, and eight men wanted to switch but were unsure as to what they would like to switch to (not shown in tables). Of the four men who reported that they would like to switch to injectable PrEP, they reported that they would do so because they believed that injectable PrEP is more convenient (n = 3) and has fewer long-term effects (n = 3) than daily PrEP, and because they did not have sex often enough to warrant daily pills (n = 1). On-demand PrEP users were asked similar questions, with most (n = 7 of 8 responses) stating that they would like to stay with an on-demand strategy and one participant stating that he would like to switch but was unsure as to what. None of the men using an on-demand strategy stated that they would like to switch to injectable PrEP.

Discussion

The availability of different PrEP formulations and dosing strategies provides greater flexibility for GBMSM to adopt, modify, and sustain use of, PrEP. While a number of studies examined the use of non-daily (in nearly all cases on-demand) PrEP strategies in European countries and Australia [15–18, 20, 21], studies describing the use of on-demand PrEP strategies among US-based GBMSM are scant [14, 19], and none have been in the context of a PrEP adherence intervention. In the current analyses, we primarily examined the use of oral daily and oral on-demand PrEP dosing strategies (although intentions to switch to injectable PrEP were also examined) among US-based GBMSM enrolled in a PrEP adherence pilot intervention. In general, most participants used a daily dosing strategy, although one third used an on-demand dosing strategy on one or more occasions during the study period. Other than consistent daily dosing across the four study timepoints, no prominent patterns emerged among those participants who used on-demand dosing at least once during the study. With new approved formulations (i.e., injectable PrEP) and dosing strategy (daily, on-demand) options, patients on PrEP now have a variety of options that can be used in conjunction with one another as life circumstances change over time. For example, if a participant is going through a low “season of risk,” an on-demand option may be preferable to a daily dosing strategy until sexual frequency increases. In addition, if a patient on injectable PrEP misses an appointment, providing a backup prescription of oral PrEP until their next injection may be an important strategy to reduce HIV risk. Future studies should closely monitor patient use of PrEP as it relates to changing life circumstances to inform how they may benefit from shifting formulations or dosing strategies to optimize the benefits of PrEP.

Among participants who had complete dosing data across all study timepoints, we found significant differences in the dosing strategy by race/ethnicity and outness in the bivariate analyses. It was noteworthy that all Hispanic/Latino GBMSM in this sample used a consistent daily dosing strategy across all time points, which may reflect their awareness of the higher rate of HIV among Hispanic/Latino GBMSM compared to non-Hispanic White GBMSM in the US [27, 28]. While warranting further investigation, Hispanic/Latino GBMSM may perceive that the use of a daily dosing strategy leads to greater protection from HIV.

A higher percentage of Asian/Pacific Islander GBMSM in this sample used a non-daily PrEP strategy compared to other racial and ethnic groups in both the bivariate and multivariable analyses. This finding may be, in part, explained by the overall lower level of outness to family, peers, and medical providers among Asian/Pacific Islander men compared to other racial and ethnic groups in this sample. A prior study [29] found an association between PrEP use among young adults and comfort with parent/child sex communication, while a second study of 1,541 GBMSM in the US showed that PrEP knowledge was associated with disclosure of sexual orientation to anyone and with disclosure of sexual orientation to their healthcare provider [30]. Outness may be an important factor in PrEP awareness and use through access to social and sexual networks that may be more or less knowledgeable and supportive of PrEP use and providing (or limiting) opportunities for providers to hold discussions with GBMSM about PrEP as a prevention option. On-demand and other non-daily PrEP strategies may offer a way for GBMSM who are concerned about disclosure to family members or peers to reduce the chances that others will witness them taking PrEP, as may be the case for Asian/Pacific Islander men in this sample. In the same way, injectable PrEP may be preferred over oral medications to address concerns about stigma and disclosure among GBMSM, which has been demonstrated in prior research [31]. Alternatively, Asian/Pacific Islander men in this sample may hold beliefs that make non-daily dosing strategies preferable to daily dosing, including beliefs about the role of medication in preventing and treating HIV or the long-term effects of daily PrEP. Finally, it may be that Asian/Pacific Islander GBMSM in our sample were exposed to more information about on-demand PrEP compared to other racial and ethnic groups, and therefore felt more comfortable using an on-demand dosing strategy. Studies using larger and more diverse samples are needed to longitudinally assess patterns of PrEP use over time, as well as collect information about beliefs about PrEP options and cultural attitudes that may impact PrEP use.

GBMSM in this sample who used an on-demand PrEP strategy at least once during the trial reported less positive attitudes about PrEP than those who consistently used PrEP daily, even after controlling for sociodemographic factors and intervention arm. Although this finding will need to be further explored since this subscale had marginal internal validity in our sample of GBMSM, this suggests that the clinical provision of PrEP care should include discussions with GBMSM about the benefits of PrEP to increase positive attitudes about it. Combined with a discussion of how different PrEP formulations and/or dosing strategies might allay widely-held concerns (e.g., the side effects of daily PrEP might be reduced by taking PrEP on demand), highlighting the benefits of PrEP may encourage its uptake among GBMSM who are initially resistant to starting it.

The level and type of sexual activity may influence GBMSM’s decision about what PrEP option to choose that optimizes protection against HIV. Our results are consistent with prior studies showing that the number of sexual partners and engaging in high-risk sexual activity are higher among daily PrEP using-GBMSM compared to GBMSM who use an on-demand PrEP strategy [16]. In particular, some of the GBMSM in the current sample may have adjusted their PrEP use based on the number of recent sexual partners. This was especially evident at the final assessment, during which some GBMSM who reported low numbers of sexual partners in the past three months chose to use an on demand PrEP dosing strategy, while other men with low numbers of sexual partners chose to discontinue PrEP. It is noteworthy that GBMSM in this sample who discontinued PrEP reported approximately the same number of sex partners as those who used an on-demand PrEP strategy across study timepoints, although a smaller proportion of those who discontinued PrEP reported engaging in condomless sex than those who used an on-demand strategy. This may suggest that factors other than the number of sex partners may influence GBMSM’s decision to discontinue PrEP. A recent study [32] that assessed reasons that GBMSM in New York City attending a large sexual health clinic discontinued PrEP, which was primarily due to problems paying for PrEP, reduced perception of being at risk for HIV, and concerns about side effects. Larger studies examining why GBMSM discontinue PrEP are needed to better illuminate drivers of PrEP discontinuation to better inform future intervention approaches to re-engage prior PrEP users.

At the final assessment, 88% of GBMSM in this sample were using a daily dosing strategy currently or prior to discontinuing PrEP, with fewer participants (n = 9) opting for a current or recent on-demand PrEP dosing strategy. This is consistent with studies in other Western countries demonstrating that GBMSM opt to use a daily dosing strategy when given the choice between that and on demand PrEP [15–18]. Using an on-demand PrEP strategy requires a higher level of awareness and foresight, compared to daily PrEP use, about one’s sexual activity and vigilance for taking PrEP more than two hours before sexual activity that may be difficult to put into practice for younger GBMSM. In a study to assess the degree to which young adult GBMSM (N = 1,013) planned for sex, half of participants reported no planning prior to their most recent sexual encounter, while 8% reported only planning minutes in advance [33]. If on-demand PrEP strategies were to be successful, it may be critical to assist GBMSM in increasing their ability to anticipate sex and provide a range of easy-to-implement adherence strategies for taking doses both before and after a sex event. Mobile technologies that accommodate complex algorithms that can track and provide timely intervention for on-demand PrEP strategies may be an important tool, which are already in development [34, 35].

Although most GBMSM in our sample chose to stay on a daily PrEP dosing schedule by the end of the study, the primary reason that participants who were on a daily dosing schedule in this sample stated that they would switch to on-demand is if they experienced a decrease in sexual activity, which was also a primary reason why men who were using or recently used an on-demand strategy continued to use it. To optimize the benefits of PrEP, it may be critical to regularly assess GBMSM’s satisfaction with their current PrEP formulation and dosing strategy to sustain engagement in PrEP use over time. Among 131 young adult GBMSM in Atlanta who initiated PrEP, 60% persisted taking PrEP after approximately 4 months [36]. Especially since new routes of administration (i.e., injectable PrEP) and dosing strategies are becoming more available, interventions and PrEP programs may benefit from more regular assessments of GBMSM’s attitudes toward their current PrEP regimen and encouraging alternatives when appropriate.

Most GBMSM currently or recently on a daily or on-demand PrEP strategy reported that they intended to stay on it. Among those who intended to switch, most were unsure as to what strategy they would want to switch to, and only four participants who were currently or recently on a daily PrEP strategy reported that they would like to switch to injectable PrEP. Notably, GBMSM in this study completed the final assessment shortly after injectable PrEP was approved by the FDA in the US. Thus, many of the men may not have been aware of this option, and therefore may be reluctant to switch to injectable PrEP without more knowledge of its effectiveness and requirements (e.g., bi-monthly visits). No doubt, as injectable PrEP becomes more widely available, interest in and uptake of this option will increase among GBMSM and other populations at risk for HIV who may prefer an every two-month injection to daily or on-demand oral PrEP [8]. As injectable PrEP adoption increases, it may be important to monitor how prior experience with oral PrEP is associated with interest in and uptake of injectable PrEP. Injectable PrEP may be preferred by GBMSM who are starting PrEP for the first time, who may have been reluctant to take oral PrEP, perceiving injectable PrEP to be a less onerous option than oral PrEP. In addition, GBMSM who have a history of using oral PrEP and report being dissatisfied with it may be persuaded to adopt injectable PrEP, while those successfully on oral PrEP may be less inclined to change their regimen. Studies with larger sample sizes are needed to better understand who and under what circumstances injectable PrEP is adopted and maintained over time.

This study has a number of limitations. First, the sample size is relatively small, and participants were a part of a pilot intervention study and were required to be on PrEP at screening. Therefore, participants may not represent all GBMSM, or GBMSM who are unwilling to join an intervention study or who have no history of PrEP use. That said, this study is one of the first to assess changes in PrEP dosing strategy as part of an intervention trial in the US and provides insights into how participants are using PrEP and the potential challenges in interpreting the results of PrEP outcomes that arise from these changing strategies. Larger studies of diverse GBMSM in the US are needed to address these limitations. Second, information presented on participants’ use and beliefs about injectable PrEP is limited in this study given that it was primarily conducted prior to FDA approval in December 2021. Larger longitudinal studies assessing patient preference and switching PrEP among strategies, including in the context of injectable PrEP use, are needed to optimize packages of PrEP that can be tailored to individual circumstances.

Conclusions

Despite these limitations, this study demonstrates the use of both daily and on-demand PrEP strategies across time among US-based GBMSM, with some participants switching between strategies during the intervention trial. Our findings support that participant race and ethnicity, the degree of outness, and sexual practices are associated with GBMSMs’ choice of PrEP strategy. As options for PrEP formulations and dosing strategies grow, continued monitoring of how PrEP is being used is critical to optimize prevention plans for GBMSM and other high-risk persons (e.g., transgender women), as well as to better assess the impacts of PrEP interventions for communities most impacted by the HIV epidemic.