Genitourin Med 1997;73:161-168 161

Public Sex workers and the control of sexually health series transmitted disease

no 6 Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from Series editor Sophie Day, Helen Ward Graham Nielsen

Objectives: To describe and assess measures to control sexually transmitted diseases (STDs) among sex workers and their partners. Methods: A review of medical, historical and social literature, focusing on selected cases. Results: Measures to control disease in sex workers today are often prompted by concerns about HIV transmission. However, the literature shows that varies from one place and time to another, together with the risk of sexually transmitted disease. A broad social definition of prostitution rather than a narrow reference to levels of sexual activity is important for effective dis- ease control, as an understanding of the relation between social disadvantage and sexual activity enables the provision of occupational services that sex workers actually want and use. Social prejudice and legal sanctions cause some sex workers and their partners to avoid even the most appropriate and accessible specialist services. Therefore targeted programmes can only comple- ment, and not replace, general measures to control STDs, which are developed for other social groups or the local population as a whole. Conclusions: Sex workers and differ from one place to another and so a single model for STD control is inappropriate. None the less, occupational health risks suggest a general need for specialist services. Where these do not compound the disadvantages that sex workers already suffer, medical services are likely to offer significant benefits in prevention, early diagnosis, and treatment of STDs. As the stigma of prostitution leads many people to remain invisible to services, a general health infrastructure and anti-discriminatory measures will be equally important to effective disease control. (Genitourin Med 1997;73:161-168)

Keywords: sex workers; sexually transmitted disease; law

Introduction core groups are important to the design of dis- In medicine, prostitutes or sex workers are ease control strategies today. http://sti.bmj.com/ generally defined in terms of their high levels However, views of prostitutes are also of sexual activity: sexual services are informed by conventional prejudice and state exchanged for money or items of monetary policies. Historians have suggested that stereo- value, such as drugs. While reliable figures are types guided 19th century state policies and hard to obtain, it is clear that the majority of public health programmes which, in fact, cre-

sex workers are poor, urban women. In ated modern prostitution.3 In Britain, for on September 29, 2021 by guest. Protected copyright. Europe, it has been reported that nine in 10 example, the of the sex workers in Belgium and 39 of every 40 in last century introduced screening for prosti- the Netherlands are women.' tutes in garrison towns and detention for those Sex workers are often the focus of sexually who were thought to be infected, in order to Department of Epidemiology and transmitted disease (STD) control measures reduce venereal diseases in their customers, Public Health, because they are traditionally seen as a source notably, members of the armed forces. Imperial College of disease. A relatively high rate of partner Historical records show that prostitutes subse- School ofMedicine at St Mary's, London change suggests that sex workers will be more quently began to work full time for a longer W2 1PG and vulnerable to infection and, if infected, they period and away from their local communi- Department of may be more likely to transmit disease than ties.34 As they were singled out for specific Anthropology, Goldsmith's College, individuals with fewer partners. In some parts forms of social control, so they became a more University of London of the world, sex workers have very high levels distinct section of the population. They saw S Day of STD. Epidemiological research has focused more clients and were thus placed potentially Departnent of on transmission from prostitutes to their part- at a higher risk of infection than when prosti- Epidemiology and ners and a notional general population. tution was a more widespread and transient Public Health, Imperial College Models with respect to the population as a activity for poor women. These policies, there- School ofMedicine at whole have been formulated for gonococcal fore, must have contributed to the situation St Mary's, London transmission in particular.2 It is suggested that that they were intended to resolve; increasing W2 1PG H Ward a small number of people having high num- sex workers' vulnerability to sexually transmit- Correspondence to: bers of sexual partners could be seen as a core ted disease and their ability to transmit infec- Dr Sophie Day, Department group, who will play a disproportionate role in tion. of Epidemiology and Public Health, Imperial College transmission. The degree to which those with This example suggests that social exclusion School of Medicine at St a high rate of partner change, such as prosti- determines risks of infection as much as multi- Mary's, Norfolk Place, London W2 1PG, UK. tutes, mix with those with a relatively low rate partner sexual activity. Lack of civil and work- Accepted for publication of partner change will have an important ers' rights deny sex workers minimum 3 March 1997 impact on the pattern of spread. Concepts of standards of occupational health and safety, 162 Day, Ward

including the ability to enforce risk reduction criminal procedures-for example, condoms measures. Poverty too is a defining feature of are no longer used as evidence of prostitution in prostitution. Growing inequalities of wealth some countries.' The general criminalisation within and between countries, rapid urbanisa- of prostitution together with widespread preju- tion, and economic have dice create acute difficulties for restructuring shaped disease pre-Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from the contemporary sex industry. The tradition- vention and treatment as many sex workers ally disadvantaged have been disproportion- will remain invisible and inaccessible to these ately recruited into prostitution. Other poor programmes. people, who do not work as prostitutes, also These differing political approaches to pros- rely increasingly upon financial help from a titution combined with data on the prevalence number of sexual partners.56 In the context of of infections among sex workers influence con- pervasive sex inequalities, prostitution cannot trol strategies. A brief review shows that, always be distinguished easily from marriage firstly, levels ofinfection associated with prosti- and/or other female careers; sex workers are tution vary and, secondly, they are difficult to also disadvantaged on the grounds of sexual interpret. identity, social class, age, and race.7

Sexually transmitted disease control Background: Sexually transmitted The basic reproductive rate of a sexually trans- disease in sex workers mitted disease-that is, the number of infec- The reported risk of STDs among people tions generated on average by an infected working in the sex industry varies in relation to person, depends on a number of factors: the geographical location and prostitution sector.'0 rate of sexual partner change, the probability This variation also results from the problems of transmission per partnership, and the dura- of interpreting results from biased samples. tion of infectiousness. Estimates ofincidence and prevalence are diffi- Currently, control programmes are based cult to establish since criminal and other sanc- on two central strategies, elaborated particu- tions make it impossible to enumerate the larly to reduce levels of HIV infection. The relevant population. Most studies report fig- first consists of a range of risk reduction mea- ures from captive groups, either the more sures involving behaviour change. These accessible sex workers-for example, those include a reduction in the number, and change who work visibly in poor neighbourhoods, or in type, of sexual partners, and the use ofbarri- those in contact with state institutions such as ers to reduce the probability of transmission. police, prisons, and health services. These The second consists of improved diagnosis problems make it difficult to generalise or and treatment for classic STDs, which facili- compare results. None the less, it is clear that tate HIV transmission. Epidemiological stud- STD rates in sex workers are often higher than

ies indicate that the risk of acquiring HIV in in other groups. Three general examples arehttp://sti.bmj.com/ the presence of ulcerative or non-ulcerative illustrative. sexually transmitted diseases is increased by a factor of 2 to 6.89 HIV infection also has an SUB-SAHARAN AFRICA impact on the clinical spectrum and manage- In urban populations of sub-Saharan Africa, ment of other infections. the classic STDs-syphilis, gonorrhoea, and Political approaches to prostitution have an chancroid-account for 7% approximately of on September 29, 2021 by guest. Protected copyright. impact on, and are frequently inseparable overall disease burden, ranking them below from, STD control programmes. In the past, measles, malaria, and gastroenteritis but above female prostitutes were commonly subject to severe malnutrition and tuberculosis." These compulsory registration, for the purposes of diseases are associated with microbial resis- screening and treatment. A number of states tance, infertility, and maternal mortality sub- continue to make compulsory health checks a sequent to ectopic pregnancies caused by work requirement, and many sex workers find pelvic inflammatory disease. that they are working illegally without access Sex workers are at considerable risk in this to health care. Over the last century, this form situation. More than 75% of female prostitutes of regulation has been replaced largely by abo- in Kenya had STDs'2; and in Kinshasa, Zaire, litionism; prostitution is criminalised in a vari- 75% of 1233 female prostitutes presented with ety of ways, with the aim of eradicating what is at least one sexually transmitted disease.'3 variously regarded as a form of slavery or During the past decade, research into HIV immorality. This leads to a formal separation infection has been a priority. In the mid- between the criminal justice system, which 1980s, high levels of HIV-1 infection were punishes sex workers, and health and social found among sex workers in some sub- services, which seek to minimise harm and Saharan African cities.'3-1 A high incidence of provide appropriate care. However, many HIV infection was demonstrated among sex assume that health services are part of the workers in Nairobi by testing sera that had criminal justice system and they fear that they been stored in the course of other studies: in will be reported to the police, as in Sweden, 1981, the prevalence of HIV-1 was 4%; by for example, if HIV infected sex workers prac- 1985, it had risen to 61%.16 Similar dramatic tice unsafe sex.' Sometimes, a formal separa- rises in HIV prevalence have been reported for tion between state agencies works to the other groups of urban prostitutes in Africa, advantage of the sex worker. Pressures for where the virus is transmitted primarily harm minimisation have led to less stringent through heterosexual contact.'7-19 Sex workers and the control ofsexually transmitted disease 163

ASIA presented with an acute STD and 44% A marked and rapid increase in the prevalence reported past infection with gonorrhoea.33 of HIV infection among sex workers in parts of Asia such as India20 and Thailand2' occurred later. Extensive HIV-1 transmission began in The politics of prostitution, the preva- Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from Thailand in 1988 and, following an initial epi- lence ofinfection, and STD control demic among Bangkok injecting drug users, It can be seen that levels of infection among came to be associated primarily with hetero- sex workers vary greatly, relating crudely to sexual transmission. The relation between the differences of wealth within countries, and two epidemics is unclear as the virus subtypes between north and south. They relate to local differed among injecting drug users and pros- population prevalences and dominant modes titutes.22 Long term state policies to increase of transmission, and to local characteristics of foreign earnings through tourism (including a the sex industry. Sex workers are themselves sex industry)23 24 may help account for the internally stratified by STD-for example, rapid increase in prevalence among sex work- according to socioeconomic status,39 injecting ers and their contacts, especially in the north. drug use, the use of non-injected drugs, and As in some African cities, sex workers have levels of condom use. Less is known about the been seen as a core group, to whom interven- risks posed by partners, although the few stud- tions are targeted. ies published suggest that distinct client Sentinel surveillance shows a recent behaviours may be relevant.404' In a British decrease in the rates of classic STDs in survey of sexual lifestyles, 6-8% of men Thailand. The prevalence of HIV may be reported having ever paid a woman for sex. decreasing among some Thais, such as mili- Such men were more likely to be unmarried, tary conscripts,25 but continues to increase to work away from home, and to have had sex- among sex workers: rising from 1A4% to ual contact with a man at some time in their 20 1 % in a 4 year period among male sex lives.42 In a London study of clients, over a workers in the north of Thailand26 and contin- third of men reported past sex with men as uing to rise steeply among female sex workers well as women.39 in the same part of the country.27 A prevalence Currently, explosive epidemics of HIV of 65% has been reported for some commer- among sex workers in some countries, together cial sex work groups in Chiang Mai.28 with epidemics of poverty and drug use in the West, have led to a renewed interest in con- EUROPE, AUSTRALIA, AND NORTH AMERICA cepts of core groups, reservoirs, and pools of In much of the developed world, HIV trans- infection,4344 which suggest that targeted pro- mission continues to be primarly associated grammes for sex workers will be crucial. These with injecting drug use and sex between men, differ from programmes in the pre-AIDS era, although heterosexual transmission has when sex workers in many countries were seen increased. Among sex workers who do not to be less central to the transmission of infec- http://sti.bmj.com/ inject drugs, other factors have been associ- tion and general programmes addressed "pro- ated with HIV infection, including a past his- fessionals" and "amateurs" alike (the term tory of STDs, and the use of crack amateur is used to describe apparently promis- (non-injected) in parts of the USA. Crack use cuous women who were not paid for sex). has been associated with an increased inci- Political approaches to prostitution together

dence of syphilis, gonorrhoea, and other sexu- with concerns about HIV transmission and/or on September 29, 2021 by guest. Protected copyright. ally transmitted infections in North American high levels of other sexually transmitted infec- cities, especially among the young.29 tions have led to three strategies for control. Low rates of HIV continue to be reported The first works through laws that mandate among many groups of non-injecting drug screening, within the regulated sector. The users. A European study from 11 centres second attempts to promote access to sex reported an HIV prevalence of 1.5% in female workers through providing services that are sex workers who did not inject drugs com- wanted. The third builds upon prostitutes' pared with 318% of injecting drug users.30 position in the sex industry to reach a wider Low rates of HIV and other STDs have been population including clients, managers, and reported among sex workers working in others outside the sex industry, who may be at Australian , where condoms are widely risk. These three approaches are not necessarily used." counterposed, and programmes frequently Rates of STDs among male, transvestite combine elements of regulation with improved and transsexual sex workers tend to be higher accessibility and the mobilisation of sex workers than among female sex workers.'2 In London; themselves. Regulation tends to highlight the 27% of 57 male sex workers compared with disproportionate role that sex workers may less than 1% of 228 female sex workers were play in the transmission of disease. Improved found to be infected with HIV."'34 This may accessibility tends to emphasise also the vul- reflect the more stigmatised and hidden nature nerability of sex workers to infection. The of male sex workers as well as the prevalence mobilisation of workers recognises equally that of infection in clients. Condoms may be used prostitutes are well placed to act as health less. 35-37 workers. The morbidity from STDs for many sex workers can remain high in these lower preva- PROGRAMMES FOR REGISTERED PROSTITUTES lence areas over the course of a working Mandatory screening and treatment provide life.33 38 In London, 14% of female sex workers access to the regulated sector of the sex indus- 164 Day, Ward

try in countries such as Greece, parts of work in a legal or formal economic sector. In Germany, and the USA. Compulsory HIV these circumstances, targeted programmes screening and the detention of those found to cannot reach the majority of workers. be infected have been introduced in many Moreover, sex workers frequently suffer high states, including parts of India, the USA, and morbidity before they come into contact with Sweden. There is no evidence that such strate- specialist clinical services." These problems Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from gies have prevented transmission; the regu- have led to a range of innovative interventions lated sector tends to be a minor part of the over the past decade akin to "pyramid selling", industry precisely because prostitutes avoid where prostitutes and, sometimes, others in the control associated with registration. the sex industry work as peers and educators. Indeed, it has been argued that repression exacerbates the problem since sex workers are EMPLOYING SEX WORKERS IN HEALTH further marginalised from health services in PROMOTION AND CHANGING THE SEX the attempt to evade state restrictions on their INDUSTRY work. Interventions have addressed conditions in the There has been much political debate about sex industry. Programmes have been devel- this approach, which is regarded by many on oped for managers, customers, and state offi- the right as tolerating, and thereby legitimis- cials who may be able to introduce health and ing, prostitution. In contrast, sex workers and safety measures or to encourage the use of human rights activists recognise this regulation condoms, as they occupy a stronger position in as an unacceptable form of social control the industry.47 In Sydney, the Australian based on unsubstantiated arguments about the Prostitutes' Collective developed a "safe house promotion of "public health". In addition, scheme" where symbols were awarded to par- effective prevention, which is particularly lours in which all clients were required to use important for viral diseases, is thought to condoms and in which working conditions require a relationship of trust between health- were adequate. The Collective also encour- care workers and prostitutes. This is not com- aged the distribution of introductory letters to patible with legal repression. A general clients of escort services which explained that discomfort with measures intended to control condoms were required for all sexual contact. disease through controlling categories of peo- Prior notification by the management means ple, together with suggestions that oppressive that responsibility for negotiating safer sex legal and public health codes are ineffectual, does not lie solely with each individual prosti- has promoted the development of services that tute.48 In Amsterdam, a group of clients was sex workers want. set up to discuss safer sex and to hand out condoms to other clients in the red light dis- PROVIDING ACCESSIBLE AND APPROPRIATE trict.49 Initiatives have lobbied for decriminali- SERVICES FOR SEX WORKERS sation so as to decrease the stigma affecting

Many services combine health education with individuals who choose to sell (or buy) sexual http://sti.bmj.com/ the provision of materials such as free con- services and to establish basic rights at work, doms and, in some countries, equipment for including occupational health and safety.5051 safer injecting as well as clinical and other ser- As noted above, law enforcement agencies vices. These have been taken into communi- have been lobbied successfully to stop using ties so as to establish contacts with the "hard condoms in court as evidence of prostitution. to reach" or "out of reach". Thus, the special- Other programmes have focused upon eco- ist clinics for sex workers which have been nomic rather than legal inequalities and intro- on September 29, 2021 by guest. Protected copyright. established145 46 often employ outreach models duced alternative ways of making a living both so that services are taken to the areas in which to sex workers and to other individuals who prostitutes work, and open at suitable hours. are financially reliant upon several partners. Attempts to maximise service coverage and Elsewhere, collectives have managed to raise facilitate risk reduction necessarily involve the fees for services, thus reducing the numbers of provision of services that sex workers want, clients a prostitute needs to see.52 either integrated with wider sexually transmit- With the partial emancipation of some sex ted disease services, or provided specifically workers as well as a growing interest in com- for them. In addition, attempts to encourage munity programmes, some sex workers have attendance imply non-judgmental attitudes been recognised as professional educators and towards sex workers. hired in STD prevention for a range of "at Reports suggest that disease is rarely the risk" categories. first priority of sex workers and so services of this kind have adopted an increasingly holistic approach in which STD control is part of a Content of the programmes-prevention, broader health and economic programme. In diagnosis, and treatment consequence, these services may reach a Sexually transmitted disease control measures greater proportion of sex workers than those for prostitutes are standard in many respects, within a regulated sector. Even so, they are comprising behavioural risk reduction on the likely to reach only a small proportion of sex one hand and early clinical diagnosis and workers. treatment on the other. In recent years, alter- Many poor women and some men combine native protective devices for women such as prostitution with other trades and move in and microbicides and the "female condom" have out of the industry over a period of years. been distributed in an attempt to introduce Many are also geographically mobile and few technologies that can be controlled by women. Sex workers and the control ofsexualy transmined disease 165

The use of nonoxynol-9 is contested; evidence peer leaders. Social relationships are consid- that it increased the risk of HIV acquisition by ered a more effective basis for change than causing disruption of the genital epithelium ideas about individual rationality which struc- has been reported53 but it appears that this ture the more traditional programmes. Role may be dose related.54 A wide range of alterna- play, focus groups, collective bargaining and tive products are being explored, primarily in negotiation with managers, clients, and state Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from the context of vaginal intercourse, including officials provide situations in which inequali- those that are already licensed.55 Previous ties are addressed and attempts made to research on contraceptive use should inform change the organisation of prostitution, which the development of appropriate forns of deliv- is seen to increase if not wholly determine ery, in situations where a range of methods workers' risks. may already be used for protection and/or con- This approach explicitly rejects programmes traception.5657 that could reinforce the social exclusion of sex Risk reduction messages have distributed workers through blaming them for sexually information on the assumption that appropri- transmitted disease and holding them responsi- ate knowledge about STD prevention will lead ble for prevention. Interventions to core to safer behaviours. This approach appears to groups associated with high rates of infection apply relatively well to many sexual contacts in are supported only if these encompass all the the AIDS and pre-AIDS era because of a pro- relevant sexual partnerships and if they avoid fessional interest in occupational health and "victim blaming." In general, community pro- safety, including condom use.i59 None the grammes advocate the provision of specialist less, prostitutes are often relatively powerless occupational health services rather than pro- to insist on safer sex; they may be unable to grammes which target sex workers in order to negotiate terms and to insist upon condom control sexually transmitted diseases in the use; women may also be unwilling to prevent population as a whole. pregnancy. Few sex workers will be in a posi- tion to respond to risk reduction messages advocating fewer partners. Do the interventions work? These factors entail specific approaches in Targeted control programmes have led to suc- clinical and other services. Regular screening cessful risk reduction and decreased levels of of women working in prostitution may be infection. Many studies report increased con- important because much infection is asympto- dom use among sex workers. Particularly dra- matic and because infected partners will not matic changes have been related to broad always notify prostitutes of the need for governmental support or to marketing through screening. At the same time, partner notifica- the private sector. For example, the Thai gov- tion is likely to be of limited use once a prosti- ernment implemented a national condom pro- tute is found to have an infection, since he/she gramme in sex establishments in 1991, is unable to notify many partners.60 Risks of accompanied by police sanction. This cam- http://sti.bmj.com/ violence from potentially infected partners are paign is reported to have been associated with also a concern.6' Since the sex industry is char- an increase in use among sex workers from acterised by relations of inequality, sex work- 15 000 to more than 50 000 condoms a month ers may be particularly unwilling and unable in one Thai province.66 It has been suggested to inform known partners for fear of violence that declining rates of the classic sexually

or dismissal. transmitted disease are due to increased con- on September 29, 2021 by guest. Protected copyright. It has also been shown that the numbers of dom use as well as other changes in behaviour, sexual partners may be less relevant to the such as widespread antibiotic use, and better transmission of infection that the type of rela- STD management.25 14Even small increases in tionship.3362 In a London cohort study, gono- condom use can decrease the transmission of coccal infection was not associated with HIV and other STDs.67 overall numbers of partners but with the num- In Zaire, it was shown that STD treatment ber of non-paying partners during the past had an independent effect on HIV-1 incidence month.33 Reports also suggest that condoms after controlling for levels of condom use and are used less extensively with regular clients63 numbers of partners.9 A clinic based interven- and by those who do not identify as profes- tion of STD care and condom promotion for sional sex workers.6 Sex workers may be able female sex workers was also associated with a to insist upon condom use in relationships declining incidence of HIV-1.44 while finding it particularly difficult to use Outcome measures are hard to interpret, condoms in their private lives, not just because even with good baseline information. It is diffi- of problems comparable with other study cult to assess the contribution of an interven- groupsT4 but also because condom use is tion in isolation from other factors. Further equated with sex.65 complications include the extensive mobility These studies confirm that knowledge and of sex workers, difficulties in interpreting attitudes bear, at most, an indirect relation to reported practices and the lack of adequate behaviour change. In consequence, traditional comparison groups. Two studies illustrate the risk reduction programmes, where it is problems. Although there were marked thought that appropriate knowledge and atti- increases in reported condom use among pros- tudes will lead to safer practices, have been titutes attending the Praed Street Project in replaced to some extent by participatory London between 1986 and 1988, it was found health programmes in which sex workers play that women new to the project in 1988 also a central role as community organisers and reported high levels of condom use. In Ghana, 166 Day, Ward

a study undertaken 3 years after a community als avoid contact with state agencies and unre- intervention showed that levels of condom use alistic programmes are designed which place were higher than they had been at enrolment responsibility for disease prevention on those but similar to those among sex workers in a who have little ability to guarantee safer sex. control cohort, who had never participated. In measures are

Accordingly, antidiscriminatory Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from both cases, risk reduction is attributed tenta- as important to programmes of disease control tively to peer networks and the diffusion of as more narrowly defined health services. new occupational norms. Other programmes may also have had an effect, for example, in modifying the behaviour of clients.'68 Conclusion There has been some indirect evaluation of Disease control measures specifically devel- the benefits for the "non-core" or "general oped for, or targeted to, sex workers have been population" of targeted interventions to sex informed by reported levels of infection and workers. Moses et al suggest that condom pro- prostitution policy. Models of core groups are motion among sex workers has been cost particularly influential in the design of current effective in reducing STDs in women and strategies. Three distinct approaches have their clients.69 In 1981-2, an epidemic of been outlined. The first is regulation. It is chancroid was described in California where widely acknowledged that punitive legislative 86% of the 271 patients with culture positive reforms are impracticable and/or ineffective. It Haemophilus ducreyi reported recent sexual is unrealistic to consider that all sex workers contact with prostitutes. Only five of the 271 can be identified through legal measures and patients were known sex workers, but it was likely that the majority will be further excluded thought that undiagnosed infection in sex from health care. Historically, repression has workers was playing an important role in the joined prostitutes with other "outcastes" continuing epidemic. The use of a prophylactic through processes of social exclusion that have mass treatment programme, reaching 287 sex created a synergy between different STD risks workers in the local jail and more than 700 such as sexual and drug using behaviours. men reporting recent contact with sex work- The second approach attempts to improve ers, was followed by a steady decline in the access to appropriate services. The positive numbers of new cases.70 results of educational and clinical programmes Programmes that target sex workers as a for sex workers are clear. In some study core group, irrespective of the very different groups, condom use has increased dramati- dynamics of transmission in particular places, cally, STD rates have decreased and transmis- may be less effective than those focusing on sion dynamics have altered. Some services wider communities. One of the few ran- appear to provide hitherto lacking health care domised interventions to assess the effects of for poor and often young people. Others STD control on HIV infection reported a 42% appear to have been avoided because sex

decline in HIV incidence among entire com- workers are held responsible for disease and http://sti.bmj.com/ munities receiving improved STD care.7' disease prevention. While this dramatic decline was not associated The third approach considers sex workers a with statistically significant differences in central component to community control pro- prevalence for a range of sexually transmitted grammes. In general, this approach tackles infections, the reduction was attributed to a problems of social exclusion directly by shortened duration of symptomatic disease. attempting to enhance the ability of sex workers This may have decreased both infectiousness to safeguard their health and improve their on September 29, 2021 by guest. Protected copyright. among HIV positive individuals and the risk of position within the sex industry. Sex workers acquisition among their partners. are enabled to insist upon reasonable work More generally, decriminalisation of prosti- conditions; they also provide channels of com- tution and anti-discriminatory measures have munication and models for risk reduction been associated with low levels of infection within the industry more generally and for and almost universal condom use. In New other groups "at risk" such as poor women, South Wales, Australia, and in the drug users, and men who have sex with men. Netherlands, legal and social changes appear Efforts to control STDs may be compro- to have paved the way for more effective health mised by the apparent ease with which the sex interventions within the sex industry.48 72 worker is defined in terms of multipartner sex- The success of specific programmes ual activity. This definition ignores relations of depends in large measure upon broader health inequality, where sex workers may be unable and prostitution policy. Laws against the provi- to insist upon the risk reduction advocated by sion of sterile injecting equipment make distri- health workers. It ignores other defining fea- bution impossible. Condom use cannot be tures of prostitution, such as poverty and prej- widely promoted if laws forbid advertising. udice. Prostitution is common, hidden, and Structural adjustment programmes distribut- criminalised. In these circumstances, the ing wealth from the poor to the rich make it majority of sex workers cannot be contacted difficult to provide accessible or appropriate through specialist services, and those who do services.73 For example, a mandate from the attend may already have suffered serious mor- World Bank led to charges for STD clinic ser- bidity. This severely limits the possible success vices in Kenya and attendance fell between of targeted programmes to control STDs. 35-60%.74 Similarly, criminal sanctions and Since the boundaries between prostitution, social stigma hamper STD control pro- sexuality, and work are often blurred and con- grammes for sex workers since many individu- tinually redefined, such programmes will also Sex workers and the control ofsexually transmitted disease 167

fail to contact those who do not work or iden- Thailand. Lancet 1993;341:1171-4. 23 Phongpaichit P. From peasant girls to Bangkok masseuses. tify as sex workers but who are at significant Geneva: ILO, 1982. risk of infection. Targeted programmes are 24 Truong TD. Sex, money and morality: prostitution and tourism in SoutheastAsia. London: Zed Books, 1990. important in the short term for those with 25 Mastro T, Limpakarnjanarat K. Condom use in Thailand: higher prevalences of infection, including how much is it slowing the HIV/AIDS epidemic? (editor- ial) AIDS 1995;9:523-5. Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from groups of prostitutes. Specialist health care is 26 Kunawararak P, Beyrer C, Natpratan C, Feng W, an important occupational service for sex Celentano DD, de Boer M, et al. The epidemiology of HIV and syphilis among male commercial sex workers in workers, regardless of the relative prevalence northern Thailand. AIDS 1995;9:517-21. of infection. However, targeted control pro- 27 Siraprapasiri T, Thanprasertsuk S, Rodklay A, Srivanichakorn S, Sawanpanyalert P, Temtanarak J. Risk grammes and specialist health services can factors for HIV among prostitutes in Chiangmai, only complement, not replace, more broadly Thailand. AIDS 199 1;5:579-82. 28 Celantano DD, Akarasewi P, Sussman L, Suprasert S, based interventions to the sex industry as a Matanasarawoot A, Wright NH, et al. HIV-1 infection whole and a general health infrastructure. among lower class commercial sex workers in Chiang Mai, Thailand. AIDS 1994;8:533-7. 29 Edlin BR, Irwin KL, Faruque S, McCoy CB, Word C, Sophie Day is supported by the Welicome Trust. Serrano Y, et al. Intersecting epidemics- use and HIV infection among inner city young adults. N EnglJ Med 1994;331:1422-7. 1 Mak R. EUROPAP: European intervention projects-AIDS 30 European Working Group on HIV infection in Female prevention forprostitutes. Gent: Academia Press, 1996. Prostitutes. HIV infection in European female sex work- 2 Yorke JA, Hethcote HW, Nold A. Dynamics and control of ers: epidemiological link with use of petroleum-based the transmission of gonorrhea. Sex Transm Dis 1978;5: lubricants. AIDS 1993;7:393-400. 51-6. 31 Harcourt C, Philpot R. Female prostitutes, AIDS, drugs 3 Walkowitz J. Prostitution and Victorian society: women, class and in New South Wales. In: Plant M, ed. AIDS, and the state. Cambridge: Cambridge University Press, drugs andprostitution. London: Routiedge, 1990. 1980. 32 McKeganey NP. Prostitution and HIV: what do we know 4 Corbin A. Women for hire: prostitution and sexuality in France and where might research be targeted in the future? (edi- after 1850. Boston, MA: Harvard University Press, 1990. torial) AIDS 1994;8: 1215-26. 5 De Zalduondo B, Bernard JM. Meanings and conse- 33 Ward H, Day S, Mezzone J, Dunlop L, Donegan C, Farrar quences of sexual-economic exchange: gender, poverty S, et al. Prostitution and risk of HIV: female prostitutes in and sexual risk behaviour in urban Haiti. In: Parker RG, London. BMJ 1993;307:356-8. Gagnon JH, eds. Conceiving sexuality: approaches to sex 34 Tomilinson DR, Hillman RJ, Harris JR, Taylor-Robinson research in a postmodern world. London and NY: D. Screening for sexually transmitted disease in London- Routledge, 1994. based male prostitutes. Genitourin Med 1991;67:103-6. 6 Kane S. Prostitution and the military: planning AIDS inter- 35 Bloor MJ, Barnard MA, Finlay A, McKeganey NP. HIV- vention in Belize. Soc Sci Med 1993;36:965-79. related risk practices among Glasgow male prostitutes: 7 Day S, Ward H. Female prostitution and sexually transmit- reframing concepts of risk behaviour. Med Anthrop Q ted disease. In: Lankinen KS, Bergstrom S, Makela PH, 1993;7: 152-69. Peltomaa M, eds. Health and disease in developing coun- 36 Morse E, Simon P, Osofsky H, Balson P, Gaumer H. The tries. London: Macmillan Press, 1994. male street prostitute: a vector for transmission of HIV 8 Cameron DW, Simonsen JN, D'Costa LJ, Ronald AR, infection into the heterosexual world. Soc Sci Med Maitha GM, Gakinya MN, et al. Female-to-male trans- 1991;32:535-9. mission of human immunodeficiency virus type 1: risk 37 Elifson KW, Boles J, Sweat M. Risk factors associated with factors for seroconversion in men. Lancet 1989;ii:403-7. HIV infection among male prostitutes. Am Jf Public 9 Laga M, Manoka AB, Kivuvu M, Malele B, Tuliza M, Health 1993;83:79-83. Nzila N, et al. Non-ulcerative STD as risk factors for 38 Philpot CR, Harcourt CL, Edwards JM. A survey of female HIV infection in women: results from a cohort study. prostitutes at risk of HIV infection and other sexually AIDS 1993;7:95-102. transmissible diseases. Genitourin Med 199 1;67:384-8. 10 Ward H, Day S. Prostitution. In: Harris JRW, Forster S, 39 Lurie P, Fernandes ME, Hughes V, Arevalo EI, Hudes ES,

eds. Recent advances in STD and AIDS. Edinburgh: Reingold A, et al. Socioeconomic status and risk of HIV- http://sti.bmj.com/ Churchill Livingstone, 1991. 1, syphilis and hepatitis B infection among sex workers in 11 Laga M, Mamadou OD, Bouve A. Inter-relationship of Sao Paolo State, Brazil. AIDS 1995;9:S31-7. sexually transmitted diseases and HIV: where are we 40 Day S, Ward H, Perrotta L. Prostitution and risk of HIV: now? AIDS 1994,8(suppl l):S119-24. male partners of female prostitutes. BMJ 1993;307: 12 D'Costa U, Plummer FA, Bowmer I, Fransen L, Piot P, 359-61. Ronald AR, et al. Prostitutes are a major reservoir of sex- 41 Faugier J, Cranfield S. Reaching male clients of female ually transmitted diseases in Nairobi, Kenya. Sex Transm prostitutes: the challenge for HIV prevention. AIDS Care Dis 1985;12:64-7. 1995;7:S21-32. 13 Nzila N, Laga M, Thiam MA, Mayimona K, Edidi B, Van 42 Johnson AM, Wadsworth J, Wellings K, Field J. Sexual atti- Dyck E, et al. HIV and other sexually transmitted dis- tudes and lifestyles. Oxford: Blackwell Scientific on September 29, 2021 by guest. Protected copyright. eases among female prostitutes in Kinshasa. AIDS 1991; Publications, 1994:135-40. 5:715-21. 43 Laga M, Nzila N, Goeman J. The interrelationship of sexu- 14 Kreiss JK, Keoch D, Plummer FA, Holmes KK, Lightfoote ally transmitted diseases and HIV infection: implications M, Piot P, et al. AIDS virus infection in Nairobi prosti- for the control of both epidemics in Africa. AIDS 1991; tutes: spread of the epidemic to east Africa. N EnglJ Med 5(Suppl 1):S55-63. 1986;314:414-8. 44 Plummer FA, Nagelkerke NJD, Moses S, Ndinya-Achola 15 Van de Perre P. Female prostitutes: a risk group for infection JO, Bwayo J, Ngugi E. The importance of core groups in with human T-cell lymphotropic virus type III. Lancet the epidemiology and control of HIV infection. AIDS 1985;ii:524. 1991;5(suppl 1):S169-76. 16 Piot P, Plummer FA, Rey MA, Ngugi EN, Rouzioux C, 45 Laga M, Alary M, Nzila N, Manoka AT, Tuliza M, Behets Ndinya-Achola JO, et al. Retrospective seroepidemiology F, et al. Condom promotion, sexually transmitted dis- of AIDS virus infection in Nairobi populations. J Infect eases treatment, and declining incidence of HIV-1 infec- Dis 1987;155:1108-12. tion in female Zairian sex workers. Lancet 1994;344: 17 De Cock KN, Porter A, Odehouri K, Barrere B, Moreau J, 246-8. Diaby L, et al. Rapid emergence of AIDS in Abidjan, 46 Scott GR, Peacock W, Cameron S. Outreach STD clinics Ivory Coast. Lancet 1989;ii:408-1 1. for prostitutes in Edinburgh. Int Jf STD AIDS 1995; 18 Djomand G, Greenberg AE, Sassan-Morokro M, Tossou 6:197-200. 0, Diallo MO, Ekpini E, et al. The epidemic of 47 Alexander P. Sex workers fight against AIDS: an interna- HIV/AIDS in Abidjan, Cote d'Ivoire: a review of data tional perspective. In: Schneider BE, ed. Women resisting collected by Projet RETRO-CI from 1987 to 1993. J AIDS: strategies of empowerment. Philadelphia: Temple Acquir Immue Defic Syndr Hum Retrovirol 1995;10: University Press, 1995. 358-65. 48 Campbell CA. Prostitution, AIDS, and preventive health 19 Martin HL, Jackson DJ, Mandaliya K, Bwayo J, RakwarJP, behavior. Soc Sci Med 199 l;32:1367-78. Nyange P, et al. Preparation for AIDS vaccine evaluation 49 Venema PU, Visser J. Safer prostitution: a new approach in in Mombasa, Kenya: establishment of seronegative Holland. In: Plant M, ed. AIDS: drugs and prostitution. cohorts of commercial sex workers and trucking company London: Routledge, 1990. employees. AIDS Res Hum Retroviruses 1994;10:S235-7. 50 Phetersen G ed. A vindication of the rights of whores. Seattle, 20 Rodrigues JJ, Mehendale SM, Shepherd ME, Divekar AD, WA: The Seal Press, 1989. Gangakhedkar RR, Quinn TC, et al. Risk factors for HIV 51 Neave M. The failure of reform. A NZ J infection in people attending clinics for sexually transmit- Crim 1988;21:202-13. ted diseases in India. BMY 1995;311:283-6. 52 Heise LL, Elias C. Transforming AIDS prevention to meet 21 Nelson KE, Beyrer C, Natpratan C, Eiumtrakul S, women's needs: a focus on developing countries. Soc Sci Celentano DD, Khamboonruang C. Preparatory studies Med 1995;40:931-43. for possible HIV vaccine trials in Northem Thailand. 53 Kreiss J, Ngugi E, Hoimes KK, Ndinya-Achola J, Waiyaki P, AIDS Res Hum Retroviruses 1994;10:S243-6. Roberts PL, et al. Efficacy of nonoxynol-9 contraceptive 22 Ou C-Y, Takebe Y, Weniger BG, Luo CC, Kalish ML, sponge use in preventing heterosexual acquisition of HIV Auwanit W, et al. Independent introduction of two major in Nairobi prostitutes. JAMA 1992,268:477-82. HIV-1 genotypes into distinct high-risk populations in 54 Weir SS, Roddy RE, Zekeng L, Feldblum PJ. Nonoxynol-9 168 Day, Ward

use, genital ulcers, and HIV infection in a cohort of sex issues in representation and experience. London: Roudledge, workers. Genitourin Med 1995;71:78-81. 1994. 55 Stone AB, Hitchcock PJ. Vaginal microbicides for preventing 66 Hanenberg RS, Rojanapithayakorn, Kunasol P, Sokal DC. the sexual transmission of HIV. AIDS 1994;8(suppl 1): Impact of Thailand's HIV-control programme as indi- S285-93. cated by the decline of sexually transmitted diseases. 56 Dallabetta GA, Miotti PG, Chiphangwi G, Liomba G, Lancet 1994;344:243-5. Canner JK, Saah AJ. Traditional vaginal agents: use and 67 Bhave G, Lindan CP, Hudes ES, Desai S, Wagle U, association with HIV infection in Malawian women. Tripathi SP, et al. Impact of an intervention on HIV, sex- Genitourin Med: first published as 10.1136/sti.73.3.161 on 1 June 1997. Downloaded from AIDS 1995;9:293-297. ually transmitted diseases, and condom use among sex 57 Elias CJ, Heise L. Challenges for the development of female- workers in Bombay, India. AIDS 1995;9:S21-30. controlled vaginal microbicides. AIDS 1994;8:1-9. 68 Asamoah-Adu A, Weir S, Pappoe M, Kanlisi N, Neequaye 58 McLeod E. Women working: prostitution now. Guildford, A, Lamptey P, et al. Evaluation of a targeted AIDS pre- England: Croom Helm, 1985. vention intervention to increase condom use among pros- 59 Day S. Anthropological perspectives on sexually transmit- titutes in Ghana. AIDS 1994;8:239-46. ted diseases. In: Job-Spira N, Spencer B, Moalti JP, 69 Moses S, Plummer FA, Ngugi EN, Nagelkerke NJD, Bouvet E, eds. Sante publique et maladies a transmission Anzala AO, Ndinya-Achola JO. Controlling HIV in sexuelle. Paris: John Libbey Eurotext, 1990. Africa: effectiveness and cost of an intervention in a high- 60 Andrus JK, Fleming DW, Harger DR, Yan Chin M, frequency STD transmitter core group. AIDS 1991;5: Bennett DV, Horan JM, et al. Partner notification: can it 407-11. control epidemic syphilis? Ann Intern Med 1990;112: 70 Blackmore CA, Limpakarnianarat K, Rigau-Perez, 539-43. Albritton WL, Greenwood JR An outbreak of chancroid 61 North R, Rothenberg K. Partner notification and the threat in Orange County, California: descriptive epidemiology of domestic violence against women with HIV infection. N and disease control measures. J7 Infect Dis 1985;151: Engl J Med 1993;329:1194-6. 840-4. 62 Centers for Disease Control. Antibody to human immuno- 71 Grosskurth H, Mosha, Todd J, Mwijarubi E, Klokke A, deficiency virus in female prostitutes. MMWR 1987;36: Senkoro K, et al. Impact of improved treatment of sexually 157-61. transmitted diseases on HIV infection in rural Tanzania: 63 Day S. Prostitute women and the ideology of work in randomised controlled trial. Lancet 1995;346:530-6. London. In: Feldman DA, ed. AIDS and culture: the 72 Perkins R. Sex and sex workers in Australia. Sydney: UNSW global pandemic. New York: Praeger, 1990. Press, 1994. 64 Holland J, Ramazanoglu C, Scott S, Sharpe S, Thomson 73 Lurie P, Hintzen P, Lowe RA. Socioeconomic obstacles to R. Between embarrassment and trust: young women and HIV prevention and treatment in developing countries: the diversity of condom use. In: Aggleton P, Hart G, the roles of the International Monetary Fund and the Davies P, eds. AIDS: responses, interventions and care. World Bank. AIDS 1995;9:539-46. London: The Falmer Press, 1991. 74 Moses S, Manji F, Bradley JE, Nagelkerke NJ, Malisa MA, 65 Day S. What counts as rape? Physical assault and broken Plummer FA. Impact of user fees on attendance at a contracts: contrasting views of rape among London sex referral centre for sexually transmitted diseases in Kenya. workers. In: Harvey P, Gow P, eds. Sex and violence: Lancet 1992;340:463-6. http://sti.bmj.com/ on September 29, 2021 by guest. Protected copyright.