Women bear a disproportionate burden of HIV, constituting over 50% of all those living with HIV worldwide and nearly 60% in sub-Saharan Africa [1]. In many countries, gender inequity means that women are unable to negotiate condom use, especially within marriage. Safe, effective, affordable HIV prevention methods that are initiated and controlled by women have therefore been seen as a high priority for HIV prevention. Vaginal microbicides are gels, cream or other topical preparations applied by a woman to reduce her risk of acquiring HIV sexually [2]. Some preparations have been designed for application immediately before and/or after each vaginal sex act (e.g. gels, dispersible tablets, films), whilst others such as intravaginal rings are intended to be placed in the vagina and to release microbicide over several weeks, providing protection independent of the timing of sexual activity. The past two decades have seen a number of trials of candidate microbicides, but until recently, none had been proven to be safe and effective in clinical trials [3–5]. A major breakthrough occurred in 2010 when the CAPRISA 004 trial in South Africa found that vaginally-administered tenofovir gel reduced HIV acquisition in women by 39% overall (and by 54% in those with high adherence), and was safe, acceptable and well tolerated [6]. Furthermore, the gel provided a similar level of protection to women against acquisition of Herpes simplex type-2 (HSV-2). Should these results be confirmed in a second trial [7], a new era in HIV prevention can truly begin.

The availability of a new prevention technology however, raises new questions for public health policy makers. In order for tenofovir gel and other tenofovir-based microbicides currently in development (such as vaginal films and dispersible pessaries [8–10]) to have the greatest impact, research is required to understand the diverse socio-cultural, behavioural and structural contexts into which such products could be introduced [11]. Information is also required on their acceptability to women and their sexual partners, and the impact that vaginal microbicide use may have on other HIV prevention strategies, particularly condom acceptability and use [11–14]. For example, research in Africa showed that although vaginal microbicides were conceived as female controlled methods that could be used covertly and not require male consent, in practice most women inform their partners, showing that acceptability and use need to be seen in the context of couple relationships [13]. The same research further showed how cultural preference for specific sexual practices interacted with microbicide acceptability [15]. A study among female sex workers (FSWs) in Uganda [16] found that foaming microbicide tablets and sponges were the most popular delivery mechanisms, and gel and film the least. Participants found the gel too messy and worried that the film would not dissolve. They liked the sponges because they could insert them in the morning and would be prepared for unexpected sex. Alternatively, in the MDP301 trial [4], married women in the same area of Uganda were very positive about study gel because it enhanced sexual enjoyment due to increased lubrication. It is possible that these differences were related to the different kinds of sexual relationships that the women in the two studies had and/or to the different physical properties of the two gels. This suggests that different products will be preferred in different contexts, and that acceptability research is necessary in different potential target groups.

Intravaginal hygiene and menstrual practices (IVP) have been recognised as possible risk factors for HIV and STI acquisition in women [17–23] and important influences on the acceptability and efficacy of vaginal microbicides in HIV prevention trials [24–28]. Research from a variety of developed and developing countries indicates that IVP are common among women at different levels of HIV/STI acquisition risk [15, 17, 24, 27–35]. For example, in a study among women working in food and recreational facilities in Mwanza, Tanzania, Allen et al. (2010) found that intravaginal cleansing using fingers, water and soap was considered necessary to remove vaginal secretions, menstrual blood and post-coital discharge, and was carried out within 2 hours of 45% of reported sex acts [24]. A household survey among 3610 women in Mozambique, South Africa, Indonesia and Thailand found that the prevalence, type, frequency, and motivations for IVP varied significantly by setting, with intravaginal cleansing and insertion of traditional products most common in African study sites and oral ingestion of products considered to have vaginal effects most common in Asia [29]. IVP have the potential to negate the protective efficacy of vaginal microbicides for HIV prevention either by creating an adverse intravaginal environment (e.g. by altering intra-luminal pH) or by mechanical means (e.g. microbicide gel is washed out immediately after sex) [24]. They are also likely to impact on the acceptability and preference of different microbicide products in different settings (e.g. gels may be preferred in settings where vaginal ‘wetness’ is desirable and where women enhance lubrication through IVP; intravaginal rings or film formulations may be preferred in settings where women are concerned about partner perceptions of ‘excessive’ vaginal lubrication and where they use IVP to induce a ‘drier’ vagina prior to sex) [27, 35, 36].

Although overall HIV prevalence is below 1% in virtually all Asia-Pacific countries, the region is second only to sub-Saharan Africa in terms of the number of people living with HIV [37]. The HIV epidemic in Asia has been largely concentrated in people who inject drugs, sex workers and their clients and men who have sex with men, but increases in heterosexually-acquired HIV infection have been reported from a number of countries, including Bangladesh [38, 39], China [40, 41], Indonesia [42–44] and Pakistan [45, 46]. Of particular note has been the increase in the proportion of women living with HIV in the region from 19% in 2000 to 35% in 2008, with striking increases observed in China [40] and India [47]. There are emerging, generalized heterosexual HIV epidemics in PNG and in neighbouring Tanah Papua Province in Indonesia, with estimates of adult HIV prevalence in the range 1-2% in these settings [48]. Other countries in the Pacific have to date experienced modest HIV epidemics but very high prevalences of sexually transmitted infections (STIs) have been reported both among FSWs and in community-based surveys [48–53]. The HIV epidemic in PNG is primarily linked to heterosexual transmission and shows significant heterogeneity, with over half the reported HIV diagnoses coming from three of its 22 provinces [54], placing women in some locations and socio-cultural contexts at considerably higher risk of HIV infection than others [48, 54–58]. Some models have forecast that under certain scenarios, adult HIV prevalence in PNG could reach 10% over the next 15 years [59, 60]. In Indonesia, Tanah Papua Province is experiencing a generalized HIV epidemic that has many parallels with the epidemic in neighbouring PNG. HIV prevalence among men and women aged 15–49 years in Tanah Papua was recently estimated at 2.4%, the highest in Indonesia and approximately 15 times the national average [61]. Among ethnic Papuans, HIV prevalence was almost twice that of other residents (2.8%, 1.5% respectively).

In these settings, new, female-initiated HIV prevention agents are potentially very appealing to both women at risk and to public authorities responsible for prevention. Apart from some limited trials of earlier candidate microbicides in two sites in Thailand [62–66] and India [66–72], virtually all of the recent trials of microbicides have been conducted in African countries. Asia-Pacific countries have been considered to have HIV incidence that is too low to justify efficacy trials, but high enough in some areas for microbicides to play an important role in prevention once they are proven safe and efficacious.

In this paper we report findings from the first study conducted in a moderate-prevalence setting in the Pacific region to investigate intravaginal practices and vaginal microbicide acceptability, and discuss the implications of our findings on future HIV prevention policy and research priorities.

This study was conducted over a nine-month period in 2010–11. Fieldwork and data collection were carried out from April – June 2010, followed by the interview data transcription, translation, coding and qualitative data analysis.

The study was conducted at Nine-Mile Sexual Health Clinic in Port Moresby, PNG (Figure 1), which is located in a disadvantaged urban area and administered by a non-governmental organization (HOPE Worldwide, PNG) on behalf of the National Capital District Department of Health. In addition to providing services to women and men from the general community, the clinic has an active community outreach and peer support program developed in collaboration with Family Health International (FHI) PNG that is designed to facilitate service uptake by people at increased risk of HIV/STI acquisition, particularly male and female sex workers and men who have sex with men. Clinic staff provided an overview of the study to potential participants as part of routine health talks (tok save) given at the start of each day to clients attending the clinic. Peer educators provided similar information to potential study participants through their community outreach activities. Those who expressed an interest in hearing more about the study and potentially taking part were invited to participate in preliminary group discussions at the clinic in which they were provided with more detailed information about the study by trained researchers and clinic staff. Clients who expressed an interest in study participation were then invited to take part in an in-depth interview (IDI), which took place at a location of their choice following the completion of informed consent procedures. This study location was selected because of the high proportion of women and men at increased HIV/STI acquisition risk who attend this clinic, including FSWs and their clients, and vulnerable individuals living in illegal urban settlement areas. Separate IDI guides were developed for interviews with women and men by the research team, in consultation with clinic staff. IDI guides included four thematic areas of inquiry: background information; sexual behaviour and sexual health; intravaginal and menstrual practices; and female-controlled methods of HIV protection. All interviews were conducted by trained, experienced and gender-appropriate interviewers in tok pisin (VF, HA).

Socio-demographic and sexual behavioural data, including experience of male and female condom use, were collected from all participants. To facilitate discussion on intravaginal hygiene and menstrual practices (IVP), women were invited to use copies of a hand-drawn template to indicate how they wash/clean their vulva and/or vagina, and to provide commonly-used, locally-appropriate names for anatomical structures by labelling the drawing. Templates were developed by the research team in consultation with clinic staff and piloted with volunteers from the clinic for appropriateness and acceptability prior to use. Women were also asked to comment on practices they had heard other women in the community were conducting and to indicate these using the template. Men were asked whether they were aware of their sexual partners conducting such practices or if they had heard about other women in the community conducting such practices, and to use the template to indicate these. Participants indicated IVP using marker pens, highlighter pens, biro and/or pencils provided by the study team and if literate, were asked to write the names of anatomical structures onto the sheet. Where participants were unable to read or write, interviewers wrote down participant’s comments / anatomical names on the sheet on their behalf, using English and/or tok pisin depending on the language used by individual participants.

Men and women were shown male and female condoms, which interviewers’ opened and encouraged participants to handle and comment on. Participants were asked to reflect on their experiences of condom use; the factors that had determined use in different contexts and behavioural scenarios; their personal views of condom acceptability; and their views about acceptability and use among women and men in the wider community.

Interviewers introduced the concept of a vaginal microbicide to participants and demonstrated a surrogate microbicide gel and applicator, prepared prior to each interview by research staff using commercially available vaginal applicators that were pre-filled with 2-3mL KY Jelly® (Figure 2 and 3). This volume was selected based on earlier microbicide efficacy trials, in particular the MDP301 trial [4]. Interviewers demonstrated how such a product would be used in practice, based on information provided to MDP301 trial participants [73]. Participants were invited to handle the surrogate microbicide and to ask any questions they may have regarding future use of such a product.

Interviewer: ‘Now I’d like to talk to you about new methods of protection being developed. We hope that women will be able to use these to protect themselves against HIV and sexually transmitted infections in future. These products are known as ‘vaginal microbicides’ and will probably be available in the form of vaginal gels or creams. We hope that an effective microbicide will become available in the next 3–5 years. I have here an example of what a vaginal microbicide gel might look like ( Hold up pre-loaded dummy applicator ). In future, women could insert a gel like this into the vagina, either daily or just before each sex act, to protect them against HIV and sexually transmitted infections.” ( Demonstrate how applicator would be inserted into the vagina; different positions she could use for insertion; gel volume and consistency. Give her the applicator to handle. Encourage her to ask questions about microbicides ).

Interviewers then asked participants to reflect on their views of personally using such a product in future, in the context of their individual sexual lives and reported IVP e.g. would they tell their husband / sexual partner(s) that they were using such a product or might they in some situations consider covert use? Similarly, they were asked to reflect on how such a product would be perceived and used by the wider community.

All interviews were digitally recorded, transcribed verbatim and where necessary translated from tok pisin (a lingua franca of PNG) into English at the PNG Institute of Medical Research (PNG IMR) in Goroka. All personal identifiers were removed from the interview transcripts and pseudonyms given to each participant. Following data cleaning and initial review of interview transcripts, a preliminary coding framework was developed by the research team to capture data relevant to each key thematic area of inquiry (e.g. respondents views on the frequency of IVP among women in their community were coded ‘IVP-freq-comm’; individual motivators for IVP were coded ‘IVP-mot-self’). The research team then hand-coded printed interview transcripts. The coding framework was periodically reviewed as data coding proceeded and additional codes and/or sub-codes added in an iterative data management process. Themes were derived from the data during preliminary analysis of the coded transcripts, which was led by senior members of the research team (LF, AV), prior to critical review and discussion with the wider group e.g. to seek alternative themes and explanations arising from the data. A summary of key findings and their interpretation was then drafted (AV, LF) and illustrative quotes selected from individual interview transcripts in order to exemplify specific issues and concepts that had emerged across the dataset.

Ethical approval was obtained from the Medical Research Advisory Committee (MRAC) in PNG and from the Human Research Ethics Committees of the University of Queensland and the University of New South Wales in Australia. All participants provided written informed consent in either English or tok pisin. All participants were advised that they could withdraw from the study at any time without prejudice. The study was conducted in accordance with the RATS guidelines on qualitative research (http://www.biomedcentral.com/ifora/rats).

Intra-vaginal hygiene and menstrual practices (IVP) were diverse in nature, socio-cultural dimensions and motivators in this setting. A dynamic combination of traditional and contemporary practices were reported, with some women engaging in new practices that they recognised as being in conflict with traditional norms and beliefs(e.g. the use of ‘virgin soap’; sex during menstruation). Broad reproductive health concerns (maintaining genital cleanliness, removal of ‘dirt’, preventing ‘bad smell’, maintaining a ‘healthy womb’) appeared to be more important motivators of vaginal practices than specific concerns focussed on enhancing sexual pleasure or on the prevention of HIV and STIs. The WHO Gender, Sexuality and Vaginal Practices (GSVP) Study included a household survey on IVP conducted among 3610 women aged 18–60 years in Mozambique, South Africa, Indonesia and Thailand [29, 30], and found that the prevalence, motivators and contexts of vaginal practices varied markedly between countries. IVP in participating African sites appeared to be motivated primarily by a desire to maintain male partner commitment and to increase male and female sexual pleasure. Conversely, in participating sites in Asia, genital hygiene and femininity were primary motivators. Researchers in a variety of settings [15, 24, 29, 30, 33, 74] have similarly distinguished ‘sexual-related’ and ‘hygiene-related’ IVP, and highlighted the potential of vaginal washing, insert and other practices (conducted at different time points, and driven by different motivators), to modulate susceptibility to HIV/STI acquisition and transmission. For example, women who use lime or lemon juice intravaginally (e.g. as a traditional contraceptive; for protection against genital infections; or for enhanced male sexual pleasure [24, 34, 75]) are likely to be at increased risk of cervical dysplasia, cervical epithelia disruption and HIV acquisition, based on recent laboratory and clinical trials data [76–80]. The use of soaps, detergents, and other products used as part of daily genital hygiene or carried out before and/or after sex, could potentially result in vaginal irritation, inflammation or epithelial disruption, thereby increasing infection risk. A recent systematic review and meta-analysis of 15 studies in Africa and the USA concluded that intravaginal practices do not appear to protect women from STIs, genital infections or HIV, and that there is inconclusive evidence whether some practices may increase acquisition risk [22].

Depending on their nature and timing, vaginal practices also have the potential to negate the protection afforded by vaginal microbicides, either by mechanical means (e.g. microbicide gel is diluted or washed out) or through chemical interaction (e.g. vaginal insert chemically reacts with microbicide product or alters intravaginal environment sufficiently to reduce product effectiveness). Although we attempted to clarify the temporal relationship between IVP and vaginal sex in this study, it was not possible to establish precise estimates during in-depth interviews with participants. Vaginal insert practices appeared uncommon in this setting and no respondents reported the insertion of lemon or lime juice. Sex during menses was however common and has the potential to increase the risk of HIV infection due to blood exposure (particularly transmission to uninfected male partners [81–85]), and to reduce microbicide effectiveness due to altered vaginal pH and other physiological changes [86–88]. Further research is warranted, using combinations of behavioural research methods as utilised in other settings [24, 74, 89–91], to better understand these inter-relationships and their implications for future microbicide uptake, adherence and effectiveness in PNG.

Notional acceptability of vaginal microbicide gels for HIV/STI prevention was high among both women and men in this study. Microbicide acceptability was perceived within a broad reproductive health framework i.e. that use would promote a ‘healthy womb’ or help ‘prevent sickness or pregnancy’, in addition to preventing HIV. This perception was held by women and men and is in contrast to advice provided by the research team. Conversely, one female respondent was concerned that gel could ‘block’ menstrual flow and cause sickness in the womb. Many women and men thought that microbicides would best be used by sex workers and considered use between ‘trusted partners’ unnecessary. Most women were confident that they would be able to negotiate gel use with their sexual partners in future and many were prepared to use microbicides covertly should their partner refuse. Other women were concerned that increased vaginal ‘wetness’ would make truly covert use difficult and could expose them to increased risk of physical or sexual violence. There were mixed views on the potential of microbicide gels to modify sexual pleasure for women and their male partners. Some women reported that their partners preferred ‘their vagina dry’ and that men would ‘ask women to clean it [out]’ should they use gel; one woman reported that increased vaginal lubrication would make sex more pleasurable for women. Women were confident that they could use a product similar to the surrogate microbicide in future but some were concerned that initial use might be problematic without appropriate counseling and support from health workers. These findings clearly have implications for future microbicides research and for the introduction of safe and effective vaginal microbicides for HIV prevention in PNG.

Low levels of awareness and lack of experience using female condoms are cause for concern, but also represent an important opportunity for comprehensive HIV prevention programs in PNG, given that notional acceptability was high among women and men in this study. Further research to clarify in-use acceptability and adherence of female condoms, including gendered sexual dynamics, socio-cultural contexts and motivators influencing use, are warranted to inform female condom promotion programs in this setting.

A challenge in countries such as PNG that have great geographic, linguistic and cultural diversity is to develop culturally-nuanced and innovative strategies for HIV prevention. This study has shown that women and men from a variety of socio-cultural backgrounds, who are at increased risk of HIV/STI acquisition, are notionally supportive of vaginal microbicides for HIV prevention in this setting. The need to engage male partners in microbicides research and in strategies for future product introduction has been recognized [13, 27] and is essential in countries such as PNG where sexual dynamics are particularly male dominated. It remains unclear how other microbicide formulations currently in development (such as intra-vaginal rings, dispersible vaginal tablets and films [8–10]) would be perceived and most effectively incorporated into people’s individual sexual lives in this setting. In-use acceptability research planned by our group will address these issues in future.

What are the implications of our research findings for HIV prevention in PNG and other Asia-Pacific countries? Firstly, our findings need to be interpreted with caution given the modest sample size, purposively selected from a population of women and men attending an urban sexual health clinic in a disadvantaged area of Port Moresby, making generalisability problematic. This population is however considered representative of people at increased risk of HIV and STI acquisition in PNG, a key potential target group for the future introduction of vaginal microbicides and other biomedical prevention technologies. Other limitations are that the notional acceptability of alternative vaginal microbicide formulations and dosing regimens were not evaluated in this study (including coitally-independent products such as intravaginal rings); and because this research was conducted prior to the successful CAPRISA 004 trial result becoming available, gel acceptability was not evaluated in the knowledge that a safe and effective vaginally-administered product had been identified. Although biomedical prevention technologies such as male circumcision and vaginal microbicides are expected to play an important role in comprehensive HIV prevention strategies in high-burden settings, their role in moderate prevalence settings is less clear. A key challenge having investigated notional acceptability and the determinants of acceptability will be to decide how future interventions could best be implemented in order to ensure the greatest public health impact. Population-level implementation of a moderately-effective vaginal microbicide is likely to be less cost-effective in lower burden settings than targeted introduction among those most at-risk of HIV acquisition [92–94], such as women in HIV sero-discordant relationships or female sex workers [95]. The ethical and moral hazards of not making such products widely available to all women clearly also need to be considered, particularly in countries where women’s sexual agency may be very limited, but should be balanced against potential risks such as condom migration [12, 14] and the emergence of antiretroviral drug resistance [96]. Comprehensive guidance on these issues for low and moderate burden settings is not currently available but is urgently required to guide HIV prevention policy outside Africa. Complementary health systems research [97, 98] and epidemiological mathematical modeling [92, 93] being conducted by our group are assisting policy makers and development partners in PNG to develop evidence-based policy. This approach may represent a model for other moderate-burden countries considering a combination of biomedical prevention technologies for HIV prevention.

Notional acceptability of a vaginal microbicide gel for HIV prevention was high among women and men at increased risk of HIV and STI acquisition attending a sexual health clinic in PNG. Further research is warranted to investigate the in-use acceptability and adherence of different microbicide formulations in this setting, and to establish the role of vaginal microbicides and other biomedical prevention technologies in countries experiencing moderate-burden HIV epidemics.

We obtained verbal consent from all those in the photograph.