Background: Health policy and systems research (HPSR) is an international public good with potential to orient investments and performance at national level. Identifying research trends and priorities at international level is therefore important. This paper offers a conceptual framework and defines the HPSR portfolio as a set of research projects under implementation. The research portfolio is influenced by factors external to the research system as well as internal to it. These last include the capacity of research institutions, the momentum of research programs, funding opportunities and the influence of stakeholder priorities and public opinion. These dimensions can vary in their degree of coordination, leading to a complementary or a fragmented research portfolio.
Objective: The main objective is to identify the themes currently being pursued in the research portfolio and agendas within developing countries and to quantify their frequency in an effort to identify current research topics and their underlying influences.
Methods: HPSR topics being pursued by developing country producer institutions and their perceived priorities were identified through a survey between 2000 and 2002. The response to a call for letters of intent issued by the Alliance in 2000 for a broad range of topics was also analyzed. The institutions that were the universe of this study consisted of the 176 institutional partners of the Alliance for Health Policy and Systems Research producing research in low and middle income countries outside Europe. HPSR topics as well as the beneficiaries or issues and the health problems addressed were content analyzed. Topics were classified into 19 categories and their frequency analyzed across groups of countries with similar per capita income. Agendas were identified by analyzing the source of funding and of project initiation for projects under implementation.
Results: The highest ranking topic at the aggregate level is "Sector analysis", followed by "Disease burden" and "Management and organization". Categories at the bottom of this ranking are "Equity", "Policy process", "Economic policy and health" and "Information systems". "Disease burden" is more often funded than other topics for which there is more demand or perceived priority. Analysis suggests few although important differences across priorities, demand for funding and actual project funding. The donors' agenda coincides most with the ranking of research topics overall. Ranking across country income groups shows important differences. Topics that gain prominence in low income countries are "Disease burden" and "Accessibility". In lower middle income countries "Insurance" gains prominence. In upper middle income countries "Decentralization/local health systems", "Equity" and "Policy process" are more prominent. "Program evaluation" is the most consistently ranked topic across income regions, showing a neutral influence by donors, governments or researchers.
Conclusions: The framework proposed offers a basis to identify and contrast research needs, projects and products at the international level and to identify the actor agendas and their influence. Research gaps are suggested when comparing topic ranking against the challenges to health system strengthening and scaling up of disease control programs. Differences across per capita income groups suggests the need for differentiated priority setting mechanisms guiding international support. Data suggests that stakeholders have different agendas, and that donors predominate in determining the research portfolio. High-level consensus building at the national and international levels is necessary to ensure that the diverse agendas play a complementary role in support of health system objectives. The Ministerial Summit for Health Research to be held in Mexico in November 2004 should be an opportunity to analyze further data and to commit funding for priorities identified through sharingand discussion of agendas.