Partnering to proceed: scaling up adolescent sexual reproductive health programmes in Tanzania. Operational research into the factors that influenced local government uptake and implementation.

BACKGROUND: Little is known about how to implement promising small-scale projects to reduce reproductive ill health and HIV vulnerability in young people on a large scale. This evaluation documents and explains how a partnership between a non-governmental organization (NGO) and local government authorities (LGAs) influenced the LGA-led scale-up of an innovative NGO programme in the wider context of a new national multisectoral AIDS strategy. METHODS: Four rounds of semi-structured interviews with 82 key informants, 8 group discussions with 49 district trainers and supervisors (DTS), 8 participatory workshops involving 52 DTS, and participant observations of 80% of LGA-led and 100% of NGO-led meetings were conducted, to ascertain views on project components, flow of communication and decision-making and amount of time DTS utilized undertaking project activities. RESULTS: Despite a successful ten-fold scale-up of intervention activities in three years, full integration into LGA systems did not materialize. LGAs contributed significant human resources but limited finances; the NGO retained control over finances and decision-making and LGAs largely continued to view activities as NGO driven. Embedding of technical assistants (TAs) in the LGAs contributed to capacity building among district implementers, but may paradoxically have hindered project integration, because TAs were unable to effectively transition from an implementing to a facilitating role. Operation of NGO administration and financial mechanisms also hindered integration into district systems. CONCLUSIONS: Sustainable intervention scale-up requires operational, financial and psychological integration into local government mechanisms. This must include substantial time for district systems to try out implementation with only minimal NGO support and modest output targets. It must therefore go beyond the typical three- to four-year project cycles. Scale-up of NGO pilot projects of this nature also need NGOs to be flexible enough to adapt to local government planning cycles and ongoing evaluation is needed to ensure strategies employed to do so really do achieve full intervention integration