Health research policy: a case study of policy change in Aboriginal and Torres Strait Islander health research

Background There is considerable potential for health research to contribute to improved health services, programs, and outcomes; the policies of health research funding agencies are critical to achieving health gains from research. The need for research to better address health disparities in Indigenous people has been widely recognised. This paper: (i) describes the policy changes made by the National Health and Medical Research Council (NHMRC) from 1997 to 2002 to improve funding of Aboriginal health research (ii) examines catalysts for the policy changes (iii) describes the extent to which policy changes were followed by new models of research and (iv) outlines issues for Indigenous health policy in the future. Methods This study had two parts: (i) semi-structured interviews were conducted over a four -month period with seven individuals who played a leading role in the policy changes at NHMRC during the period 1997–2002, to describe policy changes and to examine the catalysts for the changes; (ii) a case study was undertaken to evaluate projects by recipients of NHMRC People Support awards and NHMRC Capacity Building Grants in Population Health Research to examine the types of research being undertaken five years after the policy changes were implemented. The proposals of these researchers were assessed in terms of whether they reported intending to: evaluate interventions; engage Indigenous community members and organisations; and build research capacity among Indigenous people. Results Seven policy changes over a period of five years were identified, including those to: establish an ethical approach to working with Indigenous people; increase the influence of Indigenous people within NHMRC; encourage priority research directed at improving Indigenous health; and recognise Aboriginal and Torres Strait Islander health research as a priority area including a commitment to an expenditure target of 5% of annual funds. Seven catalysts for this change were identified. These included: a perceived lack of effective response to the health needs of Indigenous people; a changed perception of the role of NHMRC in encouraging research to maximise health gains; and leadership within the organisation. The case study analysis demonstrated that 45% of all People Support recipients intend to engage Indigenous community members and organisations in consultation, 26% included an evaluation of an intervention and two (6.5%) were granted to an individual from an Indigenous background. Six of seven Population Health Capacity Building Grants that were awarded to study Indigenous health between 2004 and 2006 included an intervention component; these grants supported 34 researchers from Indigenous backgrounds. Conclusion NHMRC made significant policy changes from 1997 to 2002 to better support Indigenous health as a result of external pressure and internal commitment. The policy changes have made some progress in supporting better research models particularly in improving engagement with Indigenous communities. However, there remains a need for further reform to optimise research outcomes for Indigenous people from research.


Methods:
This study had two parts: (i) semi-structured interviews were conducted over a fourmonth period with seven individuals who played a leading role in the policy changes at NHMRC during the period 1997-2002, to describe policy changes and to examine the catalysts for the changes; (ii) a case study was undertaken to evaluate projects by recipients of NHMRC People Support awards and NHMRC Capacity Building Grants in Population Health Research to examine the types of research being undertaken five years after the policy changes were implemented. The proposals of these researchers were assessed in terms of whether they reported intending to: evaluate interventions; engage Indigenous community members and organisations; and build research capacity among Indigenous people.
Results: Seven policy changes over a period of five years were identified, including those to: establish an ethical approach to working with Indigenous people; increase the influence of Indigenous people within NHMRC; encourage priority research directed at improving Indigenous health; and recognise Aboriginal and Torres Strait Islander health research as a priority area including a commitment to an expenditure target of 5% of annual funds. Seven catalysts for this change were identified. These included: a perceived lack of effective response to the health needs of Indigenous people; a changed perception of the role of NHMRC in encouraging research to maximise health gains; and leadership within the organisation.
The case study analysis demonstrated that 45% of all People Support recipients intend to engage Indigenous community members and organisations in consultation, 26% included an evaluation of an intervention and two (6.5%) were granted to an individual from an Indigenous background. Six of seven Population Health Capacity Building Grants that were awarded to study Indigenous health between 2004 and 2006 included an intervention component; these grants supported 34 researchers from Indigenous backgrounds.
Conclusion: NHMRC made significant policy changes from 1997 to 2002 to better support Indigenous health as a result of external pressure and internal commitment.
The policy changes have made some progress in supporting better research models particularly in improving engagement with Indigenous communities. However, there remains a need for further reform to optimise research outcomes for Indigenous people from research.

Background
There is great potential for health research to contribute to better health services, programs, and outcomes. The policies of health research funding agencies can substantially influence the kind of research conducted; there is therefore considerable interest in how the policies of research funding agencies are established, their responsiveness to government and community pressures, and the impact on research practice. For example, the recent review of health research funding in the UK-the Cooksey Report-emphasised the need for an overarching health research strategy and found that "the UK is at risk of failing to reap the full economic, health and social benefits that the UK's public investment in health research should generate" [1].
In Australia, the National Health and Medical Research Council (NHMRC) is the major funder of health research. A review of health and medical research in Australia commissioned by the government in 1998 (the Health and Medical Research Strategic Review, or Wills Committee Review) recommended changes in policy to better focus the research effort on outcomes such as health and wealth creation [2]. Subsequent changes to NHMRC policy contributed to a substantial increase in the number of patents arising from funded research and therefore potential wealth creation. However, less progress has been made in encouraging research to inform health policy and practice and produce health gains [3].
Of all the sub-groups in the Australian population who require a strengthened research effort to produce improved health outcomes, the need is clearest for Indigenous Australians. Research directed at improving the health of Indigenous people is recognised as a major priority in Australia. The life expectancy for Indigenous Australian men is 19 years less than for non-Indigenous men and 18 years less for Indigenous women than their non-Indigenous counterparts [4]. There has been little change in the mortality differential in recent years, in contrast to the progress made in comparable countries such as New Zealand and Canada, where Indigenous health has improved relative to that of the rest of the population [5].
Historically, the research effort in Indigenous health in Australia has been less than optimal [6,7]. Aboriginal populations in Australia have been the subject of research since the 19th Century; however, this research was primarily anthropological and focused on accumulating information before Aboriginal people were 'lost to science' rather than on how best to address Indigenous health problems [8]. Aboriginal Australians report feeling that they have been exploited by disrespectful experimentation-subjected to invasive examinations and procedures, objectified, scrutinised, and inaccurately representedwithout this research conferring any health benefits to Aboriginal populations [7,9,10]. In the 1970s a new dialogue began, led by Indigenous people and focused on issues of control, from community consultation and consent, to intellectual ownership and application of research findings [11][12][13][14]. In the 1980s, these debates culminated in the articulation of ethical guidelines for research in Indigenous populations. Two themes emerged from these debates: for Aboriginal communities to have ultimate ownership and control of research, a concerted effort to train Indigenous researchers will be required and to enable "useful research" to be conducted, Indigenous communities need to help identify and define research questions [6,11,12,14].
Informed by these discussions, there has been a growing consensus over the past ten years both within Australia and internationally that research is more likely to have a long-term impact in improving the health of Indigenous people if it evaluates the impact of health programs rather than simply describing health problems, involves Indigenous researchers in all stages of the research, and builds capacity among Indigenous researchers [15,16]. Research funding policies designed to stimulate research of this kind could greatly assist in creating useful evidence to improve Indigenous health.
It is clear that research funding policies over the past twenty years could have been better targeted to improve Indigenous health. Too often, research has simply described health problems without seeking to find solutions; for example, a review of published research in Indigenous health from 2001-2003 found that a mere 13% of Australian peer-reviewed papers evaluated the impact of interventions, with the remainder primarily describing health problems or their causes [17]. Further, it is widely recognised that in Australia, research has frequently failed to engage Indigenous people as equal partners in research or offered an opportunity for involvement in all stages of data collection, including planning, implementing, analysing and disseminating [6,7].
Beginning in 1997, NHMRC responded to the challenge of improving Aboriginal health with a number of substantial policy changes designed to increase funding for Indigenous research and to better target the research effort. NHMRC's response is of considerable interest in understanding the factors that can contribute to policy change in research funding. Like many research funding agencies internationally, NHMRC had historically almost exclusively funded investigator-initiated research with little capacity to strategically target funding to specific areas; its response to the challenge posed by the need to improve Indigenous health was therefore unique in its history.
This paper (i) describes policy changes NHMRC has made to improve funding of Aboriginal health research (ii) examines the catalysts for policy changes (iii) describes the extent to which policy changes were followed by new models of research and (iv) outlines issues for Indigenous health policy in the future. The paper focuses on an analysis of funding for scholarships, fellowships and other awards to pay the salaries of individual researchers (referred to by NHMRC as People Support). We decided to examine People Support because it provided an opportunity to explore the impact of the policy changes on the development of workforce capacity in Aboriginal health research and to examine support for researchers from Indigenous backgrounds. The impact of policy change on the amount of funding through People Support for Indigenous health research is described elsewhere [18].

Key informant interviews
In order to describe the policy changes and to examine the catalysts for the changes, semi-structured interviews were conducted over a four-month period with seven individuals who played a leading role in the policy changes at NHMRC during the period 1997-2002. These seven individuals were involved in the policy changes that were led by the Aboriginal and Torres Strait Islander Research Agenda Working Group (RAWG); in the course of its business, RAWG interacted with other NHMRC principal committees. Accordingly, interviews were undertaken with the former Chair of the Research Committee, as well as the Chair and other leading Indigenous and non-Indigenous members from RAWG during this period.
The interviews were conducted by one of the authors (SLB) and addressed the policy changes that had occurred; the key factors driving policy change (key evidence, individuals, and circumstances); climate and timing for policy changes; barriers to changing policy; as well as approaches and strategies used to encourage policy change. The interviews were digitally recorded and transcribed. Both the transcribed interviews and notes taken during the interviews were used to conduct the analysis.
The interviews were content-analysed according to major thematic areas and trends in current policy-making literature.

Case Study Analysis
We conducted a case study evaluating the projects undertaken by NHMRC People Support recipients of the following funding vehicles: Scholarships (for postgraduate study, most usually leading to a PhD), Training Awards (for postdoctoral researchers), Career Development Awards (for researchers two to twelve years after the award of a PhD), and Career Awards (for senior researchers). We also examined recipients of NHMRC Capacity Building Grants in Population Health Research; this new funding model was introduced in 2002 as a short-term initiative. The grants support junior researchers to work for five years within a mentoring environment with senior researchers.
A keyword search of the NHMRC Research Management Information System (RMIS) was used to identify Indigenous health researchers who received NHMRC People Support awards with the following terms in either the title, lay summary, keywords, or fields of research: Aborigines or Aboriginal; Torres Strait Islander; Indigenous; Koori. For Capacity Building Grants in Population Health Research, a standardised question on the application form (section 1.3) was used to identify "research involving Aboriginal and Torres Strait Islander Peoples." Applications that ticked 'yes' were included. The authors of this paper were granted access to original application forms for review, and double-coded all applications for 14 items including information about thematic areas of the Road Map, project design, and research practices. Operational definitions for coding the data were developed and reviewed by all the authors. Each case was coded independently by two people against a detailed operational definition; in case of disagreement, the appli-cation was jointly reviewed for consideration until the coders could agree on how the item should be coded.
The grants were assessed according to whether they: • Evaluated interventions: that is, the applications included evaluations or trials of interventions, services or programs designed to improve the health of Aboriginal people.
• Engaged Indigenous community members and organisations: that is, the application (a) described an advisory group with Indigenous membership in project design; and (b) completed a special section of their application to NHMRC outlining their commitment to engage Indigenous community members and organisations in research partnership. Completion of this section is considered mandatory by NHMRC for all health and medical research with Indigenous Australians.
• Built research capacity among Indigenous people: that is, the application (a) proposed to employ or train an Indigenous person as part of the research team; or (b) was a People Support Award to an individual who self-identified as Indigenous.

Classification of Indigenous status
NHMRC People Support applicants were classified as Indigenous if they self-identified as an Aboriginal or Torres Strait Islander person on the application form.

(i) Policy changes to improve funding of Aboriginal health research
Respondents nominated seven changes to NHMRC policy over a five-year period and spanning two triennia of the operation of RAWG; the evolution and sequence of these policy decisions is illustrated in Figure 1.
In the early part of this period, three changes occurred: • Adoption of the Darwin Criteria: Most participants nominated the adoption of the Darwin Criteria in 1997-an initiative of RAWG in its first triennium-as a key policy change. The Darwin Criteria were developed as a set of principles to guide research with Indigenous communities. They were intended to be used to assess project grant applications in terms of their level of engagement and capacity building with Indigenous communities, the significance and benefit of research proposals to Indigenous health and the transferability of the methods to other settings. NHMRC adopted the Darwin Criteria as part of the assessment process to gauge applicants' approach to working ethically and in partnerships with Indigenous communities.

• Establishment of the Indigenous Health Review Panel:
This panel was established in 1997 as part of the assessment process for project proposals in Indigenous health; it was an initiative of RAWG in its first triennium. The Indigenous Health Review Panel utilised the Darwin Criteria and the expertise of Indigenous panel members to provide advice on cultural appropriateness of applications and methods and to comment on approaches to community consultation. The Indigenous Health Review panel was also able to comment on the scientific quality of the applications and to stipulate conditions upon which funding is contingent. The reports of the Indigenous Health Review Panel were used by Grant Review Panels in making further assessment of project grant applications and final funding recommendations.
At the 144 th Council session of NHMRC held in October 2002, major policy issues addressed by RAWG in its second triennium of activity (2000)(2001)(2002) were considered. An options paper developed by RAWG was tabled outlining policy options for Aboriginal health research, including consideration of the RAWG Road Map, mechanisms to increase representation of Indigenous people in the NHMRC and consideration of efforts to increase the level of specific research funding to Indigenous health [19]. Interviewees reported that policy documents prepared by RAWG were received with "a high degree of vocal support" on Council. The following policy decisions were endorsed at that time:

• Endorsement of the NHMRC Road Map: A Strategic Framework for Improving Aboriginal and Torres Strait Islander Health Through Research:
In its second triennium, RAWG conducted a national consultation process to identify priorities for research in Indigenous health. The RAWG Road Map for research was developed through this consultation process, involving a series of four national workshops and written submissions to set out a strategic approach to Indigenous health research for NHMRC. Indigenous community members, researchers and policy-makers contributed to the consultation process. All participants acknowledged the comprehensive consultation process used to develop the Road Map, and identified the Road Map as a compelling policy document to guide future NHMRC investments.

• Acknowledgement of Aboriginal and Torres Strait Islander health research as a priority area for development:
Several participants highlighted the importance of the acknowledgement by the Council of Indigenous research as a priority in improving expenditure and Indigenous representation within the agency. One respondent reported that this acknowledgement enabled Council to allow funds to be earmarked for Indigenous health research.
• Commitment to target of 5% annual expenditure: At the October 2002 Council meeting, it was agreed that NHMRC would work towards a target of expending at least 5% of its annual budget on Indigenous health research. All participants acknowledged this funding commitment as a landmark decision by NHMRC.

• Increase Aboriginal and Torres Strait Islander representation across all NHMRC Principal Committees and Council:
In accordance with the principle to engage Indigenous community members in all stages of research, all participants emphasised the importance of Indigenous participation in decision-making within NHMRC and noted increased Indigenous participation throughout the 1997-2002 period.

(ii) Catalysts for policy change
Participants identified both external and internal factors as being influential in bringing about the policy change. Four external factors were identified by all participants: Policy timeline for Indigenous health research  Two other issues of importance were raised by participants: • Legislation: Several participants commented on the role of the NHMRC Act and differing interpretations over time.
In early discussions, there was a view in the organisation that the legislation precluded the allocation of funding for specific purposes, such as 5% for Indigenous research. This view was revised in internal discussions during the 2000-2002 triennium.
• Implementation: All participants emphasised the importance of revisiting existing policies to evaluate the implementation. Respondents noted that the Road Map had not included measurable indices for implementation and that it would therefore be difficult to assess the extent to which NHMRC had implemented the recommendations. Further, several participants also noted that a strategy for building capacity in Indigenous health [21] had been agreed by NHMRC at its Council meeting in December 2005 but had not been implemented. However, probably of most importance, the policy changes resulted in an explicit acknowledgement of Indigenous health research as a priority area for development and a commitment to a 5% expenditure target of annual funds [19].

(iii) Impact of policy change on models of research
Taken together, this represents a substantial set of policy changes to address an urgent health need through research-the magnitude of the policy response is unique in the NHMRC's history. Historically, the vast majority of research funded by NHMRC has been investigator-initiated research selected for funding primarily on the basis of scientific excellence; as noted by the Wills Review in 1998, relatively little funding prior to 1998 was directed towards strategic research focused on government or community priorities. The lack of response to Indigenous health prior to this time was therefore part of a general philosophy about research funding and indeed hampered the NHMRC's capacity to act strategically in other areas of health need as well [22]. The funding philosophy at the time is illustrated by the view inside NHMRC that its legislation may preclude allocation of funds to a specific area (e.g. 5% to Indigenous health). There was also a view during the period of reform addressed in this paper that a designated allocation of funds to Aboriginal health would necessarily involve a decline in scientific standards.
It is therefore of some interest to understand the factors that led to these policy changes. Kingdon's multiple streams model of policy-making outlines three streams that contribute to whether a policy change is adopted: the problem stream (a given situation has to be identified and explicitly formulated as a problem or issue); the policy stream (an explicit formulation of policy alternatives and proposals must be available); and the political stream (a political event or climate that affects the balance of costs and benefits) [23]. Based on interviews with key participants, three broad sets of factors corresponding to Kingdon's streams were identified. First, there was a clear identification of the problem-a lack of effective response to the urgent health needs of Aboriginal and Torres Strait Islander communities both from governments and from Indigenous communities. The data about disparities in health status were compelling enough for government to recognise that action was required.
Second, a clear action was identified for NHMRC-namely to increase the proportion of funding provided for Aboriginal health research to 5%. The identification of a simple technically feasible response by the House of Representatives, followed by a letter from the Minister, was of fundamental importance to the adoption of the policy change [20,24,25]. The RAWG Road Map also made specific recommendations that could be readily adopted by NHMRC [16]. Third, this research has highlighted the need for new initiatives to build capacity among researchers from Indigenous backgrounds. Relatively few of the People Support Awards were to researchers from Indigenous backgrounds and this has been confirmed by other analyses over a longer time frame [18]. What strategies should the NHMRC implement to increase the numbers of Aboriginal and Torres Strait Islander students in research? Evidence suggests that funding models centred around collaborative research environments with blended teams of skilled and early career professionals are more likely to work successfully on meaningful, long-term research projects and train highly skilled researchers in the process. The NHMRC Capacity Building Grants in Population Health appear to be an effective means of attracting multidisciplinary research teams to work together on highly beneficial, applied research. Some of these grants supported additional early career researchers in groups with already established strengths in Indigenous health research while others were used to establish capacity in groups interested in building new Indigenous health research teams. The reasons for the relative success of Capacity Building Grants in Population Health in supporting researchers from Indigenous backgrounds are not clear; however, it seems likely that these grants may be perceived as offering better opportunities for researchers from Indigenous backgrounds. The grants are longer (five years), are from larger teams with established infrastructure (including members of Indigenous communities and organisations), and provide greater financial support to team investigators than might be received through a Scholarship. They also offer the opportunity to work collaboratively with other early career researchers from Indigenous backgrounds.
The NHMRC might consider drawing on international models for building capacity among researchers from Indigenous backgrounds. For example, the Canadian Institutes for Health Research demonstrated a commitment to "building research capacity and infrastructure in Aboriginal health research" by establishing the Institute for Aboriginal Peoples Health to administer eight Aboriginal Capacity and Developmental Research Environments (ACADRE) centres. Each of these centres provides an array of scholarship and training opportunities to undergraduate and graduate students (the majority of which selfidentify as Aboriginal), as well as to community members and organisations interested in conducting health research [15,26]. Another example is provided by the New Zealand Centre of Research Excellence; in 2002, a fiveyear target was set to graduate a total of 500 Maori PhD scholars across all academic disciplines. In addition to active recruiting and extensive student support services, all students were provided with a mentor to guide their academic development and provide social support throughout their PhD program. The initiative has been highly successful, and New Zealand is on track to establish a critical mass of Maori scholars in disciplines including health, history, social sciences, and education [27]. As part of the current review of the Road Map, NHMRC might consider the inclusion of similar strategies.
In conclusion, it is evident that with sufficient external pressure and internal commitment, it is possible to make substantial changes to health research funding policy. The NHMRC made significant changes to its policy in 2002 to better support Indigenous health. By 2007, funding for Indigenous health research-at least through its People Support Awards-appears to be moving towards a better model of practice. However, there remains a considerable way to go before Australia could be said to have in place strategies that optimised the research effort in improving the health of Indigenous people. Government, the community and researchers should continue to advocate for improved funding and for the development of new models reflecting international best practice.