Presentation by Marlise Richter, at the National Sex Work Symposium, in the session 'What we know: evidence-based peer reviewed knowledge on sex work' (Boksburg, 22 August 2012)

1. Characteristics, sexual behaviour and risk
factors of female, male and transgender sex
workers in South Africa
National Sex Work Symposium
22 August 2012
Johannesburg
Marlise Richter, Matthew Chersich, Marleen Temmerman, Stanley
Luchters
International Centre for Reproductive Health, Ghent University
African Centre for Migration & Society,
Wits University
Marlise.richter@gmail.com

2. Background
• There is no estimation of sex worker numbers in South
Africa
– little is known about the characteristics and health needs of sex
workers in the country
• While some studies have focused on female sex workers in
urban centres in South Africa, Johannesburg, Pretoria, Cape
Town and Durban and to a lesser extent along transport
routes in KwaZulu-Natal and a gold mining area in the North
West Province, these studies are mostly a decade old.
• Male and transgender sex workers - very little information
available on these populations in Africa.

3. Background
• In 1998, HIV prevalence amongst different female sex worker
groups in South Africa ranged between 46% and 69%.
• In a 2004-2005 Durban study, 775 women at high risk for
HIV infection – 78.8% of whom self-identified as sex
workers – were screened, and 59.6% found to be HIV-
positive.
• More recent estimates of HIV burden in sex workers in
South Africa are not available.
• A recent meta-analysis emphasised the considerable risk that
HIV poses to sex workers
– They have about a 13-fold higher risk of acquiring HIV infection
compared to other women of reproductive age in low- and middle-
income countries.

4. Background
• Female condoms are one of few female-controlled HIV
prevention technologies available, with some female sex
workers even using them without clients’ knowledge.
• Acceptability of female condoms has been demonstrated in
South Africa, as has female condom reuse.
• Sex work activists have advocated for the greater availability
of female condoms in sex work settings, with little success.
• In 2010-2011, the National Department of Health
distributed:
– around 5 million female condoms (target: 6 million)
– half a billion male condoms distributed (target: 1 billion)

5. Methods
• Self-identified female, male and transgender sex workers in
Hillbrow, Sandton, Rustenburg and Cape Town were
interviewed by trained sex worker research assistants during
May–September 2010.
• University-based researchers collaborated with the Sex
Worker Education and Advocacy Taskforce (SWEAT) and
Sisonke Sex Worker Movement.
• Women, men and transgender sex workers (defined as
‘having exchange of sexual services for financial reward’) 18
years and above were eligible for participation.
• Questionnaires were translated from English into isiZulu,
isiXhosa, Afrikaans and Setswana.
• The study was approved by the University of the
Witwatersrand Human Research Ethics Committee
(Protocol number H100304).

6. Results

7. Results (socio-demographic)
• Participants were a mean 30 years old
• Just over half of female (53.7%; 878/1636) and male (55.3%;
48/87) participants, and just more than a third of transgender
people (37.9%; 22/58), were born in South Africa.
• A third of females (555/1626), a quarter of males (21/87) and
15.8% (9/57) of transgender participants noted that they had
a ‘husband/permanent partner/boyfriend or girlfriend’
(P=0.003).
• Females were responsible for a median of 4 adult and/or
child dependents – double that of male or transgender
participants (P<0.001).
• Age of sex work debut was similar across the genders, an
average of about 24 years: females 24.2 (SD=5.3), males 23.6
(SD=4.5) and transgender 24.3 years (SD=5.0).
• More than 40% of all participants had been in sex work for
more than five years

8. Results (Sexual behaviour, condom & alcohol-use)
• Median number of clients in the week preceding study
enrolment:
– 12 for females
– 10 for males
– 8 for transgender
• More women had penetrative sex with last client (92.1%;
1 522/1653) than males (81.6%; 71/87; P<0.001) or
transgender people (81.4%; 48/59; P<0.001)
• Women were less likely to have unprotected sex:
• only 5.5% (82/1 498) of women had unprotected sex with last client
in contrast to
• 27.5% (19/69; P=0.01) of men, and
• 20.0% (9/45; P<0.001) of transgender people.
• In a multivariate analysis of factors associated with
unprotected anal/vaginal sex with last clients, males were 2.9
times (AOR, 95%CI=1.6-5.3; P<0.001) more likely, and
transgender people 2.4 times (AOR, 95%CI 1.1-4.9; P=0.021)
more likely than females to have unprotected sex.

9. Results (Sexual behaviour, condom & alcohol-use)
• In univariate analysis, having fewer dependants was associated
with unprotected sex.
• Sex workers in Cape Town were 5.5 times (AOR, 95%CI
3.0-10.0; P<0.001), those in Rustenburg 2.9 times (AOR, 1.6-
5.3; P<0.001) and those in Sandton 2.7 times (AOR 95%CI
1.4-5.1; P=0.04) more likely to engage in unprotected sex
than their counterparts in Hillbrow.
• About a fifth of females (284/1566), a third (16/54) of
transgender people and over 40% (34/82) of males reported
daily binge drinking.
• More than 40.0% of females (651/1603) were drunk during
sex with last client in comparison to 59.7% of males (49/82)
and 66.1% (37/56) of transgender people.
• Participants who reported daily or weekly binge drinking
were 2.1 fold (95%CI 1.2-3.7; P=0.011) more like than those
who never engaged in binge drinking, to have unprotected
sex.

10. Results (Female condoms)
• Just less than half of female participants had ever used a
female condom (446/1 006).
• Of these,
– close to a third (116/413) “liked” female condoms, and
– almost half (189/413) “liked them a lot” (data not shown). Only 7.5%
(31/413) disliked female condoms”, with
– 77/413 (18.6%) being neutral.
• Among those female participants who did not use female
condoms and provided reasons for non-use,
– a fifth each noted that they had never been given female condoms
(99/560),
– did not know how to use them (111/560) or
– did not like them (129/560).
• A tenth each noted either they are unfamiliar with female
condoms (66/560) or that clients preclude use (47/560).
•

11. Discussion
• Sex work was the major livelihood strategy adopted by the
study populations:
– more than 40% had been in the industry for more than five years,
approximately two thirds were full-time sex workers, while over a
third had no other work experience prior to entering sex work.
• When comparing full-time sex workers’ income with data
from Statistics South Africa on monthly earnings by occupation,
sex workers in this study were earning more than clerks,
sales and services, crafts and related trades, and up to six
times more than domestic workers
• This is pertinent for some ideology-based health and social
interventions aiming to ‘rehabilitate’ sex workers or focus
solely on ‘exit programmes’

12. Discussion
• Less than half (44.3%) of female participants had ever used a
female condom.
– Of these, three quarters were in favour of female condoms.
• Studies in China and Cambodia have shown that female
condom promotion with female sex workers have increased
its acceptability and use, while a study with female sex
workers in rural Mpumalanga showed female condoms to be
highly cost-effective.
• As a female-controlled HIV prevention strategy, this should
be a vital component of sex work interventions.

13. Discussion
• It is of concern that males were 2.9 times more likely, and
transgender people 2.4 times more likely than female sex
workers to engage in unprotected sex.
– This could be a reflection of the dearth of programmes focusing on
males and transgender people within the sex industry in South
Africa or the general lack of information on anal sex, and is an area
for action.
• Of all participants, 27.0% had unprotected sex when engaged
in anal intercourse with last client – the most risky sex act
for acquiring HIV and other STIs.
• Public health interventions with female, male and transgender
sex workers and their clients should emphasise the risks
associated with anal sex and ensure that condoms and
lubrication are accessible and feely available within the sex
industry.

14. Discussion
• Sex workers in the Sandton, Rustenburg and Cape Town
sites were significantly more likely to engage in unprotected
sex than those situated within Hillbrow.
• Hillbrow had the only sex work-specific clinic and mobile
outreach clinical services for sex workers in South Africa at
the time of the study.
– A cadre of sex work peer educators disseminate information and
condoms within hotels and clubs from where sex workers operate,
while a male community health worker provides HIV/STI education
and referrals to clients within bars and nightclubs.
• This model should be duplicated in other areas of sex work
concentration in South Africa.

15. Limitations
• The study included self-reported data only and was based on
a non-random sampling design.
– Surveys were, however, conducted by trained peer interviewers,
which may have reduced the social-desirability bias in respondents’
answers.
• Some data were missing on questionnaires
• Almost all peer interviewers were female, which may have
impacted on the number of male and transgender
participants approached for participation.
• Selected research sites included two urban centres and one
semi-rural site adjacent to a mine and were purposively
selected, based on the presence of sex worker advocacy
groups and peer education work.
• Although we selected three cities aiming to obtain data on
diverse sex work settings, these findings may not apply to
other sex work areas in South Africa.

16. Conclusions
• In conclusion, sex workers remain at high risk of HIV and
other STIs in South Africa.
• This risk has been acknowledged by South African HIV/AIDS
policies and sex work-specific programmes proposed since
the first National AIDS Plan in 1994, but yet little action has
been taken.
• The ‘National Strategic Plan for HIV and AIDS, STIs and TB,
2012-2016’ contains a number of sex work-specific health
and non-discrimination provisions, and should be
implemented as a matter of urgency.

17. Acknowledgements
• Funding for this study was provided by UNFPA and Atlantic
Philanthropies.
• We would like to thank the Sex Worker Education and Advocacy
Taskforce (SWEAT) and the Sisonke Sex Worker Movements for
guidance and logistical support, and the research assistants for hard
work during data collection.
• The technical and logistical support of the African Centre for
Migration & Society and the Centre for Health Policy, Wits
University and their students was key in the conceptualisation and
development of the project, as well as the assistance of the Sex
Work Project, Wits Reproductive Health and HIV Institute within
Hillbrow.
• Special thanks for the input and support of Dudu Ndlovu, Jo Vearey,
Dianne Massawe, Carolin Kueppers, Tom Considine, Fiona Scorgie,
Elsa Oliveira, Agnieszka Flak, Marc Lewis, Ingrid Palmary, Richard
Steen, Gerrit Maritz, Francois Venter and Ziad El-Khatib.