LAUSR

Effective efforts to strengthen health systems need diverse, multi-stakeholder networks working together on complex or ‘wicked’ problems such as prevention and control of chronic diseases, solutions to which go beyond the role and capability of one organisation. The contextual complexities inherent in ‘wicked’ problems mean that solutions warrant a systems approach that encompasses innovation and new ways of thinking about, facilitating and implementing collective decision-making processes and change practices. Innovation platforms are a mechanism for facilitating communication and collaboration among diverse stakeholders, promoting joint action and stimulating innovation. Developmental evaluation is an approach that is increasingly being used to evaluate innovative and emergent programmes and projects, as it enables evaluators to provide real-time feedback so that evaluation findings can be used to guide development and adaptations. Developmental evaluation emphasises learning and adaptation, and aligns well with the implementation of innovation platforms that have continuous reflection, learning and adaptation as a specific design principle. Here, we outline our rationale for applying a developmental evaluation to enhance the formation, functioning and outcomes of an innovation platform aimed at accelerating and strengthening large-scale quality improvement efforts in Australian Aboriginal and Torres Strait Islander primary healthcare. We provide examples to explain how the developmental evaluation findings were used for adaptation of the innovation platform and assess to what extent our application of developmental evaluation was consistent with, and reflective of, its essential principles. Our evaluation aligned strongly with the principles of developmental evaluation, and the approach we took was well suited to situations with a developmental purpose, innovation niche and complexity such as innovation platforms. As a result, along with the increasing interest in multi-stakeholder platforms (e.g. innovation platforms) and the inherent challenges with evaluating these complex networks, we anticipate our use of this approach being of interest globally.