Meeting the challenge of proving impact in Andhra Pradesh, India.

Led by the International HIV/AIDS Alliance, the Frontiers Prevention Programme (FPP) was implemented in India, Ecuador, Cambodia and Morocco between 2002 and 2007. This paper focuses on the experience of evaluating this programme in Andhra Pradesh, India. The aim of the FPP was to reduce HIV infections through supporting the delivery of a comprehensive package of community-based prevention interventions, focused on saturating pre-defined geographical sites that had high densities of populations that were considered key to the epidemic, or 'key populations' - female sex workers, men who have sex with men, and people living with HIV/AIDS. Various aspects of the FPP were innovative for its time, including the implementation of a rigorous impact evaluation initially using a community randomised control trial design. The key evaluation questions posed were: (a) to what extent do HIV and AIDS prevention interventions effect positive change in the course of the epidemic within key populations and the communities in which they live? and (b) does community participation provide an additional impact beyond that of traditional prevention interventions? Some of the challenges in conducting the evaluation were surmountable - for example, those related to managing a large-scale multi-stakeholder programme with institutes and individuals located in different parts of the world - but others were less so. In particular the loss of the control sites proved problematic and resulted in the endline findings, although showing positive changes in behaviour, being largely inconclusive since these positive changes were found across all sites. Lessons include the importance of continual dialogue between programme staff and researchers and sharing and disseminating findings in appropriate formats to different kinds of stakeholders. The experience from the FPP questions the value of NGOs using experimental evaluation designs; while acknowledging their value in providing evidence of what works and therefore where to focus resources, NGOs need to be cautious before undertaking such exercises and ensure that they are fully aware of the difficulties and expense in undertaking such experimental evaluations. [ABSTRACT FROM AUTHOR]