Elsevier

The Lancet

Volume 379, Issue 9823, 7–13 April 2012, Pages 1320-1329
The Lancet

Articles
Effect of a cash transfer programme for schooling on prevalence of HIV and herpes simplex type 2 in Malawi: a cluster randomised trial

https://doi.org/10.1016/S0140-6736(11)61709-1Get rights and content

Summary

Background

Lack of education and an economic dependence on men are often suggested as important risk factors for HIV infection in women. We assessed the efficacy of a cash transfer programme to reduce the risk of sexually transmitted infections in young women.

Methods

In this cluster randomised trial, never-married women aged 13–22 years were recruited from 176 enumeration areas in the Zomba district of Malawi and randomly assigned with computer-generated random numbers by enumeration area (1:1) to receive cash payments (intervention group) or nothing (control group). Intervention enumeration areas were further randomly assigned with computer-generated random numbers to conditional (school attendance required to receive payment) and unconditional (no requirements to receive payment) groups. Participants in both intervention groups were randomly assigned by a lottery to receive monthly payments ranging from US$1 to $5, while their parents were independently assigned with computer-generated random numbers to receive $4–10. Behavioural risk assessments were done at baseline and 12 months; serology was tested at 18 months. Participants were not masked to treatment status but counsellors doing the serologic testing were. The primary outcomes were prevalence of HIV and herpes simplex virus 2 (HSV-2) at 18 months and were assessed by intention-to-treat analyses. The trial is registered, number NCT01333826.

Findings

88 enumeration areas were assigned to receive the intervention and 88 as controls. For the 1289 individuals enrolled in school at baseline with complete interview and biomarker data, weighted HIV prevalence at 18 month follow-up was 1·2% (seven of 490 participants) in the combined intervention group versus 3·0% (17 of 799 participants) in the control group (adjusted odds ratio [OR] 0·36, 95% CI 0·14–0·91); weighted HSV-2 prevalence was 0·7% (five of 488 participants) versus 3·0% (27 of 796 participants; adjusted OR 0·24, 0·09–0·65). In the intervention group, we noted no difference between conditional versus unconditional intervention groups for weighted HIV prevalence (3/235 [1%] vs 4/255 [2%]) or weighted HSV-2 prevalence (4/233 [1%] vs 1/255 [<1%]). For individuals who had already dropped out of school at baseline, we detected no significant difference between intervention and control groups for weighted HIV prevalence (23/210 [10%] vs 17/207 [8%]) or weighted HSV-2 prevalence (17/211 [8%] vs 17/208 [8%]).

Interpretation

Cash transfer programmes can reduce HIV and HSV-2 infections in adolescent schoolgirls in low-income settings. Structural interventions that do not directly target sexual behaviour change can be important components of HIV prevention strategies.

Funding

Global Development Network, Bill & Melinda Gates Foundation, National Bureau of Economic Research Africa Project, World Bank's Research Support Budget, and several World Bank trust funds (Gender Action Plan, Knowledge for Change Program, and Spanish Impact Evaluation fund).

Introduction

Two-thirds of HIV-infected people worldwide live in sub-Saharan Africa and 68% of new infections in adults occur in the region.1 Women and girls are at particularly high risk; 60% of people living with HIV in sub-Saharan Africa are women. Furthermore, people aged 15–24 years account for an estimated 45% of new infections worldwide.2 In this age group, women are more than twice as likely to be infected than are men. In Malawi, HIV prevalence in 2007 was 9·1% in young women, compared with 2·1% in young men.3 Prevention of HIV in girls is one of the most essential challenges to reach a turning point in the epidemic.4

Schooling has been suggested to be a social vaccine to prevent the spread of HIV.5 Although cross-sectional data6, 7 suggest a correlation between school attendance and HIV status, only two studies8, 9 have identified a possible causal link between school attendance and reduced risky sexual behaviour. Furthermore, although poverty, especially poverty of women, has been suggested as a major risk factor for HIV, evidence is mixed.10, 11, 12, 13 We are aware of only one cluster randomised trial14 with biological outcomes of a structural intervention to prevent HIV in women. The Intervention with Microfinance for AIDS and Gender Equity (IMAGE) study14 assessed a structural intervention that combined a microfinance programme with a gender and HIV training curriculum and showed no effect on HIV incidence. No studies causally link increased schooling or income with a reduction in HIV risk. However, this absence of clear and credible evidence is not restricted to interventions in education or social protection. A review of behaviour change interventions has shown that “the reality that current behavior change interventions, by themselves, have been limited in their ability to control HIV infection in women and girls in low- and middle-income countries”.15 No randomised controlled trial assessing a behavioural intervention has shown a significant effect on HIV incidence.16

Conditional cash transfer programmes aim to reduce current poverty and, by investment in the education of children, future poverty. As such, they have been popular in development economics since the late 1990s, especially in Latin America.17, 18 The programmes provide cash transfers to poor households with school-aged children (generally aged 6–18 years), often with the requirement that the children attend school regularly. Conditional cash transfer programmes can be an important component of social protection policy and evidence suggests that they have improved the lives of poor people.19

Because conditional cash transfer programmes increase household income and school enrolment, they are particularly suitable for investigation of the importance of education and poverty as risk factors for HIV. We assess the effectiveness of cash transfers in Malawi for reduction of the risk of HIV and herpes simplex virus type 2 (HSV-2) infections in never-married girls aged 13–22 years.

Section snippets

Study design and participants

We undertook this cluster randomised trial in Zomba district, which like most of southern Malawi is characterised by poverty, low school enrolment, and a high prevalence of HIV.3, 20 Zomba district contains 550 enumeration areas defined by the National Statistical Office of Malawi. Each enumeration area contains an average of 250 households spanning several villages. Zomba city includes 50 enumeration areas; the remaining 500 are in seven traditional authorities. Before the start of the trial

Results

Figure 2 shows the trial profile. At baseline 3796 individuals were successfully interviewed. Although 133 (7%) baseline schoolgirls and 86 (10%) baseline dropouts were lost to follow-up at 12 months, none of the enumeration areas had complete loss to follow-up. Rates were similar at the 18-month visit (figure 2). Of the 1777 individuals selected for biological testing, 71 (4%) were lost to follow-up because of either refusal to get tested (n=51) or not being located by the data collection

Discussion

The cash transfer programme decreased the prevalences of HIV and HSV-2 infection after 18 months in girls aged 13–22 years who were enrolled in school at baseline. These effects are supported by changes in self-reported sexual behaviour; we detected no effects on age of sexual debut or unprotected sex, but individuals in the intervention group chose younger partners than did those in the control group and sexual activity was less frequent with those partners.

Few randomised controlled trials

References (26)

  • S Baird et al.

    The short-term impacts of a schooling conditional cash transfer program on the sexual behavior of young women

    Health Econ

    (2010)
  • K Hallman

    Gendered socioeconomic conditions and HIV risk behaviours among young people in South Africa

    Afr J AIDS Res

    (2005)
  • J Robinson et al.

    Transactional sex as a response to risk in western Kenya

    Am Econ J Appl Econ

    (2011)
  • Cited by (470)

    View all citing articles on Scopus
    View full text