|Home | About | Journals | Submit | Contact Us | Français|
To estimate the sources of funds for health research (revenue) and the uses of these funds (expenditure).
A structured questionnaire was used to solicit financial information from health research institutions.
Forty-two sub-Saharan African countries.
Key informants in 847 health research institutions in the 42 sub-Saharan African countries.
Expenditure on health research by institutions, funders and subject areas.
An estimated total of US$ 302 million was spent on health research by institutions that responded to the survey in the World Health Organization (WHO) African Region for the biennium 2005–2006. The most notable funders for health research activities were external funding, ministries of health, other government ministries, own funds and non-profit institutions. Most types of health research performers spent significant portions of their resources on in-house research, with medical schools spending 82% and government agencies 62%. Hospitals spent 38% of their resources on management, and other institutions (universities, firms, etc.) spent 87% of their resources on capital investment. Research on human immunodeficiency virus/tuberculosis and malaria accounted for 30% of funds, followed by research on other communicable diseases and maternal, perinatal and nutritional conditions (23%).
Research on major health problems of the Region, such as communicable diseases, accounts for most of the research expenditures. However, the total expenditure is very low compared with other WHO regions.
In the past decade, there has been a push to expand the monitoring and tracking of financial resources for health research, particularly in low- and middle-income countries. Although with some constraints,1,2 the Organisation for Economic Co-operation and Development and Latin American countries can and have estimated national health research expenditures using data from their general research and development surveys. But even up to 2004, there had been limited similar measurement activities carried out in low- and middle-income countries, particularly in Africa. The Global Forum for Health Research (the Global Forum), the Council on Health Research for Development and the World Health Organization (WHO) have been working jointly to rectify the situation.
The motivation to promote the tracking of financial flows for health research in low- and middle-income countries came after many international, regional and national organisations, including the Global Forum, argued that a major mismatch exists in research funding in relation to the burden of disease and the locations of the populations who experience this burden. This imbalance is referred to as the ‘10/90 gap’ where only 10% of the world’s research funds are estimated to address 90% of the diseases or populations that experience this health burden.3
The Global Forum reported that the world spent an estimated US$ 84.9 billion in 1998 and US$ 125.8 billion in 2003 on health research.4 Of these total values, about US$ 3.6 billion in 1998 and US$ 4.1 billion in 2003 were spent in low- and middle-income countries, with the latter including mainly Latin American and Asian countries. The sources of funds for low- and middle-income countries in 2003 were estimated as follows: 59% public (including donors), 34% private for-profit firms and 7% private not-for-profit organisations. The low- and middle-income countries’ health research spending in 2003 by type of performer was estimated as follows: 41% government, 38% private for-profit firms, 20% higher education institutions and 1% private not-for-profit organisations.
This study was conducted to describe the financial dimension of health research in the WHO African Region. The data are based on a survey of health research institutions conducted by WHO in 42 sub-Saharan African countries. This survey collected financial data from ‘performers’ of health research who were then able to report on both the sources of funds for health research (revenue) and the uses of these funds (expenditure).
The methods followed to assess national health information systems are described elsewhere5 but are summarised briefly here.
The survey used Tool 6 from the Health Research System Analysis (HRSA) Initiative: Methods for Collecting Benchmarks and Systems Analysis.6 Within the institutional survey, seven questionnaires (representing separate ‘modules’) were completed by the respondent institutions. This report draws on data from two of those questionnaires:
The categories or typology of health research performers used in the survey followed closely those in the research and development surveys of the Organisation for Economic Cooperation and Development (business, government, higher education and private non-profit institutions). They were modified by the Global Forum on Health Research (2001)7 to suit health research and development and to include as specific categories hospitals, medical schools and pharmaceutical firms. The performer categories in the survey were as follows: (1) government agencies; (2) hospitals; (3) independent research institutions; (4) medical schools; (5) private non-profit institutions; (6) other higher education; (7) pharmaceutical firms; and (8) other business firms. The last three categories (6, 7 and 8) were combined under an ‘other’ category for purposes of subsequent analyses because of small numbers. We used IBM® SPSS® Statistics Version 19 statistical software to analyse the data.
The institution survey dataset included responses from up to 847 institutions in 42 countries in the Region (all except Algeria, Angola, Sierra Leone and South Africa). Half of the respondent institutions were under 30 years of age; 70.3% belonged to the public sector; 12.5% were independent research institutions and 64.3% functioned at the national level (Table 1).
Total health research expenditures for the African countries included in the survey were computed by taking the sum of itemised expenditures for the biennium 2005–2006. The estimates for total expenditure in the Region were considered ‘minimum’ levels because of the ‘partial’ coverage of the institution survey.
At the very least, an estimated total of US$ 302 million was spent for health research by institutions that responded to the survey in the Region for the biennium 2005–2006 (Table 2). By extrapolation, the estimated biennial expenditure for all health research institutions in the Region was US$ 1.5 billion (US$ 750 million per annum). This represents roughly one-sixth of the total annual expenditures reported by the Global Forum for low- and middle-income countries in 2003 of US$ 4.1 billion.
A total of 27 categories of funders for health research were used in the survey; however, for the purposes of this report, this number was reduced to just nine categories:
Table 3 shows how, for each type of institution, health research expenditures in 2005–2006 were paid for, expressed in percentage terms by type of funder.
Looking across all performer types, the most notable funders for health research activities were external funding, ministry of health, non-ministry of health government ministries, own funds and non-profit institutions. The different types of performers varied in terms of the major funder for their research activities. The patterns for each category of performer reflected the nature of their functions and financial base. Medical schools, which are service providers and receive fees, covered some of their research expenditures from their own internally generated revenues. Reliance on ministry of health funds was reported by government agencies and hospitals. Other types of research performers, including universities, relied heaviest on non-ministry of health government ministry funds. Independent research institutions, hospitals, medical schools and non-governmental organisations obtained significant funding from rest-of-the-world funds.
Use of health research funds was reported in the performer survey in terms of expense items and research topics. For the purposes of this report, the categories for expense item have been regrouped to:
Table 4 shows how, for each type of institution, health research expenditures in 2005–2006 were used by expense item, with the distribution expressed in percentage terms.
Most types of health research performers spent significant portions of their resources for research done in-house, medical schools spending 82% and government agencies 62%. Hospitals spent 38% of their resources on management, and other institutions (universities, firms, etc.) spent 87% of their resources on capital investment.
Expenditures on the 13 categories of health research topics used in the survey are shown for 2005–2006 (Table 5), with the distribution expressed in percentage terms. Over 80% of expenditures were for research on four topics: human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS), tuberculosis (TB) and malaria (30%); other communicable diseases and maternal, perinatal and nutritional conditions (23%); innovations (14%); and health impact (14%). Basic research accounted for 8% of expenditures.
Research on HIV/AIDS, TB and malaria took a significant share of the health research funds of government agencies and independent research institutions. Research on other communicable diseases and maternal, perinatal and nutritional conditions was an important component of medical schools. The study of health impacts of health policies and programmes was an important component of research in hospitals, while research on innovative practices and product development was also important to hospitals and other agencies (universities, business firms, etc.). Basic research also took a significant share of the expenditures of these other types of performers.
The study of health services delivery received the most attention by non-governmental organisations. Prevention and treatment of HIV/AIDS, TB and malaria, and other communicable diseases and maternal, perinatal and nutritional conditions were being funded across all types of institutions. Studies on the health system were generally carried out by government agencies.
At least US$ 302 million was spent on health research in the Africa Region in the biennium 2005–2006.
The different types of health research performers or institutions varied in terms of the major funders for their research activities, reflecting their functions and operational structure. For example, medical schools are service providers receiving fees for services rendered, and part of their research is paid out of their internally generated funds. Government agencies expectedly rely on government budgets. Hospitals, medical schools, independent research institutions and non-governmental organisations rely heavily on the rest of the world for donor funds for their health research activities.
Except for independent research institutions and hospitals, most performers carry out research activities in-house and through contracts to individual persons. Hospitals spend 38% of their resources on management, and other institutions such as universities and firms spend 87% on capital investment.
Four topics account for 81% of expenditures: HIV/AIDS, TB and malaria; other communicable diseases and maternal, perinatal and nutritional conditions; innovations; and health impact. Research on HIV/TB and malaria accounts for 30% of the funds, followed by research on other communicable diseases and maternal, perinatal, nutritional conditions (23%). Innovation and health impact each account for 14%, while basic research accounts for 8%.
The performer survey used in this report is the result of a start-up effort to establish a system for collecting information related to health research in the Region. The data collection instruments developed by the HRSA Initiative for the health research performer surveys were implemented for the first time in the 42 participating African countries included in this report. The pilot activity was expectedly not perfect, but it served very important purposes. First, it contributed new information about health research in Africa, some of which have been presented above. Second, it provided valuable lessons on what needs to be addressed to improve the outcome of similar activities that will be conducted in the future. The findings discussed in this section relate to some operational aspects of the conduct of the performer survey and the outcomes particularly of the financial module of the survey.
The estimated number of institutions that were identified by country teams to be performing health research activities across the 42 countries was 1496. For lack of any other information, this total number is used as a preliminary estimate, N, of the ‘universe’ or population of health research performers in the 42 countries. The number of respondents to the survey, referred to as survey participants or sample respondents (n), came to 684 institutions. However, of the 684 institutional survey participants, only 166 institutions reported 2005–2006 health research financial data, representing only about 24% of the sample (166/n=166/684) and 11% of the population (166/N=166/1496). The selection of the sample institutions, n, for the survey from among the population of institutions, N, was supposed to follow a systematic, predetermined sampling scheme. However, this requires further documentation.
Therefore, for this preliminary analysis, the relationship between the distribution of sample institutions, n (i.e. distribution according to specified institution characteristics such as type of institution and expected level of research on health topics), and the distribution of the population of institutions, N, is not known. This particular module of the survey was expected to be the most difficult for institutions to document responses and was expected to have the lowest response rates. The importance of building up transparent approaches to document the flow of research funds within institutions, and for which topics, led to the decision to maintain this module despite the expected challenges.
Building on these expectations, to improve the outcome of the financial module if and when the performer surveys are repeated in the future, a number of questions need to be answered and acted upon accordingly. These include:
Resolving these issues would increase transparency and provide better data in the future. Survey results from the financial module can be used for a wider range of analyses, including country-level analysis, if the selection of survey respondents is done systematically based on pre-set, statistically determined, country sampling schemes. Efforts should also be made for follow-ups or call-backs to maintain the original sample selected and to reduce incomplete questionnaires.
The survey results used in the report come from a start-up effort to collect Region-wide data on health research in Africa. Other findings reported on pertain to the general usefulness of the survey and lessons learned. Being a first-time effort, it contributed new information about health research in Africa, such as those presented in this paper.
The survey shows that research on major health problems of the Region, such as communicable diseases, accounts for most of the research expenditures. However, the total expenditure is very low compared with other WHO regions. Future studies could improve the quality of financial data and expand the usefulness of the data for analysis purposes.
WHO Regional Office for Africa.
Not required because the survey did not touch on ethical issues requiring individual consent.
DK wrote the paper and carried out statistical analyses; CZ, PEM, IS and WK reviewed the paper and assisted with fieldwork; PSLD reviewed the design of the study and provided support and overall leadership.
WHO Country Office focal persons for Information, Research and Knowledge Management are also acknowledged for their contribution in coordinating the surveys in countries. Their counterparts in ministries of health are also acknowledged. These surveys would not have been possible without the active participation of the head of health research institutions and their department heads who have given their time and effort to fill out and send back the completed modules and questionnaires. We also acknowledge the contribution of the consultant who prepared the background material for this paper.
Not commissioned; peer-reviewed by Vanash Patel