The Invisible Product: Preferences for sustained-release, long-acting Pre-Exposure Prophylaxis to HIV among South African youth

Abstract

Background:

Long-acting injectable and implantable approaches aim to overcome some of the documented challenges with uptake and adherence to current HIV prevention methods. Youth are a key end-user population for these methods. We used qualitative methods to examine product attributes and preferences for current and future long-acting HIV prevention approaches.

Methods:

Ninety-five South African youth aged 18–24, of whom 62 were female and 33 male, completed 50 interviews and 6 focus groups. We purposively selected for previous product experience including oral PrEP, injectable PrEP, or the vaginal ring, to ensure participants’ opinions were rooted in actual experience.

Results:

Irrespective of previous method-use experience, gender, or sexual orientation, the majority expressed a preference for prevention methods formulated as injectables or implants. Several mentioned that their top priority in any product was efficacy, and for some, this overrode other concerns; for example, even if they feared pain, an implant or an injectable would be used if fully protective. While efficacy was a top priority, there was also a clear desire across all subgroups for a product that wouldn’t interfere with sex, would stay in the system to provide protection, and that caused minimal burden, or wasn’t apparent to others; and these characteristics were most salient for long-acting methods.

Conclusions:

Narrative explanations for preferences converged thematically around different dimensions of “invisibility” including invisibility to oneself, one’s partner and household members, and community members. End-user preferences can be used to inform product development of long-acting HIV prevention approaches formulated as injections or implants to optimize adherence and impact.

Background:

The field of HIV prevention has made notable recent progress through the established safety and effectiveness of oral pre-exposure prophylaxis (PrEP) [1, 2] and the safety and partial effectiveness of a topical microbicide vaginal ring.[3, 4] While these technologies provide valuable prevention options, users of both have faced adherence challenges.[3, 5, 6] PrEP approaches that are longer-acting (e.g. 2–3 months or longer) are at earlier stages of development, and aim to overcome some of the documented challenges with uptake and adherence to current prevention methods.[7] Partially in response to clinical trial failures, [8–10] there is enhanced appreciation for end-user research during preclinical and early clinical stages of product development, to identify and potentially modify features of a method that might increase its chances for successful adoption and consequently, its public health impact.

Youth in HIV endemic settings such as South Africa are a key end-user population for long-acting methods because of their increased risk of HIV acquisition, and their documented challenges with adherence.[11] Young people have distinct risk profiles shaped by partner characteristics and partnership patterns, relationship power imbalances, and limited access to confidential, youth-friendly sexual health services.[12, 13] Furthermore, young people experience unique social and structural influences (e.g., family environment, community stigma, health delivery systems constraints) that, in addition to dynamic neurocognitive development, [14–16] shape risk perception and adoption of HIV prevention practices. In South Africa, many youth face additional challenges presented by impoverished living conditions, high unemployment, community violence and socio-cultural tensions between traditional and modern urban life.[17] Indeed, research exploring the acceptability of novel HIV prevention strategies prior to dissemination have highlighted the complexity of young people’s lives in this setting.[6, 18, 19] This paper presents findings from one of the first studies to explore the attitudes and preferences of young male and female South African end-users regarding long-acting sustained-release HIV prevention approaches formulated as injections or implants. Given that youth constitute a key target group for these products, examining their perspectives is essential to inform product development in ways that might enhance adoption and use.

Methods:

Study design and setting.

The iPrevent study is a two-stage research study employing both qualitative and quantitative methods to identify aspects of long-acting PrEP that are important to male and female South African youth aged 18–24. In-depth interviews (IDIs) and focus group discussions (FGDs) were conducted to identify individual, social, and structural factors thought to affect PrEP uptake and adherence. Participants were primarily recruited from two clinical site catchment areas in Cape Town: Emavundleni, a high-density township approximately 21 kilometers east of the City Center, and Masiphumelele, a more remote township on the Cape Peninsula, approximately 38 kilometers south of the City Center.

Study sample.

To ensure participants could provide opinions rooted in actual experience, youth who had previous experience with a range of products (either the active or placebo version of the dosing platform) were purposively selected. For IDIs, the target sample included males who had participated in an oral PrEP study; and females who had experience using oral PrEP; sustained-release antiretroviral (ARV) PrEP (e.g. the monthly dapivirine ring); or injectable PrEP (see Table 1). Participants exiting from the designated trials who were HIV negative at trial exit, met the age criterion, and gave permission to be contacted for future studies were eligible for recruitment. For enrollment, participants who reported current sexual activity (vaginal or anal sex at least 3 times in the last 3 months) and who were currently 18 to 24 years old were eligible. For FGDs, participants were purposively recruited at the two clinical sites through former study participant rosters and community outreach according to the following criteria: (1) females who had experience using a contraceptive implant (e.g. Implanon NXT^® or Jadelle), (2) females with no previous implant experience, and (3) males. Sexual orientation was not a criterion for FGD grouping, however because of recruitment source, one male FGD was comprised of MSM only (n=7) and the other was entirely heterosexual (n=10).

Table 1.

Distribution of previous PrEP and contraceptive method experience, by gender and interview type

	IDI			FGD
	Oral PrEP-experienced1	Sustained-Release (ring) PrEP-experienced2	Injectable PrEP- experienced3	Contraceptive Implant-experienced4	Contraceptive Implant-naïve5
Dosage description	Daily oral Truvada®	Monthly vaginal ring (dapivirine or placebo)	Daily oral Truvada (4wks); Two bimonthly IM injections in buttocks (rilpivirine or placebo)	Subcutaneous in upper arm (Implanon NXT® or Jadelle)	NA
Male (n)	18	-	-	-	17
Female (n)	10	10	12	12	19

¹Men recruited from iPrex OLE and CHAMPS-SA; Women from CHAMPS-SA

²Recruited from MTN-020/ASPIRE[3]

³Recruited from HPTN-076

⁴Recruited from the community. Given contraceptive implants are not available for men, males not recruited in this sub-group.

⁵Recruited from the community; 2 male and 1 female participant recruited from IDI (Oral PrEP-experienced)

Data collection.

The design of IDI guides was informed by a socio-ecological framework[20, 21], which incorporates the influence of factors at the individual, interpersonal, and structural levels, and has been used in several previous HIV prevention studies to understand product acceptability and use.[6, 22–24] Questions explored individual narratives about HIV and HIV risk perceptions, partner characteristics and patterns, living situations, and sexual behavior, and how these influenced youth’s interest in and use of PrEP. Interviews solicited impressions on long-acting injectables and implants, and perceptions of these products relative to the PrEP method(s) they previously used. In addition, youth’s perceptions of health care providers and facilities, and social norms—including stigma—around young people and sex, HIV and risk behavior were explored. During IDIs, participants were shown a pictorial tool that contained pictures of existing family planning and HIV prevention methods, with a description of key product features. This tool was used to explore knowledge of, attitudes towards, and experiences with different delivery forms, and to discuss specific product attributes of each approach. IDIs with male and female youth were conducted at research clinic sites in English or isiXhosa by trained South African social scientists. All interviewers were female; for males, comfort with a female interviewer was confirmed as part of the screening and/or informed consent processes, prior to data collection, and no males declined participation.

FGD guides concentrated on the implant as a long-acting method, with particular emphasis on obtaining feedback on prototypes of a biodegradable thin film polymer device being developed for HIV prevention.[25, 26] Topics covered included personal experience with and community perceptions of contraceptive implants, physical attributes of PrEP implants in development (e.g., size, duration, flexibility, biodegradability), insertion process, and social adoption factors (e.g., discreet use, service delivery). During FGDs, participants were shown pictures and examples of contraceptive and prototype HIV preventative implants and applicators for insertion, and were able to handle them.[27] The pictorial tool used in IDI was also shown. FGDs were co-facilitated by a biomedical engineer from the United States and a South African social scientist, using a combination of English and isiXhosa. Interview guides and pictorial tools for IDI and FGD are available in the supplementary materials.

Analysis.

Interviews were audio-recorded, transcribed and translated (if applicable). Interview transcripts were reviewed for clarity and queried by analysts on the protocol team. Once finalized, they were uploaded into Dedoose software (v 7.5.9 – v 7.6.23). IDI transcripts were coded using a structured codebook that was iteratively developed by a multinational coding team. At the beginning of coding, two transcripts were triple-coded (independently coded by each of the 3 coding team members) and discrepancies discussed. Throughout the process, the team evaluated inter-rater reliability (IRR), and had regular (e.g., weekly) meetings to resolve disagreements in code applications and to refine the codebook. FGD transcripts were coded using a modified version of the IDI codebook. Two IRR tests were taken throughout the coding process, and disagreements in code applications were resolved by consensus.

For this analysis, key codes related to study product preferences were summarized by participant group. Further iterative review of code reports was done to thematically organize and interpret the dominant themes of the data within the conceptual framework of a socio-ecological model.[21]

Ethical review committees from each collaborating institution provided oversight to the conduct of this research.

RESULTS

Study population.

A total of 95 participants, of whom 62 were female and 33 male, completed 50 interviews and 6 FGDs. FGD groups ranged in size from 4–10 participants (median 8). Participants were, on average, 22 years old, 15% of the sample identified as heterosexual male, 63% heterosexual female and 20% men who have sex with men (MSM). Ninety-one percent were in a relationship, but only 1% were married. The majority (94%) described themselves as Xhosa. Approximately two-thirds (63%) of the study sample had completed secondary school and one-third attended some college. Nevertheless, less than half (48%) received any income, including social grants. Approximately half the men and a quarter of the women also reported casual partners; half (47%) indicated that they knew or suspected that their main partner had another partner. Condoms were reportedly used during the last sex episode by 61% of respondents (Table 2).

Table 2.

Select Study Population Characteristics (n=95)

	All Participants (n=95)		By sex/sex partner preference
			Heterosexual Men (n= 14)		MSM (n=19)		Women (n=62)
	N	(%)	N	(%)	N	(%)	N	(%)
Socio-demographics
Age at time of interview**	21.5	(20,23)	21.4	(19,23)	21.6	(19,24)	21.5	(20,23)
Number of children**	0.4	(0,1)	0.2	(0,0)	-	(0,0)	0.5	(0,1)
How long lived in this location (years)**	12.7	(4,20)	10.3	(3,16)	15.2	(6,22)	12.5	(3,21)
Relationship status
In a partnership	86	(91)	12	(86)	16	(84)	58	(94)
Single	8	(8)	2	(14)	3	(16)	3	(5)
Married	1	(1)	-	-	-	-	1	(2)
Ethnic Group
Xhosa	89	(94)	11	(79)	18	(95)	60	(97)
Other	5	(5)	2	(14)	1	(5)	2	(3)
Zulu	1	(1)	1	(7)	-	-	-	-
Currently receive income	46	(48)	5	(36)	8	(42)	33	(53)
Sources of Income
Formal employment	26	(27)	4	(29)	4	(21)	18	(29)
Social grant	11	(12)	-	-	-	-	11	(18)
Other	9	(10)	1	(7)	3	(16)	5	(8)
Self-employment	6	(6)	-	-	2	(11)	4	(7)
Highest level of education
Secondary schooling, not complete	35	(37)	6	(43)	3	(16)	26	(42)
Secondary schooling, complete	29	(31)	2	(14)	10	(53)	17	(27)
Attended college or university, not complete	22	(23)	2	(14)	6	(32)	14	(23)
Attended college or university, complete	9	(10)	4	(29)	-	-	5	(8)
Sexual History and Activity
Age of first penetrative sex**	16.4	(16,18)	15.4	(13,18)	16.3	(15,18)	16.7	(16,18)
Number of sexual partners in lifetime**	6.7	(3,7)	9.1	(4,10)	12.6	(4,10)	4.3	(3,5)
Number of sexual partners in past 30 days**	1.3	(1,1)	1.1	(1,1)	1.6	(1,2)	1.2	(1,1)
Ever used condoms	88	(93)	14	(100)	18	(95)	56	(90)
Condom used last time you had sex	54	(61)	7	(50)	15	(83)	32	(57)
Currently have main partner	86	(91)	12	(86)	15	(79)	59	(95)
Living with this partner	12	(14)	3	(25)	2	(13)	7	(12)
Partner has other sexual partners
No	35	(41)	6	(50)	4	(27)	25	(42)
Yes, I suspect	31	(36)	6	(50)	4	(27)	21	(36)
Yes, I know	9	(11)	-	-	4	(27)	5	(9)
Don’t know	10	(12)	-	-	3	(20)	7	(12)
Currently have casual partner	32	(34)	7	(50)	10	(53)	15	(24)
Methods of family planning/HIV prevention ever used*
Male condom	90	(95)	13	(93)	18	(95)	59	(95)
Injectable	58	(61)	6	(43)	1	(5)	51	(82)
Pills	49	(52)	5	(36)	17	(90)	27	(44)
Implants***	23	(24)	2	(14)	-	(0)	21	(34)
Traditional/rhythm method	20	(21)	3	(21)	2	(11)	15	(24)
Other	6	(6)	-	-	-	-	6	(10)
Ever had a contraceptive implant removed	4	(31)	-	-	-	-	4	(31)
Ever experienced side effects from implant	8	(62)	-	-	-	-	8	(62)

^*Gel/spermicide, diaphragm, female condom, IUD, sterilization had been used by 0–5% and are not listed

^**Mean and Interquartile range

^***For FGD participants, implant users had only had 1 implant in use in their lifetime; those with a current implant had it use for a mean of 14.4 months (median 7, range 3–29)

By design, all IDI participants had previous PrEP research-study experience, including 18 men and 22 women who had used oral PrEP; twelve women who had used injectable PrEP and ten women who had used a vaginal ring (Table 1). Additionally, most women (82%) had experience using an injection for contraception or HIV prevention. Twenty-one (35%) women had experience using an implant for contraception.

End-user experiences with and attitudes towards short- and long-acting HIV prevention

Male and female youth in IDIs shared positive and negative feedback about attributes of each dosing formulation they had tried for HIV prevention.

Among women who had used the vaginal ring, feedback was most frequently about product efficacy, and half of the women expressed an interest to use it in the future if it was protective. Participants also commented on physical attributes (size and hardness), cost, side effects, and intravaginal administration. The majority of discussion about use experiences reflected feelings about and experiences with using the ring during sex and whether or not partners would notice it – whether she had disclosed or not disclosed ring use to a partner. Women reported that some removed the ring before sex so that partners would not feel it. One participant worried that the ring could potentially come out and cause embarrassment. She believed that an injection would overcome this issue, and that it might allow her greater sexual freedom and pleasure:

There is nothing I would change about the ring except that it is an embarrassment for example I am having sex with my boyfriend or casual partner and the ring gets attached on his penis and comes out whilst having sex, it’s an embarrassment […] Because once an injectable is injected and it won’t embarrass you. Even if you sleep with a boyfriend you will sleep with free pleasure… It won’t be attached on his penis. I will get my injection at the clinic, go out partying or sleep with whomever I like. (IDI #2003, ring-experienced female, age 23)

Oral PrEP-experienced men often described the daily dosing regimen of pills as challenging. Pills were described as forgettable and difficult to use on weekends and at the same time each day. Several men expressed a desire for something that simply lasted longer in their system.

I wouldn’t mind using it Friday when I go out and Saturday when I go out, maybe Sunday if I’m going out. …you know that maybe Friday, Saturday and Sunday you might get lucky [laughing]…. I want to use it and it stays in my system for a very long time…. As a hundred percent… (But) I hate the daily intake. (IDI #1105, male oral PrEP-experienced, age 24)

The fact that the pill dissolves and gets into your system (vs. male condoms) was viewed as a benefit; however, the pill size was problematic because it was perceived to take a long time to absorb and be effective, and to be challenging to swallow. One man suggested the PrEP pill should be made to dissolve in liquid, both to minimize the challenge of swallowing, and make it less like an ARV pill for treatment. In some cases the pills were described as embarrassing and linked to being HIV positive, necessitating secrecy to avoid stigmatization.

By contrast, some female oral PrEP users stated that pills were more discreet than other methods, e.g. they don’t leave a mark like an injection. Like some oral PrEP-experienced men, women who had used these products highlighted the benefit of the pill working systemically; it won’t break like a condom, it doesn’t interfere with sexual pleasure, nor do you have to think about it during sex. Nevertheless, others disfavored the comparatively short duration of protection that the pill offered.

Injectable-experienced women appreciated not having a daily product, and felt that receiving the injection every two to three months was feasible. The injection was perceived to provide greater protection because it would be absorbed in their bodies, they would not forget to take it (like pills) and it would not burst (like condoms). Experience of side effects (e.g. headache, dizziness, nausea) was raised as the main complaint with using injectable PrEP. Some participants stated that injection site pain was also a negative aspect, but acknowledged that the pain was temporary, and soreness subsided within a day or two.

Preferences for HIV prevention: the Invisible Product

Irrespective of previous method-use experience, gender, or sexual orientation, the majority of participants expressed a preference for prevention methods formulated as injectables or implants. Several participants mentioned that their top priority in any product was efficacy. For some, high efficacy would override the importance of all other attributes. For example, even if they feared the pain, they would use an implant or an injectable as long as it was fully protective. While efficacy was a top priority, there was also a clear desire across all subgroups for an “invisible” product that wouldn’t interfere with sex, at would stay in the system to provide protection, that caused minimal user burden, or that wasn’t apparent to others; and these characteristics were most salient for long-acting methods. As explored below and depicted in Table 3 with illustrative quotes, narrative explanations for preferences converged thematically around different dimensions of “invisibility” that reflect a socio-ecological framework, including invisibility to oneself, one’s partner and household members, community members, and outsiders.

Table 3.

Preferences for Sustained Release Long-Acting Products by theme, with illustrative quotations

Themes	Sub-theme (s)	Illustrative Quotes
Invisible to Oneself	Systemic flow: provides longer lasting protection and is forgiving of user-error, perceived as more effective Pain-free/no side-effects Not visible or palpable Forgettable to user	Because with the injection you are injected and it is absorbed inside you. A condom is something that you use and can burst anytime. So there is nothing 100% sure about it. (IDI #3009, female injection-experienced, age 22).
		I forget the pills. At least with the injection if you have injected it, you are safe in the following month when you had the injected. You can skip the pills when you have to take them for a certain day or the week. There are some things that take place when you didn’t take the pills, and you were supposed to take them. The injection is the best to me (IDI #3004, female injection-experienced, age 24).
		If it can be changed and not be painful. It should also be smaller, err […] Oh, it was not that big, it was right but the pain! No it stings. (IDI #3002, female injection-experienced, age 20)
		It should be small … Not something that will be visible and giving you nerves you see? … You should forget sometimes that you even have it. (IDI#1116, male oral PrEP-experienced, age 18).
		Once I know that I’ve got something underneath your arm even though I’m not going to think about it all the time but then every time it’s going to come that…Then I’ll be thinking, ‘O! I’ve got something under my arm’. And I’m going to be like, ‘This is so irritating’ (FGD #601, male, implant-naïve, age 24).
Invisible to others	Not visible or palpable Discreet, does not require disclosure to household members or partner Private	I want it to be inserted but to be unseen that it is in you… Not to be something that will be associated with sewing (stitches)… Yes, leave me with no pains and not to be felt. (IDI #2010, female ring-experienced, age 21)
		It was a problem for him to keep the pills at home, so I had to keep the pills for him. Every time he had to come and take them from my house… Although his parents knew about the pill but they didn’t want the pill at home…. I think it will be good if they just go one day to get the injection because they won’t see anything…. To carry around… yes. Or to be seen taking it… Yes, for three months and you would be just sitting with the injection in you. (IDI #1104, male, oral-PrEP experienced, age 22)
		The implant will be inserted and then you are done. The pill, as I have also said, it is right to take the pill during the week because everyone is at home. But there is no such time during the weekend … Maybe someone will go and sleep over at the boyfriend’s place … And then the person would be shy to take the pill with (him) because the boyfriend would ask what the pill is for … I think the implant is the best because you will insert it and you are done… I think the majority would prefer the implant. (IDI #1106, male PrEP-experienced, age 23)
		The injection is better because some people didn’t use the ring. They removed it for their partners because of what I told you… So the injection would be fine, something that will be in the blood. (IDI #2006, female, ring-experienced, age 24)
		It’d [her ideal PrEP implant] be much thinner than this [PrEP implant prototype]… [as thin] as a piece of paper. Because I don’t want when people hold me in the arm they feel that there is something inside my body. (FGD #502, female, implant-experienced, age 21)
Invisible to community	Rumors and stigma Threat of robbery	Just like with PrEP, if you were to tell someone there that you’re using these pills for HIV prevention, a person will say, “Whatever, that person has AIDS, they’re taking AIDS pills.” And then to tell them that you’ve inserted that implant, they’ll also say that you have AIDS, you understand? You’re taking an AIDS medicine to them. People from the community are always negative so I wouldn’t share with them that I’m using these things, they always have that negative thing to say. When they don’t know something, they will think of something the way that they want to think about it.(FGD #501, female, implant-experienced, age19)
		Considering first the stigma around the HIV disease […] Because knowing that I’m HIV negative, I’ll come to Lelona as a family member or a friend and let him know that okay I implanted an HIV device in my body. He’ll just hear one word, “I’ve got an HIV device in my body.” He’ll start thinking that, “Okay, Jeffrey* now has got HIV”… And that now will circulate in the location [township] that, “Yeah Jeffrey, is now HIV positive”, as they didn’t understand that I am not HIV positive as yet. (FGD #602, male, implant-naïve, age 24)
		It [contraceptive implant] apparently has a powder inside, and they smoke it. (IDI #3001, female injection-experienced, age 20)
		Like I am scared because it’s said that people with implants are at a high risk of people who use drugs because they are robbed. (FGD #402, female, implant-naïve, age 19)

^*Pseudonyms were used in all qualitative data collection and reporting activities.

Invisible to oneself.

Data about invisibility to oneself included preferences for the systemic nature of long-acting methods: that they were relatively simple and effortless (low user burden); did not require frequent administration (longer acting), and were perceived as more efficacious, which, as mentioned above, was the attribute of overarching importance to these young people. Participants favored a method that was pain free with no side effects – another aspect of being invisible or unnoticeable. Additionally, several participants expressed appreciation for a product that remains in the system, because it provides more reliable coverage (or full protection) without worry of missed doses, a common occurrence among youth.

Hence, having a product that was physically invisible to oneself was important. When discussing the implant, for example, men in particular, described wanting a more flexible and less palpable implant so that they would not be constantly reminded that a foreign object was in their body. On the other hand, when discussing the biodegradable implant under development, many cautioned that this concept may not be easily understood and that people would be concerned about where the dissolving implant goes in their body.

Invisible to others.

The invisibility of long-acting products to others was also a dominant theme; preferably they would not be seen or felt by friends, family, or partners. Rather than the commonly used location of the upper arm, youth recommended administration to locations such as the thigh or stomach or another “hidden place” – somewhere less visible or palpable. Pertaining to the implant, participants commented that “It has to be small so it doesn’t show”, and one woman said she would use the implant in the future for HIV prevention “as long as it is not going to swell and make a caterpillar [under the skin]”. Another participant clarified that her preferred method would not require stitches. For some, implant rods were perceived as less “invisible” than injectables, because someone might grab one’s arm and feel it – thereby unintentionally disclosing its use (Table 3). FGD participants, who had more opportunity for interactive discussion and learning about implants, preferred a flexible (vs. stiff) rod to enhance discretion.

An HIV prevention method that did not necessitate disclosure to family or household members, and that was private, was another salient aspect of preference. In contrast to oral PrEP, long-acting methods were seen to provide greater discretion because a product did not have to be carried around, stored covertly in crowded households, or used in settings where parents were unsupportive. Methods were similarly regarded as offering greater discretion if product use had not been disclosed to partners. For some women, the systemic application and discretion of an injectable or implantable was described as preferable within partnerships where use of the ring had not been disclosed, or was disfavored. MSM described similar situations where a long-acting method would overcome “shyness” or fear to disclose pill use to a partner (Table 3).

Invisible to community.

Similar to fears about partners or family members discovering use of HIV prevention and the consequent implications about sexual behavior and/or sexuality, several participants expressed fears about community members learning of prevention method use. Long-acting methods offer enhanced discretion from community stigma because they required fewer trips to the clinic or pharmacy (which might suggest an HIV-positive status); and did not require transportation and storage of products.

Of note, community invisibility was also highly salient due to commonly cited stories about implants inciting robbery. Participants across all subgroups shared stories about contraceptive implants being physically cut out and removed from women’s arms by gang members who wanted to smoke the drug in the implants (Table 3). One woman felt that the injection would be better for the community as a whole due to this fear, others felt that softer and less visible rods would help to address these concerns.

Discussion:

Sustained release long-acting methods of HIV prevention, including ARVs formulated as injectables and novel biodegradable or removable implants, are a promising set of strategies in pre-clinical and clinical development. In addition to exceptional safety and efficacy, a critical hope and priority is that they expand the toolbox of options for individuals requiring or preferring less frequent dosing. Young men and women in South Africa are arguably one of the most important end-user groups globally [11], and this research offers several important insights into their priorities and preferences for novel HIV prevention approaches. In this qualitative study, long-acting injectable and implantable methods were favored over oral PrEP and the dapivirine ring. Assuming that a novel product is highly efficacious, the most salient attributes favored among all HIV prevention methods, most strongly expressed by long-acting approaches, were related to invisibility. Youth wanted a product that necessitated the least amount of work and awareness; they didn’t want it to be seen or felt; they didn’t want to have to talk about it; and they wanted it to require minimal effort to maintain. Put more bluntly: while these end-users may have recognized the importance of HIV prevention, they didn’t want it to burden their daily life.

A desire for invisibility was largely a function of lifestyle. Many youth described busy, active, lives. They are not yet “settled down”, and move around – especially on weekends – to visit, socialize and party. Most did not live with their partners and many have multiple partners. Injections and implants were described as suitable for this dynamic lifestyle where daily methods might be skipped, forgotten, or raise suspicions. This finding is consistent with regional increases in the uptake of long-acting contraceptives among women (e.g. increasing from 37% to 46% of modern method mix in Sub-Saharan Africa from 2003 to 2012). [28, 29] Long-acting HIV prevention was also positively portrayed as endowing one with freedom to spontaneously have sex, worry-free, with whoever one might want. While this type of sexual behavior has been a cause for concern in previous prevention method research, for example men disapproving of microbicides for fear women will become “promiscuous” [30–33], it offers a realistic picture of the desires and activities of young people in this setting. It is pertinent to recognize that these young South Africans have lived their entire lives under the specter of an HIV epidemic, and that these youth did not express a desire to sleep around without consequence, rather, a desire for greater freedom and reduced anxiety around HIV.

The importance of community invisibility highlighted the fact that taboos around sex and HIV-related stigma continue to persist. Despite generations of people affected with HIV, youth were still wary of disclosing ring or oral PrEP use to partners, family or community members because of what it might imply about their sexual behavior, their sexual orientation, or their disease status. Stigma is related to another aspect of this African research setting: many live in densely populated areas with minimal privacy in the community and in their homes.[17] There is not the same degree of anonymity that may exist in other Western urban settings.[34]

The implant robbery anecdotes offer some important insights into introduction of novel technologies and the context of community violence in which this research was conducted. Rumors – whether based in truth or reality – are powerful. Implant robberies in Cape Town were reported in one online news story that we could identify;[35] several FGD participants reported personally knowing a victim. Nevertheless, many IDI and FGD participants – both male and female - mentioned the fear of implant robbery by gang members. The township in which many of these participants resided, Nyanga, has high rates of reported murder, physical and sexual assault, and crime.[36] While here participants expressed concern that the implant might induce violence, other research with MSM and women from these areas has highlighted that other novel technologies (e.g. rectal microbicides, rings) are perceived to confer safety from HIV in a context of high sexual assault.[18] These data reinforce the importance of evaluating, and potentially addressing, the role of rumors and other structural factors such as community violence, and their impacts, when doing research and programmatic activities with novel products.

Consideration of end-users is important during product development. Both efficacy trials and rollout of effective biomedical HIV prevention technologies have encountered setbacks that – theoretically - may have been averted through research into end-user and health system determinants of uptake and use.[37] When designing next generation PrEP dosage forms, product developers could prioritize features that the potential end-users in this study identified as important, such as a product that is long-acting and invisible by sight, by feel, and by packaging. In addition to traditional qualitative methods like these, other methodological approaches, e.g. human-centered design, social marketing, can help to elucidate ways in which PrEP can be made more attractive and appealing, less burdensome to daily activities, and better aligned with youth’s lifestyles.

There are several limitations to this research. IDI participants were purposively recruited from former PrEP clinical trials, and are likely to be more knowledgeable and accepting about PrEP and HIV prevention than other South African youth. While these participants candidly shared many concerns and complaints about current and future prevention approaches, it is unknown whether “research-naïve” youth in the community would have shared the same attitudes. It is also unknown how acceptability for these products will change when actual products are available Some FGD participants were recruited from the community-at-large, but most were recruited from former HIV or contraception-related studies. Participant reactions to implants may have been impacted by the length and content of the implant description in IDIs, which varied widely depending on the depth of questions raised by the participant. FGDs were facilitated by a bioengineer with laboratory experience in designing new PrEP technologies, and participants may have been biased (social desirability) to express favorable views towards implants. That said, participants expressed many of the same opinions in the absence of this facilitator. FGDs were homogenous by gender and previous experience of implants; however, males were not stratified by sexual orientation, which may have impacted responses regarding an HIV prevention product. That said, the FGD discussion topics focused more on HIV prevention product features than sexual behavior. Another potential limitation is that a male interviewer was not available; however, no males declined to participate because of this. As with all qualitative studies, our results may be biased by the exclusion of questions or probes that were not explored, and there may be misinterpretations of the data through various stages of analysis, including translation, coding and summarization. We did not conduct member checking, whereby researchers validate their results with study participants. These limitations were mitigated by the inclusion of local scientists on the analysis and writing teams.

In conclusion, the young South Africans in this study expressed preferences for an invisible HIV prevention product – one that required minimal user burden; did not cause side effects; and could not be noticed by partners, family, friends or community members. Long-acting sustained release injectable and implantable methods were favored by most because of these desired features. An examination of specific product attributes influential to product preference in a generalizable sample of youth could inform product design, and build on youth’s interest in a product that is perceived to offer and achieve invisibility.