Abstract

When HIV or STI prevention programmes (eg, large core group intervention, human papillomavirus vaccination programme) are being scaled up to entire populations, it often becomes difficult to conduct randomised experiments to evaluate their impact. In this talk, I will discuss some of the challenges often present when evaluating large scale interventions and how and why transmission dynamics models, in combination with surveillance data, should be used to address some of these issues and to help estimate the intervention impact more objectively. Mathematical models should also be used at an early stage, not only to choose the most optimal prevention package, but to optimally design and validate evaluation studies in a cost-effective manner.