Background: HIA can be used to judge the potential health effects of a policy, programme or project on a population, and the distribution of those effects. Progress has been made in incorporating HIA into routine practice, especially (in the UK) at local level. However, these advances have mainly been restricted to process issues, including policy engagement and community involvement, while the evidence base has been relatively neglected.
Relating policies to their impact on health: The key distinctive feature of HIA is that determinants of health are not taken as given, but rather as factors that themselves have determinants. Nine ways are distinguished in which evidence on health and its determinants can be related to policy, and examples are given from the literature. The most complete of these is an analysis of health effects in the context of a comparison of options. A simple model, the policy/risk assessment model (PRAM), is introduced as a framework that relates changes in levels of exposures or other risk factors to changes in health status. This approach allows a distinction to be made between the technical process of HIA and the political process of decision making, which involves lines of accountability. Extension of the PRAM model to complex policy areas and its adaptation to non-quantitative examples are discussed.
Issues for the future: A sound evidence base is essential to the long term reputation of HIA. Research gaps are discussed, especially the need for evidence connecting policy options with changes in determinants of health. It is proposed that policy options could be considered as "exposure" variables in research. The methodology needs to be developed in the course of work on specific issues, concentrated in policy areas that are relatively tractable.
Conclusions: A system of coordination needs to be established, at national or supranational level, building on existing initiatives. The framework suggested in this paper can be used to collate and evaluate what is already known, both to identify gaps where research is required and to enable an informed judgement to be made about the potential health impacts of policy options. These judgements should be made widely available for policy makers and for those undertaking health impact assessment.