Behavioral interventions to reduce the transmission of HIV infection among sex workers and their...

Behavioral interventions to reduce the transmission of HIV

infection among sex workers and their clients in low- and

middle-income countries (Review)

Wariki WMV, Ota E, Mori R, Koyanagi A, Hori N, Shibuya K

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library2012, Issue 2

http://www.thecochranelibrary.com

Behavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income

countries (Review)

Copyright © 2012 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

http://www.thecochranelibrary.com

T A B L E O F C O N T E N T S

1HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2SUMMARY OF FINDINGS FOR THE MAIN COMPARISON . . . . . . . . . . . . . . . . . . .

6BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

17ADDITIONAL SUMMARY OF FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . .

27DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

56DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Analysis 1.1. Comparison 1 Social cognitive theory versus standard counseling for promotion of condom use, Outcome 1

HIV incidence among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . 64


STIs incidence among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . 64


Consistent condom use by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . 65


Mean number of condom use by FSWs at past month. . . . . . . . . . . . . . . . . . . . . 66


Mean number of protected sex among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . 67


Sexual activity outcomes by FSWs at past month. . . . . . . . . . . . . . . . . . . . . . 67


Injected drug use among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . 68


Alcohol use among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . 68


Drug use among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Analysis 2.1. Comparison 2 Social cognitive theory versus no intervention for promotion of condom use, Outcome 1

Consistent condom use by FSWs. . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Analysis 2.2. Comparison 2 Social cognitive theory versus no intervention for promotion of condom use, Outcome 2 HIV

knowledge among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . 70

Analysis 2.3. Comparison 2 Social cognitive theory versus no intervention for promotion of condom use, Outcome 3

Psychosocial barriers to condom use among FSWs at 6-month. . . . . . . . . . . . . . . . . . 70

Analysis 3.1. Comparison 3 Community empowerment versus standard care for promotion of condom use, Outcome 1

STIs prevalence among FSWs at 36-month. . . . . . . . . . . . . . . . . . . . . . . . 71


Consistent condom use by FSWs with regular partners. . . . . . . . . . . . . . . . . . . . 71


Consistent condom use by FSWs with clients. . . . . . . . . . . . . . . . . . . . . . . . 72


Change in 100% condom use by FSWs. . . . . . . . . . . . . . . . . . . . . . . . . . 72

iBehavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income

countries (Review)



Change in any condom use by FSWs. . . . . . . . . . . . . . . . . . . . . . . . . . 73


HIV knowledge among FSWs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73


HIV testing by FSWs at 36-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 74


Skills of risk and protective factors among FSWs at 16-month. . . . . . . . . . . . . . . . . . 74

Analysis 4.1. Comparison 4 Microenterprise plus education intervention versus education alone for reducing the number

of sex partners, Outcome 1 Consistent condom use by FSWs at 6-month. . . . . . . . . . . . . . 75

Analysis 4.2. Comparison 4 Microenterprise plus education intervention versus education alone for reducing the number

of sex partners, Outcome 2 Number of FSWs’ partners at 6-month. . . . . . . . . . . . . . . . 75

Analysis 5.1. Comparison 5 Peer education versus standard care for promoting HIV testing and condom use, Outcome 1

Mean number of consistent condom use by FSWs at 6-month. . . . . . . . . . . . . . . . . . 76


Mean score of HIV knowledge among FSWs at 6-month. . . . . . . . . . . . . . . . . . . 76


HIV testing by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . . 77


AIDS perceived control by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . 77


AIDS perceived severity by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . 78

Analysis 6.1. Comparison 6 Peer education plus clinic based counseling versus peer education only for promotion of

condom use, Outcome 1 STIs prevalence among FSWs. . . . . . . . . . . . . . . . . . . . 78


condom use, Outcome 2 Male condom use with clients. . . . . . . . . . . . . . . . . . . . 79


condom use, Outcome 3 Male condom use with non-paying partners. . . . . . . . . . . . . . . 80


condom use, Outcome 4 Female condom use by FSWs with clients. . . . . . . . . . . . . . . . 80


condom use, Outcome 5 Male and female condom use with clients. . . . . . . . . . . . . . . . 81


condom use, Outcome 6 Male and female condom use in the last sex with non-paying partner. . . . . . . 81

Analysis 7.1. Comparison 7 Peer education plus manager training versus standard care for promoting HIV testing and

condom use, Outcome 1 Mean number of consistent condom use by FSWs at 6-month. . . . . . . . . 82


condom use, Outcome 2 Mean score of HIV knowledge among FSWs at 6-month. . . . . . . . . . . 82


condom use, Outcome 3 HIV testing by FSWs at 6-month. . . . . . . . . . . . . . . . . . . 83


condom use, Outcome 4 AIDS perceived control by FSWs at 6-month. . . . . . . . . . . . . . . 83


condom use, Outcome 5 AIDS perceived severity by FSWs at 6-month. . . . . . . . . . . . . . . 84

Analysis 8.1. Comparison 8 Manager training versus standard care for promoting HIV testing and condom use, Outcome

1 Mean number of consistent condom use by FSWs at 6-month. . . . . . . . . . . . . . . . . 84


2 Mean score of HIV knowledge among FSWs at 6-month. . . . . . . . . . . . . . . . . . . 85


3 HIV testing by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 85


4 AIDS perceived control by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . 86

iiBehavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income

countries (Review)



5 Mean score of perceived themselves at greater risk (by FSWs at 6-month). . . . . . . . . . . . . 86

Analysis 9.1. Comparison 9 Promotion of female and male condom versus promotion of male condom, Outcome 1 HIV

incidence among FSWs at 3-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Analysis 9.2. Comparison 9 Promotion of female and male condom versus promotion of male condom, Outcome 2 STIs

incidence among FSWs at 3-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Analysis 9.3. Comparison 9 Promotion of female and male condom versus promotion of male condom, Outcome 3

Consistent male condom use at 3-months. . . . . . . . . . . . . . . . . . . . . . . . . 88


Consistent female condom use by FSWs at 3-month. . . . . . . . . . . . . . . . . . . . . 88


Consistent female condom use by FSWs at 24-month. . . . . . . . . . . . . . . . . . . . . 89


Consistent male condom use at 24-month. . . . . . . . . . . . . . . . . . . . . . . . . 89

Analysis 10.1. Comparison 10 Intensive STI screening versus basic STI screening to control STI, Outcome 1 HIV incidence

among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Analysis 10.2. Comparison 10 Intensive STI screening versus basic STI screening to control STI, Outcome 2 STIs

prevalence among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Analysis 10.3. Comparison 10 Intensive STI screening versus basic STI screening to control STI, Outcome 3 Consistent

condom use by FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Analysis 11.1. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 1 STIs prevalence

among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Analysis 11.2. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 2 Consistent condom

use by FSWs with clients at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . 92

Analysis 11.3. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 3 HIV/STI

knowledge among FSWs at 6-month. . . . . . . . . . . . . . . . . . . . . . . . . . 92

92APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96DIFFERENCES BETWEEN PROTOCOL AND REVIEW . . . . . . . . . . . . . . . . . . . . .

96INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iiiBehavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income

countries (Review)


[Intervention Review]

Behavioral interventions to reduce the transmission of HIVinfection among sex workers and their clients in low- andmiddle-income countries

Windy MV Wariki1, Erika Ota1, Rintaro Mori2, Ai Koyanagi1 , Narumi Hori3, Kenji Shibuya1

1Department of Global Health Policy, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan. 2Collaboration for

Research in Global Women’s and Children’s Health, Tokyo, Japan. 3Faculty of Nursing, St. Luke’s College of Nursing, Tokyo, Japan

Contact address: Windy MV Wariki, Department of Global Health Policy, Graduate School of Medicine, The University of Tokyo, 7-

3-1 Hongo Bunkyo-ku, Tokyo, 113-0033, Japan. [email protected].

Editorial group: Cochrane HIV/AIDS Group.

Publication status and date: New, published in Issue 2, 2012.

Review content assessed as up-to-date: 15 March 2011.

Citation: Wariki WMV, Ota E, Mori R, Koyanagi A, Hori N, Shibuya K. Behavioral interventions to reduce the transmission of HIV

infection among sex workers and their clients in low- and middle-income countries. Cochrane Database of Systematic Reviews 2012,

Issue 2. Art. No.: CD005272. DOI: 10.1002/14651858.CD005272.pub3.


A B S T R A C T

Background

Various interventions have been adopted to reduce HIV transmission among sex workers and their clients but the effectiveness of these

strategies has yet to be investigated using meta-analytic techniques.

Objectives

To evaluate the effectiveness of behavioral interventions to reduce the transmission of HIV infection among sex workers and their

clients in low- and middle-income countries.

Search methods

The Cochrane Central Register of Controlled Trials (CENTRAL), the Cochrane HIV/AIDS group specialized register, the Cochrane

Database of Systematic Reviews, MEDLINE, PsycINFO, Sociological Abstracts, CINAHL, Dissertation Abstract International (DAI),

EMBASE, LILACS, BIOSIS, SciSearch, INDMED, Proquest, and various South Asian abstracting databases were included in the

database list. The publication sites of the World Health Organization, the US Centers for Disease Control and Prevention, and other

international research and non-governmental organizations also appeared in the database list.

Selection criteria

Randomized controlled trials (RCTs) and quasi-RCTs examining the effects on HIV transmission risk of different behavioral inter-

ventions or comparing behavioral interventions with no intervention, where described any one of the outcome measures, such as HIV

incidence and prevalence, STI incidence and prevalence, change in self-reported of condom use, and other HIV-related outcome.

Data collection and analysis

Two authors independently assessed trials, extracted data and assessed the risk bias. Heterogeneity amongst trials was also tested.

1Behavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income

countries (Review)


mailto:[email protected]

Main results

A total of 13 trials with 8,698 participants were included. Primary outcomes (HIV and STI prevalence and incidence) were reported

in seven trials. Of these, HIV incidence was reported in only three trials. After a 6-month follow-up assessment, there was no evidence

that social cognitive behavioral intervention was effective in reducing HIV incidence (RR 0.12, 95% CI 0.01 to 2.22). However, there

was a reduction in HIV incidence at 3-month follow-up assessment of promotion of female and male condom (RR 0.07, 95% CI 0.00

to 1.38). Social cognitive interventions and promotion of female and male condom use were significantly reduced STIs incidence (RR

0.57, 95% CI 0.34 to 0.96) and (RR 0.63, 95% CI 0.45 to 0.88), respectively. Secondary outcomes were identified in 13 trials. Meta-

analyses showed evidence that interventions to promote the use of female and male condoms do reduce non-condom use (RR 0.83,

95% CI 0.65 to 1.05) compared to promotion of male condoms alone, and that social cognitive interventions reduced drug use among

sex workers (RR 0.65, 95% CI 0.36 to 1.16) compared to standard care.

Authors’ conclusions

Available evidence nevertheless suggests that compared with standard care or no intervention, behavioral interventions are effective in

reducing HIV and the incidence of STIs amongst female sex workers (FSWs). Given the benefits of social cognitive theory and the

promotion of condom use in reducing HIV/STI and the public health need to control transmission amongst FSWs, there is a clear

finding in favour of behavioral interventions. However, it should be recognized that there is a lack of information about most other

outcomes and target populations, and that all of the trials were conducted in low- and middle-income countries.

P L A I N L A N G U A G E S U M M A R Y

Behavioral interventions to reduce HIV incidence and HIV/STI prevalence among female sex workers in low- and middle-

income countries

The rates of human immunodeficiency virus (HIV) and sexually transmitted infection (STI) transmission continue to increase, partic-

ularly among sex workers and their clients in low- and middle-income countries. Prevention efforts directed towards these infections

in this at-risk population may have had an effect in reducing the overall transmission of HIV/STIs in the general population. Several

successful behavioral interventions have been reported including interventions to reduce HIV/STI incidence and prevalence, change

behavior, promote condom use, improve condom availability, and increase sexual health knowledge. The review found seven individual

randomised controlled trials (RCTs), two cluster-RCTs and four quasi-RCTs involving 8,698 participants examining a variety of be-

havioral interventions to evaluate whether they reduced HIV/STIs rates or resulted in changed behavior among sex workers and their

clients. Results showed that the interventions were effective in HIV/STI prevention, including reducing the incidence and prevalence

of HIV and STIs. Furthermore, there were some differences in self-reported behavior including increased condom use and a reduction

in the risk of drug use. However, these trials were small and generally had few participants. As a result, evidence for the effectiveness

of social cognitive theory and promoting condom use in reducing HIV/STI incidence compared to other behavioral interventions was

limited, because no RCTs examined the effects of these interventions on HIV prevalence or on sex workers other than FSWs. In future

research and program agendas therefore it is important to assess other potentially more potent behavioral change strategies.


countries (Review)


SU

MM

AR

YO

FF

IN

DI

NG

SF

OR

TH

EM

AI

NC

OM

PA

RI

SO

N[E

xpla

nati

on]

Socialcognitivetheorycomparedtostandardcounselingforpromotionofcondomuseforsexworkers

Patientorpopulation:sexworkers

Settings:low-andmiddle-incomecountries

Intervention:Socialcognitivetheory

Comparison:standardcounselingforprom

otionofcondom

use

Outcomes

Illustrative

comparativerisks*

(95%CI)

Relativeeffect

(95%CI)

NoofParticipants

(studies)

Qualityoftheevidence

(GRADE)

Com

ments

Assumed

risk

Correspondingrisk

standard

counselingfor

promotion

ofcondom

use

Socialcognitivetheory

HIV

incidence

among

FSWsat6-month

Studypopulation

RR0.12

(0.01to2.22)

709

(1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

11per1000

1per1000

(0to24)

Mediumriskpopulation

11per1000

1per1000

(0to24)

Syphilis

incidence

amongFSWsat6-month

Studypopulation

RR0.46

(0.18to1.19)

709

(1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

38per1000

17per1000

(7to45)


38per1000

17per1000

(7to45)


countries (Review)


http://www.thecochranelibrary.com/view/0/SummaryFindings.html

Gonorrhoea

incidence

amongFSWsat6-month

Studypopulation

RR0.69

(0.3to1.58)

709

(1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

38per1000

26per1000

(11to60)


38per1000

26per1000

(11to60)

Any

STIs

incidence

amongFSWsat6-month

Studypopulation

RR0.57

(0.34to0.96)

709

(1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

103per1000

59per1000

(35to99)


103per1000

59per1000

(35to99)

Consistentcondom

use

byFSWsat6-month

Studypopulation

RR1.14

(1.07to1.21)

804

(2studies)

⊕©

©©

verylow

6,7,8,9,10

767per1000

867per1000

(805

to920)


789per1000

892per1000

(828

to947)

Druguse

amongFSWs

at6-month

Studypopulation

RR0.65

(0.36to1.16)

772

(2studies)

⊕⊕

©©

low

3,6,7,10,11

141per1000

93per1000

(65to134)



countries (Review)


236per1000

156per1000

(109

to224)

Alcohol

use

among

FSWsat6-month

Studypopulation

RR0.68

(0.46to1.00)

772

(2studies)

⊕⊕

©©

low

3,6,7,10,11

144per1000

98per1000

(66to144)


128per1000

87per1000

(59to128)

*The

basisfortheassumedrisk

(e.g.themediancontrolgroupriskacross

studies)isprovided

infootnotes.Thecorrespondingrisk(and

its95%confidence

interval)isbasedon

the

assumedriskinthecomparison

groupandtherelativeeffectoftheintervention(andits95%CI).

CI:Confidenceinterval;RR:Riskratio;

GRADEWorkingGroupgradesofevidence

Highquality:Furtherresearchisveryunlikelytochangeourconfidenceintheestimateofeffect.

Moderatequality:Furtherresearchislikelytohaveanimportantimpactonourconfidenceintheestimateofeffectandmaychangetheestimate.

Lowquality:Furtherresearchisverylikelytohaveanimportantimpactonourconfidenceintheestimateofeffectandislikelytochangetheestimate.

Verylowquality:Weareveryuncertainabouttheestimate.

1Noseriouslimitations:Allocationconcealmentwasjudgedtobeat‘‘lowriskofbias’’inthistrial.

2Singlestudy.

3Noseriousindirectness:Resultislikelytobereliable.

4Seriousimprecision:The95%CIofestimatecrossesthelineofno

effect.

5Biaswasjudgedtobeat‘‘lowrisk’’inthistrial.

6Seriouslimitations:Allocationconcealmentwasjudged

tobe

at‘‘lowriskofbias’’inPatterson

studyand‘‘unclear’’inWechsberg

study.

7Noseriousinconsistency:Heterogeneitywaslow.

8Serious

indirectness:Thereisconsiderablevariabilityintheeffectofcontrolwhich

makes

extrapolationofresulttoothersetting

unreliable.

9Veryseriousimprecision:The95%CIofpooledestimateincludesappreciablebenefittocontrolgroupoverinterventiongroup.

10Biaswasjudgedtobeat‘‘high

risk’’atW

echsbergtrial.

11Noseriousimprecision:The95%CIofestimateincludesappreciablebenefittointerventiongroupovercontrolgroup.


countries (Review)


B A C K G R O U N D

Description of the condition

In 2008, the Joint United Nations Programme on HIV/AIDS

(UNAIDS) reported that an estimated 2.7 million people (range

2.4-3.0 million) became newly infected with HIV, bringing the

total number of people living with HIV to 33.4 million (range

31.1-35.8 million) worldwide. Overall, 2.0 million (range 1.7-

2.4 million) AIDS-related deaths occurred throughout the world

(UNAIDS 2009). Most newly acquired infections occurred in low-

and middle-income countries, and sub-Saharan Africa remains the

most heavily affected region, mainly as a result of heterosexual

transmission (WHO 2009).

The high rates of HIV infection among sex workers as compared

to most other population groups has affected rates of heterosex-

ual transmission of HIV particularly in low- and middle-income

countries (UNAIDS 2009). The term “sex workers” constitutes a

meaningful single population for epidemiological purposes, but it

encompasses female sex workers, male sex workers, and male-to-

female transgender sex workers in a wide variety of settings (e.g.

brothels, massage parlours, informal settings, and on the street).

Such individuals are often the victims of discrimination (which can

sometimes lead to violence), trafficking, legal persecution and soci-

etal ambivalence (UNAIDS 2002), are socially stigmatised (Padilla

2008) and often have a precarious lifestyle due to economic dif-

ficulties (Gu 2008). Sex workers and their clients, who typically

include people such as truck drivers, security workers, sailors, dock

workers, police, migrant workers and businessmen, are at high risk

for HIV exposure. Moreover, previous research suggests that the

numbers of their clients per night can have an effect on heterosex-

ual transmission of HIV particularly in low- and middle-income

countries (Ruxrungtham 2004). The lifetime probability of a sex

worker becoming infected with HIV is high due to multiple risk

factors, including a large number and high turnover of partners,

low levels of condom use, a high prevalence of STIs, and unsafe

practices such as douching and use of inappropriate lubricants

(UNAIDS 2002).

The number of countries reporting on indicators relating to sex

workers significantly increased between 2005 and 2009. In sub-

Saharan Africa, HIV prevalence among sex workers ranged from

zero in Comoros and Sierra Leone to 49% in Guinea-Bissau. Seven

African countries (Benin, Burundi, Cameroon, Ghana, Guinea-

Bissau, Mali and Nigeria) have reported that more than 30% of all

sex workers had HIV (UNAIDS 2009; Gomes do Espirito Santo

2005). Recently, about one quarter of all sex workers (26%) in

Lesotho were reported to have had a symptomatic STI (Khobotlo

2009). In Swaziland, transmission during heterosexual contact

(including sex within stable couples, casual sex and sex work)

is estimated to account for 94% of incident infections (Mngadi

2009). In 2008, Lowndes et al reported that between 13% and

29% of men in West Africa may have paid for sex in the previous

year (Lowndes 2008). Surveys in Kenya (Gelmon 2009), Uganda

(Wabwire-Mangen 2009) and Rwanda (Asiimwe 2009) suggested

that sex workers and their clients accounted for an estimated 14%,

10% and 46% of incident HIV infections, respectively. Results

from an RCT in Jamaica have shown that 25% of AIDS patients

had had exposure to sex workers (Weir 2008).

Unprotected commercial sex is the most important risk factor for

the spread of HIV in several parts of Asia. In Vietnam, 33% of

male sex workers recruited from more than 70 sites in Ho Chi

Minh City tested positive for HIV (Nguyen 2008). The preva-

lence of HIV among male sex workers is more than twice that of

their female counterparts and is currently rising in Thailand and

Indonesia (UNAIDS 2009). Sex work is common among male-to-

female transgender people in Pakistan (Khan 2008) and high HIV

prevalence has also been reported among transgender sex workers

in studies conducted in Phnom Penh, Cambodia (22%) in 2003

(Girault 2004) and Jakarta, Indonesia (59%) (Pisani 2004). In In-

dia, high HIV and STI prevalence were found among street-based

FSWs (30% and 27%, respectively) and among those who work in

brothels (34% and 13%, respectively) (Buzdugan 2010). Clients

of commercial sex workers are also at high risk of transmission.

Given their high mobility and frequent sexual encounters with sex

workers in other parts of Indonesia, these men could be the agents

of the rapid spread of the HIV virus throughout Indonesia (Fajans

1995).

In the Middle East and North Africa, surveys of bar-based sex

workers in Djibouti have found HIV prevalence as high as 26%,

while in Yemen it has ranged from 1% to 7% (UNAIDS 2009); in

Egypt, 1% (Shawky 2009); while in Algeria, Morocco and Yemen,

the figures are 4%, 2% and 2% of their national populations re-

spectively (UNAIDS 2009). The percentage of sex workers who

report having used a condom during the most recent episode of in-

tercourse ranged from 44% in Jordan to 61% in Yemen (UNAIDS

2009).

Surveys in the Caribbean undertaken in 2005 identified high in-

fection rates in Guyana and Jamaica (27% and 9%, respectively)

(UNAIDS 2009). In Latin America, particularly in Peru, 44% of

men reported having had sex with a sex worker (Caceres 2009).

Surveys in Guatemala and El Salvador have determined HIV

prevalence among FSWs to be 4% and 3%, respectively (Soto

2007).

Description of the intervention

Several successful interventions have been reported among sex

workers and their clients to reduce the heterosexual spread of HIV,

including interventions to change behavior, promote the use of

condoms, improve condom availability, introduce voluntary HIV

counselling and testing (VCT), and educate about sexual health

and the effective management of STIs. The effectiveness of be-

havioral interventions could be increased by aiming for important

goals, such as a delay in the onset of first intercourse, a reduction in

number of sexual partners, an increase in condom use, that could


countries (Review)


be achieved using multi-level approaches (e.g. targeting couples,

families, social and sexual networks, institutions, and entire com-

munities) with both HIV-uninfected and -infected populations

(Coates 2008).

Female condom interventions may help empower women to pro-

tect themselves when they are unable to avoid sexual relations with

HIV-infected partners or cannot persuade their partners to use a

condom.

Management of STIs was based on clinical diagnosis and serologic

tests for herpes simplex virus type 2 (HSV-2) (Kamali 2003).

Peer education enlists members of a specific group to encourage ef-

fective behavioral change among their peers (Cornish 2009, Steen

2009). Its initial goal is usually to modify individuals’ knowledge,

attitudes and beliefs to bring about healthy behavior.

Structural interventions which mobilize sex workers to engage

in HIV prevention may address other factors, such as economic

security, reducing stigmatization, community-based organizing

(Ghose 2008) and rights-based advocacy (Wolffers 2003).

How the intervention might work

A decline in HIV prevalence in Cambodia occurred when rates

of consistent condom use during commercial sex rose from 53%

in 1997 to 96% in 2003 (Gorbach 2006); and in China, con-

sistent use of condoms resulted in a 70% reduction in HIV in-

fections (Wang 2009). Other countries with epidemics driven by

the sex work industry, such as Kenya and Uganda (Morris 2006),

Chile (Barrientos 2007) and India (Basu 2004), have experienced

declines in HIV prevalence when sex workers and their clients

used condoms consistently. Increasing condom use, however, de-

pends very much on condom availability (Bradley 2010). There-

fore, making condoms available in rooms where commercial sex

occurs is the most effective strategy to increase condom use (Egger

2000). In addition, consistent condom use was significantly greater

among males who perceived that some or all of the members of

their male social networks used condoms consistently (Barrington

2009). These results suggest that interventions to increase con-

dom use are more effective when implemented in conjunction

with social, network- and/or community-targeted interventions

that change the environment in which decisions about safe sex

behavior are taken.

Treatment for STIs was found to be more effective in reducing HIV

and STI transmission when combined with the consistent and

correct use of condoms (Laga 1994; Ghys 2001), suggesting that

behavioral interventions for primary prevention may also serve to

enhance the effectiveness of secondary prevention activities.

Peer education has resulted in substantial increases in STI and

HIV knowledge and use of condoms, and in the reduction of HIV

and STIs (Ford 2000).

Why it is important to do this review

Research in some countries has suggested that prevention projects

resulting in increased condom use during paid sex could signifi-

cantly reduce HIV transmission.

Various intervention strategies have been adopted to reduce HIV

transmission among sex workers and their clients in low- and mid-

dle-income countries, and their results suggest that these interven-

tions may play a significant role in preventing the spread of HIV;

however, the effectiveness of these strategies has not been assessed

rigorously through meta-analysis of randomised controlled trials.

Given the potential effectiveness and low cost of these interven-

tions, it is important to conduct a systematic review of their im-

plementation in low- and middle income countries.

O B J E C T I V E S

1. To evaluate the studies conducted on behavioral interventions

for reducing the transmission of HIV and STIs among sex work-

ers (male, female, and transgender) and their clients in low- and

middle-income countries.

2. To assess whether behavioral interventions such as condom use

and behavior modification are effective in reducing the HIV/STI

incidence and prevalence when the interventions are delivered in

sex worker settings.

3. To investigate behavioral interventions to reduce risk of HIV/

STIs transmission among sexual workers that have been tested in

RCTs.

M E T H O D S

Criteria for considering studies for this review

Types of studies

Randomized or quasi-randomized trials were considered for in-

clusion if they described behavioral interventions on any one of

the outcome measures specified below, occurring in sex worker

settings in low- and middle-income countries. Studies relevant to

outcome measures, whether primary or secondary outcomes were

included. The unit of randomization could be individual or clus-

ter level. Studies that were not eligible, for example studies that

did not randomly assign the sex workers in to the behavioral inter-

ventions and control groups, targeted not specifically sex workers

population, assessed other interventions rather than behavioral in-

terventions, or studies those not related to specified primary and

secondary outcome measures were excluded.

Low- and middle-income countries were defined based on World

Bank criteria and there are currently about 125 of these low- and

middle-income countries with populations of over one million;


countries (Review)


in 1997, their combined population was more than 4.89 billion

(World Bank).

Types of participants

Sex workers and their clients regardless of age, ethnicity, gender

identity, language and nationality are the target population:

Sex workers are defined as female, male and transgender, whether

adults or young people (including adolescents), who receive

money, goods or protection directly or indirectly in exchange for

indoor or outdoor sexual services, such as in a brothel, street or

home, either regularly or occasionally, and who may or may not

consciously define those activities as income-generating“.

Clients of sex workers were defined as female, male and transgender

adults or young people who give money, goods or protection in

exchange for sexual services to sex workers (defined above), either

regularly or occasionally.

Types of interventions

Intervention: Behavioral interventions, including social or policy

interventions to reduce the transmission of HIV in sex worker

settings. Behavioral interventions were defined in this systematic

review as interventions that aim to change not only individual be-

havior to prevent HIV infection but also peer and social norms,

including strategies such as community mobilization and struc-

tural and resource support, and through administrative or legal

decisions, such as promoting condom availability.

Control: Studies having no interventions or any other behavioral

interventions for preventing the spread of HIV among sex workers

and their clients.

Types of outcome measures

Primary outcomes

Change in biological variables for HIV/STI prevention among sex

workers and their clients, including:

a) HIV incidence

b) HIV prevalence

c) STI incidence

d) STI prevalence

Secondary outcomes

Change in self-reported or observed behavior, including:

a) Condom use (male/female)

b) HIV/STI-related knowledge

c) Types of sexual practice, such as vaginal, oral, or anal

d) Sexual risk behavior, such as having unprotected sex and mul-

tiple partners

e) Frequency of sexual encounters

f ) STIs treatment-seeking behavior

g) Psychosocial barriers to condom use

h) Drug and alcohol risk behavior

Search methods for identification of studies

Many different sources of published and unpublished research lit-

erature were searched for studies of behavioral interventions to

reduce HIV infection among sex workers and their clients. Re-

porting strategies of the effect of these interventions might not

be uniform, and there may be much grey literature and many lo-

cal publications reporting on this issue. The following databases

and conference proceedings were searched using a comprehensive

search strategy without restrictions on the language or country or

publication status of relevant trials. The date range of the search

was 01 January 1980 to the search date. The search was conducted

on 28 September 2010.

1) Electronic databases

Relevant databases were identified in consultation with the HIV/

AIDS Review Group Coordinator and the Trials Search Coordi-

nator for the Cochrane HIV/AIDS Group’s Trials Register, experts

in HIV/AIDS research and service projects working in low- and

middle-income countries, policy makers and healthcare adminis-

trators. This list served as the key document for the extraction of

data from electronic databases.

Cochrane Central Register of Controlled Trials (CENTRAL), the

Cochrane HIV/AIDS group specialized register, the Cochrane

Database of Systematic Reviews, MEDLINE, PsycINFO, Socio-

logical Abstracts, CINAHL, Dissertation Abstract International

(DAI), EMBASE, LILACS, BIOSIS, SciSearch, INDMED, Pro-

quest, and various South Asian abstracting databases were included

in the database list. The publication sites of the World Health Or-

ganization, the US Centers for Disease Control and Prevention,

and other international research and non-governmental organiza-

tions also appeared in the database list.

An extensive search strategy string was developed in consulta-

tion with the Trial Search Coordinator of the HIV/AIDS Review

Group. All possible keywords were included in the string to enable

an exhaustive electronic literature search. For further details see

the Appendices.

2) Hand searching

A hand search of key HIV/AIDS research journals was conducted

because many publications from low- and middle-income coun-

tries might not have appeared in electronic databases.

3) Personal communication

Key personnel and organizations working in HIV/AIDS interven-

tion programs in low- and middle-income countries were con-

tacted for published and unpublished references and data.

4) Conferences proceedings

Conference proceedings were searched for relevant abstracts. Con-

ferences included the Conference on Retroviruses and Oppor-

tunistic Infections (CROI), 1996-2010; International AIDS Con-

ference (IAC), 1985-2010; and International AIDS Society Con-

ference on HIV Pathogenesis, Treatment and Prevention (IAS),

2001-2009.

5) Cross-references

The bibliographies of studies identified by the procedures de-


countries (Review)


scribed above were further scrutinized to locate additional stud-

ies. The search strategy was iterative, in that bibliographies of the

included studies were searched for additional references.

Data collection and analysis

The methodology for data collection and analysis was based on

the Cochrane Handbook of Systematic Reviews of Interventions

(Higgins 2009).

Selection of studies

Using the inclusion criteria described above, two of the review

authors (WW and EO) independently assessed the potential stud-

ies that were identified as a result of the search strategy. Scrutiny

for inclusion was based on the type of study, type of participants,

type of interventions, and outcome measures. Any disagreements

were resolved through discussion and a third reviewer (RM) was

consulted. For all excluded studies a summary statement about

the exclusion reasons was made. Four authors (KS, RM, AK and

NH) who are experts were informed of the included studies and

conducted the data extraction independently.

Data extraction and management

Data was extracted independently using data collection forms. For

eligible studies, two review authors (WW and EO) extracted data

on trial characteristics including methods, participants, interven-

tions, and outcomes. Discrepancies were resolved through discus-

sion or by consulting with another review author (RM). Data

were entered into the Review Manager software (Revman 2008)

and checked for accuracy. When information regarding any of the

above was unclear, contact with the authors of the original reports

was attempted in order to elicit further details.

Assessment of risk of bias in included studies

Two review authors (WW and EO) independently assessed the

risk of bias for each study using the criteria outlined in the

Cochrane Handbook for Systematic Reviews of Interventions

(Higgins 2009). Discrepancies were resolved by discussion or by

involving an additional assessor (RM). Standard guidance was fol-

lowed for the assessment of the adequacy of methods to reduce

the risk of bias across seven domains:

1) Sequence generation (checking for possible selection bias)

For each included study the method used to generate the allocation

sequence was described in sufficient detail to allow an assessment

to be made of whether it would have produced comparable groups.

2) Allocation concealment (checking for possible selection bias)

For each included study the method used to conceal the allocation

sequence was described and a judgment made as to whether the

intervention allocation could have been foreseen in advance of or

during recruitment, or changed after assignment.

3) Blinding (checking for possible performance bias)

A description was provided of the methods used, if any, to blind

study participants and personnel from knowledge of which inter-

vention a participant received. Studies were judged to be at low

risk of bias if they were blind, or if the lack of blinding could

not have affected the results. Blinding was assessed separately for

different outcomes or classes of outcomes.

4) Incomplete outcome data (checking for possible attrition bias

through withdrawals, dropouts, protocol deviations)

For each included study, and for each outcome or class of outcome,

completeness of the data was assessed including checking attrition

and exclusions was noted, along with the numbers included in

the analysis at each stage (compared with the total number of

randomized participants), reasons for attrition or exclusions where

reported, and whether missing data were balanced across groups

or were related to outcomes. Where sufficient information was

reported, or was supplied by the trial authors, missing data was

included in the analyses.

5) Selective reporting bias

For each included study the possibility of selective outcome re-

porting bias was investigated and a conclusion reported.

6) Other sources of bias

For each included study all other possible sources of bias, includ-

ing study design and early trial cessation due to data-dependent

processes or extreme baseline imbalance were reported.

7) Overall risk of bias

Explicit judgments were made about whether studies were at high

risk of bias, according to the criteria given in the Handbook

(Higgins 2009). With reference to (1) to (6) above, the likely mag-

nitude and direction of the bias and its likely impact on the find-

ings was assessed and reported.

Measures of treatment effect

1) Dichotomous data

For dichotomous data, results are presented as summary risk ratios

(RR) with a 95% Confidence Interval (CI).

2) Continuous data

For continuous data, the mean difference (MD) was used if out-

comes were measured in the same way among trials. Standardized

mean differences were used to combine trials that measured the

same outcome with different methods.

Unit of analysis issues

All RCTs, cluster-RCTs, and quasi-RCTs were identified.

Dealing with missing data

For included trials, attrition levels were noted and the impact of

including trials with high levels of missing data in the overall as-

sessment of the treatment effect was checked through a sensitivity

analysis. For all outcomes analyses were conducted on an inten-

tion-to-treat basis. The denominator for each outcome in each

trial was the number randomized minus any participants whose

outcomes were known to be missing.

Assessment of heterogeneity

Heterogeneity amongst trials was tested using an I2 statistic in

each analysis. A value of I2=0% indicates no observed heterogene-

ity. Where substantial heterogeneity (I2 >50%) was identified, a

prespecified subgroup analysis was conducted. Subgroup analyses

were planned based on the extent of control for selection bias.

However, subgroup analysis considered the effects of different du-


countries (Review)


rations of follow-up or different routes that were not included due

to a lack of available data.

Assessment of reporting biases

Where reporting bias was suspected (see ”Selective reporting bias“

above), attempts were made to contact study authors, asking them

to provide missing outcome data. Where this was not possible and

the missing data were thought to introduce serious bias impact

of including such studies in the overall assessment of results was

explored through sensitivity analysis.

Data synthesis

A meta-analysis was conducted using the Review Manager soft-

ware (Revman 2008). Fixed-effect inverse variance meta-analysis

was used for combining data where trials were examining the same

intervention and the trials’ populations and methods were judged

sufficiently similar. Where heterogeneity between trials’ treatment

effects was suspected, random-effect meta-analysis was used. The

criteria of the Grading of Recommendations Assessment, Devel-

opment and Evaluation (GRADE) to evaluate the quality of the

evidence by outcome was performed (Guyatt 2008).

Subgroup analysis and investigation of heterogeneity

Subgroup analysis was conducted for the primary outcomes of

HIV incidence, HIV prevalence, STI incidence and STI preva-

lence. For the fixed-effect meta-analysis, a planned subgroup anal-

ysis was conducted, classifying whole trials by interaction tests as

described by Deeks (Deeks 2001).

Sensitivity analysis

Sensitivity analyses were not performed.

Results

R E S U L T S

Description of studies

See: Characteristics of included studies; Characteristics of excluded

studies.

See Characteristics of included studies; Characteristics of excluded

studies

A total of 2667 citations published between 1980 and 2010 were

obtained. Of these, 53 potential studies were identified for consid-

eration. Overall, 13 trials with 8,698 participants were analysed.

Of these, two studies in Madagascar (Feldblum 2005, Hoke 2007)

used the same population, as did two studies in Mexico (Patterson

2006, Patterson 2008). Two studies using previously published

data (Hoke 2007, Patterson 2008) were included because they

included outcomes, which were not previously reported. The re-

maining 40 not eligible studies were excluded (Figure 1).


countries (Review)


Figure 1. Study flow diagram.


countries (Review)


Seven RCTs (Ghys 2001, Ray 2001, Feldblum 2005-Hoke 2007,

Patterson 2006- Patterson 2008, Wechsberg 2006, Markosyan

2010 and Sherman 2010), two cluster-RCTs (Fontanet 1998

and Gutierrez 2010) and four quasi-RCTs (Basu 2004, Li 2006,

Swendeman 2009 and Chiao 2009) met the inclusion criteria.

The process of randomizations in the two cluster-RCTs (Fontanet

1998 and Gutierrez 2010) was done by using clusters rather than

individuals. However, we nevertheless included them because in-

dividual data was used as the unit of data analysis.

Seven out of 13 trials were conducted in Asia (four trials in India

(Basu 2004, Swendeman 2009, Gutierrez 2010, and Sherman

2010), one in Thailand (Fontanet 1998), and one in China (Li

2006), one in the Philippines (Chiao 2009), and one in Armenia

(Markosyan 2010). Four trials were evaluated in Africa (one in

Madagascar (Feldblum 2005-Hoke 2007), one in Cote d’Ivore

(Ghys 2001), one in Zimbabwe (Ray 2001), and one in South

Africa (Wechsberg 2006). The remaining one trial was conducted

in Latin America (Patterson 2006-Patterson 2008).

Primary outcomes were identified in seven trials (Patterson 2008,

Gutierrez 2010, Feldblum 2005-Hoke 2007, Ray 2001, Fontanet

1998, Ghys 2001, and Li 2006). Of these, three trials measured

HIV incidence (Patterson 2008, Ghys 2001, and Ray 2001),

two trials measured the incidence of STIs ( Patterson 2008 and

Fontanet 1998) and four trials measured the prevalence of STIs

(Gutierrez 2010, Feldblum 2005-Hoke 2007, Ghys 2001 and Li

2006). There were no trials that measured HIV prevalence. Sec-

ondary outcomes were identified in all studies.

Although most trials provided a follow-up period of at least 6

months to assess outcomes (Ghys 2001, Feldblum 2005, Patterson

2008, Markosyan 2010, Sherman 2010) the duration of the fol-

low-up varied across studies. The shortest was a 3-month trial that

occurred in Thailand (Fontanet 1998) whereas the longest follow-

up was up to three years in a trial conducted in India (Gutierrez

2010).

Interventions

1) Social cognitive theory

Social cognitive theory identifies human behavior as a personal,

behavior, and the environment factors interact (Bandura 1986).

Person and environment interaction involves human beliefs and

cognitive competencies. Social influences and structures within the

environment were developed and modified by this interaction. The

third interaction, between the environment and behavior, involves

a person’s behavior determining the aspects of their environment.

The behavior also being modified by that environment. Despite

behavior may varies from situation to situation, behavior is not

controlled by situation itself rather than person. It means different

responses can come from different people or from the same person

at different times (Jones 1989). Therefore, this theory is useful to

understand and predict both individual and group behavior, and

identify strategies in which behavior can be changed.

2) Community empowerment

Community empowerment (also called community organization,

-mobilization, or -action research) is a strategy used in public

health, behavioral sciences, and social movements, that can mo-

bilize people for recognizing structural barriers to improve health,

and empower them to change these barriers. This method is distin-

guished from behavioral change, because it addresses behaviours

through the active involvement of community members to de-

sign, execute, and evaluate their empowerment projects. Com-

munity involvement in the evaluation directly contributes to ca-

pacity-building and increased awareness in the target community

(Stevens 1998).

3) Microenterprise

Microenterprise intervention empowers individuals through mi-

cro-finance, teaching basic literacy, and training in specific eco-

nomic livelihoods (e.g., tailoring, computer literacy). Microenter-

prise programs (e.g., micro-finance, micro-credit) have been re-

ported effective in increasing women’s economic well-being, in

reducing interpersonal violence, in increasing reproductive health

decision-making power, and broaden their role in household de-

cision making (Kim 2007, Hashemi 1996, Schuler 1997).

4) Peer education

This intervention provides training and support to community

members and is usually conducted to effect change at both the

individual and the group level, with the aim to modify the person’s

knowledge, attitudes, beliefs, or behaviours. This strategy may

effect change by modifying norms that contribute to individual

change as well as changes in programs and policies.

5) Manager training

Manager training focuses on providing specialized, HIV-based in-

formation and education, reinforcing positive STI prevention be-

havior that addresses the managerial class, to increase their influ-

ence through providing positive reinforcement of their employee’s

safer sex practices.

6) Promotion of female condom use

This intervention promotes the use of female condoms properly,

enhancing condom self-efficacy and negotiation skills and knowl-

edge of HIV/STI transmission, HIV/STI risk assessment, and safer

sex options. Early female condom introduction efforts were tar-

geted to FSWs, because they are at high risk for HIV and other

STIs and have an obvious need for a female- initiated method of

protection. Introduction of the female condom has been more suc-

cessful in shorter-term programs, particularly among sex workers

and members of women’s groups (PATH and UNFPA 2006).

7) STI screening

The systematic strategy of providing STI antibody testing to a

population of apparently healthy people without signs and symp-

toms of that disease for the purpose of detecting the number of

people (or blood samples) infected with an STI. The primary aim


countries (Review)


is not to diagnose STI in a specific person. Screening programs

might direct additional resources to groups with the most risky

sexual behavior -e.g., sex workers, their clients, migrant workers,

and those attending STI clinics (WHO 2006).

8) VCT

VCT provides individuals with knowledge of their HIV infection

status, risk assessment, and promotes risk reduction. This inter-

vention may effective in reducing the risk of further transmission

of HIV by promoting safer sex (Nelson 2007).

Risk of bias in included studies

Assessment of the risk of bias in the included studies is summarized

in Figure 2 and Figure 3.


countries (Review)


Figure 2. Risk of bias summary: review authors’ judgements about each risk of bias item for each included

study.


countries (Review)


Figure 3. Risk of bias graph: review authors’ judgements about each risk of bias item presented as

percentages across all included studies.

Allocation

Generation of the randomisation sequence was judged to be at low

risk of bias for all trials. Allocation concealment was judged to be

at low risk of bias in 13 trials and unclear in four trials (Ghys 2001,

Li 2006, Wechsberg 2006, Gutierrez 2010). Descriptions which

included the following details were accepted as being adequate for

concealment: a random number table (Fontanet 1998, Basu 2004

and Swendeman 2009), the flip of a coin (Markosyan 2010), a

computer generated random allocation list (Feldblum 2005-Hoke

2007, Patterson 2006-Patterson 2008), opaque sealed envelopes

(Sherman 2010), numbered sealed envelopes (Ray 2001), and

numbers placed in a box (Chiao 2009).

Blinding

Of the included studies only three trials (Feldblum 2005-Hoke

2007, Markosyan 2010, Sherman 2010) were judged to be at low

risk of bias due to adequate blinding.

Incomplete outcome data

The proportion of participants in each intervention arm for whom

an outcome was not available was reported. Five trials (Basu

2004, Feldblum 2005-Hoke 2007, Patterson 2008, Chiao 2009,

Markosyan 2010) were judged to be at high risk of bias due to

moderate missing data that had the potential to alter the results,

while most of the trials did not address the reasons for the incom-

plete outcome data.

Selective reporting

Eleven trials were assessed as being free of selective reporting bias.

The remaining two studies (Gutierrez 2010, Sherman 2010) did

not present the study outcome completely and were judged to be

at high risk of bias.

Other potential sources of bias

Three studies (Wechsberg 2006, Li 2006, Chiao 2009) were

judged to be at high risk of bias due to an imbalance of base-

line characteristics between the intervention group and the control

group.

Effects of interventions

See: Summary of findings for the main comparison Social

cognitive theory compared to standard counseling for promotion

of condom use; Summary of findings 2 Social cognitive theory

for promotion of condom use compared to no intervention;

Summary of findings 3 Promotion of female and male condom

compared to promotion of male condom; Summary of findings

4 VCT compared to standard care of STI for increasing condom

use

Comparison 1. Social cognitive theory versus standard care

(Analysis 1)

This comparison included two trials (Wechsberg 2006 and

Patterson 2006-Patterson 2008) comprising 1017 FSWs.

1. Primary outcomes

One trial conducted in Mexico (Patterson 2008) reported HIV

incidence. After a 6-month follow-up period, the risk ratio of HIV

incidence was 0.12 (95% CI 0.01 to 2.22; Analysis 1.1) indicating

a reduction in HIV incidence. Patterson 2008 also reported STI


countries (Review)


incidence and showed a significant decrease in the cumulative

incidence of any STIs among the intervention group (RR 0.57,

95% CI 0.34 to 0.96; Analysis 1.2) compared to the control group.

2. Secondary outcomes

Two trials (Wechsberg 2006, Patterson 2008) reported consistent

condom use and one trial (Patterson 2006) reported the mean

frequency of condom use. Patterson 2006 and Patterson 2008 re-

ported sexual activity outcomes (e.g. total protected sex, had vagi-

nal sex, had anal sex, and had oral sex), and two trials (Patterson

2006, Wechsberg 2006) reported drug and alcohol use. However,

heterogeneity of study designs prevented any meta-analyses of con-

sistent condom use and alcohol risk behavior variables.

There were no differences in the increase in consistent condom

use (Analysis 1.3; Analysis 1.4; Analysis 1.5), in changes in FSWs

sexual activities (Analysis 1.6), or in injecting drug use and alco-

hol risk behavior (Analysis 1.7; Analysis 1.8), indicating that the

intervention had no impact. However, this intervention did show

a decrease in drug use among FSW (RR 0.65, 95% CI 0.36 to

1.16, p for heterogeneity=0.15, I2=52%; Analysis 1.9).

Comparison 2. Social cognitive theory versus no intervention

(Analysis 2)

Only one trial of 120 FSWs examined this comparison (

Markosyan 2010).

1. Primary outcomes

None reported.


Consistent condom use, HIV knowledge, and psychosocial medi-

ators of HIV prevention behavior were assessed. However, social

cognitive theory was found not to increase reported condom use,

including barriers to condom use and condom use self-efficacy, or

HIV-related knowledge (Analysis 2.1; Analysis 2.2; Analysis 2.3)

Comparison 3. Community empowerment versus standard

care (Analysis 3)

There were three trials assessing this comparison (Basu 2004,

Swendeman 2009, and Gutierrez 2010) involving a total of 3,858

FSWs.

1. Primary outcomes

One trial reported the prevalence of STIs (Gutierrez 2010). How-

ever, after a follow-up assessment the risk ratio of syphilis and

HSV-2 were 1.26 and 1.38, respectively (Analysis 3.1), indicating

that there was no effect in reducing the prevalence of STIs. More-

over, none of the studies reported on the other primary outcomes.


Both trials (Basu 2004, Gutierrez 2010) reported on consistent

condom use with clients and HIV-related knowledge. One trial

(Gutierrez 2010) reported consistent condom use with regular

partners, one trial (Basu 2004) reported changes in 100% condom

use and changes in any condom use, condom availability, type of

sexual practices, and drug and alcohol use. One trial (Gutierrez

2010) reported HIV testing.

A trial in India (Gutierrez 2010) was successful at reduction in

non-condom use with regular partners over a 3-year follow-up

period (RR 0.55, 95% CI 0.48 to 0.64; Analysis 3.2). However,

a community empowerment intervention showed no increase in

condom use with clients (Analysis 3.3; Analysis 3.4; Analysis 3.5),

HIV-related knowledge (Swendeman 2009; Analysis 3.6), HIV

testing (Gutierrez 2010; Analysis 3.7), risk management skills and

protective factors (Swendeman 2009; Analysis 3.8).

Comparison 4. Microenterprise plus education intervention

versus education alone (Analysis 4)

There was only one trial (Sherman 2010) consisting of 128 FSWs,

which assessed this comparison.

1. Primary outcomes

None.


There was no increase in condom use after a 6 month follow-up

period (Analysis 4.1). This trial reported that the mean number

of sex partners (MD -6.90, 95% CI -10,02 to -3,78) and number

of exchange sex partners (MD -2.00, 95% CI -3.01 to -0.99) was

smaller for those who experienced community empowerment plus

education rather than education alone (Analysis 4.2).

Comparison 5. Peer education versus standard care (Analysis

5)

There was only one trial (Chiao 2009) of 980 FSWs that assessed

the effect of peer education compared with standard care.

1. Primary outcomes

This study did not report primary outcomes.


This intervention had no effect on self-reported consistent con-

dom use (Analysis 5.1), HIV-related knowledge (Analysis 5.2) or

perceived need for HIV testing (Analysis 5.3). However, a peer-

led intervention was effective and increased self-regulatory factors,

such as perceived control over risk of HIV acquisition (RR 0.74,

95% 0.84 to 0.99; Analysis 5.4) and perceived severity (Analysis

5.5), whether in relation to the chance of acquiring HIV (MD -

1.47, 95% CI -1.83 to -1.11) or worry about acquiring HIV (MD

-1.02, 95% -1.50 to -1.00).

Comparison 6. Peer education plus clinic-based counselling

versus peer education only (Analysis 6)

There was one trial (Feldblum 2005, Hoke 2007) involving 1000

FSWs.

1. Primary outcomes

The addition of clinic-based counselling to a peer education in-

tervention decreased the prevalence of STIs (Analysis 6.1), partic-

ularly Chlamydia at 6-month follow-up (RR 0.70, 95% CI 0.50

to 0.97) and any STIs at 6-month follow-up (RR 0.78, 95% CI

0.65 to 0.93), but the intervention was found to have no effect in

reducing the prevalence of all types of STIs after 12 months’ and

18 months’ follow-up. Other primary outcomes were not assessed

by this intervention.


Feldblum 2005 reported that FSWs who were administered peer

education plus clinic-based counselling were less likely to report

overall male condom use, whether with clients (Analysis 6.2) or


countries (Review)


non-paying partners (Analysis 6.3) compared with FSWs assigned

to the peer education intervention only. In addition, Hoke 2007

reported there was no increase in the use of female condoms after

12 months’ and 18 months’ follow-up (Analysis 6.4), indicating

that there was no effect of this intervention. Hoke 2007 also found

that FSWs assigned to the combined interventions were less likely

to use male or female condoms with clients (Analysis 6.5) and

non-paying partners (Analysis 6.6).

Comparison 7. Peer education plus manager training versus

standard care (Analysis 7)

One study (Chiao 2009) of 980 FSWs.

1. Primary outcomes

None of the primary outcomes were reported.

2. Other outcomes

This trial examined consistent condom use, HIV-related knowl-

edge, HIV testing and protective behavior, such as perceived con-

trol over AIDS and perceived AIDS severity. However, the inter-

vention had no effect in increasing condom use (Analysis 7.1) and

HIV-related knowledge (Analysis 7.2). At a 3-month follow up

assessment, manager training in addition to peer education in-

creased the likelihood of seeking HIV testing by 67% (95% CI

59% to 73%; Analysis 7.3).

Comparison 8. Manager training versus standard care (Anal-

ysis 8)

One study (Chiao 2009) of 980 FSWs.

1. Primary outcomes

None.


This trial reported on consistent condom use, HIV-related knowl-

edge, HIV testing and protective behavior, such as perceived AIDS

control (Analysis 7.4) and severity (Analysis 7.5). The interven-

tion had no effect on condom use (Analysis 8.1), HIV-related

knowledge (Analysis 8.2), and the likelihood of having an HIV

test (Analysis 8.3), and had not reduce the AIDS perceived control

(Analysis 8.4). However, the manager training intervention was

found to increase the FSWs perceived severity of acquiring HIV

(MD -0.43, 95% CI -0.80 to -0.06; Analysis 8.5).

Comparison 9. Promotion of female and male condom versus

promotion of male condom (Analysis 9)

Two studies (Fontanet 1998; Ray 2001) involving 653 FSWs.

1. Primary outcomes

Ray 2001 reported HIV incidence. After a 3-month follow-up,

the risk ratio of HIV incidence was 0.07 (95% CI 0.00 to

1.38; Analysis 9.1) indicating a large reduction in HIV inci-

dence. Fontanet 1998 reported a decrease in the incidence of STIs

(Analysis 9.2), in particular, the incidence of chlamydia and gon-

orrhoea (RR 0.71, 95% CI 0.52 to 0.98 and RR 0.63, 95% CI

0.45 to 0.88, respectively).


Two trials (Fontanet 1998, Ray 2001) reported on consistent fe-

male and male condom use. Meta-analyses showed a decrease in

male condom at 3-month follow-up (RR 0.83, 95% CI 0.65 to

1.05, P for heterogeneity=0.09, I2=65%; Analysis 9.3) and over-

all non-usage of the female (RR 0.12, 95% CI 0.09 to 0.17, p=

<0.001, P for heterogeneity=0.78, I2=0%; Analysis 9.4). A de-

crease was also seen in non-usage of the female condom at 24-

months follow-up (RR 0.15, 95% CI 0.09 to 0.28; Analysis 9.5),

but there was no effect for the male condom at 24-months follow-

up (RR 0.91, 95% CI 0.82 to 1.02; Analysis 9.6).

Comparison 10. Intensive STI screening versus basic STI

screening (Analysis 10)

One study (Ghys 2001) including 542 FSWs.

1. Primary outcomes

After a 6-month follow-up period, the risk ratio of HIV incidence

was 0.65 (95% CI 0.24 to 1.73; Analysis 10.1) indicating a reduc-

tion in HIV incidence. This trial also reported a positive effect of

intensive STI screening in reducing STI prevalence at 6-months

follow-up (Analysis 10.2) compared with basic STI screening. In

particular, it reduced the prevalence of gonorrhoea and chlamydia

trachomatis (RR 0.38, 95% CI 0.19 to 0.77, and RR 0.23, 95%

CI 0.07 to 0.79, respectively). Other primary outcomes were not

reported.


There was no increase in consistent condom use after a 6-month

follow-up assessment (RR 1.04, 95% CI 0.96 to 1.13; Analysis

10.3).

Comparison 11. Voluntary HIV counselling and testing versus

the standard care of STIs (Analysis 11)

We identified one trial (Li 2006) involving 400 FSWs.

1. Primary outcome

This trial reported a positive effect of VCT compared with stan-

dard STI management in reducing STI prevalence at 6-months

follow-up (Analysis 11.1). In particular, the intervention was ef-

fective in reducing the prevalence of genital warts (RR 0.22, 95%

CI 0.05 to 0.98). Other primary outcomes were not reported.


This trial assessed the effect of the intervention on consistent

condom use and HIV/STI related knowledge. However, after 6-

months follow-up, there was no increase in either consistent con-

dom use (RR 2.05, 95% CI 1.48 to 2.85; Analysis 11.2) or HIV/

STI related knowledge (RR 1.28, 95% CI 0.81 to 2.04; Analysis

11.3).


countries (Review)


AD

DI

TI

ON

AL

SU

MM

AR

YO

FF

IN

DI

NG

S[E

xpla

nati

on]

Socialcognitivetheoryforpromotionofcondomusecomparedtonointerventionforsexworkers



Intervention:Socialcognitivetheoryforprom

otionofcondom

use

Comparison:no

intervention

Outcomes

Illustrative

comparativerisks*

(95%CI)

Relativeeffect

(95%CI)

NoofParticipants

(studies)


(GRADE)

Com

ments

Assumed

risk

Correspondingrisk

nointervention

Socialcognitivetheory

forpromotionofcondom

use

Consistentcondom

use

byFSWswithclientsat

6-month

Studypopulation

RR0.89

(0.61to1.30)

98 (1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

560per1000

498per1000

(342

to728)


560per1000

498per1000

(342

to728)

Consistentcondom

use

byFSWswithclientsin

thepast7-day

Studypopulation

RR0.92

(0.68to1.24)

98 (1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

660per1000

607per1000

(449

to818)


660per1000

607per1000

(449

to818)


countries (Review)


http://www.thecochranelibrary.com/view/0/SummaryFindings.html

Consistentapplicationof

condombyFSWs

Studypopulation

RR0.65

(0.23to1.85)

98 (1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

160per1000

104per1000

(37to296)


160per1000

104per1000

(37to296)

HIV

knowledgeamong

FSWsat6-month

Studypopulation

RR2.08

(0.4to10.85)

98 (1study)

⊕©

©©

verylow

1,2,5,6,7

40per1000

83per1000

(16to434)


Psychosocialbarriersto

condom

use

among

FSWsat6-month

Studypopulation

RR0.83

(0.36to1.93)

98 (1study)

⊕⊕

⊕©

moderate

1,2,3,4,5

200per1000

166per1000

(72to386)


200per1000

166per1000

(72to386)

Psychosocialbarriersto

condom

use

self-effi-

cacy

amongFSWsat6-

month

Studypopulation

RR1.56

(0.85to2.89)

98 (1study)

⊕©

©©

verylow

1,2,5,6,7

240per1000

374per1000

(204

to694)



countries (Review)


240per1000

374per1000

(204

to694)

*The



studies)isprovided


its95%confidence

interval)isbasedon

the









1Noseriouslimitations:Allocationconcealmentwasjudgedtobeat’’low

riskofbias‘‘

inthistrial.

2Singlestudy.



effect.

5Biaswasjudgedtobeat’’low

risk‘‘inthistrial.

6Serious


makes

extrapolationofresulttoothersetting

unreliable.

7Veryseriousimprecision:The95%CIofestimateincludesappreciablebenefittocontrolgroup

overinterventiongroup.


countries (Review)


Promotion

offemaleand

malecondomcomparedtoprom

otionofmalecondomforsexworkers



Intervention:Promotionoffemaleandmalecondom

Comparison:prom

otionofmalecondom

Outcomes

Illustrative

comparativerisks*

(95%CI)

Relativeeffect

(95%CI)

NoofParticipants

(studies)


(GRADE)

Com

ments

Assumed

risk

Correspondingrisk

promotionofmalecon-

dom

Promotionoffemaleand

malecondom

HIV

incidence

among

FSWsat3-month

Studypopulation

RR0.07

(0to1.38)

149

(1study)

⊕⊕

©©

low

1,2,3,4,5

60per1000

4per1000

(0to83)

Lowriskpopulation

Chlamydia

incidence

amongFSWsat3-month

Studypopulation

RR0.71

(0.52to0.98)

504

(1study)

⊕⊕

⊕⊕

high1

,2,3,5,6

282per1000

200per1000

(147

to276)


282per1000

200per1000

(147

to276)

Gonorrhoea

incidence

amongFSWsat3-month

Studypopulation

RR0.63

(0.45to0.88)

504

(1study)

⊕⊕

⊕⊕

high1

,2,3,5,6


countries (Review)


275per1000

173per1000

(124

to242)


275per1000

173per1000

(124

to242)

Consistentmalecondom

useat3-month

Studypopulation

RR0.88

(0.83to0.92)

572

(2studies)

⊕⊕

⊕⊕

high3

,6,7,8,9

968per1000

852per1000

(803

to891)


913per1000

803per1000

(758

to840)

Consistentfemalecon-

domuse

byFSWsat3-

month

Studypopulation

RR0.12

(0.09to0.17)

564

(2studies)

⊕⊕

⊕⊕

high3

,6,7,9,10

968per1000

116per1000

(87to165)


913per1000

110per1000

(82to155)

Consistentfemalecon-

domusebyFSWsat24-

month

Studypopulation

RR0.15

(0.09to0.28)

128

(1study)

⊕⊕

⊕⊕

high1

,2,3,5,6

950per1000

143per1000

(86to266)

Lowriskpopulation

147per1000

22per1000

(13to41)


countries (Review)


Consistentmalecondom

useat24-month

Studypopulation

RR0.91

(0.82to1.02)

128

(1study)

⊕⊕

⊕©

moderate

1,2,3,5,11

950per1000

865per1000

(779

to969)

Lowriskpopulation

868per1000

790per1000

(712

to885)

*The



studies)isprovided


its95%confidence

interval)isbasedon

the










riskofbias‘‘

inthistrial.

2Singlestudy.


4Veryseriousimprecision:The95%CIofestimateiswideandcrossesthelineofno

effect.


risk‘‘inthistrial.



riskofbias‘‘

intwotrials.

8Noseriousinconsistency:Mayrepresentsubstantialheterogeneity(65%).


risk‘‘intwotrials.

10Noseriousinconsistency:Heterogeneity(0%)mightnotbeimportant.


effect.


countries (Review)


VCTcomparedtostandardcareofSTIforincreasingcondomuseforsexworkers


Settings:low-andmiddleincomecountries

Intervention:VCT

Comparison:standardcareofSTIforincreasing

condom

use

Outcomes

Illustrative

comparativerisks*

(95%CI)

Relativeeffect

(95%CI)

NoofParticipants

(studies)


(GRADE)

Com

ments

Assumed

risk

Correspondingrisk

standardcareofSTIfor

increasingcondomuse

VCT

Syphilis

prevalence

amongFSWsat6-month

Studypopulation

RR0.52

(0.23to1.18)

278

(1study)

⊕©

©©

verylow

1,2,3,4,5

109per1000

57per1000

(25to129)


110per1000

57per1000

(25to130)

Gonorrhoeaprevalence

amongFSWsat6-month

Studypopulation

RR0.83

(0.4to1.74)

278

(1study)

⊕©

©©

verylow

1,2,3,4,5

102per1000

85per1000

(41to177)


102per1000

85per1000

(41to177)

Chlamydia

prevalence

amongFSWsat6-month

Studypopulation

RR0.67

(0.4to1.13)

278

(1study)

⊕©

©©

verylow

1,2,3,4,5


countries (Review)


212per1000

142per1000

(85to240)


212per1000

142per1000

(85to240)

Trichomonas

preva-

lenceamongFSWsat6-

month

Studypopulation

RR0.52

(0.23to1.18)

278

(1study)

⊕©

©©

verylow

1,2,3,4,5

109per1000

57per1000

(25to129)


110per1000

57per1000

(25to130)

Genital

warts

preva-

lenceamongFSWsat6-

month

Studypopulation

RR0.22

(0.05to0.98)

278

(1study)

⊕⊕

©©

low

1,2,3,5,6

66per1000

15per1000

(3to65)


66per1000

15per1000

(3to65)

Consistentcondom

use

byFSWswithclientsat

6-month

Studypopulation

RR2.05

(1.48to2.85)

278

(1study)

⊕©

©©

verylow

1,2,5,7,8

255per1000

523per1000

(377

to727)

Lowriskpopulation

255per1000

523per1000

(377

to727)


countries (Review)


HIV/STI

knowledgeam

ongFSWs

at6-month

Studypopulation

RR1.28

(0.81to2.04)

278

(1study)

⊕©

©©

verylow

1,2,5,7,8

182per1000

233per1000

(147

to371)

Lowriskpopulation

182per1000

233per1000

(147

to371)

*The



studies)isprovided


its95%confidence

interval)isbasedon

the









1Seriouslimitations:Allocationconcealmentwasjudged’’unclear‘‘

inthistrial.

2Singlestudy.



effect.

5Biaswasjudgedtobeat’’highrisk‘‘inthistrial.


7Serious


makes

extrapolationofresulttoothersettings

unreliable.

8Veryseriousimprecision:The95%CIofestimateincludeappreciablebenefittocontrolgroupoverinterventiongroup.


countries (Review)


D I S C U S S I O N

Summary of main results

The current review addressed a number of HIV behavioral pre-

vention interventions intended to reduce the transmission of HIV

among FSWs in low- and middle-income countries. The review

used comprehensive search strategies encompassing both pub-

lished and non-published studies. In addition, strict inclusion cri-

teria were followed to ensure that the review was focused on behav-

ioral interventions among sex workers. The review’s conclusions

are weakened by the small number and size of the potential trials

and by the insufficiency of the reported data for the pre-specified

primary and secondary outcomes.

Thirteen trials were identified which fulfilled the inclusion crite-

ria for this review, involving 8,698 participants. Of the interven-

tions assessed, seven studies identified the primary outcomes but

only three trials examined HIV incidence. Moreover, there was

no strong evidence that these interventions reduced HIV preva-

lence. Nonetheless, results of meta-analyses of primary outcomes

found that the interventions did have an effect in reducing HIV

incidence among FSWs assigned to a social cognitive intervention

compared with those received standard counselling for HIV/STIs

(Patterson 2008) and interventions promoting the use of female

and male condoms were more effective compared than those that

promoted male condoms only (Ray 2001). Furthermore, other

primary outcomes, such as the incidence of STIs, were assessed by

various interventions such as:

• Social cognitive theory, which was found to be effective in

reducing the incidence of any STIs in India by 43% (Patterson

2008);

• Promotion of condom use, which reduced gonorrhoea and

chlamydia incidence in Thailand by 37% and 29%, respectively

(Fontanet 1998)

• Peer education intervention in addition to clinic-based

counselling conducted which reduced prevalence of chlamydia

and of any STIs by 30% and 20%, respectively in Madagascar

(Feldblum 2005)

• Intensive STI screening, which reduced HIV incidence by

35%, and chlamydia trachomatis and gonorrhoea prevalence by

77% and 62%, respectively in Cote d’Ivore (Ghys 2001)

• VCT intervention, which led to a 78% reduction in

prevalence of genital warts among FSWs in China (Li 2006)

• Prevalence of STIs was also reported by a community

empowerment intervention in India (Gutierrez 2010), but no

intervention effect was found.

Meta-analyses also were performed for two secondary outcomes;

changes in behavior outcomes such as consistent condom use,

including female and male condom use; and drug use risk be-

havior. These meta-analyses showed strong effects of the promo-

tion of condom use in intervention and control groups, includ-

ing increases in female condom use rather than male condom use.

However, this was positively correlated with increased condom

use by clients, a finding not repeated in other results from the

meta-analysis. There was a difference between social cognitive the-

ory and standard care in the reduction of illicit drug use among

FSWs. Substantial heterogeneity (I2>50%) in this review were

identified in two meta-analyses assessed consistent condom use

through community empowerment intervention (81%; Analysis

3.3) and through promotion of female and male condom use

(65%; Analysis 3.4) indicated that variability in methodological,

implementation of interventions, settings or populations may be

differ. No observed heterogeneity were observed among studies

whether evaluated consistent condom use through social cognitive

intervention (Analysis 1.3) or through promotion of condom use

(Analysis 9.4) indicated studies contributed for meta-analysis is a

similar way.

The effectiveness of behavioral interventions for other secondary

outcomes was analysed. Results were

• A positive increase in the rate of consistent use of condoms

with regular partners by 45% as a result of a community

empowerment intervention (Gutierrez 2010)

• The addition of an education intervention to a

microenterprise intervention conducted in India was associated

with a reduction in the number of sex exchange partners

(Sherman 2010)

• The addition of manager training to peer education was

found to increase the likelihood of HIV testing among FSWs in

the Philippines (Chiao 2009)

It is not possible to conclude definitively that no positive effect has

been achieved by some interventions, either due to small sample

size or failure in the design of the studies, or both. This clearly

highlights of the need to clarify these issues for policy-makers and

researchers considering affordable options for improving the qual-

ity of interventions to reduce the transmission of HIV in resource-

limited settings. It also indicates the importance of good study

design when implementing behavioral intervention programs of

previously unknown quality.

This review was limited to examining the available literature

on female sex workers. Given their high mobility and high risk

of HIV exposure, male and transgender sex workers and their

clients should also be taken into account when considering the

spread of STIs. In spite of limited data, a number of non-RCTs

have determined the effect of HIV/AIDS behavioral interven-

tions in preventing HIV/STIs among targeted male sex work-

ers (Maticka-Tyndale 1997 in Thailand), male transgender sex


countries (Review)


workers (Khan 2008 in Pakistan), and male clients of sex workers

(Leonard 2000 in Senegal, Barrington 2009 in Dominican Re-

public, and Hoque 2009 in Bangladesh), however, this systematic

review did not include non-RCTs, so these studies and the evi-

dence from them were not available to support the results in this

review.

Although some trials provided only minimal details of follow-up

length, the length or intensity of the interventions varied across

trials. It is important to note that follow-up assessments over six

months might have seen a reduction in the incidence rates of STIs

in Thailand, Madagascar and Mexico (Fontanet 1998, Feldblum

2005, Patterson 2008) and STI prevalence in China (Li 2006),

and in the number of sex exchange partners in India (Sherman

2010) compared with shorter or longer follow-up assessments.

Conversely, the results from the current review of consistent con-

dom use did not depend on the duration of the follow-up period.

For example, results from an RCT in Thailand (Fontanet 1998)

showed effective use of female condoms after a 6-month follow up

period and after a 24-month follow up period, but the intervention

failed to increase male condom use after a 24-month assessment

period, which means this phenomenon was not specific enough

to measure the reduction of HIV/STI incidence and prevalence.

This indicates that effective behavioral change interventions to in-

crease condom usage may be long-lasting and sustainable, which is

an important consideration in dealing with hard-to-reach, hidden

and mobile populations such as sex workers.

Despite its limitations, our review provides scientific evidence to

support social cognitive theory intervention and the promotion of

male and female condom use in order to reduce the transmission

of HIV/STIs in vulnerable populations, especially FSWs, when

compared with other behavioral approaches or no interventions.

This meta-analysis also could inform the design and implemen-

tation of future studies, especially in terms of the choice of tar-

get population, setting, intensity of assessment, and the content

of the intervention, and the findings of the studies included in

this review, as well as the methods that were successful, should

be considered by policy-makers, health workers and researchers in

designing new interventions for these communities.

Overall completeness and applicability of evidence

A total of 13 trials were selected for review. However, the available

data are limited for some outcomes and the results should thus

be interpreted with caution. Although the data is available, it is

not clear how easy it would be to apply the interventions to other

settings, or to generalize the findings to other population groups.

Studies identifying the effects of behavioral interventions on HIV

prevalence were not considered in this review although this out-

come is undoubtedly important in the reduction of HIV trans-

mission among sex workers. The interventions for all sex work-

ers were reviewed but only FSW programs were eligible based on

randomised methods. Male and transgender sex workers and their

clients were not included in the review due to the unavailability

of data and due to these factors the results of the meta-analyses

can only be applied to FSWs. Also, there was heterogeneity in the

intervention effects for many of the outcomes, and there is no in-

formation as to the specific factors that might have caused the dif-

ferences between the intervention and control groups. However,

with regard to our objective, the findings of the review indicate

that the use of social cognitive-based theory and the promotion of

female and male condom use as HIV behavioral prevention strate-

gies play an important role in reducing HIV/STI incidence, along

with increasing condom use consistently and reducing illegal drug

use risk behavior among FSWs in low- and middle-income coun-

tries. Because these interventions are likely to be low-cost, capable

of being implemented by relatively unskilled workers, and can be

easily maintained over long periods, they provide an important

tool for controlling the spread of HIV amongst high-risk groups

in low- and middle-income countries.

Quality of the evidence

The quality of the evidence has been assessed using the GRADE

process (Guyatt 2008), and the results are presented in the

Summary of findings for the main comparison, Summary of

findings 2, Summary of findings 3, and Summary of findings 4.

The GRADE approach uses four levels of quality (very low, low,

moderate and high) over several domains covering limitations in

the design and implementation of studies, indirectness of evidence,

unexplained heterogeneity or inconsistency in results, imprecision

of results and high probability of publication bias. The highest

quality rating was found in six trials (Fontanet 1998, Ray 2001,

Hoke 2007, Patterson 2008, Chiao 2009, Sherman 2010). The

assessment of the risk of bias in these trials suggests concern re-

garding insufficient information of sequence generation, alloca-

tion concealment and failure to adequately address incomplete

outcome data. Methodological quality was rated as ”high“ for only

four of the thirteen trials (Fontanet 1998, Ray 2001, Patterson

2006, Sherman 2010), .

Potential biases in the review process

All authors have an interest in HIV behavior prevention interven-

tions, and each brings a different perspective to the methods used.

However, we made efforts to limit the bias in several ways: two

review authors assessed eligibility for inclusion and assessed the

risk of bias independently. Although the authors’ views varied, af-

ter extensive discussion and achieving a consensus, we decided to

accept the final conclusions. We knew that a meta-analysis might

be vulnerable to publication bias in the review process. However,

we were not able to use funnel plots due to the heterogeneity of

the study designs. But we did try to ensure that our eventual con-

clusion arose solely from the data. Feedback from readers interna-

tionally will serve to improve the next review update.

Agreements and disagreements with other studies or reviews


countries (Review)


These results indicate that combined HIV behavioral prevention

strategies have a greater potential for positive results than single

interventions such as peer education or manager training alone

(Chiao 2009). These results are consistent with those from cluster

non-RCTs that tested the combined effects of educational inter-

ventions and the promotion of condom use in reducing HIV and

STI incidence in India (Bhave 1995) and Indonesia (Ford 2002).

However, when a health education intervention was combined

with the promotion of condoms, it resulted in the intervention

having either no effect or in an increase in condom use among

FSWs in Nicaragua (Egger 2000). The effectiveness of these in-

terventions, moreover, may not be assumed when the objective is,

for example, to improve the knowledge of HIV/STI, or to change

self-reported or observed behavior, such as sexual risk behavior

and the frequency of sexual encounters. Our systematic review in

low- and middle income countries shows that not all behavioral

interventions have a positive impact when it comes to preventing

HIV/STI infections or other secondary outcomes. This finding

is consistent with the systematic review of interventions for sex

workers in high-income countries, in which the effect of behav-

ioral interventions did not significantly reduce the STI incidence

and increase the use of condoms (Ota 2011).

A U T H O R S ’ C O N C L U S I O N S

Implications for practice

HIV transmission among sex workers and their clients helps

to drive a broader epidemic of heterosexually-acquired HIV

(UNAIDS 2009), resulting in transmission even among individ-

uals who engage in low levels of risk behavior. Interventions to re-

duce the risk of HIV transmission amongst these high-risk groups

are therefore an important part of the international response to

the HIV epidemic, and should be assessed carefully for their suit-

ability, cost and effectiveness. This review strongly considers the

effectiveness of several ongoing behavioral interventions that are

likely to be available to policy makers and healthcare workers in

low- and middle-income nations. Policy makers, practitioners and

researchers therefore have a greater range of choice of potential

interventions to reduce HIV transmission risk among high-risk

groups. Before making a practice decision based on the current re-

view, further information from other reviews considering how the

role of combining sexual risk reduction, condom promotion and

improved access to STI treatment reduces HIV and STI acqui-

sition in sex workers receiving these interventions (Shahmanesh

2008) should to be taken into account. Nevertheless, given the

observed effectiveness of social cognitive theory and the promo-

tion of condom use in reducing transmission of HIV/STIs, and

the clear need for a reduction in transmission, these behavioral in-

terventions should be considered for implementation in high-risk

FSW communities in low- and middle-income countries. This

decision should be supported, however, by the knowledge that

there is lack of information about most other outcomes and target

populations, and although there is evidence in favour of their im-

plementation amongst other high-risk populations (such as trans-

gender sex workers) the quality of this evidence is not currently

sufficient to conclude that they will be effective amongst these

populations. The studies reviewed here did not cover the full range

of developing nations, but were spread across three continents and

so be sufficient to inform policy-makers from nations not repre-

sented in the current review. This review finds that the effects of

some interventions were sustained in radically different national

and local settings, indicating that the interventions be suitable in

most low- and middle-income countries.

Implications for research

Researchers designing new RCTs in to measure intervention effects

should strongly consider measuring HIV incidence and prevalence

as outcomes, and assess risk-factors for the spread of this infec-

tion. More high quality RCTs are urgently needed to further clar-

ify which behavioral interventions are most effective in reducing

the transmission of HIV among high-risk, hidden or marginalized

populations in low- and middle-income countries. Moreover, be-

cause sub-Saharan Africa remains the most heavily affected region

of heterosexual transmission, including among sex workers, critical

evaluation of interventions there should be conducted. Further-

more, the cost-effectiveness of these interventions, particularly so-

cial cognitive behavioral interventions and the promotion of con-

dom use as effective strategies targeted at vulnerable populations

should be a priority area for research. Therefore, detailed data on

the cost of interventions should be collected and analysed in a

standardized format, and future implementation programs should

consider enhanced study design and collection methods, to ensure

that key findings from those studies are valid, generalizable and of

high quality.

A C K N O W L E D G E M E N T S

We are grateful to Nagaraj Kondagunta, Sreekumaran Nair, Vinod

Bhat, Debashis Dutt, and Nalini Sathiakumar who developed the

original protocol. We are also grateful to Stuart Gilmor and An-

drew Stickley for their comments and suggestions on the draft of

the review.


countries (Review)


R E F E R E N C E S

References to studies included in this review

Basu 2004 {published data only}

Basu I, Jana S, Rotheram-Borus MJ, Swendeman D, Lee S,

Newman P, et al.HIV prevention among sex workers in

India. Journal Acquired Immune Deficiency Syndrome 2004;

36:845–852.

Chiao 2009 {published data only}

Chiao C, Morisky DE, Ksobiech K, Malow RM. Promoting

HIV testing and condom use among Filipina commercial sex

workers: findings from a quasi-experimental intervention

study. AIDS Behav 2009;13:892–901.

Feldblum 2005 {published data only}

Feldblum PJ, Hatzell T, Van Damme K, Nasution M,

Rasamindrakotroka A, Grey TW. Results of a randomised

trial of male condom promotion among Madagascar sex

workers. Sex Transm Infect 2005;81:166–172.

Fontanet 1998 {published data only}

Fontanet AL, Saba J, Chandelying V, Sakondhavat C,

Bhiraleus P, Rugpao S, et al.Protection against sexually

transmitted diseases by granting sex workers in Thailand

the choice from using the male or female condom: results

from a randomized controlled trial. AIDS 1998;12(14):

1851–1859.

Ghys 2001 {published data only}

Ghys PD, Diallo MO, Ettiegne-Traore V, Satten GA,

Anoma CK, Maurice C, et al.Effect of interventions to

control sexually transmitted disease on the incidence of HIV

infection in female sex workers. AIDS 2001;15:1421–1431.

Gutierrez 2010 {published data only}

Gutierrez J, McPherson S, Fakoya A, Matheou A, Bertozzi

SM. Community-based prevention leads to an increase

in condom use and a reduction in sexually transmitted

infections (STIs) among men who have sex with men

(MSM) and female sex workers (FSW): the Frontiers

Prevention Project (FPP) evaluation results. BMC Public

Health 2010;10(497):1–12.

Hoke 2007 {published data only}

Hoke TH, Feldblum PJ, Van Damme K, Nasution MD,

Grey TW, Wong EL, et al.Randomised controlled trial of

alternative male and female condom promotion strategies

targeting sex workers in Madagascar. Sex Transm Infect

2007;83:448–453.

Li 2006 {published data only}

Li X, Wang B, Fang X, Zhao R, Stanton B, Hong Y, et

all. Short-term effect of a cultural adaptation of voluntary

counseling and testing among female sex workers in China:

a quasi-experimental trial. AIDS Educ Prev 2006;18(5):

406–419.

Markosyan 2010 {published data only}

Markosyan K, Lang DL, Salazar LF, DiClemente RJ, Hardin

JW, Darbinyan N, et al.A randomized controlled trial of

an HIV prevention intervention for street-based female

sex workers in Yerevan, Armenia: preliminary evidence of

efficacy. AIDS Behaviour 2010;14:530–537.

Patterson 2006 {published data only}

Patterson TL, Semple SJ, Fraga M, Bucardo J, De la Torre

A, Salazar-Reyna J, et al.A sexual risk reduction intervention

for female sex workers in Mexico: design and baseline

characteristics. Journal of HIV/AIDS & Social Services 2006;

5(2):115–137.

Patterson 2008 {published data only}

Patterson TL, Mausbach B, Lozada R, Staines-Orozco

H, Semple S, Fraga-Vallejo M, et al.Efficacy of a brief

behavioral intervention to promote condom use among

female se workers in Tijuana and Ciudad Juarez, Mexico.

American Journal of Public Health 2008;98(11):2051–2057.

Ray 2001 {published data only}

Ray S, Van de Wijgert J, Mason P, Ndowa F, Maposhere C.

Constraints faced by sex workers in use of female and male

condoms for safer sex in urban Zimbabwe. Journal of UrbanHealth 2001;74(4):581–592.

Sherman 2010 {published data only}

Sherman SG, Srikrishnan AK, Rivett KA, Liu S, Solomon

S, Celentano DD. Acceptability of a microenterprise

intervention among female sex workers in Chennai, India.

AIDS Behaviour 2010;14:649–657.

Swendeman 2009 {published data only}

Swendeman D, Basu I, Das S, Jana S, Jane M, Rotheram-

Borus. Empowering sex workers in India to reduce

vulnerability to HIV and sexually transmitted diseases.

Social Science and Medicine 2009;69:1157–1166.

Wechsberg 2006 {published data only}

Wechsberg WM, Luseno WK, Lam WKK, Parry CDH,

Morojele NK. Substance use, sexual risk, and violence: HIV

prevention intervention with sex workers in Pretoria. AIDS

and Behavior 2006;10(2):131–137.

References to studies excluded from this review

Abellanosa 1996 {published data only}

Abellanosa I, Nichter M. Antiobiotic prophylaxis among

commercial sex workers in Cebu city, Philippines. SexuallyTransmitted Diseases 1996;23(5):407–412.

Bakare 2002 {published data only}

Bakare RA, Oni AA, Umar US, Adewole IF, Shokunbi WA,

Fayemiwo SA, et al.Pattern of sexually transmitted diseases

among commercial sex workers (CSWs) in Ibadan, Nigeria.

Afr J Med. med. Sci 2002;31:243–246.

Barrington 2009 {published data only}

Barrington C, Latkin C, Sweat MD, Moreno L, Ellen J,

Kerrigan D. Talking the talk, walking the walk: social

network norms, communication patterns, and condom

use among the male partners of female sex workers in La

Romana, Dominican Republic. Social Science & Medicine

2009;68(11):2037–2044.


countries (Review)


Beattie 2010 {published data only}

Beattie TSH, Bhattacharjee P, Ramesh BM, Gurnani V,

Anthony J, Isac S, et al.Violence against female sex workers

in Karnataka state, South India: impact on health, and

reductions in violence following an intervention program.

BMC Public Health 2010;10(476):1–11.

Bhave 1995 {published data only}

Bhave G, Lindan CP, Hudes ES, Desai S, Wagle U, Tripathi

SP, et all. Impact of an intervention on HIV, sexually

transmitted diseases, and condom use among sex workers in

Bombay, India. AIDS 1995;9(1):21–30.

Dandona 2005 {published data only}

Dandona R, Dandona L, Gutierrez JP, Kumar AG,

McPherson S, Samuels F, et al.High risk of HIV in non-

brothel based female sex workers in India. BMC Public

Health 2005;5(87):1–10.

Egger 2000 {published data only}

Egger M, Pauw J, Lopatatzidis A, Medrano D, Paccaud PF,

Smith PF. Promotion of condom use in a high-risk setting

in Nicaragua: a randomised controlled trial. Lancet 2000;

360(9338):971–977.

Feldblum 2007 {published data only}

Feldblum PJ, Nasution MD, Hoke TH, Van Damme K,

Turner AN, Gmach R, et al.Pregnancy among sex workers

participating in a condom intervention trial highlights the

need for dual protection. Contraception 2007;76:105–110.

Fonck 2000 {published data only}

Fonck K, Kaul R, Kimani J, Keli F, MacDonald KS,

Ronald AR, et all. A randomized, placebo-controlled trial

of monthly azithromycyn prophylaxis to prevent sexually

trnasmitted infections and HIV-1 in Kenyan sex workers:

study design and baseline findings. International Journal of

STD and AIDS 2000;11:804–811.

Ford 2002 {published data only}

Ford K, Wirawan DN, Reed BD, Muliawan P, Wolfe R.

Evaluation of an intervention for sex workers. Sexually

Transmitted Diseases 2002;29(1):50–58.

Gangopadhyay 2005 {published data only}

Gangopadhyay DN, Chanda M, Sarkar K, Niyogi SK,

Chakraborty S, Saha MK, et all. Evaluation of sexually

transmitted diseases/Human immunodeficiency virus

intervention programs for sex workers in Calcutta, India.

Sexually Transmitted Diseases 2005;32(11):680–684.

Gorbach 2000 {published data only}

Gorbach PM, Sopheab H, Phalla T, Leng HR, Mills S,

Bennett A, et all. Sexual bridging by Cambodian men.

Sexually Transmitted Diseases 2000;27(6):320–326.

Guerena-Burgueno 1991 {published data only}

Guerena-Burgueno F, Benenson AS, Sepulveda-Amor J.

HIV-1 prevalence in selected Tijuana sub-populations. Am

J Public Health 1991;81:623–625.

Hoque 2009 {published data only}

Hoque MHE, Ono-Kihara, Zamani S, Ravari SM, Kihara

M. HIV-related risk behaviour and the correlates among

rickshaw pullers of Kamrangirchar, Daka, Bangladesh: a

cross-sectional study using probability sampling. BMC

Public Health 2009;9(80):1–11.

Kamali 2003 {published data only}

Kamali A, Quigley M, Nakiyingi J, Kinsman J, Kengeya-

Kayando J, Gopal R. Syndromic management of sexually-

transmitted infections and behavior change interventions

on transmission of HIV-1 in rural Uganda: a community

randomised trial. Lancet 2003;361:645–652.

Kaul 2002 {published data only}

Kaul R, Kimani J, Nagelkerke J, Fonck K, Keli F,

MacDonald KS, et all. Reduced HIV risk-taking and

low HIV incidence after enrolment and risk-reduction

counseling in a sexually transmitted disease prevention trial

in Nairobi, Kenya. J Acquir Immune Defic Syndr 2002;30:

69–72.


Kaul R, Kimani J, Nagelkerke NJ, Fonck K, Ngugi EN,

Keli F, et all. Monthly antibiotic chemoprophylaxis and

incidence of sexually transmitted infections and HIV-1

infection in Kenyan sex workers. JAMA 2004;291(21):

2555–2562.


Kaul R, Nagelkerke NJ, Kimani J, Ngugi E, Bwayo JJ,

MacDonald KS, et all. Prevalent herpes simplex virus type

2 infection is associated with altered vaginal floral and

increased susceptibility to multiple sexually transmitted

infections. JID 2007;196:1692–1697.

Kerrigan 2006 {published data only}

Kerrigan D, Moreno L, Rosario S, Gomez B, Jerez H,

Barrington C, et al.Environmental-structural interventions

to reduce HIV/STI risk among female sex workers in the

Dominican Republic. American Journal of Public Health

2006;96(1):120–125.

Khan 2008 {published data only}

Khan AA, Rehan N, Qayyum K, Khan A. Correlates and

prevalence of HIV and sexually transmitted infections

among hijras (male transgender) in Pakistan. International

Journal of STD and AIDS 2008;19:817–820.

Larios 2009 {published data only}

Larios SE, Lozada R, Strathdee SA, Semple SJ, Roesch S,

Staines H, et all. An exploration of contextual factors that

influence HIV risk in female sex workers in Mexico: the

social ecological model applied to HIV risk behaviors. AIDSCare 2009;21(10):1335–1342.

Lau 2008 {published data only}

Lau JTF, Choi KC, Tsui HY, Zhang L, Zhang J, Lan Y,

et al.Changes in HIV-related behaviours over time and

associations with rates of HIV-related services coverage

among female sex workers in Sichuan, China. Sex Transm

Infect 2008;84:212–216.

Leonard 2000 {published data only}

Leonard L, Ndiaye I, Kapadia A, Eisen G, Diop O, Mboup

S, et al.HIV prevention among male clients of female sex

workers in Kaolack, Senegal: results of a peer education

program. AIDS Education Prevention 2000;12(1):21–37.


countries (Review)


Maticka-Tyndale 1997 {published data only}

Maticka-Tyndale E, Elkins D, Haswell-Elkins M,

Rujkarakorn D, Kuyyakanond T, Stam K. Contexts

and patterns of men’s commercial sexual partnerships in

Northeastern Thailnd: implications and prevention. Soc Sci

Med 1997;44(2):199–213.

McClelland 2008 {published data only}

McClelland RS, Richradson BA, Hassan WM, Chohan

V, Lavreys L, Mandaliya K, et al.Improvment of vaginal

health for Kenyan women at risk for acquisition of human

immunodeficiency virus type 1: results of a randomised

trial. JID 2008;197:1361–1368.

Ngugi 1988 {published data only}

Ngugi EN, Plummer FA, Simonsen JN, Cameron DW,

Bosire M, Waiyaki P, et al.Prevention of transmission of

human immunodeficiency virus in Africa: effectiveness

of condom promotion and health education among

prostitutes. Lancet 1988;2(8616):887–890.


Ngugi EN, Wilson D, Sebstad J, Plummer FA, Moses

S. Focused Peer-Mediated Educational Programs among

Female Sex Workers to Reduce Sexually Transmitted

Disease and Human Immunodeficiency Virus Transmission

in Kenya and Zimbabwe. The Journal of Infectious Diseases1996;174(Suppl 2):S240–S207.


Ngugi EN, Chakkalackal M, Sharma A, Bukusi E, Njoroge

B, Kimani J, et all. Suistained changes in sexual behavior by

female sex workers after completion of a randomized HIV

prevention trial. J Acquir Immune Defic Syndr 2007;45:

588–594.

Nzila 1991 {published data only}

Nzila N, Laga M, Thiam MA, Mayimona K, Edidi B, Van

Dyck E, et al.HIV and other sexually transmitted diseases

among female prostitutes in Kinshasa. AIDS 1991;5(6):

715–721.

Ramesh 2010 {published data only}

Ramesh BM, Beattie TSH, Shajy I, Washington R,

Jagannathan L, Reza-Paul S, et al.Changes in risk behaviours

and prevalence of sexually transmitted infections following

HIV preventive intervetnions among female sex workers in

five districts in Karnataka state, south India. Sex Transm

Infect 2010;86:i17–i24.

Reza-Paul 2008 {published data only}

Reza-Paul S, Beattie T, Syed HUR, Venukumar KT,

Venugopal MS, Fathima MP, et al.Declines in risk behaviour

and sexually transmitted infection prevalence following a

community-led HIV preventive intervention among female

sex workers in Mysore, India. AIDS 2008, 22 (suppl 5):S91-

S100 2008;22(Suppl 5):S91–S100.

Rou 2007 {published data only}

Rou K, Wu Z, Sullivan SG, Li F, Guan J, Xu C, et al.A five-

city trial of a behavioural intervention to reduce sexually

transmitted disease/HIV risk among sex workers in China.

AIDS 2007;21(Suppl 8):S95–S101.

Schwandt 2006 {published data only}

Schwandt M, Morris C, Ferguson, Ngugi, Moses S. Anal

and dry sex in commercial sex work, and relation to risk for

sexually transmitted infections and HIV in Meru, Kenya.

Sex Transm Infect 2006;82:392–396.

Strathdee 2009 {published data only}

Strathdee SA, Mausbach B, Lozada R, Staines-Orozco H,

Semple SJ, Abramovitz D, et al.Predictors of sexual risk

reduction among Mexican female sex workers enrolled in a

behavioral intervention study. J Acqui Immune Defic Syndr

2009;51(Suppl 1):S42–S46.

Stulhofer 2009 {published data only}

Stulhofer A, Bacak V, Drglin T, Puljiz M, Miklin M. Female

sex work and HIV risks in Croatia. AIDS Care 2009;21

(11):1439–1446.

Walden 1999 {published data only}

Walden VM, Mwangulube K, Makhumula-Nkhoma P.

Measuring the impact of a behaviour change intervention

for commercial sex workers and their potential clients in

Malawi. Health Education Research 1999;14(4):545–554.

Weir 1998 {published data only}

Weir S, Roddy RE, Zekeng L, Ryan KA, Wong EL.

Measuring condom use: asking”do you or don’t you“ isn’t

enough. AIDS Education and Prevention 1998;10(4):

293–302.

Weir 1999 {published data only}

Weir SS, Roddy RE, Zekeng L, Ryan KA. Association

between condom use and HIV infection: a randomised

study of self reported condom use measures. J EpidemiolCommunity Health 1999;53:417–422.

Yadav 2005 {published data only}

Yadav G, Saskin R, Ngugi E, Kimani J, Keli F, Fonck K, et

al.Associations of sexual risk taking among Kenyan female

sex workers after enrollment in an HIV-1 prevention trial. JAcquir Immune Defic Syndr 2005;38:329–334.

Yi 2010 {published data only}

Yi H, Mantell JE, Wu R, Lu Z, Zeng J, Wan Y. A profile of

HIV risk factors in the context of sex work environments

among migrant female sex workers in Beijing, China.

Psychol Health Med 2010;15(2):172–187.

Additional references

Asiimwe 2009

Asiimwe A, Koleros A, Chapman J. Understanding the

dynamics of the HIV epidemic in Rwanda: modelling the

expected distribution of new HIV infections by exposure

group. Kigali, National AIDS Control Commission,

MEASURE evaluation 2009.

Bandura 1986

Bandura A. Social foundations of thought and action: A social

cognitive theory. Englewood Cliffs, NJ: Prentice-Hall, 1986.

Barrientos 2007

Barrientos JE, Bozon M, Ortiz E, Arredondo A. HIV

prevalence, AIDS knowledge, and condom use among


countries (Review)


female sex workers in Santiago, Chile. Cad. Saude Publica,

Rio de Janeiro 2007;23(8):1777–1784.

Bradley 2010

Bradley J, Moses S, Blanchard JF, Rajaram S, Ramesh BM,

Verma S, et al.Assessing reported condom use among female

sex workers in southern India through examination of

condom availability. Sexually Transmission Infection 2010;

86(Suppl 1):144–148.

Buzdugan 2010

Buzdugan R, Copas A, Moses S, Blanchard J, Isac S, Ramesh

BM, et al.Devising a female sex work typology using data

from Kamataka, India. International Journal of Epidemiology

2010;39(2):439–448.

Caceres 2009

Caceres C, Walter M. The National response to the HIV/

AIDS epidemic in Peru: accomplishments and gaps--A

review. Journal of Acquired Immune Deficiency Syndomes

2009;51:S60–S66.

Coates 2008

Coates TJ, Richter L, Caceres C. Behavioral strategies to

reduce HIV transmission: how to make them work better.

Lancet 2008;372(9639):669–684.

Cornish 2009

Cornish F, Campbell C. The social conditions for successful

peer education: a comparison of two HIV prevention

programs run by sex workers in India and South Africa.

American Journal of Community Psychology 2009;44(1-2):

123–135.

Deeks 2001

Deeks JJ, Altman DG, Bradbum MJ. Statistical methods for

examining heterogeneity and combining results from several

studies in meta-analysis. In: Egger M, Smith GD, Altman

DG editor(s). Systematic reviews in health care: meta-analysis

in context. London: BMJ Books, 2001.

Fajans 1995

Fajans P, Ford K, Wirawan DN. AIDS knowledge and risk

behaviors among domestic clients of female sex workers

in Bali, Indonesia. Social Science & Medicine 1995;41(3):

409–417.

Ford 2000

Ford K, Wirawan DN, Suastina SS, Reed BD, Muliawan P.

Evaluation of a peer education programme for female sex

workers in Bali, Indonesia. International Journal STD AIDS2000;11:1731–1733.

Gelmon 2009

Gelmon L, Kenya P, Oguya F, Cheluget B, Haile G. Kenya:

HIV prevention response and modes of transmission

analysis. Kenya National AIDS Control Council 2009.

Ghose 2008

Ghose T, Swendeman D, George S, Chowdhury D.

Mobilizing collective identity to reduce HIV risk among

sex workers in Sonagachi, India: The boundaries,

consciousness, negotiation framework. Social Science &

MEdicine 2008;67(2):311–320.

Girault 2004

Girault P, Saidel T, Song N, de Lind Van Wijngaarden JW,

Dallabetta G, et al.HIV, STIs and sexual behaviors among

men who have sex with men in Phnom Penh, Cambodia.

AIDS Education and Prevention 2004;6(1):31–44.

Gomes do Espirito Santo 2005

Gomes do Espirito Santo ME, Etheredge GD. Male clients

of brothel prostitutes are a bridge for HIV infection between

high risk and low risk groups of women in Senegal. SexuallyTransmitted Infections 2005;81:342–344.

Gorbach 2006

Gorbach PM, Sopheab H, Chhorvann C, Weiss RE, Vun

MC. Changing behaviors and patterns among Cambodian

sex workers 1997-2003. Journal of Acquired ImmuneDeficiency Syndromes 2006;42(2):242–247.

Gu 2008

Gu J, Chen H, Chen X, Lau JTF, Wang R, Liu C, et

al.Severity of drug dependence, economic pressure and

HIV-related risk behaviors among non-institutionalized

female injecting drug users who are also sex workers in

China. Drug and alcohol dependence 2008;97(3):257–267.

Guyatt 2008

Guyatt GH, Oxman AD, Kunz R, Vist GE, Falck-Ytter Y,

Schunemann HJ. GRADE Working Group. Rating quality

of evidence and strength of recommendations: What is

”quality of evidence“ and why is it important to clinicians?.

BMJ 2008;336(7651):995–998.

Hashemi 1996

Hashemi S, Schuler S, Riley A. Rural credit programs and

women’s empowerment in Bangladesh.. World Development

1996;24(4):635–653.

Higgins 2009

Higgins JPT, Green S. Cochrane handbook for systematic

reviews of interventions version 5.0.0 (updated 2009). Wiley-

Blackwell, 2009.

Jones 1989

Jones JW. Personality and epistemology: Cognitive social

learning theory as a philosophy of science. Zygon 1989;24

(1):23–38.

Khobotlo 2009

Khobotlo M, Tshehlo R, Nkonyana J, Ramoseme M,

Khobotle M, Chitoshia A, et al.Lesotho: HIV prevention

response and modes of transmission analysis. Lesotho

National AIDS Commission 2009.

Kim 2007

Kim J, Watts C, Hargreaves J, et al.Understanding the

impact of a microfinance-based intervention on women’s

empowerment and the reduction of intimate partner

violence in South Africa. American Journal of Public Health2007;97(10):1794–802.

Laga 1994

Laga M, Alary M, Nzila N, Manoka AT, Tuliza M, Behets

F, et al.Condom promotion, sexually transmitted disease

treatment, and declining incidence of HIV-1 infection in

female Zairian sex workers. Lancet 1994;344:246–248.


countries (Review)


Lowndes 2008

Lowndes CM, Alary M, Belleau M, Bosu WK, Kintin DF,

Nnorom JA, et al.West Africa HIV/AIDS epidemiology and

response synthesis: implication for prevention. West HIV/AIDS epidemiology and response synthesis: implications for

prevention. Washington DC: World Bank, 2008.

Mngadi 2009

Mngadi S, Fraser N, Mkhatshwa H, Lapidos T, Khumalo

T, Tsela S, et al.Swaziland: HIV prevention response

and modes of transmission analysis. Mbabane National

Emergency Response Council on HIV/AIDS 2009.

Morris 2006

Morris CN, Ferguson AG. Estimation of the sexual

transmission of HIV in Kenya and Uganda on the trans-

Africa highway: the continuing role for prevention in

high risk groups. Sexually Transmited Infections 2006;82:

368–371.

Nelson 2007

Nelson KE, Williams CM. Infectious disease epidemiology:theory and practice. Second Edition. Massachusetts: Jones

& Bartlett learning, 2007.

Nguyen 2008

Nguyen TA, Nguyen HT, Le GT, Detels R. Prevalence

and risk factors associated with HIV infection among men

having sex with men in Ho Chi Minh City, Vietnam. AIDSand behavior 2008;12(3):476–482.

Ota 2011

Ota E, Wariki WMV, Mori R, Hori N, Shibuya K.

Behavioral interventions to reduce the transmission of HIV

infection among sex workers and their clients in high-income

countries. Cochrane Database of Systematic Reviews 2011,

Issue 12. [DOI: 10.1002/14651858.CD006045.pub3]

Padilla 2008

Padilla M, Castellanos D, Guilano-Ramos V, Reyes AM,

Marte LES, Soriano M. Stigma, social inequity, and HIV

risk disclosure among Dominican male sex workers. SocialScience & Medicine 2008;67(3):380–388.

PATH and UNFPA 2006

Program for Appropriate Technology in Health and United

Nations Population Fund. Female condom: a powerful

tool protection. http://www.unfpa.org/webdav/site/global/

shared/documents/publications/2006/female˙condom.pdf

2006.

Pisani 2004

Pisani E, Girault P, Gultom M, Sukartini N, Kumalawati

J, Jazan S, et al.HIV, syphilis infection and sexual practices

among transgender, male sex workers, and other men

who have sex with men in Jakarta, Indonesia. Sexually

Transmitted Infections 2004;80(6):536–540.

Revman 2008

The Cochrane Collaboration. Review Manager (Revman

5.0). Copenhagen: The Nordic Cochrane Centre, 2008.

Ruxrungtham 2004

Ruxrungtham K, Brown T, Phanuphak P. HIV/AIDS in

Asia. Lancet 2004;364:69–82.

Schuler 1997

Schuler S, Hashemi S, Riley A. The influence of women’s

changing roles and status in Bangladesh’s fertility transition:

evidence from a study of credit programs and contraceptive

use. World Development 1997;25(4):563–575.

Shahmanesh 2008

Shahmanesh M, Patel V, Mabey D, Cowan F. Effectiveness

of interventions for the prevention of HIV and other

sexually transmitted infections in female sex workers in

resource poor setting: a systematic review. Tropical Medicine& International Health 2008;13(5):659–679.

Shawky 2009

Shawky S, Soliman C, Kassak KM, Oraby D, El-Khoury

D, Kabore I. HIV surveillance and epidemic profile in the

Middle East and North Africa. Journal of Acquired Immune

Deficiency Syndromes 2009;51:S83–S95.

Soto 2007

Soto RJ, Ghee AE, Nunez CA, Mayorga R, Tapia KA,

Astete SG, et al.Sentinel surveillance of sexually transmitted

infections/HIV and risk behaviors in vulnerable populations

in 5 Central American Countries. Journal of AcquiredImmune Deficiency Syndromes 2007;46(1):101–111.

Steen 2009

Steen R, Wi TE, Kamali A, Ndowa F. Control of sexually

transmitted infections and prevention of HIV transmission:

mending a fractured paradigm. Bulletin World Health

Organization 2009;87(11):858–865.

Stevens 1998

Stevens PE, Hall JM. Participatory action research for

sustaining individual and community change: A model

of HIV prevention education. AIDS Education Prevention

1998;10:387–402.

UNAIDS 2002

UNAIDS. Sex works and HIV/AIDS. Geneva: Joint United

Nations Programme on HIV/AIDS 2002.

UNAIDS 2009

UNAIDS. AIDS epidemic update. http://data.unaids.org/

pub/Report/2009/JC1700˙Epi˙Update˙2009˙en.pdf 2009.

Wabwire-Mangen 2009

Wabwire-Mangen F, Odiit M, Kirungi W, Kisitu DK,

Wanyama JO. Uganda HIV modes of transmission and

prevention response analysis. Uganda National AIDS

Commission 2009.

Wang 2009

Wang H, Chen RY, Ding G, Ma Y, Ma J, Jiao JH, et

al.Prevalence and predictors of HIV infection among female

sex workers in Kaiyuan City, Yunnan Province, China.

International Journal of Infectious Diseases 2009;13(2):

162–169.

Weir 2008

Weir SS, Figueroa JP, Byfield L, Hall A, Cummings S,

Suchindran C. Randomized controlled trial to investigate

impact of site-based safer sex programmes in Kingston,

Jamaica: trial design, methods and baseline findings.


countries (Review)


Tropical Medicine & International Health 2008;13(6):

801–813.

WHO 2006

Low N, Broutet N, Adu-Sarkodie Y, Barton P, Hossain M,

Hawkes S. Global control of sexually transmitted disease.

http://www.who.int/reproductivehealth/publications/

general/lancet˙5.pdf 2006.

WHO 2009

World Health Organization. Toward universal access:

scaling up priority of HIV/AIDS interventions in the health

sector. Progress report 2009. Geneva:WHO 2009.

Wolffers 2003

Wolffers I, Van Beelen N. Public health and human rights

of sex workers. Lancet 2003;361(9373):1981.

World Bank

Developing country. http//youthink.worldbank.org/

glossary.php#ddd.∗ Indicates the major publication for the study


countries (Review)


C H A R A C T E R I S T I C S O F S T U D I E S

Characteristics of included studies [ordered by study ID]

Basu 2004

Methods Design: Quasi-randomized controlled trial

Random allocation: means of a random number table

Date: unclear (15 months)

Participants Number: Each community consisted 350 sex workers. 100 FSWs were randomly selected

in each area (n=200) using a two-stage randomisation process. 10 sex workers in each

area were over sampled at baseline to account for refusal and loss at the first follow-up

Inclusion criteria: FSWs those who live and work in selected communities and provided

informed consent

Setting: Community-based

Country: India

Interventions ”Community empowerment“ versus ”standard care“

Intervention: FSWs received basic STI information. Additionally, health clinics were

established in the centre of both communities’ red light areas to improve existing care

for sex workers. They also received training by a team of local peer educators to build

skills and confidence in providing education and to foster empowerment and advocacy

for local sex workers. The training comprised empowerment and advocacy, included

sustained engagement with local sex workers, showed interest in sex workers’ health and

well-being and that of their children, nurtured group solidarity among them, and raised

consciousness about sex worker rights

Control: FSWs did not receive the enhanced intervention to prevent HIV and STIs

Outcomes 1. Consistent condom utilization

2. Change in condom use

Notes

Risk of bias

Bias Authors’ judgement Support for judgement

Random sequence generation (selection

bias)

Low risk ”Communities were randomly assigned to

an intervention and control condition“

Allocation concealment (selection bias) Low risk ”First, the brothel rooms in each red-light.

...using a random number table“

Blinding (performance bias and detection

bias)

All outcomes

Unclear risk The study did not address this

Incomplete outcome data (attrition bias)

All outcomes

Unclear risk Insufficient reporting of attrition/exclu-

sions; no reasons for missing data provided


countries (Review)


Basu 2004 (Continued)

Selective reporting (reporting bias) Low risk The study protocol is available and the

study’s outcome have been reported

Other bias Low risk Baseline characteristics were comparable

Chiao 2009

Methods Design: Quasi-randomized trial

Random allocation: sequential random selection of study sites

Date: unclear

Participants Number: 980 FSWs at baseline and 903 at posttest FSWs.

Inclusion criteria: FSWs who were employed at establishment and who were self-reported

ever engaging in commercial sex

Exclusion criteria: FSWs who failed to provide key information about HIV testing and

condom use behavior

Setting: Establishment-based

Coutry: Philippines

Interventions Intervention: (1) peer education, (2) manager training, (3) a combination of peer edu-

cation and manager training

1) The peer education intervention was implemented in Legaspi. A 5-day training in-

cluded basic information on STIs and HIV, modes of transmission, interpersonal rela-

tionship with peers and clients in the work establishment, sexual negotiation, and role

playing/modelling was implemented to all participants

2) Manager training intervention was conducted in Cagayan de Oro, consisted of the

same topic as the peer education intervention, and training on the manager’s social

influence role was added

3) The combined intervention of peer education and manager training was implemented

to all participants in two contiguous cities in Cebu

Control: Usual care was conducted in Ilo-Ilo. Participants received standard treatment

which consisted of regular examinations at the health centers

Outcomes 1. HIV testing changes

2. Consistent condom use

3. HIV-related knowledge

4. AIDS perceived control

5. AIDS perceived severity

Notes

Risk of bias



bias)

Low risk ”The research design was a simple ran-

domised quasi-experimental approach...“


countries (Review)


Chiao 2009 (Continued)

Allocation concealment (selection bias) Low risk ”...number for each of the four sites were

placed in a box....“


bias)

All outcomes

High risk Either participants or some key study per-

sonnel were not blinded


All outcomes

Unclear risk ”24/1,004 (2%) missing data“ was reported

but reasons for missing data were not de-

scribed

Selective reporting (reporting bias) Low risk The study protocol is available and all of

the study’s outcome have been reported

Other bias Unclear risk ”The participant rates of FSWs at baseline

and posttest assessments were over 95%,

reducing the likelihood of a biased partici-

pant pool“ but Table 1 shows the baseline

and posttest were imbalance

Feldblum 2005

Methods Design: Randomized controlled trial

Random allocation: Computer-generated list stratified by study site was created by using

a block permuted approach with block size of 20, 10 and 4

Date: February to August 2001, and follow-up visits extended to June 2002

Participants Number: 1000 FSWs (500 FSWs at each site)

Inclusion criteria: 18 years or older, self identified active sex workers with no signs of

pregnancy

Setting: Dispensary-based

Country: Madagascar

Interventions ”Peer education plus clinic based counselling“ versus ”peer counselling only“

Intervention: FSWs received three bimonthly clinic counselling sessions. A 15-minute

counselling session involved a two way exchange of information on the following com-

ponents: individual risk assessment; transmission and verification of basic knowledge

about STIs and HIV; dual protection; demonstration of condom use with opportunity

to practice using models; counselling tailored to individual circumstances; reinforcement

of skill for negotiating condom use; and promotion of the ”no condom = no sex“ policy

Control: FSWs received condom promotion and risk reduction counselling delivered by

peer educators

Outcomes 1. STI prevalence

2. Consistent male/female or male and female condom use

Notes


countries (Review)


Feldblum 2005 (Continued)

Risk of bias



bias)

Low risk ”Participants were individually randomised

to receive condom counselling....“

Allocation concealment (selection bias) Low risk A computer generated random allocation

list stratified by study site was created at

FHI using a block permuted approach with

block sizes of 20, 10 and 4


bias)

All outcomes

Low risk ”....STI testing and reporting were done by

laboratory staff blinded to group assign-

ment“


All outcomes

Unclear risk No reasons for missing data provided

Selective reporting (reporting bias) Low risk The study protocol is available and all the


Other bias Low risk The intervention and control groups are

balance

Fontanet 1998


Random allocation: numbered table

Date: October 1994 to January 1995

Participants Number: 504 FSWs working at 71 sex establishments

Inclusion criteria: Women were at least 18 years of age, were not using a diaphragm,

a cervical cap nor vaginal spermicides, had no evidence of intravenous drug use, were

willing to use the condoms as instructed, and were willing to keep a pictorial coital log

recording all sexual acts with clients and non-commercial sexual partners

Setting: Sex establishments

Country: Thailand

Interventions ”Promotion of female and male condom use“ and ”promotion of male condom use“

Intervention: FSWs proposed a male condom to their clients as a first choice, but have the

option of using a female condom when clients refused or could not use a male condom

Control: FSWs proposed a male condom to their clients, and, if clients refused or could

not use male condoms, they were instructed not to have sex

Outcomes 1. STIs incidence



countries (Review)


Fontanet 1998 (Continued)

Notes

Risk of bias



bias)

Low risk ”The randomisation was into two groups..

..“

Allocation concealment (selection bias) Low risk ”The randomisation procedure relied on a

table of random numbers matched with the

list of sex establishments, and took into ac-

count...“


bias)

All outcomes

Unclear risk The study did not address this.


All outcomes

Low risk 4.3% (11) in female/male group and 11.

7% (33) in male condom group were loss

to follow up, but no statistically significant

differences on the baseline characteristics

of the FSWs without follow-up in the two

trial groups

Selective reporting (reporting bias) Low risk Study procedure and all study’s outcome

have been reported

Other bias Low risk Intervention and control groups compara-

ble at baseline are balance

Ghys 2001


Random allocation: unclear

Date: June 1994 to November 1997

Participants Number: 542 FSWs

Inclusion criteria: HIV negative or HIV-2 seropositive when tested in the first the screen-

ing

Setting: HIV/STD clinic

Country: Cote d’Ivoire

Interventions ”Intensive STI screening“ versus ”basic STI screening“

Intervention: FSWs randomised to the intensive STI screening and treatment strategy

were examined every month, regardless of the presence of symptoms

Control: FSWs randomised to the basic STI screening and treatment strategy were

examined only if they reported following symptoms: vaginal discharge, lower abdominal


countries (Review)


Ghys 2001 (Continued)

pain or a genital ulcer

Outcomes 1. HIV incidence

2. STI prevalence


Notes

Risk of bias



bias)

Low risk ”....women were randomised to either an

intensive or a basic STI screening and treat-

ment strategy“

Allocation concealment (selection bias) Unclear risk The study did not address this


bias)

All outcomes

Unclear risk The study did not address this


All outcomes

Low risk Reason for missing outcome data unlikely

to be related to true outcome



Other bias Low risk The intervention and control groups com-

parable at baseline are balance

Gutierrez 2010

Methods Design: Cluster-randomized controlled trial


Date: Between 2003/2004 and 2007

Participants Number: 3,442 FSws and 2,786 MSM; we excluded MSM data.

Setting: Community-based.

Country: India.

Interventions ”Community empowerment“ versus ”standard care“

Intervention: The community-based (Frontiers Prevention Project) set out to empower

target populations by improving advocacy within these groups, and increasing commu-

nity awareness, by combining the provision of a comprehensive package of prevention

interventions implemented on the appropriate scale

Control: Participants in the control group received intervention but not intensive


countries (Review)


Gutierrez 2010 (Continued)

Outcomes 1. STIs prevalence


3. HIV knowledge

4. HIV testing

Notes

Risk of bias



bias)

Low risk 24 geographically distinct sites randomised

as either FPP or no FPP

Allocation concealment (selection bias) Unclear risk The study did not address this.


bias)

All outcomes

High risk No blinding


All outcomes

High risk The study did not address the missing data

Selective reporting (reporting bias) Unclear risk The study protocol was described, but not

all the outcomes were reported

Other bias Low risk Intervention and control groups are bal-

ance

Hoke 2007

Methods Design: Randomised controlled trial

Random allocation: Computer-generated list stratified by study site was created by using

a block permuted approach with block size of 20, 10 and 4

Date: unclear


Inclusion criteria: FSWs who completed the 6-month visit of the trial’s first phase

Setting: Public dispensaries

Country: Madagascar

Interventions ”Peer education plus clinic-based counselling“ versus ”peer education alone“

Intervention: Participants randomly assigned to the peer+clinic arm additionally received

bi-monthly clinic counseling sessions. The counseling sessions involved two-way ex-

change of information concerning individual risk assessment; transmission and verifica-

tion of basic knowledge about STI/HIV; dual protection, demonstration of use of both

types of condoms with opportunity to practice using model, reinforcement of skills for

negotiating condom use, and promotion of the ”no condom=no sex“ policy


countries (Review)


Hoke 2007 (Continued)

Control: Did not receive the sessions as intensive as intervention group



Notes

Risk of bias



bias)

Low risk The 901 participants who completed the

6-month visit of the trial’s first phase were

randomised again to receive in Phase 2 ei-

ther prevention counselling by peer educa-

tors or peer counselling plus clinic-based

counselling

Allocation concealment (selection bias) Low risk Group assignment was guided by a com-

puter-generated random allocation list

stratified by study site created at FHI us-

ing a block permuted approach with block

sizes of 20, 10 and 4


bias)

All outcomes

Low risk Clinic staff and participants were not

blinded to group assignment, but labora-

tory staff were blinded


All outcomes

Unclear risk Peer-only promotion group: 19 FSWs were

lost to follow up, and peer+clinic promo-

tion group: 26 FSWs were lost to follow up.

Reasons for missing data are not described




balance

Li 2006



Date: unclear


Inclusion criteria: Women who worked in the establishments

Setting: Restaurants, barbershop and hair-washing rooms in suburban area

Country: China


countries (Review)


Li 2006 (Continued)

Interventions ”VCT intervention“ versus ”standard care of STIs“

Intervention: Two sessions of VCT were conducted; 1) a 25-minute pretest session before

STI screening to achieve risk reduction goals, 2) a 20-minute posttest session after STI

screening to discuss STI test result, and assess the study outcomes

Control: a-15 minute standard care, including messages about STI testing, treatment

and prevention



3. HIV/STIs knowledge

Notes

Risk of bias



bias)

Low risk Random selected samples were assigned to

either the intervention group or the control

group

Allocation concealment (selection bias) Unclear risk The study did not address this allocation

concealment


bias)

All outcomes



All outcomes

Low risk 54 women were not enrolled in the trial

and 122 women were lost to follow up, but

reasons for missing outcome data unlikely

to be related to true outcome

Selective reporting (reporting bias) Low risk The study procedure is available and all


Other bias High risk Demographic characteristics and baseline

outcome measures were significantly differ-

ent


countries (Review)


Markosyan 2010

Methods Design: Randomised controlled trial

Random allocation: Flip of a coin

Date: Between August 2007 and July 2008

Participants Number: 120 FSWs; 54 in intervention group and 66 in control group

Inclusion criteria: Female, 18 years of age or older, trading sex for money in the past 7

days, and written informed consent

Setting: Snowball and street-based outreach

Country: Armenia

Interventions ”Social cognitive theory“ versus ”no intervention“

Intervention: FSWs were asked to complete the questionnaires including the Informa-

tion-Motivation-Behavioural Skills Model, the Health Belief Model, and the theory of

Gender and Power. 2-hour intervention emphasized empowerment, self-efficacy to per-

suade clients to use condoms, and condom application skills implemented by health

educator

Control: FSWs did not receive any intervention but were scheduled for their follow-up

appointment and were free to leave the study site

Outcomes 1. Consistent condom use

2. HIV knowledge

3. Psychosocial barriers to condom use

Notes

Risk of bias



bias)

Low risk FSWs were randomly assigned to either the

HIV prevention or a wait-list control con-

dition through the flip of coin

Allocation concealment (selection bias) Low risk ”...through the flip of coin“


bias)

All outcomes

Low risk At baseline, 3- and 6-month follow-up, the

interviewer remained blind to participant’s

group assignment


All outcomes

Unclear risk The study did not address this outcome

Selective reporting (reporting bias) Low risk All outcomes reported

Other bias Low risk Randomization has essentially balanced the

characteristics of the intervention and con-

trol groups


countries (Review)


Patterson 2006


Random allocation: computer generated

Date: Between January 2004 and January 2006

Participants Number: 270 FSWs, 137 of whom were randomised to the intervention (denoted SSS)

and 133 to the comparison condition (CDC). An additional 342 women were recruited

in Ciudad Juarez and randomised to either SSS (N=172) or CDC (N=170)

Inclusion criteria: Self-identified as sex workers, at least 18 years old, reported had un-

protected vaginal, oral or anal sex with a client at least once in the previous month

Exclusion criteria: consistent condom use or dental dam for vaginal, oral and anal sex

with clients in the previous month, employed as a FSWs for less than one month, and

under 18 years of age

Setting: Venue- and street-based outreach. The interventions were carried out at a com-

munity-based clinic, mobile-clinic, private clinic and a government-operated Municipal

Medical Clinic

Country: Mexico

Interventions ”Social cognitive theory“ versus ”standard care“

Intervention: FSWs were interviewed face-to-face, to complete and covered a range of

topic including sexual risk behavior, working conditions, financial need, victimization

and trauma, use of alcohol and illicit drugs, social support, social influence, life expe-

riences, mood, self-esteem, social cognitive theory, social-demographic characteristics,

physical health variables, and psychiatric health variables

Control: 30 to 40 minutes face-to-face standard counselling focused on personal risk

assessment, cultural identity assessment, and strategies for reducing personal risk


2. Sexual activity outcomes

3. Drug use

4. Alcohol use

Notes

Risk of bias



bias)

Low risk ”....participants were randomly assigned to

either Proyecto Comparte Sexo Mas Seguro

or time-equivalent voluntary HIV coun-

selling and testing“

Allocation concealment (selection bias) Low risk ”....using a fixed, computer generated ran-

domisation scheme“


bias)

All outcomes

Unclear risk No information


countries (Review)


Patterson 2006 (Continued)


All outcomes

Low risk No missing data

Selective reporting (reporting bias) Low risk All outcome have been reported

Other bias Low risk There were no differences between the in-

tervention and comparison conditions

Patterson 2008


Random allocation: Computer-generated randomisation scheme.

Date: Between January 2004 and January 2006.

Participants Number: 924 FSWs (474 in Tijuana and 450 in Ciudad Juarez)

Inclusion criteria: 18 years old or older, giving informed consent, and having traded sex

for drugs, money or other material benefit within the previous two months and were

also required to have had unprotected vaginal sex with at least a client in the previous

two months

Exclusion criteria: FSWs those who reported had previously tested HIV positive

Setting: Venue- and street-based.

Country: Mexico

Interventions ”Social cognitive theory“ versus ”standard counselling“

Intervention: Motivational interviewing techniques to elicit information on the partic-

ipants’ current situation and motivations, and to increase their motivation to practice

safer sex. Four main areas were addressed: 1) motivations for practicing safer sex versus

those for practicing unsafe sex; 2) barriers to condom use; 3) techniques for negotiating

safer sex with clients; 4) enhancement of social supports

Control: A face-to face, time equivalent didactic presentation of prevention materials

extracted from the US Centers for Disease Control and Prevention revised guidelines

for HIV counselling, testing, and referral and from Mexico’s National Center for AIDS

Studies. Counseling session was focus on personal risk assessment, cultural identity

assessment, and strategies for reducing personal risk


2. STIs incidence


4. Sexual activity outcomes

5. Injected drug use

Notes

Risk of bias



countries (Review)


Patterson 2008 (Continued)


bias)

Low risk ”We randomised the participants to either

the intervention using a fixed, computer-

generated randomization scheme“

Allocation concealment (selection bias) Low risk ”......computer-generated randomisation

scheme“


bias)

All outcomes



All outcomes

Unclear risk No reasons for missing data provided

Selective reporting (reporting bias) Low risk The study protocol is available and all of

the study’s pre-specified outcomes that are

of interest in the review have been reported

Other bias Low risk Baseline characteristics are balance

Ray 2001


Random allocation: Numbered sealed envelopes

Date: unclear


Inclusion criteria: were residents of Harare, over 18 years old, had at least three different

paying clients in the previous month, written informed consent

Setting: Brothels

Country: Zimbabwe

Interventions ”Promotion of female and male condom use“ versus ”promotion of male condom use

alone“

Intervention: FSWs received a 30-minute education sessions on how to use female and

male condoms, and asked to practice how to use that. Additionally, a 10-hour focus

group discussion were held with 3-4 FSWs per group

Control: FSWs were shown how to use male condoms.


2. STIs incidence

3. Consistent condom

Notes

Risk of bias



countries (Review)


Ray 2001 (Continued)


bias)

Low risk ”....were used to assign women to group A

or B randomly“

Allocation concealment (selection bias) Low risk ”Consecutively numbered sealed envelopes

containing group assignments (generated

by random number tables and balanced in

blocks of 10)....“


bias)

All outcomes

Unclear risk The study did not address this .


All outcomes

Low risk Only 48% completed five follow-up visits,

but there were no differences in follow-up

rates between group A and group B

Selective reporting (reporting bias) Low risk Study procedures and frequency of follow-

up were identical for both study groups. All


Other bias Low risk There were no differences in baseline de-

mographic

Sherman 2010


Random allocation: by random sequences of block sizes of 24 participants

Date: Between August and September 2008

Participants Number: 128 FSWs were screened for study eligibility; 104 (81%) were eligible for

randomization

Inclusion criteria: aged over 18, reported having participated in transactional sex over

the past 12 months, and able to provide written informed consent

Exclusion criteria: FSWs who participated in an HIV prevention program or a study in

the prior 12 months

Setting: Street-based

Country: India

Interventions ”Microenterprise plus education“ versus ”education alone“

Intervention: 8 hours of HIV prevention education taught by health educators plus 100

hours of tailoring training taught by master tailors. The training occurred over the course

of a month

Control: 8 hours prevention training (the same curriculum as the intervention arm) and

delivered by the same facilitator


2. Number of sex partners


countries (Review)


Sherman 2010 (Continued)

Notes

Risk of bias



bias)

Low risk ”The randomised trial was...“

Allocation concealment (selection bias) Low risk ”...participants were given an opaque,

sealed envelope containing computer-gen-

erated random assignments, prepared by

the study’s statistical team prior to the

beginning of recruitment. Randomization

was blocked with random sequences of

block sizes of 24 participants to ensure bal-

ance of 8-12 between the two arms“


bias)

All outcomes

Low risk ”Interviewers were blind to the partici-

pants’ study arm“


All outcomes

Low risk In the 6-month follow-up visit, control arm

participants were 48 (96%)

Selective reporting (reporting bias) Unclear risk Drug/alcohol outcomes have not been re-

ported


comparable at baseline

Swendeman 2009


Random allocation: random number table

Date: 2000-2001


Inclusion criteria: FSWs those who live and work in selected communities and provided

informed consent

Setting: Community-based

Country: India

Interventions ”Empowerment intervention (Sonagachi project)“ versus ”standard care“

Outcomes 1) HIV knowledge

2) Skills of risk and protective factors


countries (Review)


Swendeman 2009 (Continued)

Notes

Risk of bias



bias)

Low risk Participants were randomly assigned to an

intervention and control condition

Allocation concealment (selection bias) Low risk Allocation using a random number table


bias)

All outcomes

Unclear risk The study did not address this outcome


All outcomes

High risk The missing data was likely related to out-

comes


study’s outcome have bee reported

Other bias Low risk “Since this loss to follow-up was likely re-

lated to outcomes, it would not violate the

less stringent missing at random (MAR)

requirement for unbiased estimates in re-

peated-measure analysis”

Wechsberg 2006



Date: unclear


Inclusion criteria: self-identified as a Black South African women, 18 years old or older,

either had a positive urine test for cocaine or reported recent substance use and sex

trading, had multiple partners and were provided written informed consent

Setting: Hotels, apartments, and informal settlements.

Country: South Africa

Interventions ”Social cognitive theory“ versus ”standard counselling“

Intervention: The intervention was based on principles of social cognitive theory, gender

theory, and empowerment. In additional, HIV-related education was tailored, and a risk

reduction and toiletry kit were given

Control: An adapted version of the revised NIDA standard intervention, which consisted

of two private one-hour education and skills building sessions


countries (Review)


Wechsberg 2006 (Continued)


2. Drug use

3. Alcohol use

Notes

Risk of bias



bias)

Low risk ”....were randomly assigned to a modified.

...“

Allocation concealment (selection bias) Unclear risk The study did not address.


bias)

All outcomes

High risk Not blinding


All outcomes

High risk The study did not address the missing data

(13 of 93)


study’s outcome have bee reported

Other bias Unclear risk The is no information about the compari-

son of demographic characteristics between

intervention group and control group

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

Abellanosa 1996 Pre-posttest study to examine the effectiveness of antibiotics to prevent STIs and HIV, but the participants

were not randomly assigned

Bakare 2002 Participants of this study were random selected of FSWs as an intervention group but the control group

was not FSWs

Barrington 2009 The study explored the relationship between social network norms and condom use among male partners

of FSWs in La Romana, Dominican Reprublic, but the participants were not randomly assigned to the

intervention and control groups

Beattie 2010 A baseline-follow up study examined if violence against FSWs is associated with reduced condom use and

increased STI/HIV risk by using polling booth surveys and integrated behavioral-biological assessment,


countries (Review)


(Continued)

but the participants were not randomly assigned to these two interventions

Bhave 1995 A controlled intervention trial, with measurements before and after an HIV intervention underwent a 6-

month program of educational videos targeting sex workers and madams in the brothels of Bombay, but

the participants were not randomly assigned to the intervention and control groups

Dandona 2005 Not randomly assigned

Egger 2000 Not all of the participants were sex workers

Feldblum 2007 A randomised controlled trial to estimate the pregnancy incidence rate among Madagascar FSWs par-

ticipating in an intervention trial promoting use of male and female condoms and assess the influence

of various predictive factors on pregnancy risk. The data of this study was the secondary data collected

during a trial conducted in the previous study (Feldblum 2005)

Fonck 2000 This study was the baseline findings of a trial of antibiotic prophylaxis to prevent STIs and HIV-1 in a

cohort of Nairobi FSWs. The participants were not randomly assigned to peer education intervention

Ford 2002 This study assessed the impact of the educational intervention among FSWs in Bali, Indonesia. The

participants were divided by areas, high and low program effort to receive whether the health belief model

or the social cognitive theory

Gangopadhyay 2005 A comparative cross-sectional study of randomly selected FSWs in Calcutta, India compared rates of

STDs between the Sonagachi Project and National AIDS Control Organization (NACO) interventions,

but the participants were not randomly assigned

Gorbach 2000 This study quantified persons randomly selected from clusters of military, police, and motorcycle taxi

drivers in five cities of Cambodia, associations between behaviours, social context and active bridging

Guerena-Burgueno 1991 A pre-post test counselling assessed the prevalence of HIV-1 infection among high risk populations in

Tijuana, Mexico and information on risk behavior was obtained from 1,069 individuals

Hoque 2009 This study described behavioral profile of rickshaw pullers in Dhaka city and identified the correlates

for having sex with FSWs in order to focus HIV prevention intervention, but the participants were not

randomly assigned to receive the interventions

Kamali 2003 A community randomised trial which assessed the effect of behavioral intervention on incidence of HIV-

1, condom use and other STIs, but the participants not particularly FSWs

Kaul 2002 Baseline result of RCT study of treatment intervention

Kaul 2004 HIV treatment intervention

Kaul 2007 Placebo controlled trial of monthly oral azythromycin therapy (Azythromycin versus placebo), without

any additional behavior intervention

Kerrigan 2006 A pre-intervention post-intervention of two environmental-structural interventions of increases in con-

dom use with clients in reducing risks of HIV and STIs among FSWs in the Dominican Republic


countries (Review)


(Continued)

Khan 2008 This study assessed HIV behavioral intervention among male transgender, however the participants were

not randomly assigned

Larios 2009 This study examined the applicability of the Social Ecological Model for explaining condom use in a

sample of FSWs participating in a behavioral intervention to increase condom use in Tijuana, Mexico.

However, the participants were not randomly assigned to the intervention and control groups

Lau 2008 Behavioural surveillance study that investigated changes in HIV-related behaviours including condom

use and HIV-related knowledge among randomly selected FSWs, but the participants were not randomly

assigned to the intervention

Leonard 2000 Preintervention and postintervention study of a peer-led HIV prevention education and condom pro-

motion program among transport workers in Kaolack, Senegal

Maticka-Tyndale 1997 This study explored the social-cultural contexts and patterns of Thai men’s commercial sexual activity

including to use a condom

McClelland 2008 A randomised controlled study to test the efficacy of monthly periodic presumptive oral treatment with 2 g

of metronidazole plus 150 mg of flunocazole. Otherwise, this intervention did not include any behavioral

intervention

Ngugi 1988 A controlled clinical trial consisting a three arms in a situation where a true randomized controlled trial

would have been unethical. This study described the effect of an AIDS education program and the

distribution of free condoms and condom use among the prostitutes

Ngugi 1998 Baseline-follow up study that shown the effect of peer mediated intervention program among random

sample of FSWs to reduce STD and HIV transmission in Kenya and Zimbabwe

Ngugi 2007 Community-based post-trial resurvey of RCT among FSWs in Nairobi examined the sustainability of

such interventions

Nzila 1991 This study confirmed the HIV and STDs prevalence decreased by using condom, but the FSWs were

not randomly assigned to the intervention

Ramesh 2010 A probability-based sampling; conventional cluster sampling

Reza-Paul 2008 This study comprised two cross-sectional surveys, rather than an RCT

Rou 2007 Pre/post intervention trial set to evaluate the changes in condom use with the last three clients, and the

prevalence of chlamydia and gonorrhea

Schwandt 2006 A peer education intervention examined the practices of anal intercourse and dry sex within a cohort of

FSWs in Kenya, but the participants were not randomly assigned to the intervention and control groups

Strathdee 2009 The study used Social Cognitive Theory to determine whether increases in condom use were predicted

by social cognitive theory and injection drug user status among women randomised to this intervention


countries (Review)


(Continued)

Stulhofer 2009 The study compared and analysed the correlates of consistent condom use with clients among two clusters

FSWs in Croatia, but the participants were not randomly assigned

Walden 1999 The study measured the impact of peer educational and condom use among FSWs and truck drivers in

Malawi, but the participants were not randomly assigned to intervention and control groups

Weir 1998 An RCT used structured closed-ended questionnaires of estimating condom use, but cannot include

because there is no any follow-up data

Weir 1999 A randomized controlled study used structured closed-ended questionnaires of estimating condom use

without any follow-up

Yadav 2005 A cohort study examined the association of baseline social-demographic variables with the uptake and

durability of an HIV-1 behavioral risk reduction program in high-risk seronegative Kenyan FSW. This

study similar to Kaul 2004

Yi 2010 The study compared the risk factor and condom use among three groups of FSWs but subjects not

randomly assigned


countries (Review)


D A T A A N D A N A L Y S E S

Comparison 1. Social cognitive theory versus standard counseling for promotion of condom use

Outcome or subgroup titleNo. of

studies

No. of

participants Statistical method Effect size

1 HIV incidence among FSWs at

6-month

1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only

2 STIs incidence among FSWs at

6-month

1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

2.1 Syphilis incidence among

FSWs at 6-month

1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]

2.2 Gonorrhoea incidence

among FSWs at 6-month


2.3 Chlamydia incidence



2.4 Any STIs among FSWs at

6-month


3 Consistent condom use by FSWs

at 6-month

2 804 Risk Ratio (M-H, Random, 95% CI) 1.14 [1.07, 1.21]

4 Mean number of condom use by

FSWs at past month

1 Mean Difference (IV, Fixed, 95% CI) Totals not selected

4.1 Vaginal sex using condoms

by FSWs with regular clients at

past month

1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]

4.2 Anal sex using condoms


past month


4.3 Oral sex using condoms


past month


4.4 Vaginal sex using condoms

by FSWs with non-regular

clients at past month


4.5 Anal sex using condoms




4.6 Oral sex using condoms




4.7 Vaginal sex using a

condom by FSWs with spouse

or steady partner


4.8 Anal sex using a condom

by FSWs with spouse or steady

partner at past month



countries (Review)


4.9 Oral sex using a condom

by FSWs with spouse or steady

partner at past month


5 Mean number of protected sex


1 Mean Difference (IV, Fixed, 95% CI) Subtotals only

6 Sexual activity outcomes by

FSWs at past month


6.1 Had vaginal sex at past

month


6.2 Had anal sex at past

month


6.3 Had vaginal and anal sex

at past month


6.4 Had oral sex at past month 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]

7 Injected drug use among FSWs

at 6-month


8 Alcohol use among FSWs at

6-month


9 Drug use among FSWs at

6-month


Comparison 2. Social cognitive theory versus no intervention for promotion of condom use


studies

No. of


1 Consistent condom use by FSWs 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

1.1 Consistent condom use by

FSWs with clients at 6-month


1.2 Consistent condom use by

FSWs with clients in the past

7-day


1.3 Consistent application of

condom by FSWs


2 HIV knowledge among FSWs at

6-month


3 Psychosocial barriers to condom

use among FSWs at 6-month


3.1 Condom barriers among

FSWs at 6-month


3.2 Condom use self-efficacy




countries (Review)


Comparison 3. Community empowerment versus standard care for promotion of condom use


studies

No. of


1 STIs prevalence among FSWs at

36-month


1.1 Syphilis prevalence at

6-month


1.2 HSV-2 prevalence at

6-month



with regular partners



with clients


4 Change in 100% condom use by

FSWs


4.1 Change in condom use:

baseline to 6 months








5 Change in any condom use by

FSWs











6 HIV knowledge among FSWs 2 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

6.1 Knowledge that condom

could prevent STI (at

16-month)


6.2 Knowledge that condom

could prevent AIDS (at

16-month)


6.3 Knowledge about condom

(at 36-month)


7 HIV testing by FSWs at

36-month


8 Skills of risk and protective

factors among FSWs at

16-month


8.1 Know that she at risk for

STIs


8.2 Refused a client by FSWs

for a particular sex



countries (Review)


8.3 Refused sex by FSWs with

a client who refused condom


Comparison 4. Microenterprise plus education intervention versus education alone for reducing the number of

sex partners


studies

No. of



at 6-month


2 Number of FSWs’ partners at

6-month


2.1 Number of sex partners at

6-month


2.2 Number of sex exchange

partners at 6-month


Comparison 5. Peer education versus standard care for promoting HIV testing and condom use


studies

No. of


1 Mean number of consistent

condom use by FSWs at

6-month


2 Mean score of HIV knowledge




6-month


4 AIDS perceived control by

FSWs at 6-month


4.1 To avoid AIDS

contraction


4.2 To control AIDS

contraction


4.3 To lower chance of getting

AIDS


5 AIDS perceived severity by

FSWs at 6-month


5.1 Chance of getting AIDS 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]

5.2 Worry about getting AIDS 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]


countries (Review)


Comparison 6. Peer education plus clinic based counseling versus peer education only for promotion of condom

use


studies

No. of


1 STIs prevalence among FSWs 2 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

1.1 Chlamydia prevalence









1.4 Gonorrhoea prevalence









1.7 Trichomonas prevalence









1.10 Any STIs prevalence







among at 18-month


2 Male condom use with clients 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

2.1 Protected coital acts at

2-month



4-month



6-month


3 Male condom use with

non-paying partners


3.1 Male condom use in the

last coital act at 2-month








4 Female condom use by FSWs

with clients


4.1 Female condom use by




countries (Review)


4.2 Female condom use by



5 Male and female condom use

with clients


5.1 Male and female condom

use with clients at 12-month


5.2 Male and female condom

use at 18-month


6 Male and female condom use in

the last sex with non-paying

partner


6.1 Male and female

condom use in the last sex

with non-paying partner at

12-month


6.2 Male and female

condom use in the last sex

with non-paying partner at

18-month


Comparison 7. Peer education plus manager training versus standard care for promoting HIV testing and condom

use


studies

No. of




6-month






6-month



FSWs at 6-month


4.1 To avoid AIDS

contraction


4.2 To control AIDS

contraction


4.3 To lower chance of getting

AIDS


5 AIDS perceived severity by

FSWs at 6-month





countries (Review)


Comparison 8. Manager training versus standard care for promoting HIV testing and condom use


studies

No. of




6-month






6-month



FSWs at 6-month


4.1 Perception that condom

can avoid AIDS contraction



can control AIDS contraction



can lower chance of getting

AIDS


5 Mean score of perceived

themselves at greater risk (by

FSWs at 6-month)




Comparison 9. Promotion of female and male condom versus promotion of male condom


studies

No. of



3-month


2 STIs incidence among FSWs at

3-month


2.1 Trichomoniasis incidence



2.2 Chlamydia incidence



2.3 Gonorrhea incidence



3 Consistent male condom use at

3-months


4 Consistent female condom use

by FSWs at 3-month


5 Consistent female condom use

by FSWs at 24-month



countries (Review)


6 Consistent male condom use at

24-month


Comparison 10. Intensive STI screening versus basic STI screening to control STI


studies

No. of



6-month



6-month


2.1 Trachomonas vaginalis

prevalence among FSWs at

6-month


2.2 Gonorrhoeae prevalence



2.3 Chlamydia trachomatis

prevalence among FSWs at

6-month



at 6-month


Comparison 11. VCT versus standard care of STI for increasing condom use


studies

No. of



6-month


1.1 Syphilis prevalence among

FSWs at 6-month











1.5 Genital warts prevalence




with clients at 6-month


3 HIV/STI knowledge among

FSWs at 6-month



countries (Review)


Analysis 1.1. Comparison 1 Social cognitive theory versus standard counseling for promotion of condom

use, Outcome 1 HIV incidence among FSWs at 6-month.

Review: Behavioral interventions to reduce the transmission of HIV infection among sex workers and their clients in low- and middle-income countries

Comparison: 1 Social cognitive theory versus standard counseling for promotion of condom use

Outcome: 1 HIV incidence among FSWs at 6-month

Study or subgroup Experimental Control Risk Ratio Risk Ratio

n/N n/N M-H,Fixed,95% CI M-H,Fixed,95% CI

Patterson 2008 0/341 4/368 0.12 [ 0.01, 2.22 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]

Total events: 0 (Experimental), 4 (Control)

Heterogeneity: not applicable

Test for overall effect: Z = 0.0 (P < 0.00001)

0.01 0.1 1 10 100

Favours experimental Favours control


use, Outcome 2 STIs incidence among FSWs at 6-month.



Outcome: 2 STIs incidence among FSWs at 6-month



1 Syphilis incidence among FSWs at 6-month

Patterson 2008 6/341 14/368 0.46 [ 0.18, 1.19 ]

2 Gonorrhoea incidence among FSWs at 6-month

Patterson 2008 9/341 14/368 0.69 [ 0.30, 1.58 ]

3 Chlamydia incidence among FSWs at 6-month

Patterson 2008 13/341 17/368 0.83 [ 0.41, 1.67 ]

4 Any STIs among FSWs at 6-month

Patterson 2008 20/341 38/368 0.57 [ 0.34, 0.96 ]

0.01 0.1 1 10 100



countries (Review)



use, Outcome 3 Consistent condom use by FSWs at 6-month.



Outcome: 3 Consistent condom use by FSWs at 6-month

Study or subgroup Favours experimental Control Risk Ratio Weight Risk Ratio

n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Patterson 2008 257/307 254/337 65.0 % 1.11 [ 1.03, 1.20 ]

Wechsberg 2006 78/80 66/80 35.0 % 1.18 [ 1.06, 1.32 ]

Total (95% CI) 387 417 100.0 % 1.14 [ 1.07, 1.21 ]

Total events: 335 (Favours experimental), 320 (Control)

Heterogeneity: Tau2 = 0.0; Chi2 = 0.91, df = 1 (P = 0.34); I2 =0.0%

Test for overall effect: Z = 3.93 (P = 0.000086)

Test for subgroup differences: Not applicable

0.01 0.1 1 10 100



countries (Review)



use, Outcome 4 Mean number of condom use by FSWs at past month.



Outcome: 4 Mean number of condom use by FSWs at past month

Study or subgroup Experimental ControlMean

DifferenceMean

Difference

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI

1 Vaginal sex using condoms by FSWs with regular clients at past month

Patterson 2006 309 11 (14) 303 9 (13) 2.00 [ -0.14, 4.14 ]

2 Anal sex using condoms by FSWs with regular clients at past month

Patterson 2006 309 0 (1) 303 0 (1) 0.0 [ -0.16, 0.16 ]

3 Oral sex using condoms by FSWs with regular clients at past month

Patterson 2006 309 4 (10) 303 3 (6) 1.00 [ -0.30, 2.30 ]

4 Vaginal sex using condoms by FSWs with non-regular clients at past month

Patterson 2006 309 38 (47) 303 34 (48) 4.00 [ -3.53, 11.53 ]

5 Anal sex using condoms by FSWs with non-regular clients at past month

Patterson 2006 309 0 (3) 303 0 (2) 0.0 [ -0.40, 0.40 ]

6 Oral sex using condoms by FSWs with non-regular clients at past month

Patterson 2006 309 12 (34) 303 8 (22) 4.00 [ -0.53, 8.53 ]

7 Vaginal sex using a condom by FSWs with spouse or steady partner

Patterson 2006 113 1 (3) 106 1 (2) 0.0 [ -0.67, 0.67 ]

8 Anal sex using a condom by FSWs with spouse or steady partner at past month

Patterson 2006 113 0 (1) 106 0 (1) 0.0 [ -0.27, 0.27 ]

9 Oral sex using a condom by FSWs with spouse or steady partner at past month

Patterson 2006 113 1 (4) 106 0 (1) 1.00 [ 0.24, 1.76 ]

-100 -50 0 50 100



countries (Review)



use, Outcome 5 Mean number of protected sex among FSWs at 6-month.



Outcome: 5 Mean number of protected sex among FSWs at 6-month


DifferenceMean

Difference


Patterson 2008 307 51.6 (49.1) 337 43.9 (49.1) 7.70 [ 0.11, 15.29 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



use, Outcome 6 Sexual activity outcomes by FSWs at past month.



Outcome: 6 Sexual activity outcomes by FSWs at past month


DifferenceMean

Difference


1 Had vaginal sex at past month

Patterson 2008 409 57 (59.8) 460 54.8 (55.2) 2.20 [ -5.48, 9.88 ]

2 Had anal sex at past month

Patterson 2008 409 8.4 (13.3) 460 9.6 (16.5) -1.20 [ -3.18, 0.78 ]

3 Had vaginal and anal sex at past month

Patterson 2008 409 58.6 (60.7) 460 56.8 (56.6) 1.80 [ -6.03, 9.63 ]

4 Had oral sex at past month

Patterson 2006 309 18 (42) 303 12 (30) 6.00 [ 0.23, 11.77 ]

-100 -50 0 50 100



countries (Review)



use, Outcome 7 Injected drug use among FSWs at 6-month.



Outcome: 7 Injected drug use among FSWs at 6-month



Patterson 2008 66/409 88/460 0.84 [ 0.63, 1.13 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



use, Outcome 8 Alcohol use among FSWs at 6-month.



Outcome: 8 Alcohol use among FSWs at 6-month

Study or subgroup Experimental Control Risk Ratio Weight Risk Ratio

n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Patterson 2006 34/309 47/303 88.8 % 0.71 [ 0.47, 1.07 ]

Wechsberg 2006 4/80 8/80 11.2 % 0.50 [ 0.16, 1.59 ]

Total (95% CI) 389 383 100.0 % 0.68 [ 0.46, 1.01 ]





0.01 0.1 1 10 100



countries (Review)



use, Outcome 9 Drug use among FSWs at 6-month.



Outcome: 9 Drug use among FSWs at 6-month


n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Patterson 2006 10/309 22/303 37.6 % 0.45 [ 0.21, 0.93 ]

Wechsberg 2006 26/80 32/80 62.4 % 0.81 [ 0.54, 1.23 ]

Total (95% CI) 389 383 100.0 % 0.65 [ 0.36, 1.16 ]


Heterogeneity: Tau2 = 0.10; Chi2 = 2.06, df = 1 (P = 0.15); I2 =52%



0.01 0.1 1 10 100


Analysis 2.1. Comparison 2 Social cognitive theory versus no intervention for promotion of condom use,

Outcome 1 Consistent condom use by FSWs.


Comparison: 2 Social cognitive theory versus no intervention for promotion of condom use

Outcome: 1 Consistent condom use by FSWs



1 Consistent condom use by FSWs with clients at 6-month

Markosyan 2010 24/48 28/50 0.89 [ 0.61, 1.30 ]

2 Consistent condom use by FSWs with clients in the past 7-day

Markosyan 2010 29/48 33/50 0.92 [ 0.68, 1.24 ]

3 Consistent application of condom by FSWs

Markosyan 2010 5/48 8/50 0.65 [ 0.23, 1.85 ]

0.01 0.1 1 10 100



countries (Review)



Outcome 2 HIV knowledge among FSWs at 6-month.



Outcome: 2 HIV knowledge among FSWs at 6-month



Markosyan 2010 4/48 2/50 2.08 [ 0.40, 10.85 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



Outcome 3 Psychosocial barriers to condom use among FSWs at 6-month.



Outcome: 3 Psychosocial barriers to condom use among FSWs at 6-month



1 Condom barriers among FSWs at 6-month

Markosyan 2010 8/48 10/50 0.83 [ 0.36, 1.93 ]

2 Condom use self-efficacy among FSWs at 6-month

Markosyan 2010 18/48 12/50 1.56 [ 0.85, 2.89 ]

0.01 0.1 1 10 100



countries (Review)


Analysis 3.1. Comparison 3 Community empowerment versus standard care for promotion of condom use,

Outcome 1 STIs prevalence among FSWs at 36-month.


Comparison: 3 Community empowerment versus standard care for promotion of condom use

Outcome: 1 STIs prevalence among FSWs at 36-month



1 Syphilis prevalence at 6-month

Gutierrez 2010 129/1292 68/855 1.26 [ 0.95, 1.66 ]

2 HSV-2 prevalence at 6-month

Gutierrez 2010 375/1292 180/855 1.38 [ 1.18, 1.61 ]

0.01 0.1 1 10 100



Outcome 2 Consistent condom use by FSWs with regular partners.



Outcome: 2 Consistent condom use by FSWs with regular partners



Gutierrez 2010 180/855 491/1292 0.55 [ 0.48, 0.64 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



countries (Review)



Outcome 3 Consistent condom use by FSWs with clients.



Outcome: 3 Consistent condom use by FSWs with clients


n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Basu 2004 67/90 58/93 40.8 % 1.19 [ 0.98, 1.46 ]

Gutierrez 2010 1279/1292 838/855 59.2 % 1.01 [ 1.00, 1.02 ]

Total (95% CI) 1382 948 100.0 % 1.08 [ 0.86, 1.36 ]





0.01 0.1 1 10 100



Outcome 4 Change in 100% condom use by FSWs.



Outcome: 4 Change in 100% condom use by FSWs



1 Change in condom use: baseline to 6 months

Basu 2004 35/90 17/93 2.13 [ 1.29, 3.51 ]


Basu 2004 35/85 16/88 2.26 [ 1.36, 3.77 ]


Basu 2004 37/92 19/80 1.69 [ 1.06, 2.70 ]

0.01 0.1 1 10 100



countries (Review)



Outcome 5 Change in any condom use by FSWs.



Outcome: 5 Change in any condom use by FSWs




Basu 2004 21/90 21/93 1.03 [ 0.61, 1.76 ]


Basu 2004 21/85 18/88 1.21 [ 0.69, 2.10 ]


Basu 2004 36/92 16/80 1.96 [ 1.18, 3.25 ]

0.01 0.1 1 10 100



Outcome 6 HIV knowledge among FSWs.



Outcome: 6 HIV knowledge among FSWs

Study or subgroup Favours experimental Control Risk Ratio Risk Ratio


1 Knowledge that condom could prevent STI (at 16-month)

Swendeman 2009 79/100 82/100 0.96 [ 0.84, 1.10 ]

2 Knowledge that condom could prevent AIDS (at 16-month)

Swendeman 2009 67/100 41/100 1.63 [ 1.24, 2.15 ]

3 Knowledge about condom (at 36-month)

Gutierrez 2010 116/1292 85/855 0.90 [ 0.69, 1.18 ]

0.01 0.1 1 10 100



countries (Review)



Outcome 7 HIV testing by FSWs at 36-month.



Outcome: 7 HIV testing by FSWs at 36-month



Gutierrez 2010 1163/1292 744/855 1.03 [ 1.00, 1.07 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



Outcome 8 Skills of risk and protective factors among FSWs at 16-month.



Outcome: 8 Skills of risk and protective factors among FSWs at 16-month



1 Know that she at risk for STIs

Swendeman 2009 59/100 55/100 1.07 [ 0.84, 1.37 ]

2 Refused a client by FSWs for a particular sex

Swendeman 2009 62/100 53/100 1.17 [ 0.92, 1.49 ]

3 Refused sex by FSWs with a client who refused condom

Swendeman 2009 42/100 38/100 1.11 [ 0.79, 1.55 ]

0.01 0.1 1 10 100



countries (Review)


Analysis 4.1. Comparison 4 Microenterprise plus education intervention versus education alone for

reducing the number of sex partners, Outcome 1 Consistent condom use by FSWs at 6-month.


Comparison: 4 Microenterprise plus education intervention versus education alone for reducing the number of sex partners




Sherman 2010 48/50 47/49 1.00 [ 0.92, 1.09 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100


Analysis 4.2. Comparison 4 Microenterprise plus education intervention versus education alone for

reducing the number of sex partners, Outcome 2 Number of FSWs’ partners at 6-month.


Comparison: 4 Microenterprise plus education intervention versus education alone for reducing the number of sex partners

Outcome: 2 Number of FSWs’ partners at 6-month


DifferenceMean

Difference


1 Number of sex partners at 6-month

Sherman 2010 50 5 (5) 49 11.9 (10) -6.90 [ -10.02, -3.78 ]

2 Number of sex exchange partners at 6-month

Sherman 2010 50 3.1 (2) 49 5.1 (3) -2.00 [ -3.01, -0.99 ]

-100 -50 0 50 100



countries (Review)


Analysis 5.1. Comparison 5 Peer education versus standard care for promoting HIV testing and condom

use, Outcome 1 Mean number of consistent condom use by FSWs at 6-month.


Comparison: 5 Peer education versus standard care for promoting HIV testing and condom use

Outcome: 1 Mean number of consistent condom use by FSWs at 6-month


DifferenceMean

Difference


Chiao 2009 149 1.61 (1.24) 141 1.7 (0.97) -0.09 [ -0.35, 0.17 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



use, Outcome 2 Mean score of HIV knowledge among FSWs at 6-month.



Outcome: 2 Mean score of HIV knowledge among FSWs at 6-month


DifferenceMean

Difference


Chiao 2009 149 6.98 (1.3) 141 5.9 (1.43) 1.08 [ 0.76, 1.40 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



countries (Review)



use, Outcome 3 HIV testing by FSWs at 6-month.






Chiao 2009 134/149 96/141 1.32 [ 1.17, 1.50 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



use, Outcome 4 AIDS perceived control by FSWs at 6-month.



Outcome: 4 AIDS perceived control by FSWs at 6-month



1 To avoid AIDS contraction

Chiao 2009 113/149 106/141 1.01 [ 0.88, 1.15 ]

2 To control AIDS contraction

Chiao 2009 117/149 119/141 0.93 [ 0.83, 1.04 ]

3 To lower chance of getting AIDS

Chiao 2009 59/149 75/141 0.74 [ 0.58, 0.96 ]

0.01 0.1 1 10 100



countries (Review)



use, Outcome 5 AIDS perceived severity by FSWs at 6-month.



Outcome: 5 AIDS perceived severity by FSWs at 6-month


DifferenceMean

Difference


1 Chance of getting AIDS

Chiao 2009 149 1.76 (1.37) 141 3.23 (1.69) -1.47 [ -1.83, -1.11 ]

2 Worry about getting AIDS

Chiao 2009 149 2.83 (1.67) 141 3.85 (1.47) -1.02 [ -1.38, -0.66 ]

-100 -50 0 50 100


Analysis 6.1. Comparison 6 Peer education plus clinic based counseling versus peer education only for

promotion of condom use, Outcome 1 STIs prevalence among FSWs.


Comparison: 6 Peer education plus clinic based counseling versus peer education only for promotion of condom use

Outcome: 1 STIs prevalence among FSWs



1 Chlamydia prevalence among FSWs at 6-month

Feldblum 2005 52/456 73/445 0.70 [ 0.50, 0.97 ]


Hoke 2007 57/435 74/428 0.76 [ 0.55, 1.04 ]


Hoke 2007 46/409 48/409 0.96 [ 0.66, 1.40 ]

4 Gonorrhoea prevalence among FSWs at 6-month

Feldblum 2005 73/456 91/445 0.78 [ 0.59, 1.03 ]


0.01 0.1 1 10 100


(Continued . . . )


countries (Review)


(. . . Continued)Study or subgroup Experimental Control Risk Ratio Risk Ratio


Hoke 2007 41/435 36/428 1.12 [ 0.73, 1.72 ]


Hoke 2007 41/409 45/409 0.91 [ 0.61, 1.36 ]

7 Trichomonas prevalence among FSWs at 6-month

Feldblum 2005 69/456 79/445 0.85 [ 0.63, 1.15 ]


Hoke 2007 120/435 126/428 0.94 [ 0.76, 1.16 ]


Hoke 2007 130/409 108/409 1.20 [ 0.97, 1.49 ]

10 Any STIs prevalence among FSWs at 6-month

Feldblum 2005 144/456 181/445 0.78 [ 0.65, 0.93 ]

11 Any STIs prevalence among FSWs at 12-month

Hoke 2007 172/428 180/435 0.97 [ 0.83, 1.14 ]

12 Any STIs prevalence among at 18-month

Hoke 2007 166/409 156/409 1.06 [ 0.90, 1.26 ]

0.01 0.1 1 10 100



promotion of condom use, Outcome 2 Male condom use with clients.



Outcome: 2 Male condom use with clients

Study or subgroup Favours experimental Control Risk Ratio Risk Ratio


1 Protected coital acts at 2-month

Feldblum 2005 358/480 364/483 0.99 [ 0.92, 1.06 ]


Feldblum 2005 383/476 348/465 1.08 [ 1.00, 1.15 ]


Feldblum 2005 370/447 334/435 1.08 [ 1.01, 1.15 ]

0.01 0.1 1 10 100



countries (Review)



promotion of condom use, Outcome 3 Male condom use with non-paying partners.



Outcome: 3 Male condom use with non-paying partners



1 Male condom use in the last coital act at 2-month

Feldblum 2005 79/243 80/268 1.09 [ 0.84, 1.41 ]


Feldblum 2005 76/236 58/254 1.41 [ 1.05, 1.89 ]


Feldblum 2005 70/258 58/264 1.23 [ 0.91, 1.67 ]

0.01 0.1 1 10 100



promotion of condom use, Outcome 4 Female condom use by FSWs with clients.



Outcome: 4 Female condom use by FSWs with clients



1 Female condom use by FSWs with clients at 12-month

Hoke 2007 278/435 291/428 0.94 [ 0.85, 1.03 ]

2 Female condom use by FSWs with clients at 18-month

Hoke 2007 235/409 252/409 0.93 [ 0.83, 1.04 ]

0.01 0.1 1 10 100



countries (Review)



promotion of condom use, Outcome 5 Male and female condom use with clients.



Outcome: 5 Male and female condom use with clients



1 Male and female condom use with clients at 12-month

Hoke 2007 370/435 344/428 1.06 [ 1.00, 1.13 ]

2 Male and female condom use at 18-month

Hoke 2007 360/409 350/409 1.03 [ 0.97, 1.09 ]

0.01 0.1 1 10 100



promotion of condom use, Outcome 6 Male and female condom use in the last sex with non-paying partner.



Outcome: 6 Male and female condom use in the last sex with non-paying partner



1 Male and female condom use in the last sex with non-paying partner at 12-month

Hoke 2007 124/435 94/428 1.30 [ 1.03, 1.64 ]

2 Male and female condom use in the last sex with non-paying partner at 18-month

Hoke 2007 119/409 86/409 1.38 [ 1.09, 1.76 ]

0.01 0.1 1 10 100



countries (Review)


Analysis 7.1. Comparison 7 Peer education plus manager training versus standard care for promoting HIV

testing and condom use, Outcome 1 Mean number of consistent condom use by FSWs at 6-month.


Comparison: 7 Peer education plus manager training versus standard care for promoting HIV testing and condom use



DifferenceMean

Difference


Chiao 2009 415 3.04 (1.87) 141 1.7 (0.97) 1.34 [ 1.10, 1.58 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



testing and condom use, Outcome 2 Mean score of HIV knowledge among FSWs at 6-month.





DifferenceMean

Difference


Chiao 2009 415 6.79 (1.39) 141 5.9 (1.43) 0.89 [ 0.62, 1.16 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



countries (Review)



testing and condom use, Outcome 3 HIV testing by FSWs at 6-month.






Chiao 2009 94/415 96/141 0.33 [ 0.27, 0.41 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



testing and condom use, Outcome 4 AIDS perceived control by FSWs at 6-month.






1 To avoid AIDS contraction

Chiao 2009 311/415 106/141 1.00 [ 0.89, 1.11 ]

2 To control AIDS contraction

Chiao 2009 292/415 119/141 0.83 [ 0.76, 0.92 ]

3 To lower chance of getting AIDS

Chiao 2009 367/415 75/141 1.66 [ 1.42, 1.95 ]

0.01 0.1 1 10 100



countries (Review)



testing and condom use, Outcome 5 AIDS perceived severity by FSWs at 6-month.



Outcome: 5 AIDS perceived severity by FSWs at 6-month


DifferenceMean

Difference



Chiao 2009 415 2.38 (1.25) 141 3.23 (1.69) -0.85 [ -1.15, -0.55 ]


Chiao 2009 415 3.35 (1.34) 141 3.85 (1.47) -0.50 [ -0.77, -0.23 ]

-100 -50 0 50 100


Analysis 8.1. Comparison 8 Manager training versus standard care for promoting HIV testing and condom

use, Outcome 1 Mean number of consistent condom use by FSWs at 6-month.


Comparison: 8 Manager training versus standard care for promoting HIV testing and condom use



DifferenceMean

Difference


Chiao 2009 198 1.79 (1.26) 141 1.7 (0.97) 0.09 [ -0.15, 0.33 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



countries (Review)



use, Outcome 2 Mean score of HIV knowledge among FSWs at 6-month.





DifferenceMean

Difference


Chiao 2009 198 6.31 (1.44) 141 5.9 (1.43) 0.41 [ 0.10, 0.72 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




-100 -50 0 50 100



use, Outcome 3 HIV testing by FSWs at 6-month.






Chiao 2009 191/198 96/141 1.42 [ 1.26, 1.59 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



countries (Review)



use, Outcome 4 AIDS perceived control by FSWs at 6-month.






1 Perception that condom can avoid AIDS contraction

Chiao 2009 165/198 106/141 1.11 [ 0.99, 1.24 ]

2 Perception that condom can control AIDS contraction

Chiao 2009 177/198 119/141 1.06 [ 0.97, 1.15 ]

3 Perception that condom can lower chance of getting AIDS

Chiao 2009 124/198 75/141 1.18 [ 0.98, 1.42 ]

0.01 0.1 1 10 100



use, Outcome 5 Mean score of perceived themselves at greater risk (by FSWs at 6-month).



Outcome: 5 Mean score of perceived themselves at greater risk (by FSWs at 6-month)


DifferenceMean

Difference



Chiao 2009 198 2.8 (1.72) 141 3.23 (1.69) -0.43 [ -0.80, -0.06 ]


Chiao 2009 198 3.61 (1.62) 141 3.85 (1.47) -0.24 [ -0.57, 0.09 ]

-100 -50 0 50 100



countries (Review)


Analysis 9.1. Comparison 9 Promotion of female and male condom versus promotion of male condom,

Outcome 1 HIV incidence among FSWs at 3-month.


Comparison: 9 Promotion of female and male condom versus promotion of male condom




Ray 2001 0/99 3/50 0.07 [ 0.00, 1.38 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



Outcome 2 STIs incidence among FSWs at 3-month.



Outcome: 2 STIs incidence among FSWs at 3-month



1 Trichomoniasis incidence among FSWs at 3-month

Fontanet 1998 11/249 8/255 1.41 [ 0.58, 3.44 ]

2 Chlamydia incidence among FSWs at 3-month

Fontanet 1998 50/249 72/255 0.71 [ 0.52, 0.98 ]

3 Gonorrhea incidence among FSWs at 3-month

Fontanet 1998 43/249 70/255 0.63 [ 0.45, 0.88 ]

0.01 0.1 1 10 100



countries (Review)



Outcome 3 Consistent male condom use at 3-months.



Outcome: 3 Consistent male condom use at 3-months


n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Fontanet 1998 219/249 250/255 66.6 % 0.90 [ 0.85, 0.94 ]

Ray 2001 25/42 22/26 33.4 % 0.70 [ 0.52, 0.95 ]

Total (95% CI) 291 281 100.0 % 0.83 [ 0.65, 1.05 ]





0.01 0.1 1 10 100



Outcome 4 Consistent female condom use by FSWs at 3-month.



Outcome: 4 Consistent female condom use by FSWs at 3-month


n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Fontanet 1998 30/249 250/255 91.4 % 0.12 [ 0.09, 0.17 ]

Ray 2001 3/34 22/26 8.6 % 0.10 [ 0.03, 0.31 ]

Total (95% CI) 283 281 100.0 % 0.12 [ 0.09, 0.17 ]





0.01 0.1 1 10 100



countries (Review)



Outcome 5 Consistent female condom use by FSWs at 24-month.



Outcome: 5 Consistent female condom use by FSWs at 24-month



Fontanet 1998 10/68 57/60 0.15 [ 0.09, 0.28 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



Outcome 6 Consistent male condom use at 24-month.



Outcome: 6 Consistent male condom use at 24-month



Fontanet 1998 59/68 57/60 0.91 [ 0.82, 1.02 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



countries (Review)


Analysis 10.1. Comparison 10 Intensive STI screening versus basic STI screening to control STI, Outcome 1

HIV incidence among FSWs at 6-month.


Comparison: 10 Intensive STI screening versus basic STI screening to control STI




Ghys 2001 6/108 10/117 0.65 [ 0.24, 1.73 ]

0.01 0.1 1 10 100



STIs prevalence among FSWs at 6-month.






1 Trachomonas vaginalis prevalence among FSWs at 6-month

Ghys 2001 29/273 26/269 1.10 [ 0.67, 1.82 ]

2 Gonorrhoeae prevalence among FSWs at 6-month

Ghys 2001 10/273 26/269 0.38 [ 0.19, 0.77 ]

3 Chlamydia trachomatis prevalence among FSWs at 6-month

Ghys 2001 3/273 13/269 0.23 [ 0.07, 0.79 ]

0.01 0.1 1 10 100



countries (Review)



Consistent condom use by FSWs at 6-month.






Ghys 2001 225/273 213/269 1.04 [ 0.96, 1.13 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100


Analysis 11.1. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 1 STIs

prevalence among FSWs at 6-month.


Comparison: 11 VCT versus standard care of STI for increasing condom use




1 Syphilis prevalence among FSWs at 6-month

Li 2006 8/141 15/137 0.52 [ 0.23, 1.18 ]


Li 2006 12/141 14/137 0.83 [ 0.40, 1.74 ]


Li 2006 20/141 29/137 0.67 [ 0.40, 1.13 ]


Li 2006 8/141 15/137 0.52 [ 0.23, 1.18 ]

5 Genital warts prevalence among FSWs at 6-month

Li 2006 2/141 9/137 0.22 [ 0.05, 0.98 ]

0.01 0.1 1 10 100



countries (Review)


Analysis 11.2. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 2

Consistent condom use by FSWs with clients at 6-month.



Outcome: 2 Consistent condom use by FSWs with clients at 6-month



Li 2006 74/141 35/137 2.05 [ 1.48, 2.85 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100


Analysis 11.3. Comparison 11 VCT versus standard care of STI for increasing condom use, Outcome 3

HIV/STI knowledge among FSWs at 6-month.



Outcome: 3 HIV/STI knowledge among FSWs at 6-month



Li 2006 33/141 25/137 1.28 [ 0.81, 2.04 ]

Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]




0.01 0.1 1 10 100



countries (Review)


A P P E N D I C E S

Appendix 1. PUBMED SEARCH STRATEGY

Search Most Recent Queries Time Result

#44 Search #39 AND #40 AND #41 AND #

42 Limits: Publication Date from 1980/

01/01 to 2010/09/28

08:29:05 1005

#43 Search #39 AND #40 AND #41 AND

#42

08:27:24 1007

#42 Search intervention[tiab] OR interven-

tions[tiab] OR risk reduction behav-

ior[mh] OR risk reduction[tiab] OR

risk reducing[tiab] OR Health Knowl-

edge, Attitudes, Practice[mh] OR in-

cidence[mh] OR incidence[tiab] OR

prevalence[mh] OR prevalence[tiab]

OR sexual behavior[mh] OR sexual be-

havior[tiab] OR sexual behaviour[tiab]

OR intervention studies[mh]

08:27:05 1153751

#41 Search prostitute[tiab] OR prosti-

tutes[tiab] OR sex worker[tiab] OR sex

workers[tiab] OR prostitution[mh] OR

prostitution[tiab]

08:26:54 6708

#40 Search (randomized controlled trial [pt]

OR controlled clinical trial [pt] OR ran-

domized [tiab] OR placebo [tiab] OR

drug therapy [sh] OR randomly [tiab]

OR trial [tiab] OR groups [tiab]) NOT

(animals [mh] NOT humans [mh])

08:26:34 2267836

#39 Search HIV Infections[MeSH] OR

HIV[MeSH]

OR hiv[tw] OR hiv-1*[tw] OR hiv-

2*[tw] OR hiv1[tw] OR hiv2[tw] OR

hiv infect*[tw] OR human immun-

odeficiency virus[tw] OR human im-

munedeficiency virus[tw] OR human

immuno-deficiency virus[tw] OR hu-

man immune-deficiency virus[tw] OR

((human immun*) AND (deficiency

virus[tw])) OR acquired immunodefi-

ciency syndrome[tw] OR acquired im-

munedeficiency syndrome[tw] OR ac-

quired immuno-

08:26:21 267057


countries (Review)


http://www.ncbi.nlm.nih.gov/pubmed/advanced?querykey=44%26dbase=pubmed%26querytype=eSearch%26

http://www.ncbi.nlm.nih.gov/pubmed/advanced?cmd=HistorySearch%26querykey=44%26











(Continued)

deficiency syndrome[tw] OR acquired

immune-deficiency syndrome[tw] OR

((acquired immun*) AND (deficiency

syndrome[tw])) OR ”sexually transmit-

ted diseases, viral“[MH]

Appendix 2. EMBASE SEARCH STRATEGY

No. Query Results Date

#6 #1 AND #2 AND #3 AND #4 AND

[humans]/lim AND [embase]/lim AND

[1980-2010]/py

146 28 Sept 2010

#5 #1 AND #2 AND #3 AND #4 181 28 Sept 2010

#4 intervention OR interventions OR ’risk

reduction’/de OR ’risk reduction’ OR

’risk reducing’ OR ’attitudes to health’

OR ’prevalence’/de OR prevalence OR

’incidence’/de OR incidence

1321732 28 Sept 2010

#3 ’prostitute’/de OR prostitute OR pros-

titutes OR ’prostitution’/de OR prosti-

tution OR ’sex worker’ OR ’sex workers’

OR ’callgirl’/de OR callgirl OR callgirls

7014 28 Sept 2010

#2 random*:ti OR random*:ab OR facto-

rial*:ti OR factorial*:ab OR cross?over*:

ti OR cross?over*:ab OR crossover*:ti

OR crossover*:ab OR placebo*:ti OR

placebo*:ab OR (doubl*:ti AND blind*:

ti) OR (doubl*:ab AND blind*:ab) OR

(singl*:ti AND blind*:ti) OR (singl*:ab

AND blind*:ab) OR assign*:ti OR as-

sign*:ab OR allocat*:ti OR allocat*:ab

OR volunteer*:ti OR volunteer*:ab OR

’crossover procedure’/exp OR ’crossover

procedure’/de OR ’crossover procedure’

OR ’double-blind procedure’/exp OR

’double-blind procedure’/de OR ’double-

blind procedure’ OR ’single-blind pro-

cedure’/exp OR ’single-blind procedure’/

de OR ’single-blind procedure’ OR ’ran-

domized controlled trial’/exp OR ’ran-

999275 28 Sept 2010


countries (Review)


(Continued)

domized controlled trial’/de OR ’ran-

domized controlled trial’

#1 ’human immunodeficiency virus in-

fection’/exp OR ’human immunodefi-

ciency virus infection’/de OR ’human

immunodeficiency virus infection’ OR

’human immunodeficiency virus’/exp

OR ’human immunodeficiency virus’/

de OR ’human immunodeficiency virus’

OR hiv:ti OR hiv:ab OR ’hiv-1’:ti

OR ’hiv-1’:ab OR ’hiv-2’:ti OR ’hiv-

2’:ab OR ’human immunodeficiency

virus’:ti OR ’human immunodeficiency

virus’:ab OR ’human immuno-defi-

ciency virus’:ti OR ’human immuno-de-

ficiency virus’:ab OR ’human immuned-

eficiency virus’:ti OR ’human immuned-

eficiency virus’:ab OR ’human immune-

deficiency virus’:ti OR ’human immune-

deficiency virus’:ab OR ’acquired im-

mune-deficiency syndrome’:ti OR ’ac-

quired immune-deficiency syndrome’:ab

OR ’acquired immunedeficiency syn-

drome’:ti OR ’acquired immunedefi-

ciency syndrome’:ab OR ’acquired im-

munodeficiency syndrome’:ti OR ’ac-

quired immunodeficiency syndrome’:ab

OR ’acquired immuno-deficiency syn-

drome’:ti OR ’acquired immuno-defi-

ciency syndrome’:ab

315753 28 Sept 2010

H I S T O R Y

Protocol first published: Issue 2, 2005

Review first published: Issue 2, 2012

Date Event Description

15 February 2010 New citation required and major changes Made protocol a ”clean slate“ for new author team.

11 November 2008 Amended Converted to RevMan 5, and re-published without new citation


countries (Review)


C O N T R I B U T I O N S O F A U T H O R S

WW designed, set up, analysed, interpreted the data and drafted the review. EO designed, set up, and analysed the data. RM, AK, NH,

and KS revised the article. All authors read and approved the final version to be published.

D E C L A R A T I O N S O F I N T E R E S T

We declare that we have no conflict of interest.

S O U R C E S O F S U P P O R T

Internal sources

• Health Labour Sciences Research Grant, Japan.

External sources

• No sources of support supplied

D I F F E R E N C E S B E T W E E N P R O T O C O L A N D R E V I E W

Ai Koyanagi was involved as a co-author. She revised the draft and approved the final draft.

HIV/STI-related knowledge, sexual risk behavior, psychosocial barriers to condom use, drug and alcohol risk behavior were not listed

in the protocol. These were added as secondary outcomes as relevance to the review question regarding change individual behaviours

to prevent HIV infection.

Sensitivity analyses were not performed as there were too few trials under each comparison.

I N D E X T E R M S

Medical Subject Headings (MeSH)

∗Developing Countries [statistics & numerical data]; ∗Safe Sex; ∗Sex Workers [statistics & numerical data]; Cognitive Therapy

[∗methods]; Condoms [utilization]; HIV Infections [epidemiology; ∗prevention & control; transmission]; Incidence; Randomized

Controlled Trials as Topic; Social Behavior

MeSH check words

Adult; Female; Humans; Male


countries (Review)


Behavioral interventions to reduce the transmission of HIV infection among sex workers and their...

Documents

Transcript of Behavioral interventions to reduce the transmission of HIV infection among sex workers and their...