Background: Disseminating research findings to policy-makers is a major interest of health researchers and funding organizations. The consensus is that some types of interaction are necessary conditions for effective knowledge transfer. An earlier presentation by the author at the Canadian Cochrane Centre's annual meeting described Health Canada as a black box from the outside. It was argued that outsiders find it difficult to know who to speak with in the organization and if one's research findings were taken into consideration as part of a policy decision-making process. Researchers and organizations who wish to disseminate their findings to government officials need to know who their intended audience is so they can interact with and understand each other. How, then, does one identify the person or persons to interact with? The answer lies in identifying and using policy networks.

Objectives: The intent of the presentation is to focus on policy networks as a means to understand the context for knowledge brokering activities and to inspire creative thinking around who to dissemination to. Furthermore, understanding policy networks means loosening our normal assumption that the transfer of research findings is a rational activity. What is required is a better understanding of how people use information and an appreciation that policies are social constructions. There are different types of evidence that are considered as different people interact with their respective agendas and interests.

Methods: The presentation is not research based. Rather it is theoretical reflection on the value of looking at policy networks in order to improve dissemination activities, including knowledge brokering.

Results: The presentation will describe the concept of policy networks and provide examples of frameworks and existing policy networks. The concept of policy networks invites researchers to identify the scope and content of a network, and in doing so to identify the key social actors and their relationships. Then a set of tools for mapping policy networks will be suggested. Some of these tools are software-based, while others are systematic ways of thinking about how individuals are embedded within different networks. In both instances, it is important to decide on the relevant characteristics of users in the network. For example, how often do they communicate with each other? What do they communicate about? Finally, some preliminary findings from a study of tobacco control policy networks in Canada will be presented.

Conclusions: Audience members will learn about the concept of policy networks and how it can enrich their knowledge transfer activities.

Acknowledgements: The author is a co-PI on a study which is examining the Canadian tobacco policy community using a policy networks approach. The other two PIs are Drs. Anita Kothari and Michelle Driedger.