Table 1

Characteristics of included studies

(first author, year)
Additional sourcesCountryYear startedDesignAimPopulationN (FSWs) at baselineAge (median)†Current contraceptive use‡ (%)Consistent condom use§Number of sex partners/frequency of sex¶Gender-based violence  (GBV)/alcohol/other risk factorHIV/STI prevalence
Outcome 1: unintended pregnancy
 Behets,39 2005Madagascar2004Prospective cohort (with intervention).Assess acceptability and feasibility of diaphragm use.FSWs who use condoms inconsistently.9128Any: 47%
LARC or permanent: <1%.
0% with clients in last month (inconsistent use was an inclusion criterion).Five partners.
Six sex acts.
N/A.Vaginitis/Pelvic inflammatory disease (PID): 8%.
TP (RPR): 27%.
 Behets,40 2008Author††
Khan et al 4
Penman-Aguilar et al 72
Madagascar2005RCT (pilot).Assess acceptability and feasibility of diaphragm and microbicide use for STI prevention.Women with high-risk sex behaviours (sex work self-reported: 81% current, 100% ever).19229Any (excl. condoms): 24%.0% in last 2 weeks (inconsistent use was an inclusion criterion).Six casual partners.
10 sex acts.
Ever violence from casual partner for suggesting condom: 21%.
Ever received more money for no condom: 38%.
 Braunstein,32 2011Braunstein et al 73 Rwanda2006Prospective cohort.Measure HIV incidence
(secondary aim: measure pregnancy incidence).
HIV-uninfected women at high risk of HIV exposure (94% reported current sex work).39724Any: 91%.
LARC or permanent: 0%.
21% with clients
18% with non-paying partners.
90 partners in past 3 months.
10 clients per week.
40 vaginal sex acts in last month.
Forced sex ever: 19%.
Alcohol before sex: 52%.
CT: 5%.
GN: 12%.
TV: 17%.
TP (RPR+TPHA pos): 7%.
HSV2: 54%.
 Chersich,34 2014Author††
Luchters et al 5
Kenya (Mombasa).2006Prospective cohort.Assess HIV incidence and microbicide trial feasibility.
This substudy: investigate links between alcohol use, and unsafe sex and incident HIV infection.
HIV-uninfected FSWs.386Mean 25.1Any (incl. consistent condom use): 57.1%.
LARC: 3.0%.
Permanent: 0%.
21.3% in last 3 months.N/A.Hazardous or harmful drinking: 26.8%.
Ever had abortion: 21%.
 Deschamps,33 2016Deschamps et al 74 Haiti, Puerto Rico and Dominican Republic2009Prospective cohort.Assess feasibility of establishing a high-risk cohort for HIV vaccine trials.
This substudy: assess retention, HIV and pregnancy incidence and risk behaviours.
HIV-uninfected FSWs.63424‡‡Permanent: 10.0% (excluded from pregnancy analysis).
Others not reported.
0.5% in last 6 months.447 partners in last 6 months‡‡.Forced sex by client in last 6 months: 37.1%.
Heavy drinker: 38.8%.
Drug use: 14.0%.
Gaffoor,42 2013Author††
Skoler-Karpoff et al 75
South Africa (one site of a multisite trial)2004RCT (phase 3, double blind, placebo-controlled).Test safety and efficacy of the microbicide. Carraguard for HIV prevention.
This substudy: describe prevalence and associations of forced sex.
HIV-uninfected sexuallyactive women (3% FSWs).41§§§§N/A.§§N/A. §§
  Lara,41 2009Author††Dominican Republic2006Prospective cohort (with intervention).Assess acceptability of the female condom and diaphragm, determinants of use, and impact on unprotected sex.FSWs.24358.8% aged 20–29 yearsAny (excl. condoms): 22.2%.
Permanent: 0%.
66% in last month.N/A.Ever had abortion: 70%.HIV: 1%.
CT: 13%.
GN: 2%.
TP (VDRL): 8%.
 McClelland,45 2008Author††
Martin et al 76
McClelland et al 77
McClelland et al 78
Kenya (Mombasa)2003RCT (placebo-controlled, nested in an open cohort study).Test efficacy of monthly periodic presumptive antibiotic treatment at reducing incidence of vaginal infections and promoting vaginal Lactobacillus colonisation.HIV-uninfected FSWs.31032Any (excl. condoms): 35.5%.
LARC: 3.6%.
Permanent: 2.9%.
Median 100% coverage of sex acts in past week.‡‡One partner.
One sex act.‡‡
N/A.GN: 0.3%.
TV: 1%.
Cervicitis (microscopy): 0.6%.
HSV2: 74%.
BV: 34.5%.
 Peterson,46 2007Author††
Macqueen et al 66
Ghana, Cameroon and Nigeria2004RCT (phase 2, double blind, placebo controlled).Investigate safety and preliminary effectiveness of tenofovir disoproxil fumarate in preventing HIV infection.HIV-uninfected women who work in hotels, bars, markets in high HIV transmission areas (areas known for sex work).936Mean 23.6‡‡Any (excl. condoms): 7.22%.
LARC: <2%.
Permanent: <2%.
N/A.Mean 21 partners in 30 days.
Mean 12 coital acts per week.
N/A.Any STI in last 6 months (self-reported): 41.2%.
 Watson-Jones,44 2008Author††
Odutola et al 79
Tanzania2004RCT (double blind, placebo controlled).Determine whether HSV2 suppressive therapy reduces the risk of HIV acquisition and genital shedding of HIV.Female workers at food and recreational facilities at risk of HIV (38% FSWs).499§§§§§§§§ §§ §§
Outcome 2: pregnancy (intention undefined)
 Bazzi,54 2015Author††
Syvertsen et al 80
Mexico2010Prospective cohort.Identify time varying risk factors for STI acquisition within FSWs’ intimate partnerships.FSWs with drug use history and their steady male partners.21233Any (excl. condoms): 53.3%.
LARC: 12.3%.
Permanent: 25.5%.
Often or always: 56%.N/A.In last year:
physical assault by partner: 41%.
Sexual coercion in relationship: 9%.
In last 6 months:
hazardous drinking: 23%.
Intravenous drug use: 62%.
HIV: 2.6%.
CT: 5.9%.
GN: 1.2%.
TP (active): 1.4%.
Any STI: 8%.
 Page,38 2013Author††
Duff et al 55
Couture et al 81
Cambodia2009Prospective cohort.Estimate HIV and STI prevalence, incidence and associated factors.
This substudy: describe contraceptive utilisation and correlates of incident pregnancy.
Young women who practice SW and/or have multiple partners (all those recruited had practiced SW).22060.3% aged 25–29 yearsAny hormonal (not LARC): 10.8%.
LARC: <1.0%.
N/A.Four partners in last month.In last year: physical or sexual violence by client: 26.0%.
Intimate partner: 20.1%.
In last 3 months:
stimulant drug use: 27.0%.
Abortion: 11.3%.
HIV: 16.2%.
 Feldblum,35 2007Feldblum et al 57
Hoke et al 82
Madagascar2001RCT.Assess impact of two condom promotion interventions.
This substudy: estimate pregnancy incidence rate and predictive factors.
FSWs.935Mean 28.3Any highly effective (excl. condoms): 16.3%.No unprotected sex with any partners: 13.2%.Mean 5–6 partners.N/A.CT: 14.6%.
GN: 21.7%.
TV: 11.7%.
Any STI: 36.1%.
 Kaewkungwal,36 2013Rerks-Ngarm et al 83 Thailand (two provinces)2003RCT (multisite double blind placebo controlled).Assess the efficacy of two vaccines to prevent HIV.
This substudy: determine the qualities and outcomes of women’s participation.
HIV-uninfected women (5% FSWs).318N/AN/A.§§N/A.§§N/A.
 Kaul,47 2004Yadav et al 84
Fonck et al 85
Kenya (Nairobi)1998RCT (double blind placebo controlled).Assess impact of monthly PPT on HIV and STI incidence.HIV-uninfected FSWs.43028.6‡‡Any hormonal (not LARCs): 39.1%.17.2% with casual partner.‡‡15.4 partners.‡‡Daily alcohol: 47.6%.
Ever intravenous drug use: 4.1%.
CT: 9.9%.
GN: 10.3%.
TV: 12.2%.
TP: 4.4%.
HSV2: 73.9%.
BV: 51.1%.
 Liu,53 2015Author††China2009Cluster RCT.Assess the impact of a preventive intervention for FSWs on condom use with clients and partners.FSWs.750Mean 27.8‡‡LARC: 29.9%.43.6% in past month.Mean 8.3 clients.‡‡N/A.CT: 14.0%.
GN: 3.3%.
TP: 1.3%.
Any STI: 16.9%.
 McClelland,37 2011Author††
Martin et al 76
et al 86
Kenya (Mombasa)1993Open cohort.Assess HIV-1 incidence and relationships between hormonal contraception, STIs and HIV.
This substudy: examine relationship between risk behaviour and biologic outcomes (STI, pregnancy, seminal fluid deposition) among HIV-positive FSWs.
HIV-infected FSWs.89831Any (excl. condoms): 43.0%.
LARC: 2.34%.
Permanent: 2.67%.
55% in past week.One partner.
Two sex acts.
 Price,49 2012Author††Kenya (Nairobi, Kilifi)2005Prospective cohort.Describe populations at risk of HIV, including HIV incidence, in preparation for HIV trials.HIV-uninfected women and men at risk of HIV (75% of women were FSWs).515§§N/A.N/A.N/A. §§ Any non-ulcerative STI: 9.1%.
Genital ulcers: 1.5%.
TP: 0.6%.
Any STI: 10.6%.
 Priddy,56 2011Kenya (Nairobi)2008Prospective cohort.Assess HIV risk behaviour and incidence, STI prevalence, vaginal practices and retention.HIV-uninfected FSWs.200Mean 28Any non-barrier method: 52.0%.
LARC: 3.0%.
Permanent: 1.0%.
N/A (only reported sometimes/always use).Mean per day:
2.4 regular clients.
1.9 casual clients.
Sexual/physical violence related to Sex work (SW) in last month: 19.5%.
Sometimes/always paid more for no condom: 29.0%.
Sometimes/always has sex while intoxicated: 31.5%.
CT: 5.5%.
GN: 6.0%.
TV: 9.0%.
TP: 2.5%.
HSV2 (antibody): 72.0%.
BV: 38.0%.
 Robb,51 2016Author††
Rono et al 87
Kenya, Tanzania and Uganda2009Prospective cohort.Describe the trajectory of acute HIV infection.HIV-uninfected women and men at high risk for HIV (64% FSWs).1463N/AAny hormonal (incl. implant): 36.5%.
IUD: 0.5%.
Permanent: 0.5%.
32.6% with clients.
20.3% with non-paying partners.
N/A.Abortion in last 3 months: 0.43%.N/A.
 Strathdee,43 2013Author††
Vera et al 88
Gaines et al 89
Mexico2008RCT (four-arm factorial).Determine effectiveness of two behavioural interventions to reduce sexual and injecting risk.HIV-uninfected FSWs who inject drugs.58433Any (excl. condoms): 39.3%.
LARC: 25.3%.
Permanent: 17.8%.
14.9% with regular clients.
11.7% with casual clients.
30 clients per month.
51 paid sex acts per month.
N/A.CT: 12.0%.
GN: 2.2%.
TV: 33.6%.
TP (active): 8.4%.
 Van Damme,48 2002Author††
et al 90
et al 91
Benin, Cote d’Ivoire, South Africa and Thailand1996RCT (multisite triple blind placebo-controlled; open cohort design).Determine effectiveness of nonoxynol-9 microbicide in prevention of HIV-1.HIV-uninfected FSWs.89226N/A.N/A (only reported use of condom in ≥50% of sex acts).Three partners per day.N/A.CT: 4.4%.
GN: 5.1%.
TV: 3.5%.
TP: 11.2%.
 van Loggerenberg,52 2008Author††
et al 92
South Africa (Durban)2004Prospective cohort.Understand HIV-1 subtype C acquisition, pathogenesis and disease progression.
This substudy: describe cohort characteristics and HIV-incidence rates, and report challenges in establishing and maintaining the cohort.
HIV-uninfected women who practice SW (79%) and/or have multiple partners.193Mean 34.3N/A.53.9% with casual partners.
20.4% with steady partners.
Two partners per week.N/A.Any STI (CT, GN, TV, MG, TP, HSV2): 31.3%.
 Vandepitte,93 2013Author††
et al 94
Uganda (urban slum).2008Prospective cohort.Understand dynamics of HIV and STI infections among FSWs.
This substudy: investigate patterns of clearance and recurrence of untreated Mycoplasma genitalium infection.
FSWs.1027Mean 26N/A.59.8% in last month.At least daily sex for money: 50.5%.Problem drinking: 55.7%.MG: 14%.
 Vielot,50 2015Author††Kenya (Nairobi)2009Prospective cohort.Compare the duration of high risk HPV infection among FSWs by exposure to STIs, using a highly sensitive biomarker assay.FSWs.35028LARC: 15.5%.
Permanent: 2.1%.
Most of the time/always:
73.8% with clients.
24.6% with non-paying partners.
10 partners per week.N/A.HIV: 24.0%.
CT: 3.8%.
GN: 2.3%.
TV: 7.3%.
MG: 12.8%.
  • †Median unless specified.

  • ‡Any=modern contraceptive method including condoms, unless specified; LARC=long-acting reversible contraception (implants or IUDs); Permanent=any method of permanent contraception, for example, tubal ligation or hysterectomy.

  • §Always uses condoms (unless specified).

  • ¶Median number per week unless specified. Sex partners may be paying, non-paying, regular or casual, unless specified.

  • ††‘Author’ indicates additional data were obtained from the author. Other references listed here reported on the same study and were used for data extraction.

  • ‡‡Reported results segregated by sub-group; data presented here are overall estimates.

  • §§Not disaggregated by sex work status.

  • BV, bacterial vaginosis; CT, Chlamydia trachomatis; FSW, female sex worker; HSV2, herpes simplex virus type 2; MG, Mycoplasma genitalium; N/A, not measured or reported, data not available from author; NG, Neisseria  gonorrhoeae; RCT, randomised controlled trial; STI, sexually transmitted infection; TP, Treponema pallidum (syphilis); TV, Trichomonas vaginalis.