Skip to main content
. 2014 Nov 4;17(1):19228. doi: 10.7448/IAS.17.1.19228

Table 1.

Summary of effectiveness studies identified

Intervention/programme and study Location/site and study population Intervention description and study design (Grey scalea) HIV-related outcome(s)b Size and period of effect Incremental interpretation
Prevention of vertical transmission Promoting male partner participation through individual and/or couple VCT
Becker et al., 2009		 Tanzania (urban)
3 ANC clinics
1521 women attending ANC, of which 81 were HIV-positive reached at follow-up (51 in the individual VCT arm; 30 in the couples VCT arm)		 HIV+: Invitation of male partner for couples VCT
Pregnant women were given invitation letters for their husbands to come with them for couples VCT at the next ANC visit.
Randomized controlled trial in which women in the control arm received individual VCT upon recruitment
(Grey scale II)		 HIV-positive women receiving Nevirapine for (1) themselves and (2) for their infants Percentages at follow-up visit three months after delivery date (couples VCT vs. individual VCT):
1) 55% vs. 24%
2) 55% vs. 22%
(significant at p<0.10)		 The study analyzed the incremental effect of couples VCT on the use of protective measures against sexual transmission and uptake of vertical transmission prevention services.
Farquhar et al., 2004 Kenya (urban)
1 clinicAmong 2104 women accepting testing, 308 had partners participate in VCT, of whom 116 were couple counselled		 HIV+: Staff encourage return with partners and couples counselling
Cohort study with a control group, comparing HIV-positive pregnant women whose partners were invited to come to the clinic for VCT (1), those whose partners came for VCT (2) and those who were counselled as a couple (3)(Grey scale IIIa)

  1. Women reporting condom use since last visit

  2. Women returning for post-partum follow-up and reporting nevirapine use at delivery

  3. Women choosing to breastfeed their infants

 Odds ratios at six months post-partum follow-up:

  • 4.2 (1.5–11.5)

  • 3.4 (1.3–9)

  • 0.2 (0.04–0.9)

 The study analyzed both the incremental effects of partners coming for individual VCT and for couples VCT on the uptake of vertical transmission prevention services and recommendations.
Mohlala et al., 2011 South Africa
1000 pregnant women (500 in each arm)		 HIV+: Invitation of male partner for ANC and VCT
Invitation letters were provided to women in antenatal care (ANC) for their male sexual partners to attend ANC and VCT.
Randomized controlled trial in which women in the control group received invitation letters for their male sexual partners to attend ANC and pregnancy information sessions
(Grey scale II)

  1. Male sexual partner that underwent HIV testing

  2. Reported unprotected sex in previous 2 weeks

 Risk ratios at 12 weeks post-randomization follow-up:

  1. 2.82 (2.14–3.72)

  2. 0.30 (0.22–0.42)

 Not incremental to standard vertical transmission prevention programmes, since the control involved male involvement in pregnancy information sessions. The study analyzed the incremental effect of the VCT invitation letter on ANC attendance, VCT uptake and unprotected sex.
Msuya et al., 2008 Tanzania (urban)

  1. district

  2. primary healthcare clinics Of 2654 pregnant women asked to invite their partners, 332 male partners came for HIV counselling and testing at the clinics

 HIV+: Invitation of male partner for individual and couples VCT
Pregnant women in third trimester encouraged to inform and invite their partners for VCT.
Cohort study comparing pregnant women (HIV-positive and HIV-negative) whose partners came for VCT and those whose did not (control group)
(Grey scale IIIa)

  1. HIV-positive women choosing not to breastfeed (recommended at time of study)

  2. HIV-positive women adhering to infant feeding method selected at post-test

 Odds ratios at two-year follow-up:

  1. 3.72 (1.19–11.63)

  2. 5.15 (2.18–12.16)

 The study analyzed the incremental effect of partners coming for VCT (individual and/or couple) on uptake of vertical transmission prevention services.
Peer support groups for pregnant women/mothers living with HIV
Futterman et al., 2012		 South Africa (peri-urban)
2 NGO and public sites
160 pregnant women attending the clinics who were diagnosed HIV-positive		 HIV+: peer mentoring and cognitive behavioural training (Mamekhaya programme)
Mothers living with HIV (MLH) were linked to mentor mothers who were also HIV-positive and had been trained to provide support.

  1. Centre for Epidemiologic Studies Depression Scale

  2. Social support availability scale

 Random intercept regression model coefficient (standard error) after six months:

  1. 4.43 (1.62)

  2. 9.32 (3.53)

 The study analyzed the incremental effect of the Mamekhaya programme on uptake of vertical transmission prevention services and adherence to preventive practices.
Non-randomized trial with control group receiving standard vertical transmission prevention care from medical staff (Grey scale IIIa)
Nguyen et al., 2009 Vietnam (urban)
Referral network among 26 health facilities
30 women (members of the group) who had learned they were HIV-positive before or during a pregnancy and chose to complete the pregnancy		 HIV+: peer support group for HIV-positive mothers
In the self-help group, core members (peer counsellors) delivered publicity materials to create a referral network among health facilities, visited hospitals and VCT sites to make informal contact with potential members.
Pre-/post-assessment without control group (Grey scale IIIb)

  1. Women with record for health follow-up at ART sites

  2. Women receiving ART (when needed)

 Indicators upon joining group and two years after joining the group:

  1. Increase from 1/30 to 30/30

  2. Increase from 1/9 to 15/15

 The study analyzed the incremental effect of participating in the support group on access to ART.
Rotheram-Borus et al., 2014 South Africa (rural and urban)
4 control clinics with 656 WLH enrolled
4 intervention clinics with 544 WLH enrolled		 HIV+: peer mentoring and support sessions
Pregnant women living with HIV (WLH) were invited to attend 8 meetings with peers, facilitated by peer mentors, and covering various topics, such as normalizing being a WLH, healthy lifestyles, treatment adherence, infant feeding methods and bonding, couple counselling and condom use.
Cluster randomized controlled trial with WLH in control clinics receiving standard vertical transmission prevention services (Grey scale II)

  1. Infant fed using one feeding method for first six months

  2. Infant weight-for-age z-score ≥2

  3. Infant exclusively breastfed for at least six months

  4. Mothers not depressed (GHQ<7)

 Estimated odds ratio from birth to 12 months post-birth:

  1. 3.02 (1.20–7.60)

  2. 1.08 (1.01–1.16)

  3. 2.38 (1.04–5.44)

  4. 1.08 (1.03–1.13)

 The study analyzed the incremental effect of receiving peer mentoring and support on service uptake, maternal and child health outcomes.
Key populations – FSWs Gender empowerment community mobilization intervention among FSWs
Basu et al., 2004		 India
2 urban centres
200 brothel-based FSWs (100 in each arm)		 HIV+: integrated empowerment intervention (Sonagachi)
Local peer educators were trained to build skills and confidence in providing education and to foster empowerment and advocacy for local sex workers. The team engaged in ongoing advocacy activities with local stakeholders and power brokers who exerted control over the sex workers’ lives.
Randomized controlled trial (Grey scale II)

  1. Proportion reporting 100% condom use

  2. Proportion of consistent condom users

 Effect after 15 months:

  1. 39% increase vs. 11% increase

  2. 25% increase vs. 16% decrease

 The study analyzed the incremental effect of the Sonagachi model on condom use.
Markosyan et al., 2010 Armenia (urban)
1 public site
120 FSWs		 HIV+: gender empowerment intervention
A health educator implemented a 2-h intervention emphasizing gender-empowerment, self-efficacy to persuade clients to use condoms, condom application skills, and eroticizing safer sex.
Randomized controlled trial (Grey scale II)		 Consistent condom use (clients in general) Adjusted odds ratios at six-month follow-up:
2.85 (1.41–5.75)		 The study did not consider the incremental effect of the gender-responsive intervention above a standard FSW programme. Instead, it analyzed the effect of the intervention (compared to a do-nothing alterative) on condom use.
Beattie et al., 2010 India (urban)
4 NGOs
Over 60,000 FSWs;
3852 participated in IBBA surveys in 4 districts; and 7638 FSWs participated in 691 polling booth surveys in 13 districts		 HIV+: multi-layered violence prevention strategy (Avahan – India AIDS initiative)
A multi-layered district-wide strategy involving policy makers, secondary stakeholders (police officers, human rights lawyers, journalists) and primary stakeholders (FSWs) to stem and address violence against the sex worker community as part of a wider HIV prevention programme.
Pre-/post-assessment without control (Grey scale IIIb)		 Experience of violence (beaten or raped) in the past year Adjusted odds ratio after 33 to 37 months:
0.70 (0.53, 0.93)		 Not incremental to standard FSW programme. The study analyzed the association between programme exposure (contacted by a peer educator or having visited the project sexual health clinic) and experience of violence.
Relationship-based sessions to reduce violence against FSWs
Carlson et al., 2012		 Mongolia (urban)
166 FSWs engaging in harmful alcohol use		 HIV+: relationship-based sessions, including violence prevention
The interventions consisted of (1) Four-weekly relationship-based HIV/STI risk reduction sessions; (2) Four-weekly enhanced HIV/STI risk reduction intervention with two wrap-around sessions engaging motivational interviewing; (3) Four-weekly sessions on overall health and wellness knowledge and skills.
Cluster randomization, with pre-/post-data analysis by intervention group (Grey scale IIIa)		 Reported experience of physical or sexual IPV in the past 90 days Estimated odds ratios at six months follow-up, based on empirical multilevel logistic modelling with an individual-level random effect:

  1. 0.46 (0.24–0.88)

  2. 0.14 (0.03–0.61)

  3. 0.20 (0.096–0.43)

 The study does not find an incremental effect of the HIV+ approach on violence, compared to the non-HIV and HIV-specific control interventions.
Female condom promotion among FSWs
Thomsen et al., 2006		 Kenya (urban)
210 FSWs		 HIV+: female condom promotion for FSWs
Adding female condom promotion to a male condom programme providing peer education and IEC materials, as well as distributing female condoms.
Pre-/post-evaluation without control group (Grey scale IIIb)		 Consistent condom use with all sexual partners in previous 7 days At 12-month follow-up:
Increase from 59.7% before to 67.1% after (p=0.04)		 The study analyses the incremental effect of female condom promotion on consistent condom use, but given the lack of a control group the effect estimate is not reliable.
Micro-enterprise services for FSWs
Odek et al., 2009		 Kenya (urban)
227 FSWs		 HIV+: micro-enterprise intervention
The micro-enterprise component was added to the existing peer education HIV risk reduction model and consisted of: credit for small business activities, business skills training and mentorship, and promotion of a savings culture.
Pre-post design without control group (Grey scale IIIb)

  1. Self-reported weekly mean number of all sexual partners

  2. Consistent condom use with regular partners

 Mean at 18 to 23 months follow-up:

  1. 1.84 (SD 2.15) compared to 3.26 (SD 2.45) at baseline (p<0.001)

  2. 93.5% compared to 78.9% (p=0.031)

 The study analyses the incremental effect of micro-enterprise activities in FSW programmes, but given the lack of a control group the effect estimate is not reliable.
Sherman et al., 2010 India (urban)
100 FSWs		 HIV+: micro-enterprise intervention
The micro-enterprise component was added to a standard HIV prevention education intervention, and consisted of 100 h of tailoring training taught by master tailors over the course of a month.
The control arm received the same 8 h course on HIV prevention provided twice per week over 2 weeks and provided by 2 health educators. It covered topics around HIV risk, as well as gender, violence and alcohol use.
Randomized controlled trial (Grey scale II)

  1. Number of sex partners

  2. Number of paying clients per month

 Mean at six-month follow-up:

  1. 5.0 compared to 11.9 at baseline (p<0.001)

  2. 3.1 compared to 5.1 at baseline (p<0.001)

 The study analyses the incremental effect of the micro-enterprise component over and above a gender-responsive HIV prevention education intervention for FSWs.
Key populations – IDUs Woman-focussed empowerment-based intervention for high-risk women/FSWs with substance abuse
Wechsberg et al., 2006		 South Africa (urban)
1 NGO site
93 women who reported
recent substance use (cocaine) and sex trading		 HIV+: Woman-focussed sessions, including condom negotiation and violence prevention
The enhanced intervention consisted of 2 one-on-one sessions with a personalized assessment of each woman's drug and sexual risks, information and skills to negotiate condom use, violence prevention strategies and referrals to community resources.
Individually randomized controlled trial, with the control group receiving a private 1-h HIV risk reduction education session (Grey scale II)

  1. Male condom used with boyfriend during last sexual encounter

  2. Any female condom used with boyfriends in the last month

  3. Mean occurrence of victimisation reported by participants

  4. Mean number of STI symptoms since last encounter reported by participants

 Effect size after one month:

  1. RRs at baseline and follow-up were 0.64 and 1.15. Effect size is 0.51 (significant at p=0.05)

  2. RRs at baseline and follow-up were 0.15 and 1.20, respectively. effect size is 1.15 (significant at p=0.01)

  3. 4.5 (intervention) vs. 6.3 (control)

  4. 0.64 (intervention) vs. 1.07 (control)

 Although the intervention targets FSWs, it builds on a basic IDU intervention.
The studies analyzed the incremental effect of the woman-focussed intervention on condom use, daily alcohol and drug use, and experience of violence.
Wechsberg et al., 2011 South Africa
583 women

  1. HIV-positive women reporting male or female condom use at last sex

  2. Women (HIV-status unknown) reporting male or female condom use at last sex

 Odds ratios at six-month follow-up:

  1. 7.27 (1.64–32.23)

  2. 5.03 (1.26–20.11)
Couple-based prevention for IDUs
Gilbert et al., 2010		 Kazakhstan
1 site
40 couples that injected drugs in past 90 days (80 participants)		 Health +: couple-based HIV/STI approach
The couple-based HIV/STI risk-reduction intervention (CHSR) included 3 single-gender group sessions with the male and female partners.
Randomized controlled trial (Grey scale II)

  1. Proportion of condom-protected acts of vaginal and anal intercourse

  2. Proportion of injection acts in which unclean needles or syringes were used in the past 30 days

  3. Number of injection acts in which unclean needles or syringes were used in the past 30 days

 Regression coefficients (standard errors) at three-month follow-up:

  1. 0.19 (0.08)

  2. 0.33 (0.05)

  3. 12.3 (3.9)

 Not incremental effect above standard IDU intervention for HIV, since the control does not cover any HIV topics.
20 couples per intervention Random-effects regression analysis The study analyses the effects of couple-based HIV intervention compared to standard health promotion intervention for IDUs.
Condom promotion and distribution Expanded female condom promotion and distribution
Dowdy et al., 2006

  1. South Africa

  2. Brazil

    General population

 HIV+: female condom promotion and distribution over and above male condom promotion
Second-generation nitrile female condom (FC2) acquisition, distribution, training and education
Impact modelling (Grey scale IIIb)

  1. Fraction of additional sex acts protected by female condoms

  2. Incremental HIV infections averted

  1. Assumed to be 3% of the number of male condoms used (low volume), 10% (moderate volume), or 30% (high volume)

    1. 1900–32,000 HIV infections averted
    2. 100–2000 HIV infections averted
 The study models the incremental effect of an expanded country-wide distribution of the second-generation nitrile female condom, over and above existing male and female condom programmes.
Condom promotion among married women
Callegari et al., 2008		 Zimbabwe (urban)
394 sexually active, married women of reproductive age, aged 17 to 47 years		 HIV+: training for married women in condom negotiation and use
A trained counsellor provided a 30-min one-to-one intervention based on social-cognitive models of behaviour change; and a one-month booster session included content similar to enrolment.
Pre-/post-evaluation (Grey scale IIIb)

  1. Condom use at last sex

  2. Consistent condom use in the past two months

 Effect after two months:

  1. Increase from 10.1 to 87%

  2. Increase from 0.25 to 48.5%

 Not incremental to condom distribution, since women in the intervention receive male and female condoms while they may not have had them at baseline.
The study analyses the effect of condom negotiation training and condom provision among married women on consistent condom use.
Behaviour Change Participatory HIV
prevention programme
Building more gender-equitable relationships (stepping stones)
Jewkes et al., 2008		 South Africa (rural)
70 study clusters comprised 64 villages and 6 townships
1360 men and 1416 women aged 15 to 26 years, who were mostly attending schools		 HIV+: gender-transformative participatory approach with women and men
Intervention stepping stones, a 50 h programme, that aims to improve sexual health by using participatory learning approaches to build knowledge, risk awareness and communication skills, and to stimulate critical reflection.
Cluster randomized controlled trial, with the control villages receiving a 3 h intervention on HIV and safer sex (Grey scale II)

  1. Incidence of HSV-2

  2. Men reporting any transactional sex with a casual partner

  3. Men reporting problem-drinking

 Adjusted odds ratio at 24 months follow-up:

  1. 0.67 (0.47–0.97)

    Adjusted odds ratios at 12 months follow-up:

  2. 0.39 (0.17–0.92)

  3. 0.68 (0.49–0.94)

    (not significant at 24 months follow-up)

 The study analyses the incremental effect of a more intensive gender-transformative approach on HIV-related risk behaviours.
Raising HIV awareness in non-HIV-infected Indian wives (RHANI Wives)
Raj et al., 2013		 India (urban)
220 women aged 18–40 married to men engaged in heavy drinking or lifetime physical or sexual spousal violence perpetration		 HIV+: educational sessions for married women on sexual communication and empowerment
Multisession intervention focused on safer sex, marital communication, gender inequities and violence. It involved 4 household-based individual sessions and 2 small group-based community sessions delivered over 6 to 9 weeks.
Control participants were referred for HIV/STI testing and treatment, local social services for alcoholics and victims of domestic violence.
Randomized controlled trial (Grey scale II)

  1. Rate of unprotected sex

  2. Condom use at last sex

 Risk and odds ratio at 4 to 5 months follow-up:

  1. 0.83 (0.75–0.93)

  2. 2.42 (1.00–5.70)

 The study analyses the incremental effect of this intervention compared to a basic HIV prevention education and referral intervention.
Culturally adapted intervention promoting safer sex and relationship control
(SISTA South Africa)
Wingood et al., 2013		 South Africa (rural)
5 rural areas
342 isiXhosa women aged 18 to 35 years		 HIV+: educational sessions with women on safe sex and relationship control
SISTA consisted of three 2.5-h interactive group sessions delivered by 2 health educators on consecutive Saturdays at community centres and covering ethnic and gender pride, social and contextual influences that enhance HIV vulnerability and sexual communication skills.
The general health comparison condition involved two 2.5-h interactive group sessions covering HIV prevention education, healthy nutrition, hygiene and self-care.
Randomized controlled trial (Grey scale II)

  1. Frequency of vaginal sex

  2. Frequency of unprotected vaginal intercourse acts in the past 30 days

 Adjusted mean difference at six months follow-up:

  1. 1.22 (p=0.02)

  2. 1.06 (p=0.02)

 The study analyses the incremental effect of a more intensive gender-sensitive approach on HIV-related risk behaviours.
Promoting more gender-equitable norms and behaviours among young men
Pulerwitz et al., 2006 (Promundo)		 Brazil
2 sites+1 control site
508 young men		 HIV+: gender-transformative participatory approach with young men
Two models of the gender-transformative approach were evaluated: (a) interactive group education sessions for young men led by adult male facilitators and (b) group education+community-wide “lifestyle” social marketing campaign to promote condom use using gender-equitable messages.
Pre-/post-evaluation with control site
(Grey scale IIIa)

  1. Reported STI symptoms over prior three months

  2. Condom use at last sex with primary partner

 Combination intervention site at six months follow-up:
    1. Decreased from 30 to 25%
    2. Decreased from 23 to 14% compared to a decrease from 18 to 12% in control (p<0.05)
    1. Increased from 69 to 70%
    2. Increased from 58 to 79% compared to decrease from 64 to 59% in control (p<0.05)
 Not incremental to basic HIV behaviour change programme.
Verma et al., 2006
(Yaari-Dosti)		 India
Urban
Rural
1423 married and unmarried young men aged 16–29 in urban settings and aged 15–24 in rural settings

  1. Condom use at last sex in the past three months with all partners

  2. Reported violence against a partner (either sexual or non-sexual/romantic) in the past three months

 Multiple logistic regression odds ratios at six months’ follow-up:
    1. 1.913 (p<0.001)
    2. 2.776 (p<0.001)
    1. 0.176 (p<0.001)
    2. 0.502 (p<0.001)
 Not incremental to basic HIV behaviour change programme.
Kalichman et al., 2009 South Africa (urban)
2 townships
475 men living in two townships in Cape Town		 HIV+: gender-transformative training with men
The five-session intervention emphasized sexual transmission risk reduction and GBV reduction through skills building and personal goal setting, geared towards addressing gender roles, exploring meanings of masculinity and reducing adversarial attitudes towards women. Men were also trained to become vocal advocates for risk reduction behaviour changes with other men in their community.
Community randomized trial without control group
(Grey scale IIIb)

  1. Men reporting 100% condom use in the past month (or three months – unclear)

  2. Men reporting having tested for HIV among men not tested at baseline

  3. Men reporting having hit a partner in the past month

 Odds ratio at 1 month follow-up:

  1. 1.7 (1.1–2.7)

    Odds ratio at three months follow-up:

  2. 0.5 (0.3–0.9)

    Odds ratio at six months follow-up:

  3. 0.3 (0.2–0.4)

 Not incremental to basic HIV behaviour change programme.
Community mobilization SASA! activist kit for preventing violence against women and HIV
Abramsky et al., 2014		 Uganda (urban)
4 intervention and 4 control communities
Random sample of adult community members sampled at baseline (n=1583) and post-intervention (n=2532)		 HIV+: gender-transformative community mobilization approach
Community activists were trained, along with staff from selected institutions (e.g. police, health care), to deliver the intervention aimed at changing community attitudes, norms and behaviours related to the power imbalances between men and women that contribute to violence against women and increase HIV risk behaviours. The cadre of activists conducted informal activities within their social networks, using local activism, local media and advocacy, communication materials and/or training. The intervention was not rigidly proscribed but evolved in response to community priorities.
Cluster randomized controlled trial
(Grey scale II)		 Past year concurrent sexual partner among non-polygamous men partnered in the past year Adjusted risk ratio at four years follow-up:
0.57 (0.36–0.91)		 Not incremental to basic HIV behaviour change programme. Due to the movement of trained health and police staff between intervention and control communities, the study examines the added value of the intensive local intervention components, rather than the impact of the whole package.
Mass media Mass media GBV and HIV (Soul City)
Goldstein et al., 2005		 South Africa
2 national surveys of 2000 respondents each, on a sample of African and “coloured” people		 HIV+: integrated GBV messaging
The Soul City series is a multimedia health promotion intervention consisting of a 13-part prime-time television drama, a 45-part radio drama and three basic full colour booklets, distributed through 10 newspapers nationally. Integrated health messages are interwoven and include GBV and HIV/AIDS.
Pre-/post-evaluation
(Grey scale IIIa)		 Proportion of respondents reporting that they always use condoms

  1. Control

  2. 1 media type

  3. 2 media types

  4. 3 media types

 At eight months follow-up:

  1. 6% (n=373)

  2. 16% (n=592)

  3. 30% (n =522)

  4. 38% (n=437)

 Not incremental to a basic programme, but could be added to a standard HIV mass media campaign. The study analyses the effect of the Soul City series (including HIV and gender messaging) on condom use.
GBV Refentse model of comprehensive post-rape services
Kim et al., 2009		 South Africa (rural)
1 district
207 Survivors of sexual assault (almost exclusively female, average 20 years old – range three months to 94 years)		 Gender+: HIV post-exposure prophylaxis
Provision of voluntary HIV counselling and testing and post-exposure prophylaxis to survivors through a five-part model, including a sexual violence advisory committee, hospital rape management policy, training workshop for service providers, designated examining room and community awareness campaigns
Pre-/post-evaluation (Grey scale IIIb)		 Completion of 28-day course of PEP drugs Adjusted risk ratio after 26 months:
3.13 (1.10–8.93)		 Not incremental to standard post-rape services. The study analyses the effect of an integrated and comprehensive model of post-rape services on completion of PEP.
Stakeholder skills building and awareness raising to prevent GBV among female apprentices
Fawole et al., 2005		 Nigeria (urban)
350 young female apprentices (203 at follow-up)		 Gender: GBV prevention through stakeholder skills building and awareness raising The intervention consisted of skills training workshops for apprentices, sensitization training for the instructors of apprentices, police and judicial officers and the development/distribution of educational materials to reduce the incidence of violence.
Pre-/post-assessment without control (Grey scale IIIb)

  1. Prevalence of beating

  2. Been sexually harassed

  3. Seeking judicial redress (or medical care) for rape

 At six months follow-up:

  1. Dropped from 65.4 to 23% (p<0.05)

  2. Dropped from 22.9 to 19.7% (p<0.05)

  3. Increased from 30 to 46% (p<0.05)

 The study analyses the effect of the intervention on the apprentices’ experience of violence.
Poverty reduction Fonkoze Microfinance Programme
Rosenberg et al., 2010		 Haiti (urban and rural)
34 centres
192 female clients		 Gender: Microfinance loans targeting women This study assessed the relationship between experience with microfinance loans and HIV risk behaviour among female clients of the Haitian microfinance organization, Fonkoze.
Pre-/post-intervention effects (Grey scale IIIa)

  1. Reported that partner was unfaithful

  2. Condom use among those with unfaithful partner

 Adjusted odds ratios after 12 months:

  1. 0.28 (0.13–0.63)

  2. 3.95 (0.93–16.85)

 The study analyses the effect of accessing microfinance loans on HIV risk behaviour among female clients.
Intervention with microfinance for AIDS & gender equity (IMAGE)
Pronyk et al., 2006		 South Africa (rural)
8 villages
Three cohorts of women; 860, 1835 and 3881 women (aged 14 to 35 in last two cohorts)		 Gender+: participatory gender and HIV training curriculum
The sisters for life participatory curriculum of gender and HIV education was facilitated by a team of trainers. The first phase consisted of a structure training (10 sessions done within centre		 Experience of IPV in the past 12 months Adjusted risk ratio after 2 to 3 years:
0.45 (0.23–0.91)		 The studies analyse the effect of the combined microfinance and gender/HIV training on IPV, HIV risk behaviours and access to HIV services.
Pronyk et al., 2008 South Africa (rural)
8 villages
Sub-group of 262 young women (aged 14 to 35)		 meetings every 2 weeks for about six months); and the second phase involved natural leaders being selected, trained and supported to facilitated broader community mobilization.
Cluster randomized trial with matched comparison group (Grey scale II)		 1) Having accessed voluntary counselling and testing
2) Unprotected sex at last intercourse with a non-spousal partner		 Adjusted risk ratios after two years:
1) 1.64 (1.06–2.56)
2) 0.76 (0.60–0.96)
Social protection Zomba cash transfer programme to keep girls in school
Baird et al., 2012		 Malawi (rural)
2 NGO sites		 Gender: Conditional and unconditional cash transfers for schoolgirls
Cluster randomized controlled trial (Grey scale II)

  1. HIV prevalence and

  2. HSV-2 prevalence post-intervention among baseline schoolgirls

 Adjusted odds ratios after 18 months:

  1. 0.36 (0.14, 0.91)

  2. 0.24 (0.09, 0.65)

 The study analyses the effect of receiving the cash transfer on prevalent HIV and HSV-2 among girls.
Comprehensive school support to adolescent orphan girls
Cho et al., 2011		 Kenya (rural)
79 households
105 adolescent orphans (Luo students) aged 12 to 14		 Gender: School support including fees, uniforms, and school supplies. Female teachers were selected and trained as helpers (approximately one helper to 10 participants) to monitor school attendance and intervene as needed, without Likelihood to delay sexual debut Logistic regression coefficient at one year follow-up: 1.50 The studies analyse the effect of receiving the school support on early marriage rates and sexual debut.
Hallfors et al., 2011 Zimbabwe (rural)
25 public sites		 providing special HIV information or life skills training.
Randomized controlled trial (Grey scale II)		 Marriage rate Adjusted odds ratio after two years (p≤0.05):
2.92 (1.0, 8.3)
AIDS education School-based provision of information on relative HIV risk
Dupas, 2011		 Kenya (two rural districts)
328 primary schools
161 primary schools in TT arm; 71 primary schools in RR arm		 Gender+: informing girls of relative HIV risk
Students in 8th grade were provided with a one-off 40-min information session covering national HIV prevalence by sex and age, and education video about “sugar daddies” and a discussion about cross-generational sex.
Cluster randomized control trial (Grey scale II), with difference-in-difference econometric analysis, controlled for school random effects		 Incidence of childbearing At 9 to 12 months follow-up:
1.5% point reduction in childbearing (RR), compared to 5.4% mean in comparison group – i.e. 28% decrease		 Not incremental analysis to sexual education in school, which was not effectively being provided at the time of the study, despite an existing national policy. The study analyses the effect of informing girls of relative HIV risk on the incidence of childbearing (proxy of unprotected sex).
a

Grey scale I: Strong evidence from at least one systematic review of multiple well-designed, randomized controlled trials; II: Strong evidence from at least one properly designed, randomized controlled trial of appropriate size; IIIa: Evidence from well-designed trials/studies without randomization that include a control group (e.g. quasi-experimental, matched case-control studies, pre-post with control group); IIIb: Evidence from well-designed trials/studies without randomization that do not include a control group (e.g. single group pre-post, cohort, time series/interrupted time series)

b

only outcomes on which the intervention had statistically significant effect are included in this table. GBV=gender-based violence; ART=antiretroviral therapy; FSW=female sex worker; IPV=intimate partner violence.