Enabling Big Science in the Domain of Population Health

Summary

CIHR's Institute of Population and Public Health (IPPH) is developing a Big Science Population Health Initiative focused on social betterment through the study of big public policy questions, where research on interventions will be emphasized. The purpose of this report is to examine the lessons learned from other fields of research in which the big science approach has been applied for several decades (e.g. high-energy physics and astronomy), or more recently (environmental, biological and biomedical sciences), and suggest how they might be applied to future initiatives on PPH research. The information in this report is derived from the white and grey literature and from 14 interviews conducted with individuals engaged in various forms of Big Science, or who have written extensively about its application.

There are several manifestations of Big Science, including think-tanks, networks, platforms, high-throughput projects, top-down initiatives, and true Big Science experiments or projects. They address big questions, require complex infrastructure, generate copious data, and involve many people working across disciplines. They need professional management and adroit communication, exceed the normal duration and funding levels of agency grants, and require partnerships between multiple funders, who exercise vigilant oversight.

Advantages and Disadvantages of Big Science. Big Science generates information that could not be obtained in any other way. It can more rapidly assemble sample sizes large enough to yield statistically valid results. It pools the talents of the international research community, and a rich interdisciplinary milieu incubates novel insights and approaches, which seem to lead to increased productivity and greater impact. Big Science can also provide a superior training environment where trainees learn about disciplines other than their own, in a supportive atmosphere that provides a wider range of experience, including the opportunity to work in industry or the public sector. Ideally, Big Science enriches Small Science by providing infrastructure, expertise, and information.

Small Science suffers from uncoordinated agendas, disciplinary silos, passive sharing of information through the overloaded medium of scholarly publication, and resultant duplication of effort. Big Science coordinates talent to achieve a logical workflow, avoid wasted effort, and allow knowledge transfer and exchange between disciplines. Under the umbrella of Big Science, it is easier to instil a shared sense of purpose, focus on results-oriented research, and facilitate the dialogue between researchers and knowledge users. Big Science projects dedicate resources so that data generated is standard in format, skilfully curated, securely stored, available to all participants, and usually to the broader scientific community. Big Science projects devote considerable effort to communications, both internal and external.

Participation in international Big Science allows Canada to be represented in high-cost areas of investigation for a modest investment. Canada's high reputation in astronomy is a good example. In health and social science research, Big Science is generally solution-oriented and sponsored in expectation of outcomes with rapid impact on serious social, environmental or health problems. Driven by public health concerns as well as scientific opportunity, it attracts funding partners from governmental, charitable and industrial sectors, who set the research agenda in collaboration with researchers.

It is this "top-down" agenda-setting that worries opponents, who believe that scientists should be able to pursue their own curiosity without interference. They also point to the opportunity cost of Big Science, its large overheads, and perceived risk. In reality, it is not a question of Big Science or Small Science, because they are interdependent. A broad, well-funded community of Small Science provides the talent required for Big Science, and Big Science provides access to tools and information on which Small Science thrives.

Big Science in Canada faces several challenges. There is no regular process for determining how Big Science proposals are evaluated or funded, and there is fragmentation of the funding agencies that provide the various elements of support. Multiple funding partners are required, and while recruiting them is tedious, it helps to ensure that the science is oriented to the needs of knowledge-user partners. Lack of process means that decisions about Big Science may be politicised.

Although funding agencies encourage large, multidisciplinary projects, peer evaluations and career advancement are primarily based on the achievements of the independent investigator. This disconnect is a deterrent particularly for young investigators to participate in Big Science. This problem has been overcome in the physics community by careful accounting of the role of every participant in Big Science projects. Other challenges include multidisciplinary peer review, ethics review for multi-site health projects, and the need to evaluate Big Science projects in health in terms of socioeconomic impact as well as publication output.

Successful Big Science depends on visionary leadership and expert management. Leaders have to be esteemed researchers and lead by example, and adopt a consultative and conciliatory style. Much of their time will be spent building coalitions and resolving incipient disputes. Participants in Big Science, while expert in their own fields, also have to appreciate the contributions that colleagues working in other disciplines can bring to the project. Skilled project managers who have the right blend of scientific background and administrative experience are rare but essential.

The organizational structure of Big Science projects separates strategic functions of governance from the daily operational decisions of management, and eliminates potential conflicts of interest in decisions about resource allocations. Good communication between participants is essential, and although electronic media are increasingly important, face-to-face contact builds trust and smooth working relationships. Policies regarding the privileges of funders, attribution of authorship, data standards, curation, release, and ownership and exploitation of intellectual property must be established. All participants must feel that they have a voice in influencing the research agenda as the project evolves. Although expensive infrastructure, plentiful operating funds, and sophisticated data management are necessary, they are not sufficient for success in Big Science unless there are healthy interpersonal relationships between the researchers.

Big Science and PPH research. Already, many epidemiological studies involve such large cohorts that the costs, complexity, data generation, time scales, number of collaborating investigators, and funding partners qualify them as "Big". There are a number of visions for the future that have to be considered in thinking about the place of Big Science in PPH research:

• One Health, a holistic view of the interdependency of the health of all species, which involves veterinary, agricultural, and environmental scientists in addition to those concerned with human health. It is also concerned with the future societal adaptations to an unstable global ecosystem.
• Novel Application of Information Technologies, which are transforming the types of study design and data available to PPH research and the way it is collected.
• Genomics, which has not yet lived up to its billing, since most gene variants uncovered so far are only weakly predictive of disease. The most fruitful intersection between PPH and genomics probably will be in explaining why genotype variation puts certain individuals at particularly high risk from environmental exposure, using the worst-case rather than the average to prompt policy changes.
• Populomics, an emerging discipline that relies heavily on innovations in computer and information technologies, and predicts a future where the integration of genetic data, health records, and population health data will provide a complete understanding of health status, and most healthcare interventions will be preventative.
• Intervention Research, moving from research that identifies the reasons for health inequalities to research on the application and evaluation of interventions that aim to reduce them. Big Science intervention projects go beyond health promotion and disease prevention, and involve multiple domains, such as education, housing, and justice. They face "horizontality" problems because multiple, uncoordinated, ministries and agencies bear partial responsibility to implement policy changes.

These considerations give rise to three suggestions for Big Science approaches to PPH, with short-, medium- and long-term impacts:

Data Platform and Analysis. IPPH should advance the acquisition, curation, and accessibility to the data on which PPH researchers depend. This requires adoption of homogenized national and international standards for data categorization and reliability, as well as solving interoperability challenges arising from hardware, software and privacy issues so that data are available for secondary analysis. There are a number of international examples of organizations that facilitate the accessibility of population health data for research purposes, such as the Population Health Research Network¹ in Australia.

Network of Research and Policy Centres. Most of the policy drivers that can affect the determinants of health are provincial responsibilities, and the most effective relationships between research and policy will occur at the provincial level. This also helps to contextualise population health problems and their solutions, and builds on personal relationships already established. CIHR-IPPH could support regional/provincial PPH research and policy centres, linked into a national network by a coordinating centre that would interact with federal organizations. CIHR and provincial government funders would jointly support the centres, with the provincial government contribution paying for experimental interventions, and CIHR supporting the research that evaluated its effectiveness.

A "Grand Challenge in PPH" focused on one selected high-impact issue affecting the health of Canadians or the global population: informants suggested climate change, infectious disease outbreak identification and response, and aboriginal and circumpolar health as topics. Any of these could combine the ecological perspective with the populomics approach. The project would be intervention-focused, combining targeted research with social advocacy and community activism, involving a range of actors. Partners would likely be drawn from public advocacy organizations, disease-oriented charities, other civil society organizations, and industry. Grand Challenges have to be transformational in their vision and execution, and their goals have to be actionable, producing concrete results within ten years.

This recent statement is about astronomy, but it seems just as pertinent to the future of population health research: "Sooner or later the community will realise that its lone-astronomer days are over, that the subject has become too vast and complicated to be done by anything except large groups whose science can be realised only through (computer) code".²

Acknowledgements

CIHR-IPPH would like to acknowledge the strategic direction and oversight provided by the advisory group members: Jamie Blanchard, Roy Cameron, Erica Di Ruggiero, Nancy Edwards, Tim Evans, John Frank, Vivek Goel, Susan Kirkland, and Michael Wolfson. Special thanks to Dr. Mark Bisby for contributing his time and expertise to prepare this report.

The contents of this report may be reproduced in whole or in part, provided the intended use is for non-commercial purposes and full acknowledgement is given to CIHR-IPPH.

1. The Population Health Research Network.
2. Quote from Jim Gunn, Princeton University in "CultureLab: How astrocoders have mapped the universe", 2010. Accessed 2011-03-28.

Date modified:

2012-12-07