Structured Abstract
Background Novel formulations for PrEP such as injectables, implants, and intravaginal rings are emerging as long-acting alternatives to daily pills for the prevention of HIV. Sub-Saharan Africa has the highest HIV burden as well as the highest PrEP coverage globally. As long-acting formulations continue to become available, it is crucial to understand the product preferences of potential users.
Objective To conduct a scoping review focused on the region of Sub-Saharan Africa to understand which PrEP products, especially long-acting formulations, different patients and demographic groups prefer as well as the factors that influence their preferences.
Design We identified 34 publications published between 2014 and 2024 that assessed preferences regarding at least one long-acting PrEP product in the region of Sub-Saharan Africa.
Results Participants preferred longer-acting products over oral pills when given the choice in almost all studies. On-demand PrEP was commonly preferred over daily dosing, and long-acting products were preferred over on-demand dosing. A majority of studies found injectables to be most commonly preferred compared to daily oral PrEP, implants, and rings. This preference was observed across a range of demographics including men and women, youth, men who have sex with men, and female sex workers. Duration, efficacy, and discretion were the three most important factors influencing participants’ choices.
Conclusions Long-acting products, especially injectables, are acceptable for a wide range of individuals at risk of HIV in Sub-Saharan Africa and tend to be preferred over daily oral pills. Participants expressed a diversity of values and opinions regarding preferences, emphasizing the benefit of providing multiple formulations to maximize coverage over heterogeneous populations.
Strength and Limitations of this Study Some key populations, such as transgender women, were underrepresented in the literature. With most studies published before long-acting products became widely available, the hypothetical preferences of non-experienced users may differ from preferences in practice.
INTRODUCTION
Pre-exposure prophylaxis (PrEP) is a promising tool in the HIV prevention portfolio and is highly effective when used with high adherence. Although PrEP uptake has increased significantly since its introduction in 2012, coverage remains low among people at HIV risk, with 1.6 million estimated global users–well below the United Nations target of 10 million users by 2025 [1]. Additionally, PrEP retention and adherence among those who initiate is suboptimal, reducing its effectiveness [2]. Barriers to PrEP uptake and adherence include pill burden, stigma, and lack of discretion associated with oral tablets. In light of these challenges, research has focused on developing several long-acting PrEP formulations, including injectables, implants, and vaginal rings, which could increase PrEP coverage and adherence by providing more convenient and discreet options. Cabotegravir, a bimonthly antiretroviral, was the first long-acting injectable (LAI) approved for clinical use by the US Food and Drug Administration in December 2021. The intravaginal dapivirine ring (DVR), developed primarily for women in low-income countries, was recommended for use by the World Health Organization (WHO) in January 2021 [3]. An implant with six-month duration is currently in development. Finally, lenacapavir is a first-in-class twice-yearly injectable which was approved for treatment of HIV in 2022 and is currently undergoing Phase III trials for use as PrEP as well. In addition to long-acting formulations, oral on-demand PrEP (or event dosing) is an alternative dosing schedule with comparable efficacy to daily use in which patients take a double dose up to two hours before a potential exposure and then once every 24 hours the following two days.
Understanding preferences for PrEP products among subgroups at risk of HIV infection is crucial for maximizing PrEP coverage and impact. The aim of this scoping review was to synthesize the literature on PrEP perceptions and preferences in Sub-Saharan Africa (SSA), the region most impacted by the HIV epidemic. PrEP is a priority intervention for scale-up in SSA and policymakers must decide which products to implement and how to tailor demand generation strategies. Understanding the relative preference of LA PrEP compared to oral PrEP by subpopulation is also important for commodity planning.
METHODS
We conducted a scoping review of peer-reviewed literature on preferences for long-acting PrEP from 2014 to 2024. We adopted a systematic approach, following Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines [4]. We conducted a keyword search for relevant articles first on PubMed followed by Google Scholar, which returned the same relevant publications as PubMed. We included the following keywords during our search: “PrEP”, “long-acting”, “discrete choice”, “preferences”, “Africa”, “on demand” (Table S1). We also received grey literature and articles not yet published from a research collaborator.
Inclusion and exclusion criteria
Studies were eligible for inclusion if they met the following criteria: (a) original research; (b) peer-reviewed and published in English between 2014 and 2024 or provided by a research partner; (c) research conducted in SSA; (d) evaluating preferences for at least one long-acting or on-demand PrEP product alongside daily oral PrEP.
Data screening
Two reviewers (BP and AS) screened the list of references for inclusion into the study. First, titles and abstracts of articles were reviewed and then selected articles underwent full text review. Disagreements were resolved via team discussions. We extracted data on title, publication year, location of data collection, population assessed, main findings, study strengths and limitations. We categorized each study by region and demographic focus.
Quality assessment
One reviewer (BP) conducted a quality assessment of the 34 eligible studies based on generalizability to the target population, participant acceptance rate, and PrEP experience/naïveté of the sample (Table S2).
RESULTS
Characteristics of studies
Of 214 unique citations identified, 35 articles met eligibility criteria and were included in the review (Figure 1). Study characteristics are summarised in Table 1. Most studies were cross-sectional (22) including nine discrete choice experiments (DCEs). Most or all individuals in 11 studies had prior experience with oral PrEP, while nine studies were conducted among exclusively PrEP-naïve participants (Table S2). 10 studies were conducted among individuals participating in randomized clinical trials for PrEP [5–11], cohort studies [12,13], or PrEP implementation projects [14].
PRISMA flow diagram
Studies included both men (n=19) and women (n=29), including 15 studies which focused on women’s preferences specifically. Five studies focused on males specifically, including three studies assessing preferences of men who have sex with men (MSM) [15–17], and two among heterosexual men [5,18]. Transgender women were included in only one study alongside MSM [17]. A majority of studies (n=19) included or focused on youth, here referred to as adolescent boys and young men (ABYM) or adolescent girls and young women (AGYW). Female sex workers (FSWs) were assessed in five studies.
South Africa was the most represented country (n=25), followed by Kenya (n=9), Zimbabwe (n=8) and Uganda (n=8). Seven papers presented findings from three countries combined (Uganda, Zimbabwe, and South Africa). Other papers included Tanzania (n=2), Eswatini (n=2), Malawi (n=1), and Nigeria (n=1).
The majority of studies evaluated LAIs (26), followed by implants (n=18), the DVR (n=14), and other prevention methods such as condoms, films, or gels (n=10). All formulations were compared against daily oral PrEP. Three studies compared preferences between daily oral PrEP and oral on-demand dosing, and one study compared on-demand dosing to injectables and implants. Table 2 summarises the results of the studies included.
Preferred products by subpopulation
Women
Adult women and adolescent girls and young women (AGYW) most commonly preferred LAIs over other long-acting products as they were perceived as having greater efficacy and favorable duration, and as suitably discreet. Women participating in the CAPRISA 082 study in South Africa who had previous experience using an implantable or injectable contraceptive method were more likely to choose that method for PrEP [14]. In a South African focus group, despite overall preference for LAIs over oral pills and the DVR, a few AGYW were dissuaded from injectable PrEP for fear of side effects similar to injectable contraceptives, such as menstruation cessation and weight gain [16].
Interest in the DVR was mixed. The ring was regarded positively among participants in the ASPIRE study, a multinational clinical trial of the DVR. Where women could select more than one product as “most preferred,” 94% of women selected the ring, compared with 39% for implants and 33% for LAIs [11]. However, many women expressed concern about discretion, such as whether their partner could feel the ring during intercourse or whether it could come out accidentally [7]. In most studies, DVR was less frequently preferred compared to LAIs [7,12,16,19]. Most AGYW in one focus group would not use the DVR due to perceived side effects, especially pain during intercourse, as well as concern for hygiene and use during menstruation [16].
Women reported broad interest in using implantable PrEP and stated that the benefits of an effective and long-lasting product outweighed potential drawbacks such as side effects or pain on insertion. They also expressed interest in a dual use product for PrEP and contraception in surveys and interviews. Since implants are already a common modality for extended-release contraceptives, a device which combines the two could be attractive to women seeking protection from both HIV and pregnancy. [7,15,17,20–24].
In a study of pregnant and postpartum women participating in the PrEP-PP and PrIMA-X trials in South Africa and Kenya, respectively, many voiced safety concerns for the mother and infant during pregnancy and breastfeeding. These women, most of whom had recent experience with oral PrEP, tended to prefer long-acting products over daily oral for their longer duration (especially in South Africa) and increased discretion (especially in Kenya) [12]. All pregnant AGYW interviewed in a study in South Africa would not use the DVR on account of perceived side effects [16]. Pregnant and postpartum women in multi-national focus groups mentioned fears of side effects ranging from exacerbation of pregnancy discomfort to severe outcomes such as miscarriage or birth defects. However, they were overall accepting of PrEP as long as it was safe and effective, and stressed the importance of choice and the ability to choose the formulation that worked best for them [25].
Female sex workers (FSWs)
Awareness of PrEP was low among FSWs in South Africa [20] and Tanzania [26], but after learning about PrEP most participants were willing to use it, particularly those who had recent symptoms or diagnosis of a sexually transmitted infection (STI). Many FSWs stated that protection by PrEP could enable them to have sex with clients without requiring condoms [20] or to accept condomless sex at a higher price [27], highlighting a potential concern regarding risk compensation. However, others viewed PrEP as a protective complement to condoms and stated that they would continue to use them to prevent STIs and pregnancy [26]. 86% of FSWs responded that they would likely use an implant in a South African DCE [21]. This high proportion of acceptability was complemented by qualitative interviews in South Africa [20], where FSWs commended the benefit of continuous protection and noted that it would be worth a brief amount of pain during insertion. Similarly, 88% of Tanzanian sex workers preferred LAIs over oral daily, although a smaller majority (58%) felt PrEP was generally worth taking. LAIs were considerably preferred over the DVR in interviews with South African FSWs; all sex workers stated they would not use the DVR on account of perceived side effects and concern that a client might notice the device [16].
Men
MSM tended to prefer LAIs over implants or oral pills. Among Nigerian MSM, preference for LAIs was associated with single relationship status, inconsistent condom use, and having a primary care provider [15]. Duration/dosing frequency was a highly prioritized product attribute for male participants. Oral PrEP-experienced young men in South Africa reported difficulty with daily dosing [7]. Men of all orientations valued privacy and discretion, though more men who have sex with women (MSW) valued being able to use PrEP without their partner knowing compared to MSM [28].
Preferences among heterosexual men were specifically assessed in two studies [5,18]. In a survey of urban heterosexual men in South Africa, 48% preferred LAIs compared to 33% who preferred oral and 20% who preferred condoms alone. Men who had children or who were less risk-averse were more likely to prefer LAIs. Men concerned with high risk of STIs other than HIV were more likely to prefer condoms over LAIs alone. As choices were discretely ranked, it was not clear how many men who preferred condoms for their protection against other STIs would prefer to use LAIs and condoms in combination. [18] In a mixed methods study conducted in South Africa, Uganda, and Zimbabwe among participants in the CHAPS trial, a majority of heterosexual male youth (65%) preferred on-demand oral PrEP compared to daily dosing. Those who did not believe they were exposed to HIV regularly enough to warrant taking a daily pill tended to prefer on-demand dosing in qualitative interviews. [5]
Product attributes driving preference
Duration
Product duration was the most important factor driving user preferences in many studies across demographics [7,10,12,23,28]. Oral PrEP-experienced pregnant and postpartum women most commonly cited product duration as a factor for switching to LAI [12]. Oral PrEP-experienced men also frequently reported difficulty adhering to a daily dosing schedule, especially on weekends, as well as difficulty swallowing the pill itself [7]. (Note that the most common oral PrEP formulation, emtricitabine-tenofovir, is a very large tablet [19 mm] which can be difficult to swallow even for users who take other, smaller tablet medications.) Male and female youth in a South African DCE valued product duration highly and were typically willing to trade their preferred product for one with a longer dosing frequency [23].
Notably, some youth in interviews in South Africa suggested an ideal formulation as a monthly rather than a daily pill [16]. This would reduce the burden of daily dosing for users who preferred an oral tablet over other long-acting formulations, such as pain associated with injections.
Efficacy
Many participants emphasized PrEP efficacy as an important determinant of product preference. They noted that efficacy was tied to longer duration due to the difficulty of adhering to a consistent dosing schedule, which hinders the observed effectiveness of oral PrEP. PrEP effectiveness was the strongest factor driving preference in several studies, notably among pregnant and postpartum women [12] and among youth populations [7,28]. High PrEP effectiveness outweighed participant concerns about side effects such as pain upon injection or product insertion.
Discretion
Discretion (i.e., being able to use a PrEP formulation without a partner or the community knowing) was a commonly mentioned concern, particularly among women and persons who did not wish to disclose their PrEP use to their partners. In several studies evaluating oral PrEP alongside LA formulations, including women [11], FSWs [26], and adults [24], participants expressed concern about stigma associated with daily pill use. They feared that others would think they had HIV if they discovered that they were taking a daily antiretroviral pill. Similarly, a visible preventive product could be seen as a mark of sexual indiscretion or promiscuity, as voiced in one South African focus group including participants of all genders and sexual orientations [29]. According to them, using PrEP could sow distrust or signal infidelity among users with romantic partners, and a woman who used PrEP might be seen as sexually promiscuous.
Notably, product discretion was described differently by different individuals. While many participants found daily pills indiscreet, others felt oral PrEP was the most discreet since it does not leave a mark like an injection [7]. Others, especially women, were concerned about their partner knowing about their PrEP use. Many participants expressed disinterest in the DVR for fear that their partner would feel the ring during intercourse, or that it might fall out and cause embarrassment. Concerns about a partner noticing signs of PrEP use were also mentioned relating to the implant; many participants in interviews disliked the notion of a visible device under the skin and preferred one that would not be seen by others.
Implant biodegradability
In general, participants preferred biodegradable implants which would not need to be removed by a provider, alleviating the need for an extra clinic visit and pain during removal [21,30]. South African healthcare providers also stressed this feature in qualitative interviews [31]. However, in one study in Kenya, many participants (especially FSWs) preferred a nonbiodegradable implant for its reversibility, as it could be removed if needed. Some also expressed fear about the effects of the degraded materials being absorbed into the body [24].
Logistical challenges
Long-acting formulations addressed the challenges of frequent visits to clinics or pharmacies, which participants variously described as overcrowded, lacking in privacy, and inconvenient or inaccessible [7,16]. These challenges drove preference for longer-duration products and for biodegradable implants, which require fewer visits.
Injection Fear
A common barrier to LAI acceptability was dislike or fear of needles. The current formulation of cabotegravir is injected in the buttock, but both males and females in South Africa tended to dislike this location for fear of discomfort especially while sitting [23,27]. Some young women were uncomfortable having to disrobe to receive an injection in the buttocks [27].
Side effects
Fear of real or perceived side effects was a salient factor influencing preference, especially for younger participants–about half of male and female youth in a study in Malawi reported being unwilling to use PrEP if they experienced side effects. Pain at injection site was the most frequently mentioned concern of LAIs, although youth in qualitative interviews stated that efficacy was a more significant factor even if the formulation was expected to be painful [7,28]. Similarly, pain upon insertion was a concern regarding implants, but participants generally felt that the benefits of an effective and long-lasting PrEP product outweighed the potential for pain. Qualitative interviews indicated a preference for a flexible versus a stiff implant for increased comfort.
Pain or discomfort also influenced acceptability of the DVR, especially discomfort during intercourse. Among pregnant and postpartum women in South Africa and Kenya, most of the participants who preferred oral PrEP over a vaginal ring did so on account of concern for physical discomfort [12]. In another study conducted among FSWs in Tanzania, participants expressed concerns about infertility associated with the DVR [26]. Conversely, in two studies conducted in Kenya and in South Africa, Uganda, and Zimbabwe, participants appreciated the reversibility of the DVR and implants, which could be removed if side effects arose [11,24].
Oral on-demand
Three studies conducted in South Africa, Uganda, and Zimbabwe assessed preference for oral on-demand PrEP compared to daily use, and one study in Eswatini, Kenya, and South Africa compared on-demand PrEP to a long-acting product. Overall, 60% of male youth [32] and 65% of youth MSM [5] preferred on-demand PrEP over daily oral tablets, with older youth tending to have greater preference for on-demand dosing. Having more frequent sexual intercourse was associated with a lower preference for on-demand PrEP and greater preference for daily oral PrEP. Participants who preferred on-demand PrEP cited not liking daily dosing, not considering oneself at frequent enough risk to warrant taking a daily pill, stigma, and pill burden.
Conversely, those who preferred daily use cited desire for continuous and/or improved protection and comparative ease of use by dosing every day instead of having to remember to take it before sex. One study examined the effect of post-traumatic stress disorder (PTSD) on preference in SSA youth and found no significant association between PTSD symptoms and a preference for on-demand versus daily PrEP, with a 61% vs 51% preference for on-demand dosing in those with and without PTSD symptoms, respectively [33]. South African men found oral on-demand, LAIs, and implants similarly acceptable, but if only one choice was available, they preferred a once-monthly pill (32%), six-month LAIs (28%) or implants (20%) over on-demand (2%) or a two-month injectable (5%) [34].
Discussion
This scoping review evaluated preferences and acceptability for various PrEP products across populations in SSA. Overall, we found high acceptability of LA PrEP across participant demographics and geographic region, suggesting that LA modalities can expand PrEP coverage among persons at HIV risk. The primary factors driving participant preferences for LA PrEP were efficacy, duration and discretion, which most participants felt were superior in LA products compared to oral PrEP. Overall, long-acting injectables were most preferred over the other LA products evaluated (implants and the DVR). However, we identified heterogeneity in preferences among subgroups which suggests a variety of choice will likely be needed to optimize coverage and impact of HIV prevention. This can be paralleled with the observed increase in contraceptive use with the availability of more contraceptive methods in a health system [35].
A previously published review assessed values and preferences for long-acting injectable PrEP; however, most studies included were published on or before October 2021, before regulatory approval of the first long-acting ARV for use as PrEP, CAB-LA [36]. Authors found broad interest in and overall preference for LAIs and highlighted the perceived benefits of discretion and less frequent dosing. Our review adds to the literature by evaluating other long-acting modalities in addition to LAIs and including additional publications from two years after the introduction of injectable PrEP. Further, we focus on Sub-Saharan Africa, as the region with the largest HIV burden globally as well as the most widespread rollout of PrEP worldwide. In regions of SSA with high HIV prevalence, the majority of transmission occurs through heterosexual mixing. Additionally, higher rates of oral PrEP use can influence attitudes among the general population regarding HIV prevention. One example is parental attitudes toward provisioning PrEP for their children. An overwhelming majority of caregivers of adolescents in Malawi (87%) and in South Africa (85%) expressed desire for their children to take PrEP [37,38]. This can be contrasted with a study in the American south in which parents of LGBTQ adolescents, though generally positive about PrEP, expressed relatively low intention for their children to take it [39].
Interestingly, the dapivirine ring, which was designed for use in low-resource health systems as in SSA, was a less popular choice than other long-acting options even among women with experience using DVR [7]; this was largely due to perceived side effects, especially pain during intercourse, and concern about indiscretion or an impact on the male partner’s pleasure during intercourse. One study within a DVR clinical trial found overwhelming acceptability of the ring after 28 weeks of follow-up, which suggested that the DVR could be more acceptable after experience with use [11]. However, the preferences of women choosing to participate in a DVR clinical trial may not be representative of the general population. That said, Mataboge et. al. [16] highlight focus group discussions among women [40] and their male partners [41] that suggest that while a considerable proportion of partners notice the ring during sex, the impact on sexual pleasure for both partners is minimal and in some cases positive.
Across studies, the most commonly reported concern about long-acting PrEP was potential side effects. However, empiric data shows that actual side effects were generally less frequent or severe than participants anticipated. Side effects of antivirals for PrEP are typically mild and of short-term duration, yet about half of youth participants in a study in Malawi stated they would not consider using PrEP if there were side effects [38]. Similarly, some women in interviews who had experienced side effects from injectable hormonal contraception worried about similar effects from injectable PrEP [16], although these side effects have not been observed [42]. Health communication that assuages these fears may influence the choices of users who are deterred by side effects from their otherwise preferred formulation.
Pregnant and breastfeeding women were especially concerned about side effects that could harm the fetus or infant, with some expressing fear of miscarriage or birth defects due to PrEP use [25]. Though oral PrEP is widely understood to be safe during pregnancy [43,44], other long-acting formulations have been slow to establish similar safety profiles [45]. As these data emerge, it will be imperative to educate pregnant and breastfeeding populations on the safety of these formulations to enable them to make informed decisions regarding PrEP.
Only one study compared oral on-demand PrEP with a long-acting formulation, which found similar acceptability between on-demand and long-acting products but a distinct preference for long-acting products over on-demand dosing [34]. Since participants frequently preferred on-demand over daily dosing, and because the schedule may have similar patient advantages to long-acting formulations (comparable efficacy, longer dosing frequency, increased discretion, etc.), more research is warranted in order to understand alternative product preference.
One study assessed religious background and found preference differences, namely a much greater preference for implants among Muslims compared to Christians, adherents of African indigenous religions, and non-religious participants [17]. Factors driving these preferences were not examined. Similarly, ethnicity was generally captured only on the national level in the studies included. For a region as ethnically and religiously diverse as Sub-Saharan Africa, these and other markers of cultural identity may be associated with different values surrounding sexuality and HIV prevention, and could aid local health authorities to provide the most culturally appropriate products for their communities.
Our review highlighted several gaps in existing studies. Studies on preferences of transgender women in Sub-Saharan Africa were scarce despite their having up to a 13 times higher HIV risk than that of the general population globally [46]. We identified only one study evaluating preferences in this priority population [17]. Transgender women are often grouped with MSM in preference studies, yet their preferences were distinct in the one study identified, in which transgender women tended preferred implants while MSM preferred LAIs. This highlights the importance of assessing heterogeneity in preferences across demographics as preferences of cisgender women or from other sexual minority individuals assigned male at birth do not necessarily align with those of transgender women.
We also raised a concern around generalizability for PrEP-experienced participants and especially individuals participating in clinical trials of PrEP formulations. Overall, eight studies were conducted among clinical trial participants; several preference studies were conducted during trial follow-up visits, therefore only participants who continued PrEP were included and preferences of those lost to follow up were not assessed. Findings from these studies may not be generalizable to those who are not currently taking PrEP or have significant difficulty with adherence. Individuals who may experience challenges using oral PrEP due to stigma, discretion, difficulty attending frequent refill visits, or pill burden are poorly represented in PrEP clinical trials, yet they are likely the primary target population for uptake of new PrEP modalities, as current oral PrEP coverage is low. Further, individuals participating in PrEP studies may differ from the general population in that they may be more interested in the PrEP modality evaluated in the study in which they are participating. For example, in a study that assessed preferences among participants in the DVR efficacy trial (MTN-020/ASPIRE), 94% of participants selected the vaginal ring as their most preferred LA PrEP product [11]. However, in the TRIO study, in which women were assigned to use all three of injectable, oral, and ring formulations, only 12% of individuals most preferred the vaginal ring [47]. Future studies should consider investigating PrEP preferences among oral PrEP-naïve individuals not participating in PrEP studies.
Finally, the results of this review are limited by the hypothetical nature of many of the studies. Cabotegravir is the only LAI currently available, and 11 of 25 studies assessing LAIs were published before cabotegravir was approved for use as PrEP. Similarly, implantable PrEP is still in development. In both cases experience with injectable or implantable contraceptives can be a useful proxy for experience with that modality as a PrEP product [14]. However, participants analogizing injectables and implants with hormonal contraceptives often feared side effects similar to such contraceptives, even though the side effect profile for ARVs is milder than that for contraceptive hormones. Further research will be needed to understand in-practice preference for long-acting modalities outside of clinical trials as LA PrEP becomes widely available.
Conclusion
Long-acting PrEP formulations are highly acceptable across demographics in SSA and can increase PrEP coverage to meet global targets for HIV prevention. Overall, injectable PrEP was most preferred followed by biodegradable implants, with product duration playing the most salient role in preferences. The intravaginal ring was the least preferred LA product but still more preferred than daily oral PrEP. There was significant interest in on-demand oral dosing among lower-risk participants with less frequent exposures. Further research will be needed to understand enacted preference as these and other long-acting modalities become available to patients.
Data Availability
All data produced in the present work are contained in the manuscript.
Authors’ contributions
All authors contributed to developing the analysis plan, interpreting results, and commenting on manuscript drafts. MS conceived of the analysis. BP, AS, and SC conducted the literature review and synthesis. BP wrote the first draft of the manuscript. All authors had full access to the data in the study and had final responsibility for the decision to submit for publication. All authors read and approved the final manuscript.
Funding
This work was funded by the Bill and Melinda Gates Foundation (INV-038274). The funders had no role in the study design, data collection, or decision to publish this article.
Competing interests
The authors have no competing interests to disclose.
Supplement
Table S1. Search strings
Table S2. Quality assessment checklist
Acknowledgements
We thank Nicole Young, Shirley Chen, and Geoff Garnett at the Bill and Melinda Gates Foundation for their support and funding of this analysis.