A randomized trial of a brief behavioral intervention for PrEP uptake among men who have sex with men at increased risk for HIV infection

Abstract

Background:

The aim of this study was to evaluate the efficacy of a novel, brief two-session behavioral intervention to promote HIV pre-exposure prophylaxis (PrEP) uptake among men who have sex with men (MSM) who are behaviorally at risk for HIV.

Setting:

A pilot randomized controlled trial (RCT) was conducted at a sexually transmitted infection (STI) clinic to compare a brief motivational interviewing (MI) intervention to passive referral only for PrEP uptake.

Methods:

MSM who scored as “high-risk” on the HIV Incidence Risk Index for MSM (HIRI-MSM) were offered a brief (15–20 minutes) MI-based intervention at the time of STI testing to address barriers to PrEP uptake, including low risk perception, stigma, side-effects, and cost. The initial session was followed by a brief, telephone booster session that lasted <10 minutes. The primary outcome was attending a clinical PrEP appointment and accepting a prescription for PrEP.

Results:

Participants were recruited from an urban STI clinic in the USA. A total of 86 MSM who were behaviorally at risk for HIV were enrolled in the study (N=43 intervention; N=43 treatment-as-usual, “TAU”). Participants randomized to the intervention were significantly more likely to attend a clinical appointment and accept a prescription for PrEP, compared to TAU (52.3% versus 27.9%, respectively; OR=3.6; 95%CI: 1.5, 8.9; p=.005).

Conclusion:

A brief behavioral intervention focused on the initial steps in the PrEP care cascade demonstrated preliminary efficacy in promoting uptake among MSM who are behaviorally at risk for HIV.

Introduction

HIV continues to be a significant cause of morbidity and mortality in the United States (US) with a disproportionate number of infections among gay, bisexual and other men who have sex with men (MSM)¹. Pre-exposure prophylaxis (PrEP) has demonstrated efficacy in preventing HIV infection^2–5 and has the potential to significantly reduce HIV incidence in populations at risk, such as MSM. The Centers for Disease Control and Prevention (CDC) estimated in 2015 that nearly 25% of sexually active MSM meet indications for PrEP⁶. Despite significant increases in PrEP awareness among MSM since this time, PrEP use remains suboptimal, particularly for African American/Black and Hispanic/Latinx MSM⁷. Given this, novel approaches to improve PrEP uptake are needed.

Significant barriers to PrEP uptake include low perceived HIV risk, stigma, concern for side-effects and cost^8–11. Of these, low perceived HIV risk has emerged as a significant challenge. Up to 37% of MSM presenting to a sexually transmitted infection (STI) clinic who reported condomless anal sex did not perceive themselves to be at-risk for HIV infection¹². This has been observed across multiple other studies^13,14. Misperceptions about side-effects and lack of awareness of programs to defray cost and minimize out-of-pocket expenses have also led to lower PrEP uptake^15–17. Despite knowledge of these myriad barriers, few interventions have successfully increased PrEP uptake.

This goal of the study was to develop and evaluate the feasibility, acceptability, and efficacy of a novel behavioral intervention to promote PrEP uptake among MSM at increased risk for HIV at an STI clinic who were not already seeking PrEP. The intervention was designed to be brief and applicable in a busy clinical setting. We hypothesized that the intervention, relative to usual clinical care, would result in: (a) increased willingness to schedule a PrEP evaluation appointment; (b) increased likelihood to attend the appointment; and, (c) increased likelihood to accept a PrEP prescription.

Methods

From January 2019 to March 2020, MSM who were not currently taking PrEP and at elevated risk for HIV infection were recruited during routine clinical care at a walk-in STI clinic. If they could not be approached during their clinic visit, they were contacted as soon as possible and within one week of their visit. Inclusion criteria were: (a) 18 years of age or older; (b) assigned male at birth (AMAB) and/or currently identify as male; (c) scored at least 10 on the HIV Incidence Risk Index for MSM (HIRI-MSM¹⁸); and, (d) HIV-negative based on results of an antibody/antigen test. Participants were excluded if they were currently taking PrEP, had taken it in the past, or did not fluently speak English. If participants had previously received a PrEP prescription but never filled it, they were eligible to participate. Participants who reported fewer than one condomless anal sex event in the past 3 months were also excluded.

After enrolling and completing a baseline assessment, participants were randomized to motivational interviewing (MI) or treatment-as-usual (TAU) with equal allocation between conditions using a randomization scheduled generated by a non-study statistician. Follow-ups were conducted at 1-, 3-, and 6-months post-baseline to assess steps completed in the initial PrEP care cascade. The 1-month follow-up was also used to collect acceptability/satisfaction data. Participants were compensated $50 for completing the baseline and each additional assessment. No remuneration was provided for intervention session attendance; participants were paid only for completing study assessments. Follow-ups occurred in person whenever possible. However, for participants’ convenience, follow-up assessments could be completed remotely via electronic data collection methods.

The 2-session, individual MI intervention was administered using a manualized, structured protocol that was previously developed and revised in a small uncontrolled pilot trial¹⁹. The brief intervention was based on MI theory, which was originally developed for use in alcohol treatment²⁰ and was rooted in the perception that motivation to change must be internally-driven. The intervention was informed by qualitative research with the target population²¹. This qualitative research, conducted with other MSM at the clinic prior to the start of the present study, revealed that MSM requested brief interventions and patient navigation at the STI clinic. Our intervention addresses a diverse and wide range of possible barriers to PrEP uptake. These were categorized as follows: (a) clinical barriers (e.g., concerns about PrEP side-effects); (b) individual barriers (e.g., low self-perceived HIV risk); (c) social barriers (e.g., fear of disclosure); and, (d) structural barriers related to the health system (e.g., challenges associated with expenses related to copayments, deductibles, and accessing patient navigation). Consequently, the intervention had built-in flexibility to allow interventionists to discuss and address whichever barrier(s) the participant reported¹⁹. One minor revision was made to the manual in October 2019 when tenofovir alafenamide (TAF) with emtricitabine (FTC) became available as a second PrEP medication. The manual was revised to discuss tenofovir disoproxil fumarate (TDF) with emtricitabine (FTC) and TAF/FTC.

The primary goals of the first intervention session were to: (a) develop rapport and a therapeutic relationship; (b) provide psychoeducation about PrEP; (c) share normative feedback on participants’ risk for HIV infection; (d) identify and address barriers to PrEP uptake; and, (e) weigh the pros and cons of PrEP uptake. We aimed to complete session 1 in about 15 minutes. Extensive counseling about other HIV prevention methods was not included in the intervention as these were provided by the clinic’s usual care. Session 2, scheduled for approximately one week later and meant to be completed by telephone in less than 10 minutes, was designed as a “booster” session, in which no additional information was presented. Instead, the goals were to probe for understanding and resonance of session 1 material and to construct an action plan when appropriate.

The control condition was the clinic’s treatment-as-usual (TAU) procedures for MSM expressing interest in PrEP. TAU did not include any elements of MI. All participants received TAU services that included referral to PrEP services. STI clinic counselors were trained to discuss PrEP, to encourage its use among individuals at increased risk, and to facilitate PrEP uptake by addressing structural barriers whenever possible (e.g., offering bus vouchers to overcome transportation issues, help in navigating clinic and industry sponsored patient assistance programs).

Interventionists were four state-certified STI clinic counselors who provided clinical services at the recruitment site. One interventionist was a Bachelor’s level provider and the other individuals had Master’s degrees. A four-hour didactic training workshop for the interventionists educated them on the challenges of motivating MSM to take PrEP, MI perspectives on health behavior change, and skills to help patients make change plan decisions. The manual was reviewed and role-plays of standard, challenging, and emergency situations were conducted. Interventionists had weekly telephone supervision during the trial. Supervision was provided by one of the study’s principal investigators, a doctoral-level clinical psychologist.

Study participants completed quantitative assessments that included questions on demographics, medical treatment access, recent sexual behaviors and sexual health, and attitudes toward PrEP and possible barriers to PrEP uptake. Hazardous drinking was examined with the short-form Alcohol Use Disorders Identification Test-C (AUDIT-C²²) and substance use/misuse was examined with items from the EXPLORE study, which studied risky behavior among MSM²³. The EXPLORE substance items were used in lieu of the HIRI-MSM to assess 12-month substance use. Participant mental health symptoms were characterized by the Patient Health Questionnaire-9 (PHQ-9²⁴) and the Generalized Anxiety Disorder-7 (GAD-7²⁵). A PrEP stages of change measure was used to characterize perceived willingness to take PrEP and perceived need²⁷.

In a qualitative exit interview at the 1-month follow-up, we asked participants about positive and negative aspects of MI and solicited ideas for improvements. To examine MI satisfaction, the 8-item Client Satisfaction Questionnaire-Revised (CSQ-R²⁶) was administered. The CSQ-R has been extensively studied and has shown excellent internal consistency, reliability, and convergent validity, and has been shown to operate similarly across different racial and ethnic groups.

As done in prior research, including work at our site¹², we assessed the following primary outcomes of interest: (a) scheduling an appointment with a clinic prescriber to discuss PrEP; (b) attending the prescriber appointment; and, (c) accepting a PrEP prescription. These data were objectively gathered via review of participants’ electronic medical records. Thus, we obtained these data for all participants except those who were withdrawn from the study. We relied on participant self-report if individuals received PrEP from an external prescriber. This occurred in one case. The MI intervention was designed to increase participants’ perception of HIV risk and increase their openness to PrEP. These factors were assessed at baseline by self-report with the Perceived HIV Risk Scale²⁸ and items developed to assess individuals’ willingness and perceived need to take PrEP, based on work by Parsons and colleagues²⁷.

All statistical analyses were conducted in SPSS. Descriptive statistics were used to summarize feasibility and acceptability outcomes related to: MI fidelity, MI session length, MI session attendance, recruitment rates, data collection retention rates, responses to exit interviews with participants, and satisfaction ratings at treatment completion. Descriptive statistics were calculated to characterize the sample. We used t-tests and χ² tests for between group comparisons on continuous and categorical variables, respectively.

Preliminary effects of MI on PrEP uptake outcomes, compared to TAU, were assessed using logistic regression analyses. No covariates were used in these models as we did not specify any covariates a priori. Baseline comparisons revealed that MI and TAU participants were equivalent on most characteristics. Due to the pilot nature of this work, statistical power was not prioritized. All procedures were approved by The Miriam Hospital Institutional Review Board.

Results

During the study period, a total of 828 MSM were screened and 287 were approached during STI care. Of those approached, 147 (51.2%) agreed to participate. At the time of recruitment, 60 individuals expressed willingness to participate in the study but they were not ultimately enrolled or consented because they did not complete the baseline assessment. Of the remaining 87 MSM, one individual was withdrawn after randomization to the MI condition when it became clear that he had previously taken PrEP. Thus, the final sample consisted of 86 MSM. The majority of these participants (n=52; 60.5%) completed the endpoint of 6-month self-report follow-up data collection procedures. Additionally, because our primary outcomes related to PrEP uptake were measured objectively through chart review, we obtained data on all participants who did not withdraw from the study (n=82; 94.3%). Four participants withdrew from the study after baseline (4.7% of sample; 2 MI, 2 TAU). These individuals mostly discontinued due to time constraints, although one person withdrew due to privacy concerns. See Figure 1 for study flow.

Figure 1.

Study flow chart.

One participant was assigned female at birth (AFAB) but identified as male; otherwise, all other participants were AMAB. The sample mostly identified their gender as male (n=84; 97.7%), with two participants identifying as genderqueer (2.3%). Most participants described their sexual orientation as same-gender loving or gay (n=68; 79.1%), with bisexual being the second most common orientation (n=14; 16.3%). Participants were younger adults (M=32.0 years old; SD=13.0 years old). All participants surpassed the HIRI-MSM risk cut-off of 10 (M=20.0; SD=6.1). The sample was primarily non-Latinx White (n=56; 66.7%) although a sizeable group identified as being ethnically Latinx (n=22; 26.8%). The overall sample was highly educated, with 60 participants having at least a college degree.

Independent sample t-tests and χ² analyses were used to compare baseline participant characteristics between MI and TAU. The TAU group had statistically, significantly higher depressive symptoms: t(1)=2.01, p=.048, compared to individuals in MI. However, closer inspection revealed that these differences were minor, as the MI group’s mean PHQ-9 score was 3.4 (SD=2.9) whereas the TAU group’s mean score was 5.3 (SD=5.4). Both represent minimal to no depression. Although TAU participants also demonstrated statistically, significantly higher perceived need to take PrEP (t(1)=2.40, p=.019), these differences were not clinically meaningful either. Otherwise, results revealed no baseline differences on demographic or clinical variables between intervention groups. See Table 1 for demographic summary, comparing the two intervention groups.

Table 1.

Demographic, health, and sexual behavior characteristics at baseline (n=86)

	Descriptive statistics		Group comparisons
	MI (n = 43)	TAU (n = 43)	t or χ2 (p =)
Age, mean (SD)	32.3 (12.8)	31.8 (13.3)	−0.18 (.861)
Race			2.75 (.766)
White, n (%)	28 (63.6%)	28 (70%)
African American/Black, n (%)	4 (9.1%)	5 (12.5%)
Asian, n (%)	3 (6.8%)	2 (5%)
Other, n (%)	7 (15.9%)	4 (10%)
Latinx ethnicity, n (%)	11 (26.2%)	11 (27.5%)	0.02 (.894)
People born male, n (%)	43 (97.7%)	42 (100%)	0.97 (.326)
Male gender identity, n (%)	43 (97.7%)	41 (97.6%)	<0.01 (.973)
College graduate, n (%)	31 (73.8%)	29 (72.5%)	5.23 (.265)
Currently student, n (%)	11 (25%)	9 (21.4%)	0.15 (.695)
Employed part- or full-time, n (%)	34 (79.1%)	30 (71.4%)	4.67 (.458)
Single, n (%)	39 (88.6%)	35 (87.5%)	1.70 (.428)
Income (<$35000/yr.), n (%)	21 (51.2%)	21 (53.8%)	3.05 (.931)
Currently insured, n (%)	25 (59.5%)	26 (66.7%)	0.44 (.506)
Total sex partners (3 mo.), mean (SD)	5.1 (3.7)	5.5 (5.2)	0.49 (.623)
STD positive (12 mo.)			0.48 (.788)
Gonorrhea, n (%)	6 (13.6%)	8 (19%)
Syphilis, n (%)	5 (11.4%)	6 (14.3%)
Chlamydia, n (%)	6 (13.6%)	3 (7.1%)
Sex worker (12 mo.), n (%)	1 (2.3%)	3 (7.1%)	1.15 (.284)
HIV acquisition risk (HIRI), mean (SD)	19.9 (5.6)	20.0 (6.6)	0.09 (.931)
Perceived HIV risk (PRHS), mean (SD)	17.4 (2.8)	17.2 (2.6)	−0.42 (.679)
PrEP Staaes of Chanae
Willingness to take PrEP, mean (SD)	1.4 (0.5)	1.5 (0.6)	0.93 (.355)
Perceived need to take PrEP, mean (SD)	1.5 (0.6)	1.8 (0.7)	2.40 (.019)
Primary PrEP uptake barrier			6.04 (.302)
Side effects	20 (48.8%)	16 (41%)
Cost	18 (43.9%)	14 (35.9%)
Does not want daily medication	3 (7.3%)	4 (10.3%)
Believes PrEP not needed	0	3 (7.7%)
Stigma	0	1 (2.6%)
Hazardous drinking (AUDIT-C), mean (SD)	3.6 (1.7)	3.8 (2.1)	0.52 (.602)
Substance use/misuse (12 mo.), n (%)	36 (81.8%)	36 (85.7%)	0.24 (.625)
Depressive symptoms (PHQ-9) mean (SD)	3.4 (2.9)	5.3 (5.4)	2.01 (.048)
Anxiety symptoms (GAD-7), mean (SD)	3.5 (3.5)	5.1 (4.7)	1.83 (.071)

Note: MI=Motivational Interviewing; TAU=Treatment-as-Usual; STD=Sexually Transmitted Disease; HIRI=HIV Incidence Risk Index; PRHS=Perceived Risk of HIV Scale; PrEP=Pre-exposure Prophylaxis; AUDIT-C=Alcohol Use Disorders Identification Test-C; PHQ-9=Patient Health Questionnaire-9; Generalized Anxiety Disorder-7.

All sessions were audiotaped and 20% of sessions were randomly selected for fidelity evaluation using an internally developed measure. Review of these audio files was evenly split between the first and second sessions. All four study interventionists had sessions coded for fidelity by a non-study rater. According to this rater, interventionists’ adherence to the intervention manual was 99.3% in session 1 and 97.2% in session 2. Both MI sessions lasted about as long as planned. On average, session 1 was 21.1 minutes (SD=5.2 minutes) and session 2 was 6.3 minutes (SD=2.1 minutes). Interventionists had the option to provide “check-in calls” with MI participants in cases in which a scheduling issue or payment concern arose for those ultimately interested in taking PrEP. This check-in did not include any additional information or intervention. Only one participant received a check-in call during the course of the study.

All participants randomized to MI (n=43) attended session 1. Session 2 attendance was very good: n=37 (84.1%). Otherwise, two MI participants withdrew from the study during the course of follow-up (4.7%). Totals on the CSQ-R, which has a maximum score of 32, were highly positive for MI participants: mean=30.0, SD=2.7. On the study-designed exit interview measure, in which item ratings ranged from 0 to 6, with higher scores indicating increased agreement, feedback was generally positive: participants found the sessions helpful (mean=5.6, SD=0.7), sessions helped participants better understand PrEP (mean=5.6, SD=0.9), and they appreciated having the initial session at the STD clinic (mean=5.5, SD=0.9). Open-ended feedback was also supportive of the intervention. No adverse events occurred during this trial.

Logistic regression analyses showed significant efficacy for the intervention group versus TAU on all primary PrEP outcomes. In comparison to those randomized to TAU, participants in the intervention group were significantly more likely to: (a) schedule an appointment to further discuss PrEP with a prescriber (OR=6.0; 95%CI: 2.3, 15.6; p<.001); (b) attend the prescriber appointment (OR=3.6; 95%CI: 1.5, 8.9; p=.005); and, (c) receive and accept a prescription for PrEP (OR=3.6; 95%CI: 1.5, 8.9; p=.005). See Figure 2 for graphical summary of conditions at each step of the initial PrEP uptake cascade.

Figure 2.

Number of participants at initial PrEP uptake cascade steps by condition.

Discussion

This is among the first studies to demonstrate efficacy of a behavioral intervention to promote PrEP uptake for MSM who are behaviorally at significant risk for HIV acquisition. The intervention was designed to be brief and easy to implement in a busy STI clinic setting. MSM who underwent the intervention had a 6-fold increase in scheduling a PrEP clinical appointment and 3-fold increase in attending the appointment and accepting a prescription. The intervention was highly acceptable among all participants. In general, retention rates were high for MI and the study assessments considering that this was a substance-using, younger, non-PrEP seeking sample (see Figure 1). For the purposes of this analysis, alcohol and other drugs were combined when describing this sample given their high usage of a variety of substances. Results from the HIRI-MSM, used as a measure of increased risk for HIV transmission, revealed that substances like amyl nitrates were used by a large proportion of our sample (n=45; 52.3%).

These findings suggest that a brief MI-based intervention, which flexibly addresses common barriers to PrEP uptake, could be successful in promoting uptake among MSM who are at increased risk for HIV infection. In this study, the research team used the HIRI-MSM to screen for HIV risk, a highly sensitive but only moderately specific tool by which to evaluate HIV risk. A scale-up of the MI intervention could be applied to anyone who may be at-risk, and not just based on the HIRI-MSM. Additionally, although individuals with a previous history of PrEP use were not included in this study, the MI intervention can likely be applied to those with a previous history of PrEP use.

A notable advantage of MI, relative to other behaviorally-based therapies, is that it can be efficacious over a very short duration, with as few as two to four intervention sessions leading to positive outcomes²⁹. This is ideal in a busy clinical setting where contact with patients can be very brief. Thus, this intervention has the potential to be applicable and easily implemented in both clinical and community settings which engage MSM at increased risk for HIV. Moreover, as noted in a recent review³⁰, there are few existing, theoretically-driven PrEP uptake interventions. Indeed, to date, few interventions have demonstrated any efficacy for PrEP uptake³¹.

This intervention also addressed structural determinants of PrEP. Structural determinants of health “shape the distribution and availability of health-relevant resources, including systems of power and privilege in society.³²” Structural determinants related to the health system such as deductibles and copayments and assistance with accessing patient assistance programs were included as key components of this PrEP intervention; these details have been explored elsewhere³². Our results demonstrate that, taken together, this suite of intervention strategies was associated with a significant increase in PrEP uptake.

Limitations of our study include the generalizability of the sample. Our sample was somewhat small and mostly White cisgender MSM recruited from one site. Future studies should explore efficacy of the intervention in other populations with larger sample sizes. Participants were compensated during this trial, but compensation was equal for both conditions, and those participants receiving MI were not paid for attendance at MI sessions. While it is possible that the prospect of participant payment may have biased the sample, the sample chosen was unique and important given that patients were not initially PrEP-seeking. The research team felt that it was crucial to offer participants modest remuneration for their time and effort. In the future, with possible scale-up of MI for PrEP services, it may be worthwhile to consider whether paying participants to engage in MI could be a useful strategy for enhancing rates of PrEP initiation. Future research could further evaluate efficacy of the MI intervention while controlling for other predictors of PrEP use, such as condomless sex, number of partners, etc. The HIRI-MSM was used as a primary inclusion and screening tool. As noted in recent studies^33,34, the HIRI-MSM’s sensitivity in identifying individuals who are likely to acquire HIV is 63–85%, with lower sensitivity among people of color. While the HIRI-MSM is not a perfect screening tool, it identifies a sizable proportion of patients who might benefit from PrEP. We further note that only 10 participants were excluded due to a low HIRI-MSM score. The HIRI-MSM, coupled with our other intervention components, can be readily implemented in clinical settings due to its brevity.

Moreover, the control condition was the clinic’s usual care practices for patients who might benefit from PrEP. As such, intervention participants might have had increased contact and discussions with staff about PrEP. The next step would be to compare the intervention protocol to an attention-matched control condition. Although we monitored receipt of the PrEP prescription, we did not determine if participants actually filled the prescription. Expanding follow-up to assess this important next step, and potentially examining longitudinal adherence, as well, would be important in future research of this intervention. Individuals with a history of taking PrEP were excluded from this study. This was intentional due to our focus on MSM not already seeking or receptive to PrEP. However, future studies could test this approach with PrEP-experienced MSM. Lastly, a sizeable proportion of eligible patients either declined to participate or agreed, but never completed enrollment procedures. This highlights the importance of flexibly delivering an intervention, such as the one developed here, as the target population is not necessarily treatment seeking and might encounter other barriers to seeking or obtaining care. One solution might be to offer remote contact if time or travel is a barrier.

In conclusion, this is one of the first studies to demonstrate efficacy of an intervention to promote PrEP uptake among MSM at high risk for HIV acquisition. This intervention, suitable for real-world settings, coupled brief counseling, patient navigation and risk assessments to enhance PrEP uptake. The next step will be to assess the intervention’s effectiveness in larger and more diverse samples and its impact on downstream endpoints in the PrEP care continuum.