A research and policy agenda for the post-pandemic world

Novel vs established data, methods and techniques

The COVID-19 pandemic required innovation from within the academic, public, private and health sector communities. In record speed, researchers developed new theory, methods and ways of working, and these three aspects interacted to produce new knowledge about biological and social systems and processes affecting disease dynamics. In near real time, researchers sequenced the coronavirus genome, developed tests and validated vaccines via novel mechanisms including the novel mRNA platform and the vaccine platform mechanism more broadly.⁴ In terms of methods, time dynamics of testing required the use of novel data sources and valid methods for including them in both policymaking and academic theory development.^5,6 Novel datasets and new testing have been essential to identifying emergent trends quickly, informing both theory generation and policy, and enabling governments to reduce a 5-week decision cycle imposed by disease and testing dynamics. These advances were made while most researchers adapted to remote work in response to the pandemic, opening new opportunities for interaction and collaboration across institutions and borders.⁷

Novel techniques nevertheless present challenges from both ethical and validity standpoints. Leslie offers five steps for responsible data use and privacy-preserving innovation to respond to the COVID-19 pandemic; these steps fit more broadly into a research agenda going forward, providing a third way between the false choice of compromising privacy on the one hand and pandemic response on the other.⁸ From a validity standpoint, novel mechanisms such as the use of non-traditional data sources for near real-time epidemiological modelling also have their limitations, one notable example being the use of Google search queries for flu prediction.⁹ It is essential to maintain evidence standards for both theory generation and policymaking purposes, not by dismissing these techniques out of hand but by continuing to refine research methods such that they can be used responsibly within a suite of indicators to provide valid, near real-time insights. Anticipatory innovation models provide a structured process to consider the technical, biological, temporal, spatial and political uncertainties and possible futures that may arise and to systematise the technical, political and knowledge management mechanisms that have enabled digital health innovation at speed that have emerged during the pandemic. These approaches, bringing together technologists, interdisciplinary academics, policymakers and citizens, are increasingly important to engage digital transformation toward democratic ends.¹⁰

Coordination between levels: centralised vs local decision making

Across fields, hierarchical top-down/bottom-up models have given way to networked organisations. Indeed, the pandemic has demonstrated the longstanding tension between centralised and local decision making; administrations juggled the coordination and standardisation benefits of centralised decision making with the need for agility, adaptation to local conditions and experimentation that decentralised decision making brings, to varying degrees of success.^11,12 The UK government has laid out guiding principles for local authorities in the interest of a coordinated set of decision-making criteria across levels of government.¹³

In the academic sphere, research coordination quickly became an issue; researchers moved quickly to adapt their research agendas to address the pandemic but, in doing so, left important gaps, such as the study of COVID-19 in children.¹⁴ Further, the commendable rush to contribute to pressing COVID-19 research nevertheless threatened to slow important, existing research in other areas of public health.¹⁵ In the public sector, countries moved quickly to establish the data infrastructure necessary to aggregate disparate epidemiological, health systems and economic data; the different approaches they have taken to data collection, user journeys that generate data, organisations involved and aggregation, and analysis will indicate best practices going forward for the aggregation and use of disparate, real-time data in health policy.^16,17 The design and use of intersectoral data systems contributes to the ‘situational awareness’ necessary for the precision policy decisions that represent a third way between overly permissive policies that allow the virus to propagate and burden health systems and overly restrictive policies that burden the economy and constrain civil liberties.¹⁸

Funding mechanisms like the EU Coronavirus Pledging Conference created necessary support and coordinating signals that helped align the disparate international research groups contributing to COVID-19 research to ensure research was allocated to necessary areas.¹⁹ Data portals similarly enabled researchers to centralise data to enable open science and rapid advancement of the knowledge generation and peer review processes.²⁰ Whole-genome sequences were available to international researchers on GISAID as early as January 2020 and the open access it provides to over 450,000 genome sequences was essential to vaccine development; similarly, the COVID-19 Genomics UK Consortium centralises over 370,000 sequences identified through UK testing programmes and coordinates genomic analysis contributing to regular reports to SAGE on virus mutation.^21,22

Coordination across organisations, disciplines, sectors and borders

Academics have greatly expanded the use of new mechanisms of collaboration across organisations and disciplines, as well as communicating research to policymakers and the public. Researchers developed ad hoc research collaborations across disciplines, institutions and borders (such as the COVID-19 Distributed Volunteer Research Network) to quickly develop cross-cutting research that addressed the biological, epidemiological, legal and ethical aspects of disease dynamics and interventions.⁷

Mechanisms, such as preprints, fast-track publishing and scientific advisory committees (eg the UK government's SAGE), also acted as portals for policymakers and journalists to quickly acquire and act on knowledge from the academic sphere. But while preprints enabled academics to collaborate and policymakers and journalists to track research findings in real time, these external communities were not necessarily equipped to understand the theoretical context and communication for an academic audience, leading to calls for strategies to govern the use of preprints, particularly as scientific debate has played out in the public in real time.^23–25

An interdisciplinary response is essential for this pandemic, in which macro policy choices were quickly framed as a trade-off between economic and health outcomes, ethical considerations of privacy and equity have come to the fore, and individual behaviour is paramount to disease dynamics. The UK government's Scientific Pandemic Insights Group on Behaviours (SPI-B) designed principles for co-production of guidance between policymakers and governed populations to equitably establish local behavioural standards and policies, increasing acceptability and effectiveness of resulting policies.²⁶ The question of vaccine challenge trials brought together researchers, health system administrators, private companies and governments to consider both the scientific and ethical merits, resulting in the approval of the Human Challenge Programme in the UK.²⁷ We would argue that a joint letter from leading researchers (including virologists, epidemiologists, economists, philosophers and ethicists) actually represents a promising achievement in its own right, as a rapid, interdisciplinary policy proposal that unites expert opinion on the scientific and ethical concerns.²⁸

However, these policy trade-offs are particularly challenging because, despite admonitions to ‘follow the science’, under evidence-based policymaking, policymakers trade off policy priorities (an inherently value-driven and political process that depends on scientist navigation of the policy process) informed by available evidence.^29,30 However, the interdisciplinary nature of these decisions means that they are also implicitly trading off silos of academic knowledge and disparate groups of scientific advisors with different expertise and norms.

Similarly, as public compliance with preventive behavioural measures is essential to mitigating this respiratory pandemic, public understanding of and confidence in scientific findings is a critical component of pandemic response. Innovative public communication and transparency tools (such as the GOV.UK dashboard) and novel approaches to data journalism were developed in short order to communicate the breadth of relevant epidemiological and operational data and visualisations interactively, in real time, and at scales that enable individuals to both hold government to account and make informed behavioural decisions.^31,32 As the public engages with these resources, effects such as 7-day moving averages, test sensitivity/specificity, timeseries lags and demographic or geographic multimodality are routinely covered, data literacy, numeracy and susceptibility to cognitive biases are likely to change. The pandemic will not itself resolve a trend of low numeracy in the general public, but advances in data journalism over the past year have gone a long way toward integrating these concepts into common policy discourse.³³ Scientists themselves have also taken to social media to establish collaborations with a wider sphere of academics and to engage directly with the public on the scientific process and their own field, correcting popular misperceptions (or deliberate misrepresentations) of the scientific record and narrowing the gap between the academic and public spheres.

Academic institutions have increasingly partnered with private sector and healthcare organisations to improve translation of their research and inform the design of promising innovations. Collaborations like the new partnership between the P4 Precision Medicine Accelerator (an ecosystem for precision medicine startups) with Nuffield Health bring together private start-ups, healthcare organisations and academics to co-design and test promising innovations, improving their relevance to practice and reducing time to development.³⁴ These initiatives, along with the advancement of the implementation science field, help systematise the design of practice-relevant technologies and their adoption across settings to close the implementation gap and improve health service quality.³⁵

At the international level, coordination across countries and regions is a key example of the limits of centralised decision-making, as multilateral institutions rely on coordination among member states in the absence of hard power. While multilateral institutions, notably the World Health Organization, had a coordinated pandemic response plan in place, they had few mechanisms or hard power to compel nations to act together.³⁶ Most recently, the EU vaccine distribution strategy further demonstrated the challenges of achieving international consensus to vet, procure and distribute specific vaccines.³⁷ While COVID-19 drew exceptional attention because it impacted the world's population, the restrictions on mobility have also impeded efforts to address continuing global health and sustainable development challenges that disproportionately or exclusively impact low- and middle-income countries, doubly impacting these populations.³⁸

Recommendations: a joint research and policy agenda

As Blair and Yiu argue, in setting the agenda for future health policy: ‘the specific technologies matter less than the shift in mindsets and approaches that the modern operating environment makes possible.’³ We recommend a cross-cutting shift in mindsets and approaches that will strengthen our capacity for health research and its relevance for public policy. The pandemic has laid bare the challenges and opportunities in applying research to policy, and a concerted effort across academia and the public sector is necessary to create the structures, mechanisms and norms for effective evidence-based policymaking.

To this end, we propose the following recommendations to shape global health research policy over the immediate, medium and long term, including target outcomes and illustrative initiatives to achieve them.