FACTORS DETERMINING THE VULNERABILITY OF WOMEN TO SEXUALLY TRANSMITTED HIV : A LITERATURE REVIEW

Gender-related vulnerability is described as a crucial factor contributing to increased susceptibility of women to HIV, accounting for more women than men being infected. At the same time, empowerment interventions are being promoted as effective strategies for increasing the ability of women to adopt protective behaviours. The aim of the review was to identify, collate and categorise the factors determining the gender-related vulnerability of women to sexually transmitted HIV. A review of literature from theoretical and empirical studies using diverse methodologies was undertaken. Reports included those identified through electronic and manual searching. Twenty factors, forming five clusters, were identified as influencing the ability of women to adopt protective behaviours. Each factor was analysed to describe its component parts and the relationship between a factor, gender-related vulnerability, HIV risk level and empowerment status. Further analysis provided a description of markers named predictors and indicators. The literature portrays markers that can be identified and used to describe gender equality status, HIV risk level and related empowerment. This provides the potential to identify factors in gender equality status and HIV risk level to address in programmes designed to empower women in order to lower their risk to sexually transmitted HIV.


INTRODUCTION
South Africa continues to have the largest number of people living with HIV/AIDS in the world with AIDS being one of the ten leading causes of death in 2005 (Statistics South Africa, 2007:19).Tuberculosis, the leading cause of death in South Africa in 2005, has a high prevalence of co-infection with HIV, and accounts for up to a third of AIDS deaths worldwide (World Health Organization, 2008:1).HIV prevalence, estimated to be 10.8% amongst persons aged two years and older, increases dramatically among young females and peaks at 33.3% of the 25-29 age group (Pettifor, Rees, Kleinschmidt, Steffenson, MacPhail, Hlongwa-Madikizela, Vermaak & Padian, 2005:1526;Shisana, Rehle, Simbayi, Parker, Zuma, Bhana, Connolly, Jooste & Pillay, 2005:124, 135).Gender issues related to a woman's place in South African society is described as an important factor contributing to the rapid spread of HIV in this region.This is due to unequal gender power relationships fuelled by negatively scripted, culturally defined gender roles and poverty.This lack of power of women within their social and economic contexts thus needs to be addressed within the HIV discourse.Empowerment is considered to be the key to enabling women to protect themselves from HIV infection (Bentley, 2004:257;Shearer, Hosterman, Gillen & Lefkowitz, 2005:312;Greig & Koopman, 2003:195;Ackermann & De Klerk, 2002:163;Buvé, Bishikwabo-Nsarhaza & Mutangadura, 2002:2013;Kim & Motsei, 2002:1248;Doyal, 2001Doyal, :1061Doyal, -1063;;Umerah-Udezulu, 2001:5;Amaro & Raj, 2000:724;Parker, Easton & Klein, 2000:S23-S24;Pettifor, Measham, Rees & Padian, 1996:2003;Preston-Whyte, 1995:218).The problem seems to be the unequal gender power relationship, which makes women more vulnerable to HIV infections and this will be examined in more detail.
This article describes a literature review undertaken to identify the factors influencing the ability of women to adopt safe sexual practices.The purpose of this review is to provide background information, as part of a larger study, for planning a holistic HIV intervention aimed at empowering women.To design such an intervention, it is important to understand the context, challenges and choices of women in South Africa (Stadler, Delany & Mntambo, 2008:189-190).Thus the framework reported in this study will be used as a tool for planning an intervention for the empowerment of women within the context of HIV in South Africa.
The literature review was based on an approach described by Kirkevold (1997:981) as a synthesis review, a method whereby isolated information is integrated into a "more comprehensive and internally consistent whole".This enables the integration of separate studies with a different focus and using a variety of methodologies that provide a comprehensive account of the phenomenon being reviewed.Kirkevold (1997:981) argues that this type of review is a powerful knowledge development tool, as it enables knowledge to be accumulated that is "beyond merely evaluating the strength and weaknesses of existing knowledge".

SEARCH METHODS
The search was of literature from 1995-2007.The SWICE database was used, which searches AMED, BNID, CINAHL, DH DATA, EMBASE, KINGS FUND, MEDLINE and PsychoINFO.The search terms used were 'HIV', 'women', 'vulnerability', 'South Africa' and 'southern Africa'.Manual searching revealed further literature and reports were added as they were published.
Primary studies, review articles, international and national reports, and national policy documents were included.A primary study was defined as primary empirical research including studies using qualitative and quantitative methods (Whittemore & Knafl, 2005:546-547;Kirkevold, 1997:983).The following inclusion criteria for literature sources were applied: • the study reported on women's ability to choose and/or adopt protective behaviours in hetero-sexual relationships; • it identifies predictors or indicators of safety or risk to sexually transmitted HIV between heterosexual couples; • it was conducted in southern Africa; • a study undertaken elsewhere which explicated the topic in more depth than the regional data; and • studies available in English.
It was decided that the value of the article in providing insight into the contributors of either safe or unsafe sexual practices would serve as the primary selection criterion.For the identification and/or confirmation of factors in the various clusters, 91 articles were consulted.

DATA ANALYSIS
Literature sources were analysed using a constant comparison method aiming to identify similarities and differences in the findings across the studies (Dellve, Abrahamsson, Trulsson & Hallberg, 2002:140-141;Creswell, 1998:57).Through this process the findings from each study were coded, grouped into categories and compared with the original data to ensure that the fit was appropriate.The categories were then refined to create a description of 20 factors determining women's vulnerability to sexually transmitted HIV.Each identified factor was analysed to consider the effect of a status of empowerment or disempowerment on the factor.The primary researcher and two co-researchers (supervisors) categorised the literature and reached consensus on the identification of factors and clusters.It is acknowledged that bias may be evident owing to the constraints of access to literature, and this may be a limitation of a review of this nature in terms of generalisation to other contexts.However, this review has value within the broader framework of this study where the aim of the identification of factors is to assist in planning an intervention for the empowerment of women within the HIV context in South Africa.
Scope for decision-making differentiates between predictors describing relationships that restrict decisionmaking compared with ones that allow for autonomous decision-making.Linked indicators describe the practice of abstinence, marital status, age at first marriage and whether marriage was arranged or chosen by the woman (Zellner, 2003:41;Gregson et al., 2002Gregson et al., :1897Gregson et al., -1899; Mbulaiteye et al., 2002:41).

School attendance
In developing countries school attendance is a predictor of the cultural value attached to being a girl and a mother's belief in education for her daughter.Non-attendance or infrequent attendance leads to economic dependence and lowers future prospects (Grown, Gupta & Pande, 2005:541).Adolescent and young women occupied by school activities are less likely to be engaged in sexual activity.However, sexual exploitation of girls at school places them at risk (Shisana & Simbayi, 2002:63;Human Rights Watch, 2001:1, 36, 61).School provides an opportunity for education on sex and HIV, therefore a lack of formal schooling can be seen as increasing vulnerability to HIV (Grown et al., 2005:541;Blake, Ledsky, Goodenow, Sawyer, Lohrman & Windsor, 2003:958-961;Blum, Halcon, Buehring, Pate, Campbell-Forester & Venema, 2003:459;Karim et al., 2003:18;Lum, Kristen, Ochoa, Judith, Hahn, Shafer, Evans & Moss, 2003:919-

Level of education
Educational level is a predictor of gender power and role (Grown et al., 2005:541).Low educational levels are associated with an increase in vulnerability and risk resulting from fewer life opportunities, poor nutrition and hygiene, higher mortality, lower fertility rates and lower economic development (Kongolo & Bamgose, 2002:86;Bowleg, Belgrave & Reisen, 2000:614).Garbus (2003:3) found that economic status correlates directly with HIV prevalence where the greatest number of persons infected is situated at the lowest socio-economic point and HIV prevalence decreases with rising economic status.Girls and women face particular risks of HIV infection as the interplay between their economic positions and social status influences the decisionmaking authority given to them.Therefore the predictor describes income status and the indicators are employment status, income bracket and ability to support dependants (Grown et al., 2005:542;Buvé et al., 2002Buvé et al., :2014Buvé et al., , 2016;;Gilbert & Walker, 2002:1103;Kehler, 2001:1;Amaro & Raj, 2000:724;Preston-Whyte, 1995:220).Financial dependence and poverty are criti-cal in determining whether sex will be exchanged as a commodity, as transactional sex limits decision-making ability and the scope for negotiating safe sexual practice.Sex trafficking is increasingly being associated with South Africa and places victims at risk of HIV.A predictor of risk is the practice of transactional sex and the indicators describe reports covering exchange of sex for gain, sex trafficking and scope for negotiating safe sexual practice (Shelton et al., 2004:892;Garbus, 2003:35;Buvé et al., 2002Buvé et al., :2014;;Jewkes & Abrahams, 2002:1232;Wojcicki & Malala, 2001:102;Gutierrez, Oh & Gillmore, 2000:582;Preston-Whyte, 1995:218).

Economic status
HIV risk linked to employment is the predictor of occupational risk.Indicators differentiate between work associated with HIV risk, such as hospital workers, and workers who, owing to the nature of their jobs, are at increased risk (e.g.migrant workers and truckers).
The religious socio-cultural context can influence women's rights through either promoting or limiting equality of opportunities.The predictors distinguish between a context that is non-discriminatory and one that discriminates against women.This is indicated through religious guidelines (doctrine) that support or disregard gender equality and indicate level of gender inclusiveness.The predictor of religious teaching distinguishes between teaching that supports/enables gender equality or contributes to gender inequality.Indicators describe religious teaching on gender and safe sex, as well as practices harmful to women.

Living arrangements
This category measures whether living arrangements offer protection from unwanted sexual assault.sure to accurate sex information (Eaton et al., 2003:160;Wood & Jewkes, 1998:12-15, 36;Swart-Kruger & Richter, 1997:958).Indicators distinguish between an environment offering protection and one that does not do so, as well as the type of dwelling in which a woman lives.Studies by Swart-Kruger and Richter (1997:964) and Eaton et al. (2003:160) provide evidence that women living with strangers, in institutions or on the street are particularly vulnerable.Females living in any setting, however, including their homes, are at risk owing to sexual assault.

Cluster 2: Cultural and societal factors
Cluster 2 comprises gender power balance and culturally scripted roles.

Gender power balance
Power balance is the predictor and the indicators describe who controls the power, the distribution of power and whether negative consequences of requests for condom use will be a barrier to safe sexual practice.

Culturally scripted roles
The cultural milieu provides the context and describes the culturally prescribed role of women that either upholds or infringes the rights of women.The value given to the female role portrays women as valued or devalued.The subordinate female position entrenched in many African societies limits the control women have over their own sexual choices and over their husbands' behaviour outside of marriage.This vulnerability to HIV is exacerbated by extramarital affairs, which may be accepted and even encouraged as a show of masculinity through early sexual initiation and many sexual conquests.Roles that emphasise innocence, virginity, submission to male prerogative, low status, erotic fantasy, myth (cure of HIV by having sex with a virgin), ignorance, lack of knowledge regarding sex and HIV are harmful to women (Shearer et al., 2005:320;Andersson et al., 2004:955;Buvé et al., 2002Buvé et al., :2014Buvé et al., -2016;;Kim & Motsei, 2002:1247;Umerah-Udezulu, 2001:4;Bujra, 2000:13-14;Jewkes, Penn-Kekana, Levin, Ratsaka & Schrieber, 1999:3, 8, 23).Therefore the indicators describe whether women are accepted as equal partners or as subordinates and how the media, a powerful shaper of opinion, portrays the value of women.
The Constitution of the Republic of South Africa 108 of 1996 upholds the rights of individuals and therefore anything limiting this is an infringement (quoted by Bentley, 2004:247).Patterson and London (2002:966) motivate for an approach to supporting human rights that also responds to a socio-economic rather than a simplistic "libertarian approach" to rights.The indicators describe the impact of rights being upheld for people who are HIV positive (Bentley, 2004:257).Stigmatisation is an infringement and fear of being stigmatised can be a barrier for people who wish to know their HIV status through HIV testing or to disclose their positive serostatus (Campbell et al., 2005:812-813;Burke, 2004:423;Parsons, Vanora, Missildine, Purcell & Gomez, 2004:459).Stigma can have serious consequences such as isolation and neglect, further increasing the burden on women.Many keep silent for fear of being identified, which prevents them from seeking treatment that could prolong life (Garbus, 2003:33;Lanouette, Noelson, Ramamonjisoa, Jacobson & Jacobson, 2003:918-919;Stein, 2003:2-3;Umerah-Udezulu, 2001:6;Mathews, Kuhn, Fransman, Hussey & Dikweni, 1999:1238).However, human rights of an individual, when respected and upheld, promote social inclusion of people who are HIV positive as well as an increase in disclosure.

Cluster 3: Communication
The third cluster refers to communication with family, friends, colleagues, sexual partners and the influence of media messages.
Communication within the primary group, including family, friends and work colleagues, determines the level of information girls and women have regarding sexual issues and HIV.The predictor therefore describes whether women are informed or uninformed.Indicators describe communication style, information covering sexual matters, HIV and knowing the partner's HIV serostatus (Bird et al., 2001:234;Amaro & Raj, 2000:736) Communication with a sexual partner is predicted by communication style with indicators being discussion on sexual matters, HIV and safe sexual practices (Bird et al., 2001:234;Amaro & Raj, 2000:736).The influence of media on female empowerment refers to the portrayal of women by the media as well as messages relayed on practices that are harmful to women and sexual practices (Keller & Brown, 2002:68).

Cluster 4:Perceptions
This cluster refers to how individuals perceive their own vulnerability to HIV and the influence of normative behaviour.

Perception of vulnerability to HIV
A determinant of sexual practice is a woman's perception of her risk.Being in a committed relationship is associated with a perception of low HIV risk (Amaro & Raj, 2000:724).HIV/AIDS complacency can be defined as minimising, discounting and discrediting the threat of HIV/AIDS and can lead to inaccurate assessments of personal risk.This may decrease motivation to adopt safe sexual practices although it might increase disclosure of an HIV-positive status to a partner (Burke, 2004:422;Parsons et al., 2004:471;Amaro & Raj, 2000:725).Knowing someone who has HIV or who has died from AIDS is hypothesised to increase awareness of HIV/AIDS and its threat to personal health (Camlin & Chimbwete, 2003:231;Shisana & Simbayi, 2002:87), yet many South Africans do not know their own serostatus despite having access to voluntary counselling and testing (Shisana & Simbayi, 2002:66).Refusal to test for HIV status is most often linked to concerns related to stigmatisation in the event of testing positive (Garbus, 2003:23).Women perceiving themselves at a low risk of HIV infection are not highly motivated to be tested for HIV (Garbus, 2003:33;Agha, 2002:113;Bird et al., 2001:234;Amaro & Raj, 2000:740;Bowleg et al., 2000:630;Gutierrez et al., 2000:583;Reid, 2000:713;Coleman & Mngomezulu, 1999:1060;Swart-Kruger & Richter, 1997:962).
Cluster 5: Self-efficacy Self-efficacy describes the confidence and competence to practise safe sex and covers the application of knowledge, the skills needed to practise safe sex and the attitudes and feelings influencing decision-making and therefore contributing to sound judgement.

Practice
Self-efficacy refers to the confidence and competence of an individual to practise safe sex.It includes the intention and application of skills (Rose, 2004:25;Eaton et al., 2003:158).Skills include the ability to resist sexual advances, negotiating condom use and correct and consistent condom use (Eaton et al., 2003:159).

Attitudes and feelings contributing to sound judgement
Attitudes relating to the self play an important role in sexual choices.Low self-esteem and feelings of poor self-worth have been linked to early sexual debut and a need to have multiple sexual partners as a way of selfaffirmation (Eaton et al., 2003:157-158).Self-esteem and self-confidence increase assertiveness (Rose, 2004:27).Rose (2004:25) states that feelings of selfworth and confidence give a feeling of being safe to make choices for behaviours that are protective.

DISCUSSION
This integrated review of the literature attempted to describe the factors determining women's vulnerability to sexually transmitted HIV.A challenge of the review was finding a method to deal with the volume of literature on the topic and the emphasis on a subset (South African) and the perceived similarity to the population being studied.
There has been a proliferation in studies pertaining to HIV.These studies cover a wide spectrum including prevention, transmission, clinical trials and social and economic aspects.It is not possible to be sure that all relevant studies have been included in this review.
Amendments to the findings were made as new information became available.For example, initially, school was classified as a protective factor, but studies indicating that in certain schools inappropriate behaviour by some teachers increased the vulnerability of girls required amendments to be made (Brookes, Shisana & Richter, 2004:35, 39;Shisana & Simbayi, 2002:83, 95).
According to Parker et al. (2000:S29-S30), it is feasible that research conducted in some parts of the developing world is also pertinent to issues encountered in inner city populations in the USA.The literature review described was undertaken as a prequel to a study in South Africa, so the emphasis was on studies from southern Africa.Therefore the findings may be most relevant to the region and would need to be tested in other settings before comment can be made as to this study's usefulness in other contexts or its generalisability.Cooper (1998:43) states that generalisability is increased in integrative reviews containing articles conducted at different times, places, with varying samples of ages and races, as well as studies applying different methodologies.In this study the references cited reflect this diversity.
This study sought to describe all the related factors associated with women having an increased susceptibility to HIV and to better understand how their resilience could be promoted.This could lead to the perception that only some people are vulnerable, but the reality is that anyone who has been exposed to the virus is at risk of contracting HIV.From the description provided it may appear that the factors are equal in all circumstances while the literature emphasises the pivotal role of gender, relationships, cultural and societal factors (examined in this study as relationship factors, gender power balance and culturally scripted roles).
The constant comparison method used in reviewing the various documents provided the categories and the predictors and indicators linked to each category.The process resulted in a description of the factors influencing the vulnerability status of women to HIV and the relationship between empowerment, as a facilitator promoting the adoption of protective behaviours and safe sex (Shearer et al., 2005:312;Greig & Koopman, 2003:197;Amaro & Raj, 2000:728).
The first aim of this study was to identify, collate and categorise the factors determining the gender-related vulnerability of women to sexually transmitted HIV.The 20 factors identified provided a comprehensive description of the factors influencing the gender-related vulnerability of women to sexually transmitted HIV, demonstrating the breadth and complexity of the factors that need to be addressed in HIV prevention programmes.
The second aim was to indicate the link between empowerment and each factor.Empowering aspects that would influence gender equality status and in turn inform HIV risk level were identified.These factors should be borne in mind when planning interventions for the empowerment of women within this context to minimise the risk of HIV infection.This finding supports the understanding that HIV prevention programmes will be most effective if they address gender empowerment issues as identified in this study rather than specific behaviours only (Di Noia & Schinke, 2007:486;Greig & Koopman, 2003:195;Petersen & Swartz, 2002:1010;Parker et al., 2000:S27-S28).

CONCLUSION
The literature demonstrates that empowerment is a facilitator promoting the ability of women to adopt safe sexual practices and the value of adopting an empowerment approach in HIV prevention programmes.The identification of the factors influencing the vulnerability of women to sexually transmitted HIV, with related predictors and indicators, enables planners to differentiate between a position of gender equality compared to gender inequality and between a high or low level of HIV risk, and so conduct an assessment of the level of HIV risk.Assessment will highlight areas of risk, thus promoting the use of targeted HIV prevention programmes.The indicators provide the potential to monitor and evaluate the effectiveness of projects or programmes designed to empower women while lowering their risk to sexually transmitted HIV.
During the process of analysis of each of the factors and the influence of empowerment/disempowerment on each one, predictors and indicators emerged.Twenty factors, grouped into the following five clusters, were identified as determining the vulnerability of women to sexually transmitted HIV: bio-psychosocial and demographic factors, cultural and societal factors, communication factors, perceptions and self-efficacy (see Figure 1).This classification was selected based on examples of categories found in research questionnaires in the studies reviewed.Each factor was analysed to reveal its component parts, which were then classified as predictors of gender equality status or indicators of HIV risk level.This demonstrated the relationship between a factor, gender-related vulnerability and HIV risk status, while at the same time distinguishing between the effects of a status of empowerment (predictors of gender equality and indicators of low HIV risk) or disempowerment (predictors of gender inequality or indicators of high HIV risk) on each component (for examples see Figure 1 and Figure 2).Each of the five clusters and their concomitant factors will be discussed in more detail in the section that follows.Cluster 1: Bio-psychosocial and demographic factors Bio-psychosocial and demographic factors, for the purposes of this review, include age, relationship status, school attendance, level of education, economic status, religious affiliation, ethnicity and race, geographical location of residence and living arrangements.
Vulnerability is linked to lack of protection and behaviours such as incest, sleeping 'rough' and being in prison (for males).Identified risk behaviours include earlier sexual debut, more sexual partners, rape and forced sexual activity as part of survival as well as lower condom use.Adolescents on the street are often non-school attendees, with higher rates of illiteracy and less expo-