Preferences Among Physicians and Men Who Have Sex with Men (MSM) for a Long-Acting, Removable Implant for HIV Prevention: A Discrete Choice Study

Abstract

A longer acting, removable implant for HIV prevention has the potential to improve uptake of HIV pre-exposure prophylaxis (PrEP) by removing the need for daily adherence to an oral tablet, reducing potential side effects, and eliminating concerns about residual drug following injections. To end the HIV epidemic, we must understand the needs and preferences of groups most affected by HIV (e.g., men who have sex with men; MSM), and the physicians who prescribe PrEP to them. This article describes a discrete choice experiment to estimate the preference share for the implant within a competitive context of other PrEP products (including the oral tablet, dissolvable implant, and injection) and evaluate the impact of potential implant attributes. Physicians who had prescribed oral PrEP (n = 75) and MSM at risk for HIV (n = 175) completed a web-based survey that prompted decision-making about PrEP product preferences. The findings from both physicians and MSM demonstrated that the removable implant could capture a meaningful portion of the preference share, making it feasible to advance in the development pipeline as an important addition to the biomedical HIV prevention toolkit. Among MSM, specifically, the cost of treatment was the most important attribute impacting product preference. Our findings inform implant developers and future payers (e.g., commercial manufacturers, insurance companies) about specific device attributes that will likely affect MSM's willingness to use and physicians' willingness to prescribe this HIV prevention strategy.

Introduction

Longer acting biomedical HIV prevention products are in various stages of clinical or preclinical development, including long-acting injectable cabotegravir (CAB-LA) and cabotegravir implants. These long-acting products have the potential to revolutionize the way systemic HIV pre-exposure prophylaxis (PrEP) is used, compared with daily tenofovir/emtricitabine (TDF or TAF/FTC, referred to as TDF/FTC throughout) pills. For example, CAB-LA appears to be more effective and to have a different side effect profile compared with oral TDF/FTC among cis-gender men and women and transgender women,¹ can be administered over wider intervals (e.g., CAB-LA is given every 8 weeks),² and does not require end-users to remember to take a daily pill.

The longer acting cabotegravir implant is a particularly exciting potential future product. Implants that are removable are one of the few HIV prevention strategies in development that allow users to discontinue use by having the device removed at any time, while also requiring no daily maintenance.^3–5 Furthermore, preclinical research shows that cabotegravir implants could potentially overcome the issue of the cabotegravir “tail” inherent to CAB-LA injections (e.g., where cabotegravir remains in the system at a low, nontherapeutic dose long after discontinuation).⁶ Understanding the needs and preferences of potential cabotegravir implant users and prescribers is critical in the preclinical phase. Attributes that users and prescribers find appealing can be leveraged to optimize the overall design and early-phase trials of these long-acting products, adherence during trials, as well as uptake and sustained use following product approval.^4,7–10

More knowledge about the cabotegravir removable implant-related needs and preferences of groups that are disproportionately affected by HIV (e.g., men who have sex with men; MSM),^4,10 and the physicians who prescribe PrEP to them could help decrease the spread of HIV. As of 2018, although MSM accounted for 69% of new HIV diagnoses in the United States,¹¹ PrEP uptake is low and increased use could help avert new infections.¹² Barriers to oral PrEP uptake and adherence are well-documented in this population, and include cost, side effects, conflicts with partners over PrEP use, low health system accessibility, competing stressors that influence the ability to use a daily product, and stigma (e.g., PrEP users may be perceived as promiscuous or mistakenly identified as a person living with HIV).^13–18

More information is also needed to understand physicians' perspectives on prescribing PrEP to eligible MSM.^19,20 If a long-acting, removable cabotegravir implant could be developed with the needs and preferences of these populations incorporated into product design, this could help to overcome these barriers. A user-friendly implant could become an additional HIV prevention option for individuals who cannot or will not use either oral TDF/FTC or CAB-LA for PrEP. This article explores end-user preferences among HIV prevention options, including a long-acting, removable implant, by conducting a discrete choice experiment (DCE).

Materials and Methods

Participants and procedures

This substudy was part of the Sustained Long-Acting Prevention against HIV (SLAP-HIV) research program,⁵ a program whose overall objective was to develop and clinically test a long-acting drug delivery system of a fourth-generation antiretroviral for prevention and to investigate the acceptability of a wide variety of PrEP designs among potential users and prescribers. The research design for this phase included two separate groups of participants: physicians and HIV-negative MSM at risk for HIV. Eligible physicians had to report seeing at least 50 patients per week (including 10 or more MSM), and having recommended or prescribed oral PrEP to 10 or more MSM in the past 3 months. Eligible MSM were 18–45 years old, self-reported HIV-negative status, aware of PrEP or Truvada, and reported having had two or more male sexual partners in the past 3 months, with at least one condomless anal sex act in the past month.

Participants were enrolled from October 23 to November 3, 2017, and recruited from an established panel that included physicians (Decision Analyst's Physicians Advisory Council^® panel) and other adults (including MSM) who previously agreed to participate in marketing research surveys. They were initially contacted through an email invitation asking them to participate in a web-based survey of their preferences for emerging long-acting HIV prevention products (e.g., long-acting PrEP injections, subdermal PrEP implants). They were informed that the survey was estimated to require ∼20 min to complete. In the email, potential respondents clicked on a hyperlink, or cut and pasted the web address into their browser, to go to a screening questionnaire to determine their eligibility for the web-based survey. The eligibility screening questionnaire was hosted on Decision Analyst's password-secured server. If the respondents met the eligibility requirements, they were invited to participate in the web-based survey.

All respondents who completed the web-based survey were paid an incentive or honorarium in appreciation for their time ($10 for MSM and $35 for physicians, based on market industry standards). All data were encrypted and stored behind a secure firewall.

Participant background

The following measures were collected to characterize MSM participants: demographic variables (age, geographic residence, household income, employment status, education attainment, health insurance coverage, and deductible), number of male sexual partners and occasions of condomless anal intercourse in the past 3 months, and sexually transmitted infection (STI) history. Data collected to characterize physicians included information about their professional practice (specialty, practice setting and location, years in practice, gender, patient volume), experience with MSM patients, and oral PrEP prescribing behavior.

Discrete choice modeling

Choice modeling was developed by economists and cognitive psychologists to study decision-making and is a well-accepted method used to simulate individual respondents' decision-making processes. It is based on the premise that people behave rationally when considering their choices.^21,22 DCEs ask respondents to choose between different products or services based on a group of diverse attributes. Respondents are thereby required to consider both preferred and less preferred attributes in each product selection. Analysis of choices made reveals the true preference for each attribute in relation to the others by allowing observation of overall like/dislike trends. Discrete choice modeling estimates preference shares.

Preference reflects what physicians/MSMs value. It is based on perfect information (100% awareness of all products and their prices, 100% distribution of all products tested, etc.).

Market share represents what physicians/MSMs actually do. A choice model normally yields preference shares; however, in theory, for a given scenario, postsimulation market share can be inferred from existing shares' data and the preference shares resulting from a discrete choice model. A choice model approximates potential market shares, given the proper calibration for advertising/promotion, competitive presence, distribution, and time.

This DCE was designed to identify specific combinations of implant attributes that are most appealing both to MSM end-users and the physicians who might prescribe these medications, and the preference share for this product within a competitive context of other HIV prevention products (e.g., longer acting PrEP injections, daily oral TDF/FTC pills) that could exist on the market together. Respondents in this study completed a DCE in which they compared different treatment options that have a future potential to coexist on the market together (pill, injection, replaceable implant). Tables 1 and 2 show the baseline treatment options presented to MSM and physician groups for each product.

Table 1.

Baseline Treatment Options (Men Who Have Sex with Men Group)

Product attribute	Pill	Injection	Removable implant
Description
Length of protection/quantity	One pill taken orally each day	One injection every 2 months	Implants are removed and new implants are inserted every 12 months
Chances of becoming infected with HIV each year	One person out of 30 on the treatment are infected	One person out of 75 on the treatment are infected	One person out of 500 on the treatment are infected
Side effects	Rare side effects such as kidney damage, bone weakening, and nausea	Rare side effects (of the pill) are mostly eliminated	Rare side effects (of the pill) are mostly eliminated
Possible problems	Forgetting to take pill daily reduces effectiveness	Mild injection-site reactions	Possibility of small scars; rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Pill can be stopped without problem anytime	Injection is not removable	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $275/month ($3,300/year)	Typical copay with insurance is $650 every other month ($3,900/year)	Typical copay with insurance is $4,400 per year

Table 2.

Baseline Treatment Options (Physician Group)

Product attribute	Pill	Injection	Removable implant
Description
Length of protection/quantity	One pill taken orally each day	One injection every 2 months	Implants are removed and new implants are inserted every 12 months
Chances of becoming infected with HIV each year	One person out of 30 on the treatment are infected	One person out of 75 on the treatment are infected	One person out of 500 on the treatment are infected
Time required to learn necessary procedures	None	None	30 min
Time required to complete procedures	None	2 min	5 min for insertion, 10 min for removal
Typical reimbursement for procedure	N/A	$300 for injection	$850 for removal of implants and insertion of new implants
Side effects	Rare side effects such as kidney damage, bone weakening, and nausea	Rare side effects (of the pill) are mostly eliminated	Rare side effects (of the pill) are mostly eliminated
Possible problems	Forgetting to take pill daily reduces effectiveness	Mild injection-site reactions	Possibility of small scars; rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Pill can be stopped without problem anytime	Injection is not removable	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $275/month ($3,300/year)	Typical copay with insurance is $650 every other month ($3,900/year)	Typical copay with insurance is $4,400 per year

N/A, not applicable.

During the DCE, each screen displayed two treatment options with various features, including length of protection/quantity, chance of becoming HIV infected, cost, side effects, possible problems, and reversibility of medications. Attributes were based on prior foundational research to explore preferred product attributes and barriers and facilitators to PrEP product use.^{3,4,6,7,9,10,23} In the physician group, participants were presented with additional features, including time required to learn necessary procedures, time required to complete procedures, and typical reimbursement for procedures. Respondents were asked to select their preferred product, based on these features. The MSM then rated their likelihood to use the preferred treatment, while physicians rated their likelihood to recommend the treatment to their patients.

Statistical analysis

Decision Analyst conducted the DCE analysis, using a hierarchical Bayesian model for MSM and a latent class choice model for physicians. The resulting model coefficients were used to develop the attribute importance, preferences, and optimal combinations. For the preference shares, overstatement was adjusted by calibrating the pill (when in the market alone, without any other products) to a 20% preference share, based on current uptake among those eligible.²⁴ The same calibration factor was applied to the injection and implant options. Decision Analyst calculated the proportion of the preference share for the long-acting removable implant in the context of market availability of oral PrEP and CAB-LA injections.

Ethical considerations

This study was reviewed and considered not to be human subjects research by the Northwestern University IRB.

Results

Demographic characteristics

Table 3 summarizes the demographic characteristics of the MSM (N = 175) and physician samples (N = 75). The MSM participants had a mean age of 31 years, and a mean household income of $63,000. More than half of the sample reported white race (56.6%), and the majority were employed full-time (75.0%), had employer-based health insurance (64.0%), and had a deductible of <$5,000 (68.0%). The sample was highly educated, where 34.0% were college or trade school graduates and another 28.0% had attended or completed graduate school. They reported having multiple sex partners (a mean of seven male sex partners in the past 3 months), and five occasions of condomless anal intercourse in the past month, on average; about one-third of participants reported ever having had a previous chlamydia infection (34.0%) and/or a prior gonorrhea diagnosis (33.0%).

Table 3.

Participant Demographics

Characteristics	Mean	N (%)
MSM (N = 175)
Age	31
Household income	$63,000
No. of male sex partners in the past 3 months	7
No. of occasions of condomless anal intercourse in the past month	5
Race
White		99 (57)
Black/African American		76 (43)
Education
High school or less		11 (6)
Some college or associate's degree		56 (32)
College or trade/technical school graduate		60 (34)
Attended or completed graduate school		48 (28)
U.S. geographic region
South		61 (35)
West		46 (26)
Midwest		36 (21)
Northeast		32 (19)
Employment status
Full-time		131 (75)
Part-time		13 (8)
Student		11 (7)
Not employed		11 (6)
Self-employed or homemaker		9 (5)
Prior diagnosis of STIs other than HIVa
Chlamydia		59 (34)
Gonorrhea		57 (33)
Syphilis		32 (18)
Genital herpes		13 (7)
HPV		8 (5)
Other		5 (3)
None of these		78 (44)
Health insurance coverage
Employer-based insurance		112 (64)
Private insurance purchased on an exchange		26 (14)
Medicaid		15 (9)
Medicare		15 (9)
No insurance/unsure		19 (11)
Health insurance deductible amount (among those with insurance)
<$5,000		106 (60)
5,000+		21 (14)
Don't know/unsure		29 (19)
Physicians (N = 75)
Years of practice since residency	15
Weekly patient volume	190
Weekly MSM patient volume	28
No. of MSM prescribed TDF/FTC in the past 3 months	31
Sex
Male		56 (75)
Female		19 (25)
Primary specialty
Internal medicine		37 (49)
Family medicine		12 (16)
Primary care		6 (8)
Infectious disease		5 (7)
Other		15 (20)
U.S. geographic region
Northeast		26 (35)
South		22 (29)
West		18 (24)
Midwest		9 (12)
Setting
Urban		38 (51)
Suburban		33 (44)
Rural		4 (5)
Primary practice setting
Private, single-specialty group practice		26 (35)
Hospital based		15 (20)
Private, multiple-specialty group practice		14 (19)
Solo private practice		12 (16)
Academic medical center		4 (5)
Community health center		4 (5)

^a Percentages add up to more than 100% because participants could select all that applied.

HPV, human papillomavirus; MSM, men who have sex with men; STI, sexually transmitted infection; TDF/FTC, tenofovir/emtricitabine.

In addition, not shown in Table 3, 2% of participants endorsed agreeing or strongly agreeing that they were worried about contracting HIV.

In the physician group, most participants were male (75.0%), reported a mean of 15 years of practice since residency, a mean weekly patient volume of 190, where an average of 28 patients were MSM. In the past 3 months, physicians reported prescribing PrEP to a mean of 31 MSM. About one-half of physicians were internal medicine practitioners (49.0%) in urban settings (51.0%). Just over one-third were located in the Northeast (35.0%), and were housed in private, single-specialty group practice (35.0%).

Estimated preference shares of PrEP product options

Table 4 summarizes estimated preference shares for all three products for MSM and physicians. Among MSM, an estimated 25.4% of the market would use one of the three presented products; the pill represents 9.1% of the preference share, the CAB-LA injection 7.0%, and the removable implant 9.2%. Among black/African-American MSM, the preference share is highest for the pill (9.5%), followed by the removable implant (8.7%) and the injection (7.6%). Among white MSM, the preference share is highest for the removable implant (9.7%), followed by the pill (8.8%) and the injection (6.5%). Estimated preference shares based on physicians' preferences were 5.4% (pill), 5.5% (injection), and 14.0% (removable implant). Remaining MSM and physicians would recommend/use none of the three tested prescription products or choose some other course of recommendation or action, such as condom use, frequent testing, or another modified behavior to prevent HIV.

Table 4.

Preference Shares for Baseline Treatment Options: Men Who Have Sex with Men and Physicians

Population	Pill (%)	Injection (%)	Removable implant (%)	No product (%)
Market share by product (MSMa)
Black/African American MSM (n = 76)	9.5	7.6	8.7	74.2
White MSM (n = 99)	8.8	6.5	9.7	74.9
Total MSM of all races (N = 175)	9.1	7.0	9.2	74.6
Market share by product (physiciansb)
Total physicians (N = 75)	5.4	5.5	14.0	75.1

^a Results adjusted for overstatement by calibrating the pill (when in the market alone) to a 20% market share.

^b Eighty-three percent of physicians self-reported that they would recommend at least one of these three products to their patients.

Optimal combinations and maximum cost configurations of the removable implant

The goal of the DCE was to compare the implant with other potential PrEP products. Table 5 shows how the preference share for the removable implant varies based on optimal attribute combinations among MSM. The combinations varied in typical annual copay cost ($4,440 vs. $440), chance of infection (1/500 vs. 1/300), and side effects (rare pill side effects mostly eliminated vs. partially reduced). Across all MSM, when the removable implant is offered at the lowest cost, it has the highest preference share, increasing by 4.0% from the baseline combination. Other combinations such as chance of infection or side effects did not appear to impact preference share.

Table 5.

Removable Implant Optimal Attribute Combinations: Men Who Have Sex with Men

Product attribute	Base implant	Implant 1	Implant 2	Implant 3	Implant 4
Description	Two 3.7 mm implants
Length of protection/quantity	Implants removed and new implants inserted every 12 months
Chances of becoming infected with HIV each yeara	One person out of 500 on implant is infected	One person out of 500 on implant is infected	One person out of 300 on implant is infected	One person out of 500 on implant is infected	One person out of 300 on implant is infected
Side effects	Rare side effects of pill mostly eliminated	Rare side effects of pill mostly eliminated	Rare side effects of pill mostly eliminated	Rare side effects of the pill are cut in half	Rare side effects of the pill are cut in half
Possible problems	Possibility of small scars, rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $4,440 per year	Typical copay with insurance is $440 per year
Preference shares (MSM)
Black/African American (%)	8.7	12.5	12.4	12.4	12.4
White (%)	9.7	14.0	14.0	13.9	13.9
Total (%)	9.2	13.3	13.3	13.2	13.2

Across MSM groups, the base version of the removable implant offered at the lowest cost has the highest preference share (∼4 points above the base case). This is Implant 1. Assumes the pill and injection treatments are included in the market at their respective base levels.

^a Modifying the chance of infection (e.g., 1 person out of 500 on implant becomes infected) or possible side effects (rare side effects of pill are mostly eliminated) had only a slight impact on preference when using the base implant.

Table 6 shows how the preference share for the removable implant varies when the maximum possible copay is held constant among MSM. Modifying the chance of infection (1/500 vs. 1/300), size and number of implants (two 3.77 mm implants vs. three 2.55 mm implants), or possible side effects (mostly eliminated vs. cut in half) had a very slight to negligible impact on preference share.

Table 6.

Removable Implant Attribute Combinations Holding Maximum Cost Constant: Men Who Have Sex with Men

Product attribute	Base implant	Implant 1	Implant 2	Implant 3	Implant 4
Description	Two 3.7 mm implants				Three 2.5 mm implants
Length of protection/quantity	Implants removed and new implants inserted every 12 months
Chances of becoming infected with HIV each year	One person out of 500 on implant is infected	One person out of 300 on implant is infected	One person out of 500 on implant is infected	One person out of 300 on implant is infected	One person out of 500 on implant is infected
Side effects	Rare side effects of pill mostly eliminated	Rare side effects of pill mostly eliminated	Rare side effects of the pill are cut in half	Rare side effects of the pill are cut in half	Rare side effects of pill mostly eliminated
Possible problems	Possibility of small scars, rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $4,440 per year
Preference shares (MSM)
Black/African American (%)	8.7	8.6	8.6	8.5	7.7
White (%)	9.7	9.7	9.6	9.6	9.1
Total (%)	9.2	9.2	9.2	9.1	8.5

Across all MSM groups, keeping the base case cost, and modifying the chance of infection or possible side effects had only a slight impact on preference.

Assumes the pill and injection treatments are included in the market at their respective base levels.

Table 7 shows how the preference share for the removable implant varies based on optimal attribute combinations among physicians. The base implant, which had the longest protection duration (12 months) but the highest typical annual copay ($4,440), had the lowest preference share (14.0%), while Implants 1–4, which had a protection duration of between 6 and 12 months and a typical annual copay of between $440 and $640, had similar preference shares of 16.0%.

Table 7.

Removable Implant Optimal Attribute Combinations: Physicians

Product attribute	Base implant	Implant 1	Implant 2	Implant 3	Implant 4
Description	Two 3.7 mm implants
Length of protection/quantity	Implants removed and new implants inserted every 12 months	Implants removed and new implants inserted every 6 months	Implants removed and new implants inserted every 6 months	Implants removed and new implants inserted every 6 months	Implants removed and new implants inserted every 12 months
Chances of becoming infected with HIV each year	One person out of 500 on implant is infected
Time required to learn necessary procedures	30 min
Time required to complete procedures	5 min (for insertion) 10 min (for removal)
Typical reimbursement for procedure	$850 for removal of implants and insertion of new implants
Side effects	Rare side effects of pill mostly eliminated
Possible problems	Possibility of small scars, rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $4,440 per year	Typical copay with insurance is $440 per year	Typical copay with insurance is $540 per year	Typical copay with insurance is $640 per year	Typical copay with insurance is $440 per year
Preference shares (physicians)
Physicians (%)	14.0	16.0	16.0	16.0	16.0

Assumes the pill and injection treatments are included in the market at their respective base levels.

Table 8 shows the preference shares for the removable implants when the maximum possible cost is held constant, among physicians. In the maximum cost configuration, the base implant, which would require two 3.7 mm implants, removed every 12 months, with rare pill side effects mostly eliminated, has the highest preference share (14.0%) by a slight margin (Implant 1 = 13.6%).

Table 8.

Removable Implant Attribute Combinations Holding Maximum Cost Constant: Physicians

Product attribute	Base implant	Implant 1	Implant 2	Implant 3	Implant 4
Description	Two 3.7 mm implants	Two 3.7 mm implants	Three 2.6 mm implants	Two 3.7 mm implants	Three 2.6 mm implants
Length of protection/quantity	Implants removed and new implants inserted every 12 months	Implants removed and new implants inserted every 6 months	Implants removed and new implants inserted every 12 months	Implants removed and new implants inserted every 12 months	Implants removed and new implants inserted every 12 months
Chances of becoming infected with HIV each year	One person out of 500 on implant is infected
Time required to learn necessary procedures	30 min
Time required to complete procedures	5 min (for insertion) 10 min (for removal)
Typical reimbursement for procedure	$850 for removal of implants and insertion of new implants
Side effects	Rare side effects of pill mostly eliminated	Rare side effects of pill mostly eliminated	Rare side effects of pill mostly eliminated	Rare side effects of the pill are cut in half	Rare side effects of the pill are cut in half
Possible problems	Possibility of small scars, rare possibility of implant moving
Reversibility of medication (in case of side effect or desire to stop medication)	Implants can be removed anytime
Cost of treatment for those with health insurance	Typical copay with insurance is $4,440 per year
Preference shares (physicians)
Physicians (%)	14.0	13.6	13.5	13.4	13.0

Assumes the pill and injection treatments are included in the market at their respective base levels.

Discussion

This study compared estimated preference shares for existing (oral PrEP pill) and potential future longer acting biomedical HIV prevention products (injection, removable implant) among MSM and physicians (Supplementary Data). In addition, it examined the impact of several attributes on variations in estimated preference share for a removable cabotegravir-eluting implant for HIV PrEP based on the preferences of MSM and physicians. The findings demonstrate that the removable HIV prevention implant has the potential to capture a meaningful portion of the preference share, making it a feasible product to move forward in the development pipeline. It should be noted that the sample size of subgroups does not allow us to detect statistically significant differences in preferences among black/African American versus white MSM.

These findings indicate that the long-acting implant could be equally or more impactful at the population level for preventing HIV in MSM as the oral pill or injection. Thus, the removable implant is a realistic and important addition to the biomedical HIV prevention toolkit. However, implant developers and future payers (e.g., insurance companies, commercial manufacturers) should remain vigilant about specific device attributes that will likely affect MSM's willingness to use and physicians' willingness to prescribe this HIV prevention strategy. Specifically, for MSM, cost of treatment appears to be the most important attribute of the removable implant. In the assessment of optimal implant configurations, reducing the cost from $4,440 annually to $440 annually increased the preference share by nearly 4 percentage points among MSM. This held true when other attributes of the implant remained identical to the baseline device and when they varied (e.g., level of protection increases, change in side effect profile decreases).

This highlights the importance of cost in the development and rollout process. Given this, payers (e.g., public and private insurance companies) and pharmaceutical companies should prioritize ensuring that implants remain low or no cost after they receive regulatory approval. There is some precedent for this, through the Gilead's Advancing Access program, which offers copay coupons (commercially insured persons can receive up to $4,800 in rebates per year) or no-cost access (uninsured individuals) for their oral TDF/FTC and/or TAF products.²⁵ These cost parameters fit well into those outlined in the current study (e.g., removable implant max cost = $4,400 per year), suggesting that this expense could potentially be covered for qualifying patients.

In this sample, the majority of MSM were employed full-time (75.0%) and had employer-based health insurance (64.0%). If covered by such insurance, this prevention tool would be accessible. However, costs may limit access among those without insurance and those needing to pay high deductibles. Youth, in particular, face additional challenges. Under the current configuration of the Advancing Access program, youth and young adults up to age 26 are unable to access copay coupons without using their parents' insurance, which poses a barrier for those who wish to keep their PrEP use private.²⁶ Future iterations of a similar program could consider forming alternate pathways for young people in this situation to confidentially access implants at low or no cost. (e.g., a third low/no-cost confidential option where individuals younger than 26 years provide documentation of plan coverage under a parent and receive no-cost implants without charging insurance).

Cost to patients was an important concern for physicians as well. When presented with optimal attribute combinations, reducing the cost from $4,440 annually to $440 annually increased the preference share by 2 percentage points (14.0%–16.0%) when all other attributes were held constant. Furthermore, cheaper implants were preferred (copays of $640–$440) even in the case that devices had to be changed every 6 months, rather than every 12. On the contrary, when the maximum copay was held constant, the preference share varied by ≤1% across all implant combinations. This shows that variations in duration of protection, number and size of implants, and likelihood of side effects did not impact the preference share to the same extent as cost, among physicians as well.

Another key consideration is that the preference share among black MSM is lower than that of white MSM for all implant combinations across optimal and maximum cost configurations. This is not unexpected. Medical mistrust of both the pharmaceutical industry and health care providers continues to hamper PrEP uptake and health care engagement in black MSM communities.^27–30 The Tuskegee Syphilis Study and other medical and scientific abuses of the black community remain an important frame of reference for how black MSM interpret the intentions and motivations of clinical researchers, health care providers, and drug companies.^27,29,31 This requires the scientific community's immediate and sustained attention.

In a study to understand how to build trust to improve engagement in HIV/STI research with black MSM, respondents reported that the following are important: (1) authentic engagement with the community, (2) increased transparency of the research process, (3) capacity building of research staff from the black community, and (4) a balance of research and action.³² Thus, if HIV researchers intend for the products they develop to be appealing and effective for black MSM, they must take steps to make a greater investment in the black MSM community as a whole. Despite the barriers created by repeated and prolonged medical mistreatment, many black MSM are willing to consider PrEP products after the drugs spend time on the market and their effects on people and groups in a nontrial, real-world setting are clear.²⁷ Given this, it is possible that the preference share among black MSM could grow over time.

Finally, another critical point to mention is that while the implant garnered a similar preference share to other PrEP products, the majority still reported choosing not to use any HIV prevention product, which is a finding that merits future attention to ensure successful rollout of the full range of PrEP products. It is possible that greater communication campaigns, cost coverage, or other factors will be needed to ensure PrEP uptake among a greater proportion of people at risk for HIV.

Limitations

The limitations of this study are that some relevant sociodemographic variables were not assessed, including relationship status/stability and self-assessment of HIV risk, thus limiting characterization of the sample in terms of HIV risk. Also, as mentioned above, the sample size did not allow us to detect differences in product preferences among subpopulations of MSM, which limits the interpretability of the nuances of preference shares by race. In addition, when these data were collected, injectable CAB-LA was not yet FDA approved. That is, participants could have perceived CAB-LA differently, before its regulatory approval (e.g., that it is being “tested” and therefore not as reliable). Furthermore, cost will vary greatly by insurance coverage, and the figures on cost were designed to reflect base differences in products that could be expected, to enable participants to compare the options. Some level of wrap-around services would be required of all products as well.

Finally, not all HIV prevention strategies were assessed in this DCE, and it is possible that other attributes in addition to those explored here could impact preferences.

Conclusions

Longer acting, removable implants for PrEP are in the early stages of development. The DCE showed that these implants occupy a preference share similar to oral and injectable PrEP strategies. While more work is needed on equitable prevention efforts among MSM as well as at-risk populations not studied here (e.g., youth, gender nonconforming individuals), these results indicate that adding removable implants to our HIV prevention toolkit is a promising potential future way to improve prevention of new HIV infections.

Authors' Contributions

C.T.R.: conceptualization (equal); formal analysis (equal); writing—original draft (lead); writing—review and editing (lead); and visualization (lead). R.G.: conceptualization (equal); formal analysis (equal); writing—original draft (lead); writing—review and editing (lead); and visualization (lead). S.S.: methodology (lead); formal analysis (lead); conceptualization (lead); and writing—review and editing (equal). E.H.: methodology (lead); formal analysis (lead); conceptualization (lead); and writing—review and editing (equal). R.J.S.: conceptualization (lead); writing—original draft (supporting); and writing—review and editing (equal). G.J.G.: conceptualization (supportive) and writing—review and editing (equal). R.D.: conceptualization (supportive) and writing—review and editing (equal). E.B.T.: conceptualization (supportive) and writing—review and editing (supportive). P.F.K.: conceptualization (supportive) and funding acquisition (supportive). T.J.H.: funding acquisition (lead); conceptualization (supportive); and writing—review and editing (supportive).

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the SLAP-HIV Program (UM1 AI120184 to T.J.H. and P.F.K.).

Supplementary Material

Supplementary Data