The sexual health of female sex workers compared with other women in England: analysis of cross-sectional data from genitourinary medicine clinics

Background While female sex workers (FSWs) are assumed to be at increased risk of sexually transmitted infections (STIs), there are limited comparative data with other population groups available. Using routine STI surveillance data, we investigated differences in sexual health between FSWs and other female attendees at genitourinary medicine (GUM) clinics in England. Methods Demographic characteristics, STI prevalence and service usage among FSWs and other attendees in 2011 were compared using logistic regression. Results In 2011, 2704 FSWs made 8411 recorded visits to 131/208 GUM clinics, (primarily large, FSW-specialist centres in London). FSWs used a variety of services, however, 10% did not have an STI/HIV test at presentation. By comparison with other female attendees, FSWs travelled further for their care and had increased risk of certain STIs (eg, gonorrhoea ORadj: 2.76, 95% CI 2.16 to 3.54, p<0.001). Migrant FSWs had better sexual health outcomes than UK-born FSWs (eg, period prevalence of chlamydia among those tested: 8.5% vs 13.5%, p<0.001) but were more likely to experience non-STI outcomes (eg, pelvic inflammatory disease ORadj: 2.92, 95% CI 1.57 to 5.41, p<0.001). Conclusions FSWs in England have access to high-quality care through the GUM clinic network, but there is evidence of geographical inequality in access to these services. A minority do not appear to access STI/HIV testing through clinics, and some STIs are more prevalent among FSWs than other female attendees. Targeted interventions aimed at improving uptake of testing in FSWs should be developed, and need to be culturally sensitive to the needs of this predominantly migrant population.


INTRODUCTION
Factors associated with sex work (eg, multiple sexual partners, violence and drug use) pose a risk to the health of female sex workers (FSWs). 1 In a number of countries the prevalence of HIVand sexually transmitted infections (STIs) are higher among FSWs than other women, [2][3][4] and some subpopulations, such as migrant FSWs, appear to have worse sexual health outcomes. 5 In England, FSWs are thought to be at increased risk of STI [6][7][8] and to experience barriers to accessing prevention and treatment services. [9][10][11] Disparities in sexual health outcomes and service usage between migrant and UK-born FSWs have also been described. 12 However, much of what is reported is based on information gathered from special studies which are usually small and reliant on self-reported information.
The Genitourinary Medicine Clinic Activity Dataset (GUMCAD) is a patient-level, electronic dataset including diagnoses made and services provided at all genitourinary medicine (GUM) clinics in England that enables analysis of associations between patients' demographic characteristics and their use of sexual health services and sexual health outcomes. Since 2011, it has included information on whether patients are sex workers (SWs) 13 presenting a unique opportunity to undertake a comprehensive analysis of SWs seeking sexual healthcare in England using routine national surveillance system data. For the first time, we compare the sexual health outcomes and service usage of FSWs with those of other females attending GUM clinics in England in order to better determine the demographic and clinical characteristics of FSWs, their risk of STI acquisition and patterns of service access. Such information could be used to inform the development of sexual health services better tailored to the needs of this population.

Source of data
Guidelines for collecting and reporting GUMCAD data have previously been published. 14 Briefly, demographic data is self-reported by patients at first attendance (ie, gender, age, ethnicity, country of birth and postcode of residence) or during a clinical consultation (ie, sexual orientation). (It should be noted that self-reported sexual orientation may not always be congruent with sexual practice, 15 ie, FSWs may engage in sex with men but identify as homosexual). For each consultation, relevant services and diagnoses are recorded for a patient using a set of uniform codes known as Sexual Health and HIV Activity Property Type (SHHAPT) codes. The SHHAPT code 'SW' is attached to visits made by SWs but is not permanently attached to a patient's clinic record; it is recorded independently at each attendance, thereby requiring active coding by healthcare workers. There are no guidelines provided to GUM clinics on how to define sex work or ascertain patients' SW status, clinics may rely on self-disclosure by patients or on healthcare staff actively asking all or selected patients. A subset of GUMCAD data containing all visits by females between 1 January and 31 December 2011 Open Access Scan to access more free content was extracted from the GUMCAD database held by Public Health England ((PHE), formerly the Health Protection Agency). This subset contained a record of tests, services and diagnoses across attendances within a clinic for each woman. Women recorded as SWs (by the application of the SHAPPT code 'SW' during one visit in 2011) were classified as such for any other visits that year. This was to ensure that visits without the 'SW' code attached, either due to inconsistent disclosure by FSWs 10 11 16 or coding errors by staff, were not excluded from the analysis.

Data analysis
The demographic characteristics of attendees, number of clinic visits and the following service use variables were compared using Pearson χ 2 tests: the proportion of patients receiving postexposure prophylaxis for HIV following sexual exposure ( Univariate associations between SW status and diagnoses and demographic factors (eg, age, ethnicity, sexual orientation, location, deprivation and migrant status) were investigated using logistic regression. The 2010 index of multiple deprivation score for a patient's postcode of residence was used as a measure of deprivation, and non-UK born FSWs were defined as migrants. Factors with p value<0.10 were included in multivariate logistic regression to explore the effect of SW status on infection and reinfection, adjusting for potential confounders. Women experiencing a second infection with an STI more than 6 weeks after their original diagnosis was recorded as reinfections. Subsequent reinfections were not included in the analysis. This analysis was repeated to compare UK-born and migrant FSW in terms of their demographic characteristics, attendance patterns, service use and STI prevalence and reinfection. Stata V.12 was used for all analyses.

Demographic characteristics
In 2011, there were 699 645 women who attended GUM clinics in England, of whom 2704 (0.4%) were identified as SWs. Table 1 shows the demographic characteristics and services used for FSWs and other female attendees. A greater proportion of FSWs than other female attendees were migrants, and FSWs tended to be older.  concentrated; they were more likely to attend a clinic in London than other attendees (74.7% vs 28.4%, p<0.001). They were also more likely to attend London clinics even when residing elsewhere; 14.5% of FSW and 6.6% of other female attendees attending clinics in London were not London residents ( p<0.001). Visits by FSWs were concentrated in large clinics providing SW-specific services; more than three-quarters (79.4%) of their visits were recorded at just 12 clinics, seven of which were in London. These 12 clinics accounted for 92.2% of migrant FSW visits compared with 57.8% of UK-born FSWs.

Service use
FSWs were more likely to use non-testing services such as contraception, smear tests and hepatitis B vaccination than other attendees (table 1), and a greater proportion of FSWs had a HIV test or sexual health screen (which tests for chlamydia, gonorrhoea and syphilis). As FSW visits were not equally distributed across all clinics we investigated whether the higher proportion of FSWs tested may be due to higher levels of testing overall at the 131 clinics they attended. However, the significant association remained when the analysis was restricted to these clinics (88.9% vs 63.5%, p<0.001 for sexual health screens; 87.8% vs 61.4%, p<0.001 for HIV tests). Migrant FSWs were more likely than UK-born FSWs to have a sexual health screen or HIV test, and to use non-testing services, with the exception of PEPSE (table 2).

Sexual health
There were no significant differences in the period prevalence of HIV or syphilis between FSWs and other female attendees. Chlamydia was the most prevalent STI diagnosed in both groups with a significantly higher period prevalence among FSWs (table 3). FSWs were almost twice as likely to be diagnosed with chlamydia, and three times more likely to be diagnosed with gonorrhoea than other female attendees, adjusting for demographic factors. As FSWs made more visits on average than other female attendees, this increased prevalence may have been linked to the higher number of opportunities FSWs had to be diagnosed. As our model did not adjust for number of visits, we also compared prevalence of STI diagnosis in those tested at first visit and found that FSWs were still twice as likely to be diagnosed with chlamydia (2.9% vs 1.3%, p<0.001) and four times more likely to be diagnosed with gonorrhoea (0.8% vs 0.2%, p<0.001) than other female attendees. Other conditions such as hepatitis B, hepatitis C, scabies, BV, candidosis, PID, UTI and abnormal cervical cytology were also higher among FSWs, though other female attendees had a higher prevalence of genital warts. While reinfections with some STIs were more common in FSWs than in other females, the only statistically significant differences were for chlamydia reinfections (6.6% of infected FSWs became reinfected in 2011 vs 3.4% of other attendees, p=0.02) and BV recurrences (7.4% vs 3.7%, p=0.004). UK-born FSWs had a higher prevalence of chlamydia and trichomoniasis than migrant FSWs (table 4); however, the prevalence of candidosis, UTI, PID and abnormal cervical cytology were higher among UK-born FSWs. Adjusting for demographic factors and other diagnoses, only differences in PID and trichomoniasis prevalence remained statistically significant. Migrant FSWs were 83% less likely to be diagnosed with trichomoniasis than UK-born FSWs but three times more likely to be diagnosed with PID. There were no statistically significant differences in reinfections between migrant and UK-born FSWs.

DISCUSSION
For the first time, we have been able to provide a comprehensive description of the characteristics and sexual health needs of FSWs receiving care at GUM clinics in England. FSWs appear to be at higher risk of infection and reinfection with certain STIs than other female attendees; however, the levels of STI among FSWs reported here are lower than in comparable studies. While services, such as STI testing, vaccination and contraception are more frequently accessed by FSWs than other attendees, there is evidence of missed opportunities. One limitation of our study is the likely underestimation of the number of FSWs attending GUM clinics, partly due to under-disclosure by FSWs but also because of an absence of guidelines on how SW status should be ascertained, or lack of coding by staff. Additionally, FSWs attending GUM clinics may not be representative of the wider FSW population, as those who attend GUM clinics for testing and treatment may have different risk behaviours than those who do not. Lack of knowledge about the different health-seeking behaviours among FSWs limits the generalisability of the conclusions that can be drawn from this analysis to FSWs who attend GUM clinics. Nonetheless, by comparison with special SW studies, our study provides a national picture of the sexual health of FSWs attending GUM clinics and their use of services, enabling geographic variations in service access to be explored and specific sexual health needs of FSWs to be investigated. In our study, FSWs were older, a greater proportion lived in London and, most notably, three times as many were migrants compared with other attendees. Sociodemographic differences such as these may be useful for tailoring interventions and services for FSWs, for example, deciding on the location of outreach services and deciding which languages to offer them in. Country of origin of migrant FSWs also has implications for service priorities at clinics, for example, ensuring HIV testing is available to the large number from Brazil where HIV prevalence among FSWs is 6.2%. 17 Visits by FSWs were not equally distributed across clinics, only 131/208 reported an attendance by FSWs in 2011, and large, specialist centres with SW-specific services appear to play a crucial role in their sexual healthcare provision. One clinic in London reported almost a quarter of all FSW visits, with some FSWs who attended this clinic living as much as 300 miles away in North West England. FSWs living outside of London were more likely to attend services in London than other female attendees which may be a reflection of FSWs working patterns. FSWs working in London but living elsewhere in the country may choose to attend a clinic near their workplace rather than their area of residence. However, the pattern of visiting clinics in London when living elsewhere, the concentration of visits at large, specialist centres and the low number, or absence, of FSW attendances in some areas of the country, could also signify geographical inequalities in terms of access to, or awareness of, suitable services for FSWs. For example, in Cumbria, there were no attendances by FSWs reported to our surveillance system in 2011 even though a study mapping the sex market in the region identified more than 180 male and female SWs. 18 It is also possible that FSWs are accessing these clinics, but the SHAPPT code 'SW' is not being applied to their consultation as they do not disclose, or are not asked about, their involvement in sex work. Using routine surveillance data to identify clinics that are frequently and rarely visited by SWs may be useful for commissioners and service planners to investigate areas of potential unmet needs, and to strategically plan services and outreach programmes.
While FSWs attending GUM clinics were significantly more likely than other attendees to have a sexual health screen or HIV test, over 10% of FSWs were not tested. These FSWs may    have been assessed not to need testing when they attended, or may access testing through primary health care providers, 11 which do not report to our surveillance system; however, it is likely that there remains a minority of FSWs who are not testing for STIs at all, as has been seen in other UK studies. 19 20 That one in eight FSWs attending a GUM clinic had no sexual health screen or HIV test despite being already engaged with healthcare services, represents major missed opportunities. Studies have suggested that migrant FSWs have poorer access and are less likely to engage with services than UK-born FSWs, 12 perhaps due to lack of knowledge or fears related to their legal/ immigration status. 21 22 However, in our study, migrant FSWs made more visits on average than UK-born FSWs, and a greater proportion had a sexual health screen or HIV test. A greater proportion of migrant FSWs also used other services, such as contraception and cervical cytology, which highlights the vital role GUM clinics play in meeting their broader sexual and reproductive healthcare needs, but may also indicate unmet need for primary care-type services, similar to that expressed by FSWs in Bristol. 11 The higher prevalence of non-sexually transmitted illness (eg, UTI) and the increased risk of PID further support this theory. This may have implications for service planning and intervention design for FSWs which may need to adopt an integrated approach to improving sexual health in the wider context of improving their general health.
FSWs have been shown to be at greater risk of STI when compared with the general population, for example, STIs were 9-60 times more common in street-based FSWs than in females aged 16-44 years from the General Household Survey (GHS). 19 In our study, FSWs were twice as likely as other female attendees to be diagnosed with chlamydia and almost three times more likely to be diagnosed with gonorrhoea, but there was no significant difference in the period prevalence of HIV or syphilis. Our calculated odds of infection for FSWs by comparison with other females may be lower than those reported elsewhere, as our comparative population of GUM clinic attendees is likely to be at higher risk than the GHS population. Furthermore, street-based FSWs are a higher risk group than other FSWs 10 23-25 and are not likely to be representative of FSWs attending GUM clinics. The increased risk of certain STIs among FSWs that we identified may not be a consequence of sex work per se, but rather of other factors associated with sex work. STI acquisition among FSWs has been shown to be associated with intravenous drug use (IDU) and the use of condoms with, and number of, non-paying casual partners 6 24 rather than the number of clients or duration of sex work. As our surveillance system does not gather data on sexual and drug-injecting behaviours, the influence of these factors on STI risk could not be determined. Currently, the collection of behavioural data (including IDU) through GUMCAD is being piloted by PHE, and so the impact of these factors may be explored in the future. More than 10% of FSWs infected with chlamydia and more than 30% diagnosed with BV in 2011 experienced a reinfection or recurrence during the year, suggesting opportunities exist for improving clinic-based interventions to prevent repeated risk exposure.
Migrant FSWs were less likely to be diagnosed with chlamydia or trichomoniasis than UK-born FSWs, but there were no significant differences for infection with other STIs between the groups. While our findings contradict those of other studies in the UK 7 12 that suggest migrant FSWs are at higher risk of STI than indigenous FSWs, they must be interpreted cautiously given the limitation in generalisability of this study to FSWs attending GUM clinics. Previous studies reporting poorer sexual health outcomes among migrant FSWs have also shown they have poorer access to services such as GPs, 12 thus, the comparatively better sexual health outcomes observed among the migrant FSWs in our study may be a reflection of their increased use of services. The differences in their working conditions could also be a contributing factor; it is thought migrant FSWs in England work almost exclusively indoors, 12 26 a lower-risk occupational environment, 23 25 however, this hypothesis was not tested in this study as this type of information is not gathered routinely through GUMCAD. Understanding the cause of the variation in sexual health between migrant and UK-born FSWs in the future will be important for effective intervention design.
Overall, FSWs in England have access to high-quality sexual health care through the GUM clinic network, but there is evidence of geographical inequality in access to these services. Though uptake is high, a minority of FSWs do not appear to access STI/HIV testing through clinics, and further quantitative and qualitative studies to identify the barriers to testing among FSWs are warranted. Some STIs are more prevalent among FSWs than other female attendees, and the high levels of reinfection with chlamydia and BV experienced by FSWs should be investigated further. Targeted interventions aimed at improving uptake of testing and reducing the risk of repeat infection in this population should be developed, and need to be culturally sensitive to the needs of this predominantly migrant population.

Key messages
▸ FSWs appear to be at higher risk of certain sexually transmitted infections and reinfections compared with other female genitourinary medicine (GUM) clinic attendees, even adjusting for demographic factors. ▸ FSWs have access to high-quality sexual healthcare through the GUM clinic network, but there is evidence of geographical inequality in access to services. ▸ Differences in service usage among FSWs by migrant status may indicate migrant FSWs experience unmet needs for primary care-type services. ▸ There is little variation in sexual health outcomes among FSWs by migrant status; however, chlamydia infection is more common in UK-born FSWs.

Handling editor Jackie A Cassell
Contributors HW and GH devised the study, advised on data analysis and participated in interpreting the data and reviewing the manuscript; KM advised on data analysis and participated in interpreting the data and reviewing the manuscript; LM carried out the statistical analysis, participated in interpreting the data and prepared the manuscript.
Funding HW is supported by the Wellcome Trust 090285/Z/09/Z and the NIHR Imperial College Healthcare NHS Trust Biomedical Research Centre.

Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 3.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/3.0/