Independent report

Chapter 3: research

Updated 10 January 2023

Research and scientific advice are cross-cutting themes that are relevant to much of this report, and so we set out a summary here of the structures and practices that have been important in this pandemic for both research and scientific advice, alongside our reflections.

Introduction

In all pandemics and major epidemics the initial response depends on sparse information, and in the case of a new pandemic such as COVID-19 there will often be no proven medical countermeasures. The key purpose of research is:

• to understand the disease itself
• to improve information for policy and clinical decision-making
• to optimise existing clinical treatment
• to provide the tools to move from social to medical countermeasures

The central role of research in supporting the response is sometimes underestimated by non-medical planners and policymakers. Since the mid-nineteenth century science has always been, and will almost always be, the exit strategy from pandemics and epidemics. Throughout the COVID-19 pandemic research has been important in informing the response.

CMOs and GCSAs had a central role in making the case to prioritise science from the earliest stages, supporting the direction and co-ordination of research as well as interpreting its outputs to policymakers and the public. This is likely to be true in future pandemics and large epidemics. Science and technical meetings in January 2020 considered research funding needs and coordination between funding agencies and across different disciplines including the social sciences. In the initial coronavirus action plan laid out in early March 2020 the government priorities were ‘contain, delay, research, mitigate’. Many policymakers were surprised that research was given this priority at that stage, but it was in our view essential from the start to undertake the research to lay the groundwork for any realistic exit from the pandemic.

The ultimate success of scientific endeavours throughout this pandemic has however relied on the collective efforts of thousands of researchers, clinical professionals and the public in undertaking research. We have been struck by the selflessness of the public in taking part in trials and observational studies in great numbers; over a million took part in the UK alone. Their efforts enabled us collectively to test and then deploy life-saving interventions throughout this pandemic at an unprecedented speed. The extraordinary efforts of scientists, and of clinicians who undertook clinical research while treating a major influx of severely ill patients was remarkable.

Unsurprisingly, in many aspects of the pandemic the response was best in those areas where the UK already had strengths pre-pandemic. Pre-pandemic research preparation was also important in enabling rapid initiation of various studies and was stronger in some areas than others.

At the outset of the pandemic the UK had:

• a strong and established clinical, public health and biomedical research sector
• broadly based but reasonably centralised processes to fund and manage publicly funded research
• a relatively large and highly skilled research-focused workforce and research infrastructure
• expertise across a range of relevant disciplines – particularly in clinical sciences

A strong industrial research and development base was also present in some areas, with a skilled and experienced workforce. Public sector research establishments such as the Office for National Statistics (ONS), the Health and Safety Executive and Public Health England (PHE, latterly the UK Health Security Agency (UKHSA)) were also important in enabling rapid commissioning and execution of research. It was also important to have a rigorous and experienced regulator for therapeutics and vaccines in the Medicines and Healthcare products Regulatory Agency (MHRA), and for research ethics in the Health Research Authority (HRA). The UK also had in the NHS and devolved equivalents a workforce with a long tradition of basing clinical decisions on trial data, and undertaking trials and observational studies.

On the other hand, there were areas where the UK was not as strong as other countries, and these are equally important to reflect upon. The UK’s diagnostics industry, for example, was not as large as some other high-income countries that were able to more rapidly step up large-scale testing operations.

In any emergency there are 4 key considerations for scientific research:

1. What are the most important questions to answer at a given point in time – and what will be the future ones research needs to start for now?
2. How can these be answered most effectively and efficiently with the tools available and involving the right people? This often required a degree of pragmatism.
3. How can the outputs of research be shared, interpreted and translated into scientific advice at speed to support practical decision-making, but without losing rigour?
4. What are the practical applications of science that will be needed and how will this be achieved?

Under normal circumstances assessing the relative priority of different disease areas and mobilising funds takes much of the time before research can be conducted. The clear international priority of combatting COVID-19 made this much faster. Similarly, many processes which normally are time consuming, including data sharing and scientific and ethical review, were made extremely rapid – often taking days rather than months. Regulators including the HRA and the MHRA turned things around at remarkable speeds, without losing rigour. This was however at the expense of a lot of otherwise excellent non-COVID-19 research which was deprioritised or stopped altogether and has proved slow to restart.

There is always some degree of tension between speed and strength of scientific methodology, but this is much more acute during a medical emergency. Methodologically weak research is potentially dangerous because it gives a false sense of certainty and can mislead. At several points in this pandemic there was pressure to agree widespread deployment of treatments before trials or other research had been undertaken and analysed based on weak (or absent) evidence. Helping to make the case for proper studies and then waiting for evidence was a key role of senior medical and scientific leaders in the system. NHS staff were extremely disciplined in randomising new treatments to trials rather than just giving them based on theory and this paid dividends in rapid results with convincing answers, whether positive (such as dexamethasone, various vaccines) or negative (such as chloroquine, HIV drugs, ivermectin). For example, even at the peak of the first wave some hospitals recruited 60% of eligible patients into the Randomised Evaluation of COVid-19 thERapY (RECOVERY) trial.^{[footnote 1]}

This chapter does not go into the details of the hundreds of studies undertaken, or even the major ones, some of which are covered in subsequent chapters. It simply aims to identify some common themes that may prove useful in subsequent pandemics and epidemics.

How the most important questions changed over time

The relative importance of different kinds of research for policy and practice changed as the pandemic evolved. In part this was due to the urgency of the decisions being taken, but largely because different study designs were inevitably going to report along different timelines. They all needed to be running from as soon as possible after COVID-19 was clearly likely to be a global threat. Many of the major studies were approved and launched within a few weeks of the first cases of COVID-19 being reported outside China, although it was accepted that some of them, especially clinical trials, would take many months to years to provide definitive results. The sections below set out some broad areas of enquiry, some of which are covered in more detail in other chapters of this report.

Understanding the virus and the disease

(For more detail on this see Chapter 1: understanding the pathogen.)

In the first 3 months as COVID-19 moved from being a localised disease in China to a pandemic, basic epidemiological and clinical data were urgently needed to inform public health and clinical advice. Key variables included:

• mortality by age and other characteristics
• the basic reproduction number (R0) and doubling time
• probable routes of transmission and their relative importance

Much of this was initially from Chinese scientists and clinicians, and then replicated in other countries, especially Italy with a more similar age structure and health service to the UK. Having the genotype publicly available early on due to the work of the Chinese and other scientists was essential to the development of polymerase chain reaction (PCR) tests and the initial work on possible vaccine candidates, including in the UK. The global sharing of genotype information has been a critical part of the response to COVID-19 throughout the pandemic to date. We consider these are likely to be common to the majority of future epidemics. The rapid establishment of COVID-19 Genomics UK consortium (COG-UK) supported viral genotyping at scale which enabled an understanding of viral spread and evolution.

Modelling data were important in helping to refine the key epidemiological variables and helped inform advice on early policy and public health decisions. In the initial phase of the response, modelling research played a critical role. Modelling is covered in Chapter 5 in more detail.

Studies on virology and immunology were important to inform an understanding of the clinical picture and potential interventions. Early establishment of sample collections was important.

Understanding the impact of the pandemic and of different interventions

Observational clinical studies were also needed both to inform early policy and clinical practice. The First Few Hundred (FF100) study was specifically designed to answer clinical questions early and something similar needs to be undertaken for any pandemic or epidemic. These have the advantage of producing early data when there are none, and the disadvantage that early cases tend to be atypical. Recognising this, several important clinical and cohort studies were conceived or launched in these early months. These included:

• the International Severe Acute Respiratory and emerging Infection Consortium’s (ISARIC) COVID-19 Clinical Information Network (CO-CIN) study of patients from across the UK with severe disease
• the SARS-CoV-2 Immunity and Reinfection Evaluation (SIREN) study of healthcare workers
• the Easter study in care homes
• the Vivaldi study

ONS data and data on disparities in outcome including by ethnicity and geography from PHE (subsequently UKHSA) became very accurate (for more details see Chapter 2: disparities). Observational data from combined studies were increasingly granular and influential as the first wave progressed and going into the second wave, meaning that the epidemiological and clinical understanding of the disease was substantially better in late 2020 than in March 2020. This also allowed for more accurate modelling. The combined effect was that the scientific advice to inform policy could be much more certain from the start of the second wave onwards. Routine data flows and interoperability improved and became an important resource for research.

Observational data helped change the scientific consensus on several key variables over this time. Important examples for public health measures included:

• the relative contribution of asymptomatic transmission
• the relative contribution of aerosols compared to droplets
• the risks for people from different ethnicities, for children and for those living with obesity

Examples for clinical practice included the clinical course of severe disease, the role of thrombosis and anticoagulation, and mechanical ventilation.

COVID-19 led to significant experimentation with different ways of deploying mass testing of different forms to try to improve clinical outcomes, reduce transmission and provide practical isolation advice. Some of this was conducted with formal scientific methodology. Mass testing was most central to thinking during the second wave, after reliable rapid tests usable by the public had been developed but before a vaccine had been deployed at scale. The development of tests and their deployment are covered in Chapter 6: testing.

Building knowledge of medical countermeasures

In the case of a variant of a known pathogen such as influenza, the normal early research would be to determine whether existing medical countermeasures to that pathogen (drugs, vaccines) work or can be adapted. In the case of COVID-19 there were no coronavirus-specific human medical countermeasures. An assessment from first principles was undertaken, which informed early drug trial candidates (most of which did not work). A decision was therefore taken to trial existing drugs with some theoretical reason they might work, largely undertaken in the public sector, while accelerating development of coronavirus-specific treatments in the pharmaceutical industry with some public sector support. The trials of existing drugs were expected to take months, and development of new drugs years. The first readouts from clinical trials of existing drugs occurred before the second wave peaked, with the most important ones being those which altered the immune reaction to COVID-19 (steroids and other rheumatology drugs) rather than antivirals. Drugs specifically for COVID-19 inevitably took longer. The development and testing of therapeutics and vaccines is covered in more detail in Chapter 9: pharmaceutical interventions.

Studies to develop a vaccine for COVID-19 started within weeks of the genotype being published. It was supported by clinical trial data within 9 months and available from midway through the second wave in the UK. The one general point it is worth making here is that the extraordinary speed of development and effectiveness of viral vector and RNA vaccines was a surprise to almost all scientists. On the positive side this demonstrates how fast a vaccine could be developed for the next pandemic, if it is achievable. There is a danger this falsely reassures some policymakers that a vaccine can be produced at this speed for the next pandemic. The last major pandemic was HIV where there is still no effective vaccine, despite decades of serious investment and scientific effort.

Mechanisms to get research studies prioritised and underway

From the start of the pandemic there were several concurrent risks which are likely to remain a theme in future pandemics including:

• the risk that research would not be undertaken because of the urgent need to act. This can lead to research being perceived as a luxury or getting in the way of action, in turn leading to an endless cycle of unevidenced intervention and the science for an exit not being undertaken. This risk is exacerbated where clinical research staff have to be reassigned to provide clinical care
• multiple studies launching together competing for resources so that none of them had sufficient statistical power to get a definitive answer in a realistic timeframe. This was seen in many countries around the world
• research would only be undertaken in the teaching hospitals, slowing its completion and raising equity and generalisability concerns
• novel interventions (such as new drugs) would be prioritised over the more easily scalable testing of existing interventions (such as steroids)
• the risk of not translating research findings into practical deliverable products – there is a tendency to underestimate the needs of development and deployment science leading to delays in the pull through and implementation

At all stages, but particularly in the earliest months, there were hundreds of potential questions to answer about the pathogen, the disease, their impacts and possible effective interventions – and these therefore required careful but rapid prioritisation. Doing so involved multidisciplinary panels and committees drawing on a range of scientific expertise – for example, in the ‘Urgent Public Health’ (UPH) badging panel which was activated in January 2020 to determine the most important COVID-19 research for priority funding and resource. This is covered more fully in Chapter 9: pharmaceutical interventions. The National Institute for Health and Care Research (NIHR), the Medical Research Council (MRC) and NHS England in particular used this mechanism in England to prioritise their resources, and this was supported by the CMOs and national clinical directors. These panels directed resources to a limited set of studies considered of national importance, at the expense of others. National panels could take account of international panel views of priorities such as those convened by the World Health Organization (WHO). Later, the 7 strands of the National Core Studies programme brought together senior experts to identify research projects, integrated teams and infrastructure needed to answer essential policy and operational questions ranging from transmission risk in specific setting or groups through to immunity and long COVID. This coordination, cross disciplinary working and focus on implementation was important. With a limited ability to test treatments, the COVID-19 Therapeutic Advisory Panel collated expert views on which drugs to bring into major trials to get answers on the most promising.

Not every observer will agree with every decision taken by these prioritisation panels, especially knowing with hindsight which studies and interventions worked. The alternative, which was potentially multiple uncompleted, underpowered or slow-to-report studies, would however almost certainly have been worse, and the relative contribution of UK science to the global stock of knowledge about COVID-19 in the first 2 years supports the overall approach. Large studies on the few most important practical questions enabled us to get results fast, though this approach does not work for all potential research interventions and so a range of approaches will be needed. The UK approach was successful for phase 3 and 4 trials but less effective for phase 1 and 2.

In addition to sifting proposals from researchers, interdisciplinary expert groups helped to review current evidence and flag gaps in the evidence base to highlight further questions that might have been missed. They were supported by a number of evidence review teams. Throughout the pandemic, ongoing collaboration between scientific experts (including the Scientific Advisory Group for Emergencies (SAGE)) and policy and operational teams, helped determine which questions were most needed to inform the response as well as what science could reasonably deliver to answer them in a given timeframe.

Close working between government experts and academics was important in targeting resource to high priority research in both directions. There were routine updates – for example, between the National Immunisation Schedule Evaluation Consortium (NISEC) and UKHSA, the Joint Committee on Vaccination and Immunisation (JCVI), the Deputy CMO (DCMO) and the Vaccine Task Force to keep NISEC’s clinical research relevant to the UK Immunisation Programme. The New and Emerging Respiratory Virus Threats Advisory Group (NERVTAG) also worked closely with the UK’s major platform trial for repurposed therapeutics, RECOVERY. It was also helpful to communicate regularly across all 4 nations of the UK, and joint UK GCSA and CMO forums supported this.

Conducting research swiftly and efficiently

Early coordination, rapid funding mechanisms from UK Research and Innovation (UKRI), MRC, NIHR and the joint CMO and GCSA fund enabled a fast start.

For clinical research, once broad prioritisation had happened, swift ethical review and regulatory review by HRA and MHRA was key.

A UK CMO letter to clinicians (1 April 2020) supported the UPH badging process for clinical studies by asking the NHS to prioritise recruitment to UPH trials, and to desist from prescribing off-licence drugs outside of trials.^{[footnote 2]} UK CMOs also supported recruitment for priority trials in the NHS by writing to doctors to encourage enrolment and by mobilising the NIHR and equivalent workforce in devolved nations in April and May 2020.^{[footnote 2]}, ^{[footnote 3]} Central direction helped clarify research priorities, and academics worked at speed and with innovative approaches to complex issues to make the prioritised questions researchable.

Sleeping protocols and contracts designed pre-pandemic for emergencies helped stand up research rapidly – for example, CO-CIN, which built on the inFLUenza Clinical Information Network (FLU-CIN) established during the 2009 to 2010 H1N1 influenza pandemic. From the outset of the pandemic, there was a need for management protocols, data collection protocols and ethics approvals to collect samples and data rapidly, as well as plans and a repository to enable data sharing and linkage. Further examples are given in Chapters 9: pharmaceutical interventions, and Chapter 10: improvements in care.

Wider practical coordination across government, the private sector, the NHS and academia was also needed – for example, to ensure sufficient tests were made available to support vaccine trials and key observational studies at a time when testing capacity was under pressure. These relationships and processes helped keep researchers apprised of policy and operational challenges so that their work adapted as necessary throughout the pandemic, and kept government clear on what research could realistically deliver, when, and where the blocks to doing this might lie.

It was, as is usual in emergencies, most efficient to adapt existing infrastructure and processes where possible to rapidly commission and undertake research once priorities were set. This included using the NIHR’s clinical research staff to support trials, engaging Health Protection Research Units (HPRUs) or using existing research consortia such as NISEC. Setting up new systems invariably takes longer.

There was a need to use multiple approaches to funding or commissioning new research in order to move things along swiftly. Broadly, these approaches in the UK were:

• open calls for research through funders. This included rapid calls for research that were set up in the first months of the pandemic, longer term research calls, and targeted calls for research on particular topics such as an NIHR and UKRI call in early summer 2020 for research to explain and mitigate the disproportionate death rate from COVID-19 among people from black, Asian and minority ethnic (BAME) backgrounds, including BAME health and social care workers
• direct funding for urgent research, which drew on existing funding release mechanisms – such as the Fighting Fund which distributed NIHR funding for research following joint agreement from both CMO (England) and GCSA. Work funded through this route included the Oxford vaccine, CO-CIN and COG-UK
• support for commercial studies – for example, by mobilising clinical research networks to support Novavax vaccine trials

A wide range of disciplines have been important in supporting the pandemic research response including biological, medical and pharmaceutical sciences, social sciences (including behavioural science), data sciences, epidemiology, immunology and engineering among others.

It should be acknowledged that not every intervention was easy to test. Established methods for testing drugs, vaccines and diagnostics, augmented by platform trials, allowed rapid progress. The established science of advanced manufacturing then enabled production of vaccines and therapeutics at speed. Testing social interventions or indeed the effects of face coverings was much harder.

Sharing, interpreting and translating research outputs into scientific advice at speed

During this pandemic there was a global shift in research practices, with open access and pre-prints widely available from early on and experts able to review evidence as soon as it was available. In March 2020, chief science advisers from 12 countries wrote an open letter to journals outlining their support for open access practices, building on experience of previous epidemics on sharing data.^{[footnote 4]}, ^{[footnote 5]} There is no doubt these practices were beneficial to pandemic response and should be supported in a future pandemic. The rapid review processes did, however, present some difficulties in some cases in interpreting the evidence, especially when rigorous peer review processes were bypassed and there was a pressing need for expert review of research evidence. Review is important not only to translate research outputs for decision-makers, but also to examine their methods and implications in depth.

The public and general media engaged with research to a degree not seen before, debating its outputs and methods in public forums and often with unprecedented levels of discussion between the media and scientific experts. Organisations such as the Science Media Centre also helped explore diverse expert views and summarise latest evidence at speed.

SAGE had a central role in interpreting the latest research evidence and its relevance to UK policy, determining confidence in research outputs, summarising where consensus views were clearest, and highlighting further questions that needed research focus. The breadth of disciplines present at various SAGE meetings where new research was considered is notable; a list of participants is publicly available.^{[footnote 6]} Alongside existing sub-groups of SAGE such as the Scientific Pandemic Influenza Group on Modelling Operations (SPI-M-O), further groups were set up to provide regular specialist advice on key topics such as:

• children and young people
• care settings
• environmental modelling

Scotland and Wales also set up national groups of experts to consider the latest evidence for their local contexts, the COVID-19 Advisory Board and the Technical Advisory Cell respectively. After a short delay SAGE minutes and papers were made publicly available from early in the pandemic and provide summaries of emerging evidence with confidence statements alongside to aid decision-makers and the public in interpreting research outputs.^{[footnote 7]}

Clinical research was generally assessed by clinical panels, existing expert groups and individual clinicians reviewing evidence relevant to their clinical practice, though the speed and volume of review needed could make this challenging. Research for vaccine scaling, high-tech manufacturing and production required industrial as well as academic scientists.

Reflections and advice for a future CMO or GCSA

Point 1

Research will always be one of the most important parts of any response.

It is fundamental to turning a response to any new pandemic or major epidemic from a very broad-based societal response to a much more focused (and therefore less potentially harmful) medical one, as well as improving clinical management.

Point 2

The main reason that a research response was possible at scale was pre-existing strengths.

These strengths included:

• the excellence and broad base of UK academic and industrial science
• the strong culture of evidence-based medicine in the NHS
• co-ordinated funding
• above all a remarkable spirit of volunteering by the public

Point 3

Pre-planning before the pandemic where possible, and adapting existing structures rather than building new ones, allowed a much faster response than would have been possible otherwise.

It was also important to prioritise key areas for an accelerated response.^{[footnote 8]}

Point 4

Rapid prioritisation and review was essential, along with a commitment to test clinical interventions rather than just deploy them.

CMOs and GCSA had to take a visible role in this along with the collective clinical and scientific leadership of the UK, as the temptation just to deploy untested clinical interventions in the face of a rising wave or to launch multiple underpowered studies was very strong.

Point 5

Several methods and processes came to the fore in this pandemic including:

• platform trials
• preprints and open access
• very rapid review

They were essential for the emergency phase, but short-cutting peer review comes with some disadvantages. Which of them should be retained for non-emergency times needs debating. Disadvantages included a potential loss of rigour in peer review and potential for early or minimally evidenced findings to be misinterpreted in the public arena.

Point 6

Multidisciplinary research increased in importance and strong cross-disciplinary teams emerged.

This was a feature of several aspects of the response and is likely to be important for any future pandemic. Funding mechanisms and coordination across disciplines, and between industry and academia were needed.

Point 7

The testing of social interventions and policies was difficult.

More work in this area would be beneficial.

Point 8

It is important to plan not only for stepping up pandemic-related research, but also reinstating other (non-pandemic) research as soon as possible.

It was possible to stop non-COVID-19 and less urgent research very rapidly due to the work of many teams of researchers and their sponsors. It has been difficult and slow to stand up this research again after the initial emergency phase. This is a concern.

References

1. CAS Alert letter from UK CMOs and NHS England National Medical Director to key health bodies, 18th August 2020. Available from: https://www.cas.mhra.gov.uk/ViewandAcknowledgment/ViewAlert.aspx?AlertID=103085 ↩

2. CAS Alert letter from UK CMOs and NHS England National Medical Director to key health bodies, 1st April 2020. Available at: https://www.cas.mhra.gov.uk/ViewandAcknowledgment/ViewAlert.aspx?AlertID=103012 ↩ ↩²

3. CAS Alert letter from UK CMOs and NHS England National Medical Director to key health bodies, 6th May 2020. Available from: https://www.cas.mhra.gov.uk/ViewandAcknowledgment/ViewAlert.aspx?AlertID=103037 ↩

4. Letter from chief scientific advisers of Australia, Brazil, Canada, Germany, India, Italy, Japan, New Zealand, Republic of Korea, Singapore, United Kingdom, United States of America to members of the scholarly publishing community, 13th March 2020. Available from: covid19-open-access-letter.pdf (wellcome.org) ↩

5. Whitty CJM, Mundel T, Farrar J, Heymann DL, Davies SC, Walport MJ. Providing incentives to share data early in health emergencies: the role of journal editors. The Lancet. 2015 Nov 7;386(10006):1797-8. ↩

6. SAGE Transparency data: List of participants of SAGE and related sub-groups for the COVID-19 pandemic. Available from: https://www.gov.uk/government/publications/scientific-advisory-group-for-emergencies-sage-coronavirus-covid-19-response-membership/list-of-participants-of-sage-and-related-sub-groups ↩

7. SAGE Collection: Scientific evidence supporting the government response to coronavirus (COVID-19), including minutes of COVID-19 SAGE meetings. Available from: https://www.gov.uk/government/collections/scientific-evidence-supporting-the-government-response-to-coronavirus-covid-19 ↩

8. Cabinet Office Guidance: ‘100 Days Mission to Respond to Future Pandemic Threats. Reducing the impact of future pandemics by making Diagnostics, Therapeutics and Vaccines available within 100 days. A report to the G7 by the pandemic preparedness partnership.’ Published 12 June 2021. Available from: https://www.gov.uk/government/publications/100-days-mission-to-respond-to-future-pandemic-threats ↩