RESUMO
BACKGROUND: In 2003, the National Health Service in England and Wales, despite its large investment in information and communication technology, had not set a national research agenda. The National Health Service has three main research and development programs: one is the Service Delivery and Organisation program, commissioned in 2003, and the others are two parallel "scoping exercises" to help set a research agenda. This paper reports on one of those projects. A parallel literature review was carried out by others and has been reported elsewhere. OBJECTIVE: The objective was to explore the concerns of stakeholders and to review relevant policy in order to produce recommendations and a conceptual map of eHealth research. METHODS: There were two parallel strands. For the stakeholder consultation, 37 professionals representing 12 "stakeholder" groups participated in focus groups or interviews. Discussion was prompted by eHealth "scenarios" and analyzed using thematic content analysis. Subsequently, 17 lay participants, in three focus groups, discussed and prioritized these themes. For the policy review, 26 policy makers were interviewed, and 95 policy documents were reviewed. Recommendations were subsequently reviewed in a conference workshop. Recommendations for research from both strands were combined into a conceptual map. RESULTS: Themes from stakeholder consultation and policy review were combined as 43 recommendations under six headings. Four of these headings (using, processing, sharing, and controlling information) describe the scope of eHealth research. The other two relate to how research should be carried out (ensuring best practice is first identified and disseminated) and to the values considered important by stakeholders (in particular, measuring improvement in health). CONCLUSIONS: The scope of eHealth research (using, processing, sharing, controlling information) derived empirically from this study corresponds with "textbook" descriptions of informatics. Stakeholders would like eHealth research to include outcomes such as improved health or quality of life, but such research may be long term while changes in information technology are rapid. Longer-term research questions need to be concerned with human behavior and our use of information, rather than particular technologies. In some cases, "modelling" longer-term costs and benefits (in terms of health) may be desirable.