Cash transfers-magic bullet or fundamental ingredient?

How many more trials will be needed to endorse the use of small regular cash amounts to improve the health of the very poor? In their study in The Lancet Global Health, Audrey Pettifor and colleagues show that cash transfers reduce key HIV risk behaviours in adolescent girls (sexual activity in the past 12 months, unprotected sex in the past 3 months, and intimate partner violence as an additional outcome). But the unexpected fi ndings from this study might have even greater implications for social protection and HIV prevention. This trial had robust statistical analyses and a carefullydevised hypothesis—ie, conditioning cash transfers on education would improve school attendance and consequently reduce HIV risk. But, as the authors of this study found, conditionality might make cash transfers both less feasible and less eff ective. Because the cash was conditioned on school attendance, the trial had to exclude adolescent girls who had left school, and those who were pregnant. Thus, some of the girls who were most vulnerable to HIV infection could not be included in the study. Participants who stopped attending school lost their cash transfer. The combination of being out of school and having no money might have made girls more vulnerable to exploitative sexual relationships. The infrastructure and cost of policing the conditions reduces reach to individuals at highest risk. The clear lessons from this study are that conditionality does not substantially advance the cause, and that cash transfers have some important eff ects in HIV risk reduction. The challenge now is to understand the mechanisms and to explore pathways to widen the scope of the eff ects. This trial provides clues for the way forward in HIV prevention. Girls who received cash transfers had reductions in some, but not all, HIV risks. However, continued school attendance did reduce HIV acquisition in both study groups. In this study, HIV infection was lower than anticipated and school attendance higher. These results might seem confusing. But in a presentation at the 21st International AIDS Conference (AIDS 2016), Pettifor and colleagues reported an important fi nding. Irrespective of trial intervention, 97% of the girls included in the study attended school compared with 86% of other girls in the area. This suggests that inclusion in the trial participant made a diff erence. What distinguished these girls from their peers? This was a rigorous study, led by highly ethical researchers with data collection over several years. All participants knew that their school attendance was being monitored. They were asked about their sexual relationships. All attended study-enrolled schools in which girls’ education was a focus. All were regularly tested for HIV, with pre-test and post-test counselling: a standard of care that is unusual in the region. Other fi ndings on adolescent risk-taking show that adult monitoring and attention can reinforce positive behaviours and reduce risk. The overall increased school attendance and reduced HIV incidence in this trial might have been due to this provision of care: the supervision, testing, and support that characterised a robust study and an ethical research team. Quasiexperimental evidence has shown that combinations of cash and adult monitoring of adolescents are associated with greater HIV risk reductions than cash alone. In this large-scale trial, we see a parallel result. Cash helped with some HIV risks, whereas care helped with other risks. Girls who received both cash and the (unintentional) care, showed an array of prevention benefi ts. The evidence of enhancing the eff ect of cash transfers by combining them with wider care seems to be growing. Pettifor and colleagues’ study comes at a time of change. Governments in countries with high prevalence of HIV are starting to realise that universal cash transfers, as well as universal school provision, are fundamental HIV prevention approaches and good policy. As such, the study provides timely and important lessons for the scale-up of social protection. As the authors say, context is essential, and we now need evidence from other countries in the region. Cash alone is an essential component, but not a magic bullet. To navigate this risky developmental stage, adolescents also need our attention, supervision, and care.