Abstract
Lehoux and colleagues plea for a health systems perspective to evaluate innovations. Since many innovations and their scale-up strategies emerge from processes that are not (centrally) steered, we plea for any assessment with a dynamic, instead of a sequential, approach. We provide further guidance on how to adopt such dynamic approach, in order to better un-derstand and steer innovations for better health systems. A systems-level challenge is constituted by interactions and feedback loops between different actors and components of the health system. It is therefore essential to explore both the entry-point of innovation and the interactions with other components. If innovation is regarded as an injection of resources and opportunities into a health system, this system needs to have the capacity to transform these into desired outputs, the 'absorption capacity.' The highly organic diffusion of innovation in complex adapative systems cannot be easily controlled, but the system behaviours can be analysed, with occurance of phenomena such as path dependence, feedback loops, scale-free networks, emergent behaviour and phase transitions. This helps to anticipate unintended consequences, and to engage key actors in ongoing problem-solving and adaptation. By adopting a prospective approach, responsible innovation could set in motion prospective policy evaluations, which on the basis of iterative learning would allow decisionmakers to continuously adapt their policies and programmes. Priority-setting for innovation is an essentially political process that is geared towards consensus-building and grounded in values.