Abstract
The objective of this cross-sectional household survey was to assess factors influencing HIV risk perception, behaviour and intervention uptake in a community characterised by high HIV prevalence and availability of antiretroviral therapy (ART). The survey was conducted in Khayelitsha, South Africa and involved two-stage sampling with self-weighting clusters and random selection of households within clusters. One man and woman between 14 and 49 years old was interviewed in each household; 696 men and 879 women were interviewed for a response rate of 84% and 92% respectively. Ninety-three percent and 94% were sexually active with median age of sexual debut 15.3 and 16.5 years. Eighty-three percent and 82% reported a partner at the time of interview and 29% and 8% had additional partner(s). Forty-one percent and 33% reported condom use during the last sexual encounter. Thirty-seven percent of men not using condoms did not as they believed their partner to be faithful, whilst 27% of women did not as their partner refused. Twenty-eight percent and 53% had been tested for HIV. Having undergone HIV testing was not associated with condom usage, whilst current relationship status was the strongest association with condom usage for both men and women. In spite of a relatively high uptake of condoms and testing as well as ART availability, the HIV epidemic has continued unabated in Khayelitsha. Even greater coverage of preventive interventions is required, together with a national social and political environment that builds on the availability of both preventive and treatment services.
Introduction
The moral, social and economic imperative of providing antiretroviral treatment (ART) for people living with AIDS in resource-poor settings is today no longer questioned. However, the duality between prevention and treatment continues. In rich countries, prevention and treatment programmes are independently funded. Treating a patient with advanced AIDS is not put in balance with resource allocation for prevention, as is often the discourse in poorer countries (Marseille, Hofmann, & Kahn, Citation2002) where a false dichotomy between treatment and prevention is created. Such dichotomies can be detrimental to policy development for both prevention and care, whilst increasingly it is argued that these should be mutually reinforcing (Berkman, Citation2001; Coetzee et al., Citation2004; De Cock & Grubb, Citation2006; Salomon et al., Citation2005). Proponents of ART have argued that turning AIDS from a death sentence into a treatable chronic disease will alleviate the surrounding stigma and that breaking the silence results in a more open approach to prevention (Médecins sans Frontières & University of Cape Town, Citation2003).
In South-Africa, numerous surveys have shown that the majority of people interviewed are aware of HIV, know its sexual transmission route and the protective benefits of using condoms (Johnson & Budlender, Citation2002; Kelly & Parker, Citation2000; Pettifor et al., Citation2005b; Shisana & Simbayi, Citation2002). This knowledge has not, however, translated into a marked reduction of new HIV infections; on the contrary national HIV prevalence has risen in South Africa from less than 1% in 1990 to almost 25% within 10 years (Department of Health, Citation2005). The aim of this survey was to assess HIV risk perception and risk behaviour in a specific context characterised by high HIV prevalence and exposure to treatment interventions, and to explore at a community level any association between the uptake of preventive interventions and the awareness of ART.
Setting
Khayelitsha is a township on the outskirts of Cape Town, with a population of approximately 400,000 (Municipal Demarcation Board, Citation2007). The HIV seroprevalence in the two public-sector antenatal clinics in this district was 15% in 1999 and 33% in 2004 (Department of Health, Citation2005). Unemployment is high at 43%. Most residents rely on health services delivered by the state. In 1999, the first district-wide public sector programme in South Africa to prevent the transmission of HIV from mother-to-child (PMTCT) was initiated in Khayelitsha (Abdullah, Young, Bitalo, Coetzee, & Myers, Citation2001). Dedicated clinics for adults and children with HIV were established in April 2000. In April 2001, the first patients were started on ART. By the end of the survey (March 2004), more than 1500 people were on ART in Khayelitsha, representing at that time one of the areas with the highest ART coverage. Other preventive interventions such as VCT, condom distribution, treatment of sexually transmitted infections and peer education have generally been available, although coverage of these interventions has never been systematically assessed.
Methods
Study design and sampling
The survey was conducted from September 2003 to March 2004, using a two-stage cluster design. The primary sampling units were enumerator areas (EA's) as defined for the national census, each consisting of around 200 households. Eighty out of a possible 622 clusters were selected on a probability proportional to size basis. Within each EA, 10 households were randomly sampled. In each household, one randomly selected man and woman aged between 14 and 49 years of age was interviewed. If eligible participants were not at home at the first visit, up to two repeat visits were made. The sample size was calculated based on previous surveys in Khayelitsha, which estimated condom use during last sexual intercourse at around 50% (Parker, Oyosi, Kelly, & Fox, Citation2002). Anticipating a design effect of 2, it was calculated that a sample of 800 men and 800 women between 14 and 49 years old would be required in order to provide a 95% confidence interval (CI) spanning 3% either side of this parameter for men and women separately.
Measurement
The questionnaire was developed based on prior qualitative work, the UNAIDS Best Practice Collection on Behaviour Monitoring and the questionnaire of a multi-centre study (Lagarde et al., Citation2001a), adapted and translated into isiXhosa and piloted in the community. The questionnaire is available on request.
Data management and statistical analysis
All the data were double entered and validated. Weighted proportions and means with 95%CI for patient characteristics, HIV/AIDS knowledge and sexual behaviour were estimated for men and women separately. Each cluster was weighted equally and CIs were calculated with cluster as primary sampling unit.
The associations between possible predictors and condom use or VCT were assessed using simple and multiple logistic regression models, adjusted for correlated outcomes at the cluster level using generalized estimating equations (GEE) (Hardin & Hilbe, Citation2003). The relationship between age and condom use or VCT was modelled using a piecewise linear effect.
Ethical approval for the study was obtained from the ethical committees of the University of Cape Town, South Africa and the Institute of Tropical Medicine in Antwerp, Belgium.
Results
Community characteristics
A total of 1576 people were interviewed, of whom 879 (55%) were women (). The response rate was 84% for men and 92% for women. The mean age of respondents was 28.3 years in men and 28.7 in women. A total of 1127 households were visited, with a mean of 3.0 adults and 1.1 children under 14 years. Two-thirds were informal structures. Although only 5% of respondents were born in Khayelitsha, more than half had been residing in Khayelitsha for 10 years or longer.
Table 1. Sociodemographic characteristics and relationships of 14–49-year-old men and women in Khayelitsha (South Africa), 2003–2004.
The literacy rate (defined as being able to read a newspaper in isiXhosa) exceeded 90% and was higher for women (98%) than for men (90%). Overall, 28% of men and 29% of women had obtained a school leaving certificate. Women were more likely to report being unemployed (45% versus 28%) and of those women who had some employment, they were less likely than men to be in full-time employment (17/34% compared to 33/51%). This is corroborated by a higher proportion of women relying on others for income. More women however were able to access social grants as a source of income (16% versus 2%).
In response to a question on circumcision, 75% of all male respondents reported having been circumcised, whilst; the median age of circumcision was 21 years.
Risk factors for HIV
In all 93% of men and 94% of women had ever been sexually active (), with an average age of sexual debut of 15.3 and 16.5 years. The total number of lifetime sexual partners reported was most frequently 5–10 for men and between 2–4 for women. In every age group men reported more lifetime sexual partners than women.
Table 2. Sexual risk factors for HIV/AIDS of 14–49-year-old men and women in Khayelitsha (South Africa), 2003–2004.
Men reported that their current partner was on average 4.4 years younger than them, accentuated when looking at men above 40 who reported partners seven years younger on average, whilst women of all ages reported partners who were 5.1 years older on average.
Of the 83% of men and 82% of women who reported a current sexual relationship at the time of the interview, 29% and 8% stated they had been sexually involved with someone outside of their regular relationship in the previous 12 months. Of these, men were more likely than women to have more than one additional partner.
Knowledge about and exposure to HIV
The majority of respondents cited sexual and blood transmission as the modes of transmission for HIV (). Almost half of women identified vertical perinatal transmission. The majority of respondents spontaneously identified condoms as a means of preventing HIV infection.
Table 3. Knowledge about HIV/AIDS transmission and prevention in 14–49-year-old men and women in Khayelitsha, 2003–2004.
Over half the respondents knew someone with HIV (54% of men and 63% of women), with women more likely to report know of a family member infected than men (19% versus 10%) and similarly more likely to report having cared for someone with HIV. Close to three-quarters knew someone who had died due to HIV (73% and 76%).
Weight loss (79% & 74%), diarrhoea (29% & 31%), immobility (29% & 21%), hair loss (44% & 30%) and lesions on the lips (36% & 20%) were the symptoms and signs respondents most commonly identified with HIV. Twenty-two percent of men and 14% of women felt one could tell if someone had HIV by looking at them.
Condom use
When asked whether a condom was used the last time they had sex with any partner, 41% of men and 33% of women responded that they had (). This was highest in the youngest age group with 68% of boys and 56% of girls reporting they used a condom at last intercourse.
Table 4a. Condom use in 14–49-year-old, sexually active men and women in Khayelitsha (South Africa), 2003–2004.
Protection against HIV and other infections were the main reasons given by respondents for using condoms. There were large gender differences however in the reasons given for not using condoms; 37% of men felt they did not need to use condoms since “their partner is faithful” (compared to 23% of women), whereas 27% of women reported not using condoms as their partner refused (compared to 3% of men), rising to 72% when limited to those women who had wanted to use a condom in their most recent sexual encounter but did not. Less than one in ten respondents (7% of men and 10% of women) reported buying condoms.
Models to explore associations with condom use confirmed that condom use was highest in the youngest age groups and decreased with increasing age until age 35 in men and 25 in women (). Condom use was higher for respondents with at least some secondary school education (OR: 1.3 and 1.9 in men and women respectively). The lowest condom use was seen in couples living together and highest in short-term relationships. For men, condom use was higher among those who knew how HIV is transmitted (OR: 1.8), who knew that condom use and abstinence prevent HIV infection (OR: 10.3 and 1.5) and among those who discuss HIV/AIDS with friends (OR: 1.7). For women, condom use was negatively associated with living in a one-roomed house (OR: 0.6) but higher in those who knew someone who had died of AIDS (OR: 1.7), in those who discussed HIV/AIDS with others (OR: 1.6) and in those with specific knowledge of ART and the need for treatment to be taken life-long (OR: 2.1).
Table 4b. Multiple logistic regression model for association with condom use during last sexual contact in 14–49-year-old, sexually active men and women in Khayelitsha (South Africa), 2003–2004.
Voluntary counselling and testing
Over three-quarters (78%) of men and almost all (97%) women said they knew where they could be tested for HIV and 28% of men and 53% of women reported having been tested (). When respondents who had tested were asked whether they would feel comfortable disclosing their HIV status to the interviewer, the minority were uncomfortable disclosing (4% of men and 1% of women). Of note, only 1% of men and 6% of women who had tested reported testing HIV-positive.
Table 5a. Uptake of VCT and knowledge of interventions in 14–49-year-old men and women in Khayelitsha (South Africa), 2003–2004.
Models restricted to sexually active individuals demonstrated a number of associations with having tested for HIV (). The strongest associations in men were a partner (OR: 9.6) or other acquaintance (OR: 3.3) having tested for HIV, increasing age until 30 (OR: 1.2/year), being full-time employed (all other categories negatively associated), being in a relationship but not living with their partner, having a partner less than five years younger (OR = 0.52 if greater) and knowing that condoms prevent HIV infection (OR: 2.3). In women, increasing age above 30 was negatively associated with VCT (OR: 0.9/yr), whilst not working or job-seeking (OR: 3.8), being born somewhere other than the Eastern Cape, having older partners (OR: 1.6), openness about HIV (OR: 1.9), having cared for someone with HIV (OR: 2.1), a partner having tested (OR: 2.5) and knowing that condoms prevent HIV (OR: 2.6) were positively associated with VCT. In addition, knowledge about PMTCT (OR: 4.2) and knowing people on ART (OR: 2.6) were strongly associated with having tested.
Table 5b. Multiple logistic regression model for association with VCT uptake in 14–49-year-old, sexually active men and women in Khayelitsha (South Africa), 2003–2004.
Knowledge of prevention and treatment interventions
Overall, 8% of men and 7% of women believed that HIV could be cured (). More women than men had heard of ART (43 versus 27%).
For all ages combined, radio was the most common sources of information on ART, followed by television. Four respondents were themselves on ART, whilst 23% of men and 36% of women knew of the Treatment Action Campaign, the most prominent organisation in the area advocating for treatment.
Discussion
Khayelitsha is characterised by extremely high levels of HIV sero-prevalence mirroring that in South Africa more generally. Living in Khayelitsha reflects many of the challenges faced by the urban poor, who are generally considered to be at higher risk of HIV infection for a myriad of reasons. Khayelitsha is a community that has been at the forefront of activism around accessing antiretroviral treatment and is looked to for evidence of the synergies between treatment and prevention interventions. This survey demonstrates that many of the best practice prevention interventions do in fact have a high uptake in Khayelitsha. There are relatively high levels of VCT uptake, condoms use is consistent with or higher than in other surveys (Pettifor et al., Citation2005b; Shisana & Simbayi, Citation2002) and knowledge about HIV appears to be good.
At the same time however, HIV prevalence has continued to rise unabated, including in the youngest age groups (Department of Health, Citation2005). A particular conglomeration of risk factors that are common to the epidemic in the region continue in Khayelitsha, including early sexual debut, missed opportunities for condom usage, concurrent relationships, especially amongst men where they are frequently multiple, age differentials between male and female partners, and the inability of women to negotiate condom usage when they would otherwise choose to use condoms. These estimates are not dissimilar to those described in other South African surveys ( Parker et al., Citation2002; Pettifor et al., Citation2005a, Citationb; Shisana &Simbayi, 2002; South African Department of Health, Citation2001).
Given the importance of consistency in condom use (Ahmed et al., Citation2001; Lagarde et al., Citation2001b), it is notable that nearly 20% of men and 30% of women would like to have used a condom the last time they had sex but did not, with lack of availability remaining an important contributing cause
The age differential between male and female partners is particularly marked in Khayelisha, in keeping with a number of studies in Southern Africa in which older male partners are postulated to place young women at greater risk of HIV acquisition (Glynn et al., Citation2001; Gregson et al., Citation2002; Laga, Schwartlander, Pisani, Sow, & Carael, Citation2001).
Given recent evidence on the potential role of circumcision in contributing to HIV prevention (Auvert, Taljaard, Lagarde, Sobngwi-Tambekou, Sitta, & Puren, Citation2005), the potential impact on the epidemic of an intervention around circumcision in this community would need to be considered in the light of the extremely high proportion of men who are for traditional reasons already circumcised in early adulthood, as well as the current age at circumcision.
In understanding the forces shaping risk behaviour, the importance of distal contextual factors in addition to personal factors is increasingly stressed (Eaton, Flisher, & Aaro, Citation2003; Mathews, Citation2005). In this study, an important negative finding was a lack of association with having undergone VCT and condom usage in the last sexual encounter, whereas trial evidence suggests an association between prior VCT and risk behaviour (Allen et al., Citation2003; Hogan & Salomon, Citation2005). In contrast, responses suggesting openness around HIV, knowledge of antiretroviral interventions and exposure to individuals on antiretroviral treatment were variously associated with condom usage and testing. The implausibly low proportions of patients who admit in the survey to being HIV-infected in spite of apparent openness, reflects however ongoing stigma associated with the diagnosis.
There are two broad conclusions from this study – firstly modest increases in coverage of current best-practice prevention interventions are limited in the extent to which they can impact on a mature generalised epidemic as is found in Khayelitsha. In the same way as a constant expansion of the clinical service platform is required in order to meet treatment needs and approach universal access, so too is a constant, aggressive and sustained expansion in the promotion of prevention interventions required in order to attain even higher levels of coverage.
Secondly, the relationship between treatment and prevention services is impossible to unravel in the constantly changing context of a maturing HIV epidemic. Due to the global commitment to treatment access, it is fortunately no longer necessary to justify treatment interventions on the putative basis of their impact on prevention. Furthermore, discussions on the synergies between treatment and prevention often focus on personal and proximal issues. The high reliance on broadcast media as a source of information in this survey point to the importance of the distal context, where South Africa has lagged behind many other countries in the region in terms of the visibility, political commitment and priority afforded to both treatment and preventive interventions for HIV in the national discourse. It is at this level perhaps that the greatest synergies might exist between treatment and preventive paradigms.
Acknowledgements
This study was conducted with the support of the AIDS IMPULSE Programme II (“Scaling up health system responses to AIDS” University of Cape Town [UCT], South Africa), funded by the Directorate General for Development Cooperation of the Belgian Government. The authors thank Prof. Anne Buvé for her appreciated input to the design of the study.
References
- Abdullah , M.F. , Young , T. , Bitalo , L. , Coetzee , N. and Myers , J.E. 2001 . Public health lessons from a pilot programme to reduce mother-to-child transmission of HIV-1 in Khayelitsha . South African Medical Journal , 91 : 579 – 583 .
- Ahmed , S. , Lutalo , T. , Wawer , M. , Serwadda , D. , Sewankambo , N.K. Nalugoda , F. 2001 . HIV incidence and sexually transmitted disease prevalence associated with condom use: A population study in Rakai, Uganda . AIDS , 15 : 2171 – 2179 .
- Allen , S. , Meinzen-Derr , L. , Kautzman , M. , Zulu , I. , Trask , S. Fideli , U. 2003 . Sexual behavior of HIV discordant couples after HIV counseling and testing . AIDS , 17 : 733 – 740 .
- Auvert , B. , Taljaard , D. , Lagarde , E. , Sobngwi-Tambekou , J. , Sitta , R. and Puren , A. 2005 . Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: The ANRS 1265 Trial . PLoS Medicine , 2 : e298
- Berkman , A. 2001 . Confronting global AIDS: Prevention and treatment . American Journal of Public Health , 91 : 1348 – 1349 .
- Coetzee , D. , Hildebrand , K. , Boulle , A. , Maartens , G. , Louis , F. Labatala , V. 2004 . Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa . AIDS , 18 : 887 – 895 .
- De Cock , K. and Grubb , I. 2006 . Towards universal access: WHO's role in HIV prevention, treatment and care . Bulletin of the World Health Organisation , 84 : 506
- Western Cape Department of Health 2005 . Antenatal HIV and syphilis prevalence survey . Cape Town : Department of Health .
- Eaton , L. , Flisher , A.J. and Aaro , L.E. 2003 . Unsafe sexual behaviour in South African youth . Social Science and Medicine , 56 : 149 – 165 .
- Glynn , J.R. , Carael , M. , Auvert , B. , Kahindo , M. , Chege , J. Musonda , R. 2001 . Why do young women have a much higher prevalence of HIV than young men? A study in Kisumu, Kenya and Ndola, Zambia . AIDS , 15 ( Suppl. 4 ) : S51 – S60 .
- Gregson , S. , Nyamukapa , C.A. , Garnett , G.P. , Mason , P.R. , Zhuwau , T. Carael , M. 2002 . Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe . Lancet , 359 : 1896 – 1903 .
- Hardin , J. and Hilbe , J. 2003 . Generalized estimating equations , Florida : Chapman & Hall .
- Hogan , D.R. and Salomon , J.A. 2005 . Prevention and treatment of human immunodeficiency virus/acquired immunodeficiency syndrome in resource-limited settings . Bulletin of the World Health Organisation , 83 : 135 – 143 .
- Johnson , L. and Budlender , D. 2002 . HIV risk factors: A review of the demographic, socio-economic, biomedical and behavioural determinants of hiv prevalence in South Africa , Cape Town : Centre for Actuarial Research .
- Kelly , K. and Parker , W. 2000 . Communities of practice: Contextual mediators of youth response to to HIV/AIDS. Sentinel site monintoring and evaluation project , Grahamstown : Centre for Aids Development, Research and Evaluation .
- Laga , M. , Schwartlander , B. , Pisani , E. , Sow , P.S. and Carael , M. 2001 . To stem HIV in Africa, prevent transmission to young women . AIDS , 15 : 931 – 934 .
- Lagarde , E. , Auvert , B. , Carael , M. , Laourou , M. , Ferry , B. Akam , E. 2001a . Concurrent sexual partnerships and HIV prevalence in five urban communities of sub-Saharan Africa . AIDS , 15 : 877 – 884 .
- Lagarde , E. , Auvert , B. , Chege , J. , Sukwa , T. , Glynn , J.R. Weiss , H.A. 2001b . Condom use and its association with HIV/sexually transmitted diseases in four urban communities of sub-Saharan Africa . AIDS , 15 ( Suppl. 4 ) : S71 – S78 .
- Marseille , E. , Hofmann , P.B. and Kahn , J.G. 2002 . HIV prevention before HAART in sub-Saharan Africa . Lancet , 359 : 1851 – 1856 .
- Mathews , C. 2005 . “ Reducing sexual risk behaviours: Theory and research, successes and challenges ” . In HIV/AIDS in South Africa , Edited by: Karim , S.S.Abdool and Abdool Karim , Q. Cape Town : Cambridge University Press .
- Médecins sans Frontières & University of Cape Town 2003 . Antiretroviral therapy in primary health care: Experience of the Khayelitsha programme in South Africa . Geneva : World Health Organisation .
- Municipal Demarcation Board 2007 . South African Census 2001, as provided by voting ward by the Municipal Demarcation Board . Available from: http://www.demarcation.org.za . Accessed 21 June 2007 .
- Parker , W. , Oyosi , S. , Kelly , K. and Fox , S. 2002 . On the move: The response of public transport commuters to HIV/AIDS in South Africa , Pretoria : Department of Health .
- Pettifor , A.E. , Kleinschmidt , I. , Levin , J. , Rees , H.V. , MacPhail , C. Madikizela-Hlongwa , L. 2005a . A community-based study to examine the effect of a youth HIV prevention intervention on young people aged 15–24 in South Africa: Results of the baseline survey . Tropical Medicine and International Health , 10 : 971 – 980 .
- Pettifor , A.E. , Rees , H.V. , Kleinschmidt , I. , Steffenson , A.E. , MacPhail , C. Hlongwa-Madikizela , L. 2005b . Young people's sexual health in South Africa: HIV prevalence and sexual behaviors from a nationally representative household survey . AIDS , 19 : 1525 – 1534 .
- Salomon , J.A. , Hogan , D.R. , Stover , J. , Stanecki , K.A. , Walker , N. Ghys , P.D. 2005 . Integrating HIV prevention and treatment: From slogans to impact . PLOS Medicine , 2 : e16
- Shisana , O. and Simbayi , L. 2002 . Nelson Mandela/HSRC Study of HIV/AIDS 2002 , Cape Town : Human Sciences Research Council .
- South African Department of Health . 2001 . South African Demographic and Health Survey 1998 , Pretoria : South African Department of Health .