Impact case study database

The Birthplace in England study [1] has been a key driver towards a stronger focus on midwife-led birthing services across the UK. The National Institute of Health and Care Excellence’s 2014 Guidelines ‘Intrapartum care for healthy women and babies.’ direct practitioners to use the data presented from [1] to ‘advise low‑risk multiparous women that planning to give birth at home or in a midwifery‑led unit (freestanding or alongside) is particularly suitable for them because the rate of interventions is lower and the outcome for the baby is no different compared with an obstetric unit’ [A]. Additional analyses by the University of Oxford researchers who led the Birthplace study [1] provided key evidence to the policy document ‘Better Births’, a five-year forward view for maternity care in England published in February 2016 [Bi]. The evidence from the Birthplace study was integral to the conclusions of ‘Better Births’, which included the recommendation that ‘ The [University of Oxford] evidence reports commissioned for this review show that midwifery care results in fewer interventions. If we can increase the proportion of births supported by midwifery care, we will be able to reduce the cost of medical interventions’ [Bi]. The Birthplace study [1] was also cited by the maternity strategies for Scotland [Bii] (January 2017), Wales [Biii] (July 2019) and Northern Ireland [Biv] (July 2012), as evidence to support increase in provision of midwifery units for low-risk women.

Approximately 750,000 women give birth in the UK each year, 70% of whom are considered ‘low risk’, and 45% are eligible for midwifery-led birth at the end of pregnancy. The Birthplace study [1] has led to substantial improvements in women’s birth experiences and access to midwifery-led settings for birth. Between 2010 and 2016, the number of midwifery-led units in England rose by 42% from 111 to 158; during this period, the proportion of births in midwifery-led settings almost trebled from 5% to 14% [C]. Women who plan birth in a midwifery-led setting have between a two- and five-fold increase in the odds of having a normal birth compared with women planning birth in an obstetric unit. In addition, the proportion of women offered birth in a midwife-led unit/birth centre in England increased from 35% in 2013 to 45% in 2018. Similarly, in Northern Ireland, the number of midwifery-led units has increased from five in 2012 to nine in 2020. By 2018, over half (59%) of UK NHS organisations with alongside maternity units operated an opt-out policy whereby the midwifery unit is the default planned place of birth for low-risk women [D].

The Birthplace programme evidence [1] helped to drive changes in maternity services in Europe, Australia, New Zealand, Canada and the USA. For instance, the New Zealand Government’s 2014 Report on Maternity states ‘ Primary birthing units and home births are recommended for well, healthy women likely to experience normal birth’ (citing [1] and [A]) [E]. Similarly, the 2017 American College of Obstetricians and Gynaecologists (ACOG) guidance on planned home birth specifically notes that the low rates of perinatal mortality reported by the Birthplace study [1] reflect ‘ provision of care by uniformly highly educated and trained certified midwives’ and on this basis ‘supports the provision of care by midwives who are certified by the American Midwifery Certification Board’ for planned home birth [Fi]. Prior to this, ACOG had a negative position towards home births, described by the American College of Nurse-Midwives [Fii].

The MBRRACE-UK and UKOSS national programmes of maternal mortality and morbidity surveillance and confidential enquiries showed that two thirds of the population attributable risk of maternal death is associated with medical co-morbidities, and highlighted maternal suicide as an important cause of death in the year postpartum. This led to the introduction of new maternal medicine networks in England, announced in November 2017 [G]. Previously, networked maternal medicine was only available in one of 12 regions of the UK (8%) despite the fact that 30% of women who give birth are ‘higher risk’. Networked maternal medicine is now available in the remaining 92% of areas and therefore this model of care is accessible to over 160,000 women for whom it was not previously available.

MBRRACE-UK and UKOSS national programmes were cited as evidence underlying the national maternity strategy, including expanding access to specialist perinatal mental health services in the NHS England Long Term Plan published in January 2019 [H], and new funding of GBP50,000,000 for perinatal mental health services in Scotland, announced in March 2019. The work also underpins many of the recommendations in recent NICE guidance on intrapartum care for high-risk women [I]. The MBRRACE-UK evidence has been used to develop a new Medical Emergencies in Obstetrics (MEmO) course which to date has been run over 40 times with 400 total participants. The evidence also helped develop a new Acute Care Toolkit for Obstetric Medicine by the Royal College of Physicians which has been downloaded over 1,400 times.

The measures to prevent maternal deaths in association with medical and mental health co-morbidities is already having an effect. There was a decrease in the maternal mortality rate from indirect (medical and mental health-related) causes from 7.2 per 100,000 women giving birth in 2009-11 to 6.0 per 100,000 women giving birth in 2015-17. This is in spite of an increasing average age at which women give birth and increasing medical complexity of the maternity population overall.

The work led to new legislation in the United States (the US Congress Preventing Maternal Deaths Act of 2018 [Ji]) regarding maternal mortality surveillance and review, and in 2019 Knight contributed insights to a workshop that led to recommendations for quality, availability and consistency of surveillance data and review in the United States [Jii, Jiii]. The work was cited as evidence driving World Health Organisation initiatives on sepsis and maternal mental health, as well as the health care of pregnant refugee and migrant women and their newborn children [K].

Zena Forster, a local independent playwright, became aware of the research findings on maternal mental health, and, working with the researchers, developed a play, 'Stitch Up'. Steeped in emotion and drama, ‘Stitch Up’ draws directly on key public health messages arising from the research, and women's lived experience of mental illness both during pregnancy and postnatally, to raise awareness, reduce stigma and encourage discussion about maternal mental health. An evaluation on behalf of the Arts Council including feedback from audience members noted that it was a ‘ brilliant way to initiate discussion on a controversial topic - approachable and open dialogue’.

The rapid response research studies prompted by the COVID-19 pandemic [5,6] identified key lessons to guide ongoing service provision during the pandemic, including planning for winter and future lockdown scenarios. The increased risk to pregnant women from BAME communities highlighted in [5] generated significant media interest including all major UK outlets (e.g. BBC News, The Times, The Telegraph) and international distributors (e.g. CNN). In response the Chief Midwife in England wrote to all the maternity units in the country calling on them to take specific actions to minimise the additional risk of COVID-19 for BAME women and their babies [L(i)]. Both the UKOSS and MBRRACE-UK studies [5,6] were the basis of the joint statement from the Royal College of Midwives and RCOG: ‘Planning for Winter 2020/21: reducing the impact of COVID-19 on maternity services in the UK’ [L(ii)]. A key principle of this is that services should show ‘particular attention to those from BAME communities or those living with medical, social or psychological conditions that make them higher risk. ’ The report recommends ‘ Services should review the following evidence and recommendations as a matter of priority, and encompass them into their planning: 1. The findings and recommendations of the MBRRACE rapid report’, citing reference [6]. The MBRRACE-UK report was also cited in the Royal College of Psychiatrists’ guidance: ‘COVID-19: Working with vulnerable people’, which listed a number of ‘ key red flags identified from the MBRRACE reports which still need to be acted upon promptly’ [L(iii)]. Even in non-pandemic situations, approximately 10% of pregnant women have a mental health problem. Safeguarding mental health provision therefore protects an estimated 75,000 vulnerable women per year from avoidable harm during pregnancy.

The reviews of neuraminidase inhibitors [2,3]; the development of new methods [1] and the TrialsTracker [4]; and the resulting insights into compliance [5,6] have underpinned campaigns for greater transparency in clinical trial reporting, influenced global changes in healthcare policy, and helped to hold public bodies to account.

Data from the UK National Audit Office confirmed that between 2006/7 and 2012/13, the Department of Health and Social Care spent GBP424,000,000 on oseltamivir (“Tamiflu”) for use in a pandemic, but had to write off GBP74,000,000 of the stockpile [A]. Expenditure in the USA was reportedly in excess of USD1,300,000,000.

Following the first publication of the Cochrane review [2] in January 2012, the UK Parliamentary Public Accounts Committee called in 2013 for an inquiry into the UK government’s policy of stockpiling antiviral drugs. Part of the remit was to explore “ the Cochrane Collaboration’s updated review of Tamiflu and the ramifications that access to clinical trial data has for the whole of medicine.” The University of Oxford lead for the review, Heneghan, gave written evidence based on this research including reflections on the difficulties encountered in obtaining full information on trials for [2]. The Committee concluded that the failure of trial sponsors to share the full results of clinical trials undermined the ability of doctors, researchers and clinicians to make informed decisions about treatments, and thus undermined the use of medicines by the NHS [A].

In 2016, the University of Oxford researchers published an audit of the registration and publication of a set of clinical trials funded by the NIHR and undertaken at two large research units in Oxford, at the request of the associated Patient Involvement Working Group [B]. In March 2017, Goldacre gave written and oral evidence to the Parliamentary Science and Technology Commons Select Committee, drawing on this proof-of-concept audit and other work from the Oxford group. In 2018, the Select Committee’s report “ Research integrity: clinical trials transparency” recommended that the NHS “ should be provided with funding to establish a national audit programme of clinical trials transparency, including the publication of a single official list of which UK trials have published results and those which are due to but have not.” [C].

Based on these recommendations, the NHS Health Research Authority (HRA) launched the “ Make it Public” campaign in June 2019, citing the 2017/18 ‘ Research integrity’ report [C]. The University of Oxford provided data from EU TrialsTracker to inform the consultation, which was acknowledged in a summary in the consultation document highlighting that ‘ around 25% of UK sponsors do not report results on time’. The consultation stated that the HRA planned to change their processes to address this challenge and included proposals on how to improve reporting. [D(i)]. The resulting policy, ‘Make it Public: transparency and openness in health and social care research’, was published in September 2020 and outlined a vision for research transparency and an expectation that results be reported within 12 months of the study end date and annual publication of compliance rates [D(ii)].

In 2015, on the basis of the work undertaken to establish trial reporting transparency for neuraminidase inhibitors [2, 3], the WHO commissioned the Centre for Evidence-Based Medicine (CEBM) at the University of Oxford to provide the background briefing document [E(i)] for the WHO consultation on data and results sharing during public health emergencies. The resulting WHO statement and consensus in September 2015 [E(ii)] resulted in the implementation of a new protocol for data sharing during the Zika outbreak of 2016 [E(iii)], that was subsequently important in the response to COVID-19. The Head of Research and Development at WHO said “ That work and the consultation it supported was very influential as the norms for sharing information prior to manuscript publication have shifted dramatically…. During the 2020 pandemic Medrxiv, Biorxiv and other preprint servers have become critical information sharing modalities; something that was called for during the 2015 meeting” [E(iv)].

In May 2017 the WHO led the development of a joint statement on the public disclosure of results from clinical trials, in which various non-commercial funders, including the MRC, the Wellcome Trust and CEPI, committed to implement key trials transparency policies [F]. The Head of Research and Development at WHO confirmed that the Oxford researchers “played key roles in helping to develop international standards around transparency in clinical research” and that “ The joint statement has led to real changes in the policies of major funders” [E(iii)].

In March 2017 an expert committee of the WHO, citing [2], recommended deleting the antiviral drug oseltamivir from the Core Essential Medicines List (the most efficacious, safe and cost-effective medicines for priority conditions) [G(i)] and transferring it to the Complementary List (medicines which are not necessarily affordable, or for which specialised health care facilities or services may be needed), where it has since remained [G(ii)]. The peer review report [G(iii)] making this recommendation noted that [2] “ includes data from the complete set of clinical study reports of clinical trials of oseltamivir”.

In 2013, the University of Oxford (as the CEBM), Goldacre and the BMJ co-founded AllTrials.net, a global campaign for trials transparency. By 2020, AllTrials had 95,000 individual supporters, 727 organisations (including patient groups, academic bodies, drug companies), and advocates for all past and present clinical trials to be registered, with their methods and results reported in full. Following their 2018 report on ‘Research Integrity’ [C] and the launch of the EU TrialsTracker in 2018 [6], in early 2019 the Parliamentary Science and Technology Commons Select Committee asked Goldacre to provide data on the reporting performance of all UK universities and NHS Hospital Trusts. The Committee reminded all universities and NHS Hospital Trusts of their responsibilities (regardless of current performance) and demanded overall compliance to improve. For a follow-on hearing, Goldacre and AllTrials provided written evidence with updates on performance across all UK universities and Trusts to the Committee [H]. This evidence showed that in the 12 months following the publication of the report (Oct 2018 to Oct 2019), reporting rates for both UK university-sponsored and NHS Trust-sponsored clinical trials increased significantly (59.7% to 72.1% and 35.4% to 56.3% respectively, by October 2019 covering 1,349 trials due to report). It also resulted in warnings from the government to institutions with poor reporting levels.

In 2015, the European Medicines Agency (EMA) published its policy on clinical data for medicinal products for human use (EMA Policy 0070), which required that previously unpublished clinical reports (including clinical study reports) from central regulatory applications be published in an anonymised format [I]. Following publication of the EU TrialsTracker [6] in September 2018, the reporting rate of EU clinical trials rose rapidly over the subsequent 12 months [6] in September 2018, from 51.2% to 61.5%, and increased further to 68.1% as of December 2020 [J(i)]. Clinical trial reporting by industry has also significantly improved as a direct result of TrialsTracker. The Director of the Clinical Disclosure Office at Novartis stated that “ *The TrialsTracker…helped allow Novartis to improve its excellent transparency compliance and achieve 100% compliance in the EU in 2020. In the US, we are proud to use these trackers to highlight our excellent compliance record.*” [K(i)]. Eli Lilly stated that TrialsTracker is used to provide monthly compliance reports and enabled them to meet EMA expectations [K(ii)].

The transparency advocacy organisation TranspariMED, citing TrialsTracker, reported that UK universities outperform European universities in clinical trial reporting [J(ii)], concluding that this was a result of combined pressure from parliament [C], research funders [F] and the media.

The findings from the Cochrane A159 review [2] were covered widely, criticising the money wasted on stockpiling unevidenced drugs. In June 2015, the UK Chief Medical Officer, concerned that widespread coverage could affect public trust negatively, asked the Academy of Medical Sciences to investigate how to restore public trust in scientific evidence for decision making. In their report (June 2017), ‘Enhancing the use of scientific evidence to judge the potential benefits and harms of medicines’, the Academy, cited the Cochrane review [2] as a case study in the introduction, stating that this emphasised “ the need for openness in decision-making processes to allow wider society to judge whether decisions are made based on sufficiently robust and relevant evidence” [L]. Recommendation 5 of the report addressed the publication of research findings.

Following the publication of the Select Committee report on Research integrity [C], the Telegraph used the EU TrialsTracker to identify three unreported vaccine trials from Public Health England [M(i)]. As a result of this coverage, and pressure from Sir Norman Lamb MP, PHE apologised, citing the EU TrialsTracker [6], and reported the results of all three trials [M(ii),(iii)]. In December 2020, PHE was at 100% compliance on the EU TrialsTracker.

TranspariMED has consistently used TrialsTracker data to pressure academic institutions in the US and Europe, resulting in increased compliance. The founder of TranspariMED states that “ TranspariMED has…used data from the Trials Trackers as the backbone of most of its advocacy work” and that they “have never encountered a data tool that has had such a dramatic and sustained positive impact on institutions’ and regulatory bodies’ behaviour” [N(i)]. TranspariMED reports of clinical trial reporting by German universities, using TrialsTracker data, resulted in increased compliance [N(ii)]. TranspariMED also affirm that after they used TrialsTracker data for a report on compliance in the US, “ the majority of the universities flagged as having compliance gaps have substantially improved their performance” [N(i)], illustrating the value of the FDAAA capabilities [5] developed by the DataLab team. Similar outcomes are described by TranspariMED in Austria, Denmark, the Netherlands and Spain [N(i)].

The dietary scenarios developed in study [4] were the centrepiece of a highly influential report by the international EAT-Lancet commission in 2019: ‘Food in the Anthropocene’ [A]. This provided a template for global food-based dietary guidelines. The report concluded that international health, sustainability and climate change objectives would only be achieved with a radical population-level shift towards plant-based diets. In addition to generating considerable media interest (see below), this report resulted in international policy action. At the C40 Mayoral Summit in October 2019, 14 global cities committed to city-level policy change to align with EAT-Lancet’s recommendations [B]. The report has also helped shape new national dietary guidelines that incorporate sustainability, including for Indonesia and Norway. EAT-Lancet also formed the basis of a new policy document from the International Association of Medical students to integrate nutrition and sustainability into global medical curriculums [C].

The research studies have also influenced United Nations (UN) climate negotiations, including meetings of the Conference of the Parties (COP) of the UN Framework Convention on Climate Change in 2017 (COP23) and 2018 (COP24). The research evidence and recommendations were cited in subsequent COP publications: for instance, the WHO’s COP24 Special Report Health and Climate Change [Di] states: ‘ A synergistic combination of supply and demand measures to increase consumption of diets with more fruit and vegetables, produced sustainably and ideally locally, will be necessary to gain the potential health and environmental benefits’, citing [4] . The research studies also informed the Intergovernmental Panel on Climate Change (IPCC) special reports on Global Warming of 1.5 degrees Celsius (2018) [Dii] and on Land (2019) [Diii]; and in several UN reports, including those of the Standing Committee on Nutrition. For example, the IPCC special report on land [Diii] states ‘Consistent evidence indicates that, in general, a dietary pattern that is higher in plant-based foods… is more health-promoting and associated with lower environmental impact…[than] current global average diets’, citing [2], [3] and [4]. The University of Oxford also jointly published with the UN Food and Agricultural Organisation the report ‘ Plates, pyramids, planet’ [Ei], citing [1], which found that only four countries worldwide integrated sustainability within their food-based dietary guidelines for health, prompting more countries to begin doing this. The report [Ei] was cited in the 2018 WHO information booklet on healthy and sustainable diets [Eii] which stated that ‘ A healthy diet should be sustainably produced and consumed, and there is growing evidence that human health and environment ‘win-wins’ are possible’. This represented a new policy direction: prior to this, WHO dietary guidelines focused on health outcomes, without integrating environmental objectives. Fiscal policies, as modelled by [5], to achieve healthy and sustainable diets have now become prominent issues for European Governments; in particular, a tax on processed and red meat was discussed by a meeting of the European Parliament and has featured in election campaigns (e.g. Denmark and Finland).

The EAT-Lancet study [A] was directly referred to during oral evidence sessions for the Parliamentary Environmental Audit Committee inquiry into Planetary Health, and the recommendations in the resulting 2017 report, Our Planet, Our Health [Fi], draw on those of the EAT-Lancet study, for example: ‘One of the key messages of the recent EAT–Lancet Commission on healthy diets from sustainable food systems was that: “Transformation to healthy diets from sustainable food systems is necessary to achieve the UN Sustainable Development Goals [SDGs] and the Paris Agreement”’. The report also cited the EAT-Lancet Commission as part of its recommendation that ‘The Government should …consider using financial incentives to promote access to, and consumption of, healthy and sustainably produced food’. The underlying research papers [2,4,5] were also cited in the Parliamentary Office of Science and Technology’s 2019 policy briefing paper on Climate Change and Agriculture [Fii] and the Committee on Climate Change (CCC)’s 2020 report on land use in the UK [Fiii] as evidence supporting a population level shift to plant-based diets to reduce greenhouse gas emissions. The CCC report made the specific recommendation to ‘ reduce… the consumption of beef, lamb and dairy by at least 20% per person’ [Fiii]. In January 2020, the Behavioural Insights Team published an evidence review outlining the 12 most effective strategies to promote healthy and sustainable diets [Fiv]. This called for carbon taxes on food products to be introduced: ‘ Many health professionals have become proponents for a meat tax to save lives’, citing publication [5].

MPs have acknowledged the role of University of Oxford research in driving forward the debate on healthy and sustainable diets in numerous parliamentary debates. For instance, during a debate in the House of Commons on ‘Environment and Climate Change’ in May 2019, Kerry McCarthy (MP for Bristol East) said ‘If we do not make changes, the food and farming system will singlehandedly use up our Paris climate agreement emissions budget within the next 30 years…There have been endless wake-up calls, including from the UN, the IPCC, EAT-Lancet, Chatham House, academics from Harvard and Oxford, and many more…’. Health and sustainability are now firmly integrated into Government food production policy, as demonstrated by the Department for Environment, Food and Rural Affairs’ 2020 Progress Report: Farming for the Future [G]. This states ‘The next 10 years are a period of significant change and 2021 represents the first step of government and farmers working together to invest in farming business, public goods and the production of safe, healthy and sustainable food.’ Several MPs have also given public support in favour of government-enforced financial incentives to increase access to healthy and sustainable food products, including Caroline Lucas, MP for Brighton Pavillion, at the 2019 Oxford Farming Conference.

The World Economic Forum (WEF) commissioned University of Oxford researchers to write a review on alternative proteins as part of their discussion series on the future of meat. The report [Hi] drew upon the modelling analyses developed for the studies outlined above. The subsequent WEF white paper Meat: The Future, A Roadmap for Delivering 21st-Century Protein [Hii] cited this commissioned report extensively as supporting evidence, for instance: ‘[Hi] highlights that replacing a single daily portion of beef (and to a lesser extent both pork and chicken) with the studied alternatives (beans, pulses, mycoproteins, nuts etc.) can reduce diet-related mortality in high-income and upper-middle-income countries by up to 5%, while simultaneously reducing greenhouse gas emissions’. This paper was sent to delegates and discussed at the World Business Council on Sustainable Development (WBCSD) in December 2018). In its 2020 report on assessment of risks in the food and agriculture sector [Hiii], the WBCSD said: ‘The report issued by the EAT-Lancet Commission [A] in February 2019 outlines how to deliver a sustainable and healthy food system for 10 billion [10,000,000,000] people within the boundaries of the planet by 2050…This is an opportunity for governments to set regional- and country-specific targets and regulations’.

In June 2020 FAIRR (Farm Animal Investment Risk & Return), a global network of investors managing over USD20,000,000,000,000 of assets, published The Livestock Levy: Progress Report [I] in which they concluded that ‘ momentum is clearly building in European capitals’ for an EU-wide meat tax. The research was invoked extensively as supporting evidence in the report, for instance: ‘Research by Oxford University concludes that a health tax on red and processed meat could save over $40 bn [USD40,000,000,000] in global healthcare costs’, as found in [5] . The report’s recommendations for investors in the animal protein sector include reducing meat content in composite products and adopting supply standards that mitigate health and environmental risks.

The EAT-Lancet report [A] has inspired pilot projects from major food service companies to move towards healthier and more sustainable products and menus. This includes the Compass Group, the world’s largest foodservice company, whose Sustainability Director wrote:

“Our sustainability strategy is focused on taking targeted actions where we believe they will have the greatest impact…This exciting pilot with EAT allows us to test key elements from the EAT-Lancet report and use it to inform our business decisions” [J].

The EAT-Lancet report [A] sparked considerable public debate about the merits of plant-based diets, having been featured in over 5,800 articles in 118 countries, and shared over 1,000,000,000 times on social media in the year following its publication. The report and the research [1–5] received widespread coverage in traditional media, documentary films and podcasts, including BBC News, Sky News, The Daily Telegraph and The Guardian. The researchers have been interviewed about their work on Radio 4’s Today programme, and other programmes including BBC’s Trust me, I’m a doctor (1,410,000 episode views) and Countryfile (4,860,000 episode views). Documentary films have included The End of Meat (9/2017, Germany) and #Powerplant (2/2019, Netherlands), as well the video How could veganism change the world? for The Economist (780,000 views). For the latter, approximately 94% of reactions (likes/dislikes) were positive. Opinion pieces in national newspapers have described how this research has motivated dietary change e.g. in the Guardian: ‘I have changed what I eat because of the now overwhelming evidence of global environmental damage caused by meat and dairy production… If the world’s diet doesn’t change, we simply can’t beat climate change [embedded link to EAT-Lancet report [A]]’ . Marco Springmann also summarised the main research findings from [1-5] in a ‘TEDx’ talk (5/2019) which has been viewed over 16,000 times, with approximately 81% of reactions (likes/dislikes) being positive.

The researchers also engaged the UK film and TV industry at screenwriter workshops including BAFTA Screenwriting for Sustainability (6/2016), BBC Eco-Enders (10/2018), and BAFTA Soap Sustainability Story Summit (6/2019). This led to greater understanding within the sector of the health and environmental rationale for plant-based diets, as testified by the Head of Industry Sustainability at BAFTA [K]: ‘ Peter Scarborough and the team at Oxford have helped the industry to get to grips with (climate change). Bringing the scientific and creative community together in this way would not have been possible without the department’s detailed understanding and passionate delivery’. In 2018, Coronation Street (viewing figures of approximately 6,000,000 people) ran a story featuring vegan diets, while in October 2020 EastEnders (viewing figures pf approximately 5,000,000 people) stopped serving meat products on screen, using vegetarian alternatives instead in order to avoid waste and reduce its carbon footprint – a decision that generated significant media attention.

Several surveys [L] indicate that greater awareness of the environmental and health consequences of animal-based foods has led to more individual action to reduce meat consumption. These indicate that the number of vegans in the UK quadrupled from 150,000 in 2014 to 600,000 in 2019 (The Vegan Society) and that in 2019, a quarter (26%) of those who describe themselves as meat-eaters (rather than plant-based flexitarians) were actively trying to reduce their meat consumption (YouGov). The UK National Diet and Nutrition Survey has shown a downward trend in consumption of red and processed meat between 2008/09 and 2016/17. Worldwide participants in the annual Veganuary campaign steadily increased from 3,000 in 2014 to over 400,000 in 2020, with around 50% saying they would continue a vegan diet at the month’s end. University of Oxford research has been cited in Veganuary’s campaign materials, for instance in the following Twitter post from December 2019: ‘ #Veganuary2020 starts in just two weeks time!!! Recent Oxford Uni research concluded that, ‘a vegan diet is probably the single biggest way to reduce your impact on planet earth’.

a. Impact on weight management guidance and practice: The work described above [1-3] was the source of all 37 evidence statements on which the 2014 NICE Public Health Guidance 53 ‘Lifestyle Weight Management’ was based [A]. The guidance, in the form of 17 recommendations, is for commissioners and providers of weight management programmes and health and social care professionals who refer people to these programmes. Six of the recommendations were directly drawn from the evidence reviews from the group. In particular, these reviews informed the key recommendations about the types of weight loss programmes that should be provided. The guidance recommends ‘[commission and provide programmes that] address dietary intake, physical activity levels and behaviour change’ and ‘Ensure specific dietary targets are agreed (for example, for a clear energy [calorie] intake or for a specific reduction in energy intake’ (Recommendations 9 and 11), which are specifically identified from the reviews of behavioural techniques and weight management.

In July 2020, the UK Prime Minister consulted directly with the team at the University of Oxford on strategies for his new policy paper on obesity, ‘Tackling obesity; empowering adults and children to live healthier lives’ [B,C]. This strategy outlines new measures to be taken to tackle obesity, prompted by the increased severity of COVID-19 in people with obesity. The policy includes plans to expand evidence-based services for weight management available for referral from primary care of the kind tested in the University of Oxford trials and reviews [1-4]. This accelerates plans in the 2019 NHS Long Term Plan to improve access to weight management interventions in primary care, which was based on evidence from these trials [D]. Justification of funding and programme rollout has been supported by evidence from the WRAP trial [4], as well as from the brief intervention trial [5], with the programme being adapted to the COVID pandemic by moving to digital interventions [D].

b. Impact on brief opportunistic interventions: As a result of the University of Oxford-based work on brief interventions [5] and subsequent conversation analysis of the GP consultation recordings, PHE approached Aveyard and colleagues to work with them to write guidelines in 2017 to support health and care professionals make brief opportunistic interventions in weight management in routine clinical care [E]. This trial is the only academic reference in that guidance which directly adopts the approach tested in the trial, while the phrases suggested for practitioners to use are taken directly from the findings of the conversation analysis work. Modelling of the cost-effectiveness of the intervention estimated that implementation would generate 5,000 additional quality-adjusted life years (QALY) annually while saving GBP9,000,000 per year for the NHS [F]. The research team also worked with the Royal College of GPs and Cancer Research UK to produce a course for GPs in 2017 on brief opportunistic interventions based on and citing the research evidence [G], which is referenced under ‘further resources’ in the PHE guidance. By December 2020 this had been accessed by 1,020 GPs. The Government’s 2020 obesity policy included plans to implement brief opportunistic interventions [5] in GPs’ Quality and Outcomes Framework from April 2021.

**c. Introducing diabetes remission programmes: The DROPLET trial of the effectiveness, safety and cost-effectiveness of total diet replacement (TDR) programmes [6] directly influenced NHS England to pilot these programmes in NHS settings for people with obesity and type 2 diabetes with the aim of achieving diabetes remission (an intention stated in the 2019 NHS Long-Term Plan) [D,H]. NHS England used data from this and another TDR trial, which applied population impact models to examine the costs, effects on weight loss and effects on incidence of disease, to justify and shape the national pilot programme of work. NHS England stated ‘ The DROPLET study undertaken by Jebb and Aveyard and their group, is one of two studies on which a national pilot programme of work is based, aiming to translate the study findings into routine NHS clinical practice’ [D]. The pilot, launched in September 2020 (postponed from April due to the COVID pandemic), is designed to assess total diet replacements in 5,000 people with recently diagnosed Type 2 diabetes. By December 2020, around 300 people had been referred into the programme [D].

Between the 2014 NICE guidance based on the University of Oxford research and 2019, 341,000 people attended tier 2 weight management services in England, which is an increase of 122,000 over the number expected if trends prior to 2014 had continued [I]. Based on modelling [4], this increase is estimated to have prevented 760 people developing diabetes, hypertension, and cardiovascular disease, leading to a gain of 785 QALYs, and saving GBP327,000 (at standard NICE discount rates) in England.

The programmes for total diet replacement commissioned by the NHS and based on DROPLET trial data were rolled out in September 2020 in a pilot for 5,000 NHS patients with goals including putting diabetes into remission [H]. Modelling of the cost-effectiveness of a total diet replacement programme estimates that this pilot alone may generate 300 quality adjusted life years gained, reducing net NHS costs by GBP6,700,000 over the lifetime of these 5,000 people [J].

The expertise of the University of Oxford team on dietary and behavioural strategies for weight loss enabled them to advise, lead and play a prominent role in a 3-part BBC TV Horizon Special ‘Right Diet’ in 2015 [K]. The series examined how multiple factors including the science of weight gain, hormonal control of appetite and emotional aspects of food could be harnessed to aid weight loss and created empathy for people’s struggles with their biology and psychological make-up. The University of Oxford researchers wrote a manual for the 75 volunteers in the programme who lost an average of 8kg each. 2,300,000 people watched the series in 2015, 8% of the audience share, with 1,200,000 using the website and iPlayer, the second highest viewing figures of that week. The researchers worked with the BBC to engage a wider audience through additional resources – taking questions about the right diet for people in a Q&A session after one of the episodes and developing an online self-help weight loss book that has been downloaded 50,000 times; an online test of which diet would suit which person taken by 1,400,000 people which was the BBC’s record for this kind of engagement; and online recipes that would support the approach that had 162,000 downloads. 575,000 people engaged with the online BBC Magazine, and there were 467,300 tweets about the programme.

The University of Oxford researchers led on and appeared in a BBC1 science special, ‘The Big Crash Diet Experiment’, which was based directly on the DROPLET trial [6] and was viewed by 1,800,000 people in 2018. Working with the BBC, the researchers developed resources to engage a wider audience [K] - an online assessment to enable people to judge their suitability for this approach (taken by 159,000 people) and recipes to allow people to do take this approach at home (viewed by 45,000). 65,000 viewed the University’s web pages providing support for people to follow a total diet replacement and for GPs in managing medication in patients using this approach.

The REF2014 case study described the impact on national and international guidelines on cardiovascular disease prevention from research up to 2011 [1,2,3,5]. New guidelines issued during the REF2021 period continue to draw on this work and be informed further by newer research [4,6], and the effects of these changes have been seen in prescribing practice and service quality standards. In particular, the CTT work has provided a unique and continuously evolving summary of the evidence for additional benefits of lowering cholesterol to very low levels, and as a result has helped to drive target cholesterol levels progressively lower in successive iterations of international guidelines during that period. In Europe, for example, whilst the 2016 guidelines recommended a target of 1.8 mmol/L for LDL cholesterol in very-high-risk patients, the most recent iteration in 2019 [A] lowered this target to 1.4 mmol/L with explicit reference to the work of the CTT.

Clinical guidelines have changed substantially based on findings from the Cholesterol Treatment Trialists’ (CTT) Collaboration and the Oxford-led randomised trials that higher statin doses yield larger reductions in major cardiovascular events. For example, the 2014 NICE lipid modification clinical guidelines [B] recommended that atorvastatin 20mg daily is used for primary prevention for those with expected risk ≥10% of developing cardiovascular disease (CVD) over 10 years, and atorvastatin 80mg is considered for use in secondary prevention. This compares to the previous NICE recommendation (in 2008, referenced in REF2014 case study) to use simvastatin 40mg daily (which produces less reduction in LDL cholesterol than atorvastatin 20mg daily) in both primary and secondary prevention. Internationally, including in Europe [A] and in the US [C], the evidence that ‘lower is better’ provided by the CTT over the past decade has resulted in recommendations that high-dose statin regimens are used to meet lower LDL cholesterol targets.

For example, in the 2019 European Society of Cardiology/European Atherosclerosis Society guidelines on the management of dyslipidaemia, the key recommendation for pharmacological LDL-lowering therapy was: ‘It is recommended that a high-intensity statin is prescribed up to the highest tolerated dose to reach the goals set for the specific level of risk’ [A] with level of evidence ‘A’ (that is, supported by meta-analyses of randomised trials, with the 2010 CTT Lancet paper [2] specifically cited).

Similarly, in the 2018 US Guideline on the Management of Blood Cholesterol, the recommendation was: ‘In patients with clinical ASCVD, reduce low-density lipoprotein cholesterol (LDL-C) with high-intensity statin therapy or maximally tolerated statin therapy. The more LDL-C is reduced on statin therapy, the greater will be subsequent risk reduction. Use a maximally tolerated statin to lower LDL-C levels by ≥50%’ [C]. Throughout that document, the CTT is cited specifically to justify recommendations. For example, on page 1092, it is stated that: ‘ The writing group used primarily the Cholesterol Treatment Trialists’ (CTT) meta-analysis of statin RCTs plus 4 other RCTs’ [2,4], and there are further citations of the CTT to support recommendations for people with diabetes [2] (page 1099), to support recommendations for statins in women [6] (page 1114) and to support recommendations that statins are used for patients with CKD [5] (page 1115).

There has been a clear shift towards prescribing higher intensity statin therapy in response to more stringent targets. For example, a retrospective cohort study using data from all 8,142 standard NHS general practices in England showed that the proportion of statins prescribed which produced LDL-lowering below the NICE-recommended 40% threshold, fell from 80% in 2011/12 to 45% in 2019 [D]. In the United States, an assessment of trends in statin therapy for secondary prevention of atherosclerotic CVD in US adults reported that the use of high intensity statin therapy approximately doubled over the period 2007-2016 [E].

The SHARP trial [5] showed that lowering LDL cholesterol reduces cardiovascular risk in patients with CKD, and remains the sole randomised trial providing evidence for the efficacy and safety of lowering LDL cholesterol in this population. The SHARP trial validated the concept of combination therapy with a statin and ezetimibe in patients for whom the statin dose cannot be increased (whether for safety reasons, or because statins are not tolerated). The SHARP trial was cited as supporting evidence for the National Institute for Health and Care Excellence (NICE) Guidelines on Lipid Modification in 2014 [B]. This brought in a new recommendation that cholesterol-lowering treatment is provided by “atorvastatin 20mg for the primary or secondary prevention of CVD in people with CKD”, with the guidelines noting that “ the evidence base for the use of statins in people with CKD stages 3b to 5 is the SHARP trial”. This led to a new NICE quality standard in 2017 [F], which stated that all adults with CKD stage 3-5 should be offered atorvastatin 20mg. Within three years of the 2014 change in NICE guidance on statin use in patients with CKD, 69% of such patients were taking a statin [G].

SHARP is also cited as the main source of evidence for cholesterol-lowering in patients with CKD stage 3-5 in the 2013 Kidney Disease: Improving Global Outcomes (KDIGO) Clinical Practice Guideline for Lipid Management in Chronic Kidney Disease - the main guideline followed by nephrologists internationally [H]. Furthermore, SHARP is also cited to support the use of statin therapy in the 2018 US Guideline on the Management of Blood Cholesterol [C].

The net effect of wider use of statin therapy worldwide, as well as – in recent years – the increasing use of high-intensity statin regimens and ezetimibe, has been to reduce mean population levels of LDL cholesterol and non-HDL cholesterol. A recent publication [I] estimated that in 2017, high non-HDL cholesterol (which is strongly correlated with LDL cholesterol) was responsible for 3,900,000 deaths from CHD and ischaemic stroke. It also estimated that about half of the reduction in non-HDL cholesterol occurring in high-income countries from 1980 to 2018 was due to the use of LDL-lowering therapy, chiefly statins.

The publication of all of CTT’s main papers has been accompanied by major media coverage, and University of Oxford researchers have played a major role in informing the public about the efficacy and safety of statins. For example, the publication of the in-depth review [4], which summarised the work of the CTT of statins and provided new data on the safety of statins, was accompanied by major international media coverage and attracted substantial interest from the public – as reflected by over 40 mentions in news media and in over 1,600 mentions in social media [J]. University of Oxford researchers have been prominent in making the case for the safety of statin therapy, and in drawing attention to the negative consequences of misinformation about statin safety deriving from non-randomised and non-blinded studies [K].

The University of Oxford’s research into health-related food and drink taxes has had significant international impact, contributing directly to the introduction of sugary drinks taxes in Mexico (2014), the UK (2018) and Ireland (2018) and influencing the introduction of similar taxes around the world. By December 2020, according to the World Cancer Research Fund, there were sugary drinks taxes in 38 countries [A].

Mexico: For over 20 years the University of Oxford has worked closely with researchers in Mexico including (during the period 2011 - 2018) the Undersecretary of Prevention and Health Promotion of the Mexican Ministry of Health. As testified by the Undersecretary, this close collaboration and discussion of the research evidence meant that ‘ even before results were published, the findings from the Mexico City Prospective Study influenced health policy decision making in matters related to non-communicable diseases such as diabetes, hypertension and obesity. In particular, the study influenced both the decision in 2014 to introduce taxes on sugary drinks and the declaration in 2016 of diabetes and obesity as epidemiologic emergencies (the first time this had been done for non-communicable diseases)’ [B]. The 2016 declaration led to the introduction of a wide range of health policies aimed at tackling and treating obesity and diabetes, including healthy eating, on-pack labelling, medical check-ups and physical activity. The introduction of the sugary drinks tax in Mexico was the catalyst for similar adoptions by other countries, as acknowledged by international agencies: ‘The passage, implementation and subsequent evaluation of Mexico’s SSB (Sugar sweetened beverage) tax (implemented 1 January 2014) acted as a tipping point for global action’, World Cancer Research Fund International [C].

UK: The introduction of a sugary drinks tax in Mexico led to the proposal among UK policy makers for a similar tax for Britain. One of the UK modelling studies conducted by Oxford University [4] was cited in Public Health England (PHE)’s policy document ‘ Sugar reduction, Responding to the challenge’ [D] (June 2014) In their subsequent report ‘ Sugar Reduction, The evidence for action’ [D] (October 2015) PHE specifically recommended the introduction of ‘ a price increase of a minimum of 10-20% on high sugar products through the use of a tax or levy such as on full sugar soft drinks, based on the emerging evidence of the impact of such measures in other countries.’

The recommendation from PHE was discussed during the UK Parliamentary Health Select Committee’s inquiry into childhood obesity in October 2015. Professors Susan Jebb and Peter Scarborough presented oral evidence to the committee, which included discussing the evidence from Mexico which indicated that such a tax would be effective in the UK. The resulting report ‘ Childhood obesity – brave and bold action’ (November 2015) [E] recommended a sugary drinks tax in the UK and stated that ‘ The evidence suggests that increasing the price of high sugar products by 10–20% or more through the use of a tax or levy would be likely to have an effect on purchasing behaviour and therefore sugar consumption at least in the short term’. During a debate in Parliament on this recommendation (November 2015), the Chair of the Select Committee (Sarah Wollaston MP), said, ‘ We know from the experience in Mexico that a 10% levy on sugary drinks has led to a 6% reduction in consumption’ [F].

This report had a direct impact on the opinions of MPs, who then spoke in support of the tax in Parliament in November 2015 [F]:

• ‘During the Select Committee’s inquiry, we heard some compelling evidence calling for brave and bold action on obesity…A few months ago I was against a sugary drinks tax, because I am against extra taxation, but the compelling evidence that we heard changed my mind.’

MP for Erewash, Maggie Throup

• ‘Following the introduction of the tax on sugar-sweetened drinks (in Mexico), purchases were reduced by 6% in 2014. We need to be aware of the evidence showing that a tax will actually reduce the purchases of sugary drinks’, MP for Heywood and Middleton, Liz McInnes

• ‘As I have looked into the issue more and more, I have reached the position where I find the evidence compelling: something needs to be done.’ MP for St Austell and Newquay, Steve Double.

This increased political support, coupled with greater public acceptance (described below), led to the announcement of the UK Soft Drinks Industry Levy (SDIL) in March 2016.

Ireland: The publication of the modelling studies [2] and [3] led to the researchers being commissioned by the Irish Government’s Department of Health in 2011 to carry out a modelling study of a 10% tax. This was incorporated within a health impact assessment (HIA) published by the Irish Institute of Public Health in 2012 and then subsequently in BMC Public Health [6]. This led to a recommendation by the Department of Health for the introduction of ‘ a graded tax on pre-packaged [sugar-sweetened soft drinks] on sale in Ireland’ in 2016 [G]. As supporting evidence, the Department of Health referenced the HIA: ‘[the HIA] found evidence that ‘taxation could work to reduce overweight and obesity’ depending on the level of tax imposed…[and] concluded that a 10% tax on the price of SSDs [sugary drinks] would reduce obesity by 1.25% among adults in Ireland’ [G]. The recommendation for a tax was acted upon by the Irish Government, which introduced a tax in April 2018.

International: The success of the tax on sugary drinks in Mexico influenced WHO’s stance on using fiscal measures to reduce sugar consumption [H]. In their influential report ‘ Tackling NCDs (Noncommunicable Diseases): Best Buys’ [H], WHO includes as a ‘best buy’ policy recommendation: ‘ Reduce sugar consumption through effective taxation on sugar-sweetened beverages.’ This report has contributed to the roll out of sugary drinks taxation policies worldwide. The World Cancer Research Fund tracks international food policies and currently notes sugary drinks taxes 39 countries including 10 in Europe [A]. Of these, 37 have been implemented since the introduction of the tax in Mexico in 2014, with many countries citing evidence of the link between sugar consumption, weight gain, overweight and obesity as a driving factor.

The work of the researchers resulted in extensive media attention and public discussion of the issue. For example, Professor Mike Rayner appeared in BBC2’s Trust Me I’m a Doctor (2014) in a feature exploring the health impacts of excess sugar. Professor Rayner was also interviewed in Jamie Oliver’s Sugar Rush Documentary (2015) which was viewed by over 1.2M people in the UK. The celebrity chef Jamie Oliver subsequently went on to be a key advocate for a sugary drinks tax in the UK, contributing to the adoption of the tax by the UK Government and public acceptance. Increased general awareness of the health impacts of sugar led to greater public acceptance of a tax on sugary drinks. In 2013 only about 36% of people in the UK supported the introduction of a tax on sugary drinks. By 2015 this had risen to around 58%, but by March 2016, just after the sugary drinks tax had been announced, 69% supported the introduction [I].

The SDIL succeeded in prompting widespread product reformulation. All major brands of soft drinks (including supermarket brands) launched new sugar-free versions. This increased access to low-sugar drinks and allowed consumers to reduce their sugar consumption without compromising their preferences or brand loyalties. In an evaluation of the UK SDIL, the Oxford University researchers demonstrated that the proportion of soft drinks containing >5g/100ml of sugar available on the UK market fell from 49% in 2015 to 15% in 2019 [J]. In addition, the majority of new products had a sugar level between 4.5 and 5.0 g per 100 mL, suggesting they the 5.0g threshold of the levy had guided the reformulation. In contrast, drinks exempt from the SDIL (such as 100% fruit juice and milk-based drinks), saw no reduction in average sugar content, reinforcing that the levy was the motivating factor for change [J].

A review by Public Health England (PHE) in October 2020 concluded that between 2015 and 2019 in England, there was a 43.7% reduction in the total sugar content per 100ml for retailer and manufacturer branded drinks subject to the SDIL [K]. This meant that although sales in all soft drinks increased by 14.9% during this time, the total sugar purchased per household from drinks subject to the SDIL decreased across all socio-economic groups (between 38.5% - 35.1%). Reduced purchasing of sugar ultimately led to reduced sugar consumption. An analysis conducted by the University of Oxford researchers found that between 2015 and 2018, total sugar consumption from soft drinks declined by 29% despite soft drink consumption rising during this period. Further preliminary results of the evaluation of the levy indicate that compared to the pre-announcement, the amount of sugar purchased in drinks by a household each week reduced by 29.5g (9.8%) one year after the introduction of the levy. The 2017 modelling study [5] suggested that the observed reformulation and reduced sugar consumption due to the levy will ultimately result in 140,000 fewer people with obesity and 19,000 fewer incident cases of type 2 diabetes per year in the UK, as well as nearly 270,000 fewer decayed, missing, or filled teeth annually. According to PHE data, the number of 0-5 year olds being admitted to hospital in England for tooth extractions fell by 20.9% between 2014/15 (16,591) and 2018/19 (13,120). Between April-October 2018, the levy raised almost GBP154,000,000 for programmes to support pupil health and well-being, including primary physical education and the Healthy Pupils Capital Fund.

An evaluation of the tax in Mexico (introduced in 2014) shows that purchases of taxed beverages fell by 7.6%, between 2014 and 2016 [L]. In a modelling study, the Oxford University researchers estimated that this would lead to a projected reduction in new cases of diabetes by 189,300 for the time period 2013 to 2022, amounting to total savings between USD769million to USD1.2billion in direct healthcare costs.

The research projects directly addressed clinical problems identified by clinicians and statisticians in the University of Oxford teams and by several hundred international breast oncologists who are EBCTCG collaborators. The meta-analyses provided conclusive evidence in areas where previously there was uncertainty because of the inconclusive nature of many of the individual trials, which were too small to detect moderate treatment effects. In addition, the work on side-effects has enabled radiotherapy to be delivered more safely, thus increasing its net benefit.

As EBCTCG engages almost all breast cancer clinical trialists across the world, it is in a highly effective position to disseminate its discoveries. For instance, all the main international European, USA and UK guideline groups (identified below) include EBCTCG members in their steering committees. The findings of these studies have led to changes in the international guidelines used by oncologists worldwide. In changing clinical practice and improving patient outcomes and survival following breast cancer, this work has directly benefitted the over 2,000,000 women who are newly diagnosed with breast cancer each year.

EBCTCG papers published between January 2005 and May 2017 were primary references in important updates to each of the major international guidelines for the management of early breast cancer. These include those produced by the National Institute for Health and Care Excellence (NICE) [A], the UK’s Royal College of Radiologists (RCR) [Bi] and the Scottish Intercollegiate Guidelines Network (SIGN) [Bii]; the European Society for Medical Oncology (ESO-ESMO) [Ci], the Japanese Breast Cancer Society [Cii] and the St Gallen international consensus on breast cancer [Ciii]; the American National Comprehensive Cancer Network (NCCN) [Di], Cancer Care Ontario and the American Society of Clinical Oncology (ASCO) [Dii], and the American Society for Radiation Oncology (ASTRO) [Diii]. These updated guidelines brought in new recommendations (detailed below) based on EBCTCG evidence, to improve the survival of breast cancer patients. Speaking of the 2014 update to the NICE guidelines on Early and locally advanced breast cancer: diagnosis and management, the Director for Guidelines for NICE wrote:

‘NICE’s recent (almost completed) major update of our guideline on early and locally advanced breast cancer has been helpfully informed by the analyses conducted by [EBCTCG] over the preceding few years. Your recent work on systemic therapies was central to determining the scope of the current update’ (April 2018) [E].

Following the publication of a meta-analysis of aromatase inhibitors versus tamoxifen [2] aromatase inhibitors have been adopted as standard-of-care: as a result, starting endocrine therapy with an aromatase inhibitor is now recommended for postmenopausal women with breast cancer whereas previously they were managed with either tamoxifen or an aromatase inhibitor. A study in 2019 found the majority of older women (55+) received aromatase inhibitors [F].

Life-saving radiotherapy treatment is now recommended to a wider range of breast cancer patients, including those whose cancer spreads to 1-3 axillary lymph nodes. This policy change was supported by the EBCTCG study [1], as reflected in the ASCO guidelines [Div]:

‘The 2014 publication of the EBCTCG meta-analysis [citing reference 1 as above] provided the signal for this focused update. Based in large part on this signal, the ASCO Breast Cancer Advisory Group ranked updating the ASCO PMRT guideline question concerning use of PMRT for patients with one to three positive lymph nodes as a high priority’. [Div]

The results of the meta-analysis [3] prompted an immediate policy change, with bisphosphonates now being globally recommended for breast cancer patients, when they were not before [A, E, D]. The joint American Society of Oncology and Clinical Care Options guidelines acknowledged the importance of the work, stating: ‘Results of the recently published Oxford Overview (Early Breast Cancer Trialists’ Collaborative Group [EBCTCG]) analysis of individual patient data have provoked particular interest in this area [citing reference 3 as above] and are a key portion of the evidence on this topic’ [Dii]. An increase in bisphosphonates use was confirmed by a survey in March 2020 by the charity Breast Cancer Now, finding that 94% of NHS Trusts who have a breast cancer service were routinely prescribing them and a further 3% in process of making them available [Gi].

The results presented in [3] were recognised by policy makers and used to lobby for improved patient access to bisphosphonates. For instance, during a 2017 debate in the House of Commons on Breast Cancer Drugs, Baroness Blackwood argued in favour of bisphosphonates being licensed as treatment for breast cancer patients, saying ‘ *research in The Lancet in 2015…found that bisphosphonates can be used to help women who are being treated for early breast cancer after the menopause by reducing the risk of the breast cancer spreading to the bone by 28%*’ [H]. At least 35,700 postmenopausal women are diagnosed with primary breast cancer in the UK each year. Routine treatment with bisphosphonates prevents 1,180 women from dying from breast cancer annually: equivalent to one in ten breast cancer deaths with an overall net NHS saving of GBP5,090,000 per annual cohort of patients [Gii].

Before the heart radiation dose-response relationship model was published [4], it was not possible to assess the risk of heart damage radiotherapy for individual patients, based on cardiac risk factors and the amount of radiation received by the heart. The model has influenced practice worldwide for all cancers treated with thoracic radiotherapy, with over 100 references in national or international guidelines [A, Bi, Div, I]. In breast cancer, this enables patients for whom the estimated risk is larger than the benefit to be identified, so that they can avoid radiotherapy, whilst radiotherapy is recommended for women in whom the benefit is larger than the risk [A].

In response to the heart radiation dose-response relationship [4], techniques were developed to spare the heart in several cancer types [J] and are now being used in the clinic. The Royal College of Radiologists’ voted unanimously that the heart should be excluded from breast radiotherapy fields, and that all UK radiotherapy departments should use a breath-hold technique [Bi].

The widespread introduction of heart-sparing radiotherapy has been welcomed by oncologists and patients. Patient representatives from the organisations Independent Cancer Patients Voice and National Cancer Research Institute contacted the research team to thank them for developing the heart dose-response model, noting, for example: ‘I welcome this research that can calculate cardiac risk and give patients the information I did not get. When I was treated the aim was ‘alive after 5 years’. Today breast cancer patients can look forward to a long life, so they need to know what their risk is and how to weigh up options.’ [K].

In the USA, heart-sparing breast cancer radiotherapy has also increased, with >80% (around 430) radiation oncologists using heart-sparing techniques in 2017. Heart sparing radiotherapy is now recommended in national and international clinical guidelines [Bi, Diii]. Worldwide average mean heart dose in left breast radiotherapy reduced by almost half (4.6 to 2.6 Gy) from 2014 to 2017.

These changes in practice have improved survival outcomes of women diagnosed with breast cancer, particularly since the benefits are largely additive. Between 2010 and 2017, the mortality rate for breast cancer fell from 37.4 per 100,000 females to 33.4 (Cancer Research UK statistics). In particular, the reductions in heart dose during radiotherapy are a significant improvement since in breast cancer survivors, heart disease is the most common cause of death (after breast cancer). In the USA, for instance, approximately 250,000 women are newly diagnosed with breast cancer each year, with around 60% receiving radiotherapy. Assuming that half of them had disease on the left side, the results of the study [4] indicate that of the approximately 875,000 women who received radiotherapy to treat breast cancer between 2013-2020, these reductions in heart dose translate to approximately 1,500 fewer deaths from heart disease over the next 30 years.

The EBCTCG meta-analyses have informed prognostic tools used to model patient outcomes, including the most widely-used decision aid, PREDICT [L]. This is recommended in the UK by NICE, and used to provide prognostic information for women with breast cancer around 20,000 times per month i.e. around 240,000 times per year. These decision aids enable clinicians to estimate the absolute effects of treatment for individual women.

Between 31 December 2019 and 31 December 2020 there were more than 83,207,000 confirmed cases of COVID-19 worldwide. The case fatality rate has been estimated at approximately 1% in high income countries, or 20-25% of all hospitalised patients. By performing the largest clinical trial of COVID-19 treatments, robustly and at speed, the RECOVERY trial led by the University of Oxford has achieved worldwide impact in guiding treatment, saving lives, and demonstrating the power of evidence-based medicine. For the first time, treatment of an epidemic disease was changed during a pandemic.

The streamlined trial design (built on research including [1, 2]) achieved unprecedented speed in initiation and recruitment, catching the first peak of infections in the UK, essential in a global pandemic with no known effective treatments. On 16 March 2020, the Chief Medical Officer and NHS England Medical Director endorsed RECOVERY and urged all NHS Trusts to adopt the trial, emphasising that it was crucial research and had been kept extremely simple [Ai]. Within 16 days, 1,000 patients had been randomised, 10,000 by 14 May, and 20,000 by 8 December 2020. Participants were recruited at 176 NHS hospital organisations and, during 2020, 10% of hospitalised UK COVID-19 patients were recruited. In August 2020, RECOVERY was selected to be the UK national platform for phase II as well as phase III COVID-19 trials, based on its unique national coverage and recruitment success [Aii].

In March 2020, Landray’s collaborator Horby (UOA1) found that corticosteroids were specifically not recommended in most COVID-19 treatment guidelines and chose to include the corticosteroid dexamethasone in the RECOVERY trial [B]. Due to the speed and scale of the trial, it was only three 3 months later that RECOVERY proved that dexamethasone reduces COVID-19 mortality by one third in ventilated patients and approximately one fifth in oxygen-treated patients [3]. In December 2020, corticosteroids remained the only globally-available drugs proven to reduce mortality in severe and critical COVID-19.

Impact on national and international policy and clinical guidelines: It is unprecedented that research results are announced at lunchtime, become policy and practice by evening, and save lives by the weekend. The results were announced on 16 June 2020, a day on which 993 people in the UK died from COVID-19. Within 4 hours, dexamethasone – the world’s first coronavirus treatment proven to reduce the risk of death – was recommended for use across the NHS, and the Chief Medical Officer instructed hospitals to act immediately [Ci], stating “ dexamethasone has been demonstrated to have a clear place in the management of hospitalised patients” and urging clinicians to use it for patients requiring oxygen or ventilation.

In the US, on 17 July 2020 the National Institutes of Health (NIH) [Cii] changed its guidance: “On the basis of the…(RECOVERY) trial,…(the Panel) recommends using dexamethasone”. COVID management protocols were revised globally to add recommend dexamethasone, including Saudi Arabia on 17^th June [Ciii], South Africa on 20 June [Civ], and India on 27 June [Cv]. On 22 June 2020, the WHO reviewed their guidance on corticosteroids for COVID-19 “ triggered…by the publication of the preliminary report of the RECOVERY trial” [Cvi]. The WHO conducted a meta-analysis of corticosteroids trials, of which RECOVERY was by far the largest, and the revised guidance, published on 2 Sep 2020, recommended “ *systemic corticosteroids rather than no corticosteroids for the treatment of patients with severe and critical COVID-19 (strong recommendation, based on moderate certainty evidence)*” based on the RECOVERY result [Cvi]. On 18 Sep 2020, the European Medicines Authority (EMA), unusually, provided a template for manufacturers to accelerate submission of amendments to their dexamethasone drug licenses to include the new indication, based on the RECOVERY results [Cvii].

Benefits to patients and global clinical care: Patients in the dexamethasone group in the trial benefited both from increased likelihood of survival and shorter duration of hospitalisation [3]. Health economics analysis [D] estimates that in the UK 12,000 lives (90% confidence interval, 4,250 - 27,000) were saved between 1 July 2020 and 31 December 2020. If dexamethasone has a similar effect size in settings where access to oxygen therapies is limited, in the same period this would translate into approximately 650,000 lives (90% confidence interval 240,000 - 1,400,000) saved globally [D].

Dexamethasone is off-patent, affordably available in most countries, and can be taken by everyone: for less than GBP50, eight patients can be treated, and one life saved. It rapidly became standard of care for the sickest patients across the world. Six days after the RECOVERY result, the drug purchaser Vizient, which supplies approximately half of US hospitals, reported a 610% increase in demand for dexamethasone [Ei]. Independent analysis of US prescribing rates by health care technology company Aetion shows dexamethasone use for COVID-19 in hospital rising from 28% on 14 June 2020 to 52% on 28 June [Eii]. Clinical data reports from International Severe Acute Respiratory and emerging Infections Consortium (ISARIC), gathered from more than 550 sites across 42 countries, show high levels of steroid use globally since the RECOVERY result [F]: for patients admitted since 16 June (until 9 Nov 2020), 70% of those on ventilation and 43% of those on oxygen received steroids.

Adoption of dexamethasone, based on RECOVERY, is widely credited with contributing to the decline in COVID-19 mortality; decreases of 18% in death rates for hospitalised COVID-19 patients have been reported between March and August 2020. For example, clinicians in the US [Gi] and India [Gii] are quoted in academic news articles as attributing decreased mortality to steroids alongside other improvements in patient care.

Benefits to patients and healthcare providers: Learning a treatment is not effective is important, as it protects patients from potential harm and avoids wasting resources. Early in the pandemic, in March 2020, both hydroxychloroquine and lopinavir-ritonavir were widely recommended [B], and hydroxychloroquine was championed by US President Donald Trump. RECOVERY announced in press releases on 5 June and 28 June 2020 that hydroxychloroquine [4] and lopinavir-ritonavir [5], respectively, are ineffective for COVID-19. The large scale of RECOVERY allowed a definitive conclusion – and certainty for clinicians – on the lack of benefit of both treatments among hospitalised patients, which had been initially suggested by smaller trials and non-randomised studies. As a direct result of RECOVERY, other clinical trials of both treatments for severe COVID-19 were rapidly halted, including these arms of the WHO’s large, international SOLIDARITY trial [Hi,ii]. NDPH researchers made a major contribution to the design and analysis of SOLIDARITY, and the hydroxychloroquine and lopinavir-ritonavir interim results from SOLIDARITY were consistent with RECOVERY [Hiii]. Both treatments have can have serious side-effects, including potentially fatal heart arrhythmias associated with hydroxychloroquine, so preventing unnecessary and ineffective prescribing reduced risks to patients, as well as avoiding raising false expectations. Proving that these drugs do not work avoided wasted resources for healthcare providers.

Impact on policy and clinical guidelines: The US Food and Drug Administration (FDA) had granted emergency use of hydroxychloroquine and chloroquine for COVID-19 on 28 March 2020, and the US government distributed millions of doses to treat patients not enrolled in clinical trials. As a direct result of the RECOVERY finding, the FDA revoked the emergency approval on 15 July 2020 [Ii], stating “Only randomized controlled trials can answer the question of whether HCQ or CQ is of clinical benefit in hospitalized patients with COVID-19, and the RECOVERY Trial results offer persuasive evidence of a lack of benefit of HCQ”. Hydroxychloroquine is not recommended by the WHO or EMA for COVID-19, with the EMA citing RECOVERY and SOLIDARITY [Iii]. The lopinavir-ritonavir drug regime is no longer recommended for COVID-19 by any international guidelines. Therefore, three RECOVERY results in little more than three weeks turned COVID clinical guidelines on their head: from widespread use of hydroxychloroquine and lopinavir-ritonavir and low use of dexamethasone in March, to the opposite pattern in July 2020. Subsequently, RECOVERY also found no benefit from azithromycin in patients hospitalised with COVID-19; this was announced on 14 Dec 2020 [Ji] and on 15 Dec the NHS recommended that azithromycin should not be used for these patients [Jii]. This change avoids inappropriate antibiotic use, which can increase antibiotic resistance.

RECOVERY has played a critical role, through media coverage, in changing the public perception of the importance of evidence-based medicine. During 2020, RECOVERY was covered 15,203 times in the media (online, print and broadcast), and #RECOVERYtrial was mentioned 19,000 times on social media. In particular, RECOVERY’s power to counter vocal claims of the beneficial effects of hydroxychloroquine has been an influential tool against fake news. The most prominent example is Twitter limiting the account of Donald Trump Jr and ordering him to delete a misleading tweet containing a video on 28 July 2020 after he made claims about the utility of hydroxychloroquine, which RECOVERY had already proved to be false [K]. Twitter also deleted several tweets shared by US President Donald Trump that contained the false claims, and added a note to its trending topics warning about the potential risks of hydroxychloroquine use [K]. In an article about RECOVERY and SOLIDARITY, expert authors including the Director of the Institute for Evidence-Based Heathcare, Bond University, Australia, commented “ it has been refreshing to see how perfectly such weakly founded claims [of efficacy] can be swept aside by evidence from properly conducted, large-scale, randomized trials” [L].

NDPH research demonstrating the value of linkage (e.g. [2]) drove the establishment, in Sep 2019, of NHS DigiTrials, led by Landray, with IBM, Microsoft and NHS Digital [M]. It facilitates the use of routine health care data by any clinical trial team in the UK to enable more and better trials.

The DiGEM trial [1,2] and subsequent studies carried out between 2008 and 2012 [3,4,5] shifted national healthcare policy in 2015 regarding use of blood glucose monitoring systems, leading to long-term and sustained impacts from changes in clinical practice. Blood glucose self-monitoring systems are used for type 1 and type 2 diabetes and in 2010 the cost to the NHS of providing blood glucose testing strips was approximately GBP150,000,000. Farmer and colleagues’ work was instrumental in challenging the assumption that glycaemic control and quality of life in type 2 diabetes patients not taking insulin (around 88% of the total type 2 population) would be improved with self-monitoring of blood glucose and has led to significant economic savings to the NHS.

At the time the DiGEM trial was carried out, self-monitoring of blood glucose was widely seen as standard of care for routine management of type 2 diabetes. This view gradually changed in the UK and internationally over time with the publication of the trial [1,2], the cost-effectiveness study [3,4] and the meta-analysis [5], culminating in the recommendation made in the NICE guidelines published in December 2015 [A] (described below). However, national and international awareness of the importance of the findings of the trial in advance of the 2015 NICE recommendations was apparent and some examples of this are provided to illustrate the pathway to impact. For example, the 2010 Scottish SIGN guidelines on management of diabetes [Bi] were updated from previous guidelines to take account of new clinical evidence, including that from the DiGEM trial in recommending that routine self-monitoring of blood glucose was not needed for people with type 2 diabetes not taking insulin or drugs that might lead to hypoglycaemia. Initial results from the trial were also cited in 2008 by the National Prescribing Centre (NPC) [Bii] and used by prescribing support teams across the country to provide up-to-date information about the place of self-monitoring blood glucose in type 2 diabetes [Biii].

An observational study in East London implemented a policy of not routinely using self-monitoring of blood glucose for people with non-insulin treated type 2 diabetes between 2010 and 2013, citing the evidence from DiGEM for lack of a meaningful clinical effect from self-monitoring [C]. In this study, local ‘do not use’ guidelines were introduced across two Clinical Commissioning Groups with a total population of 425,000. Unnecessary prescribing of self-testing strips for these patients (approximately 18,000) fell from 42.8% in 2010 to 16.5% in 2013. The authors demonstrated that, if extrapolated to a UK-wide population, this policy would have avoided unnecessary testing in 340,000 people and reduced diabetes prescribing costs in 2013 by GBP21,800,000. Removing the need to test blood glucose several times a day could also lead to an improvement in quality of life and reduced anxiety, as reported by the DiGEM trial group [3] and other studies.

Reinforcing the importance of the DiGEM trial findings, the 2015 RAND evaluation of the NIHR Health Technology Assessment programme [D] selected the trial as one which ‘ had high potential to impact on policy, practice, and ultimately health outcomes and economic benefits’. It highlighted the actions of the research team to widen impact: “ *the project team went on to synthesise evidence across multiple trials, [and] strengthened their case that self-monitoring of blood glucose offered no significant benefit, and also allowed them to demonstrate this for specific patient groups as well as across the population in general.*”

The RAND evaluation went on to note: “ From an implementation point of view, these findings are easy to adopt…However, in practice, there were some challenges associated with putting these findings into practice. A strong industry lobby continued to push back on these findings and worked to try and convince GPs that monitoring was a necessary part of treatment practice.”

Despite this, the findings from the research were incorporated into the updated NICE guidelines on management of type 2 diabetes in adults in 2015 [A]. The guidelines cite the DiGEM trial over thirty times in summarising and discussing evidence for self-monitoring of blood glucose, stating that it provides the most useful evidence with fewest limitations. Overall, a strong ‘ do not do’ recommendation was made concerning self-monitoring of blood glucose for the majority of people with type 2 diabetes, because the guideline development group agreed that self-monitoring would not be of sufficient benefit for most people.

The NICE guidelines have had a significant impact on clinical practice throughout the UK, encouraging more effective use of resources for self-monitoring blood glucose. Following their publication, Training, Research and Education for Nurses in Diabetes (TREND-UK) issued consensus guidelines in January 2017 for the healthcare community [E]. This brought together guidelines from NICE and other national and international bodies with evidence for and against self-monitoring blood glucose in type 2 diabetes and was sponsored by eight of the blood glucose meter companies. DiGEM was one of the six studies selected to provide evidence. Drawing from these sources to suggest who may benefit from self-monitoring, the document states that " self-monitoring of blood glucose is not recommended as part of routine care if HbA1c is within target". The NICE recommendations, including citation of the contribution of the DiGEM study, have also been widely disseminated in professional journals with an emphasis on implementation in day-to-day clinical practice (for example in Nursing in Practice [F]). From 2016-2018, most Clinical Commissioning Groups issued guidance to healthcare professionals on the use of self-monitoring of blood glucose testing in line with the NICE guidance and citing these and TREND guidelines [G].

A wide range of local clinical guidance now recommends that, for many people, routine use of blood glucose testing adds little, if anything, to usual care and should not be carried out unless there is a well-established system for training and support.

Since publication of the NICE guidelines in 2015, prescribing blood glucose monitoring has remained static despite an increase in numbers of people with diabetes and an increasing number of prescriptions for continuous glucose monitoring related to the management of type 1 diabetes, consistent with the background of guidance for much wider use of test strips for people with that condition. In 2011, self-monitoring prescriptions cost the NHS GBP158,000,000 in England (National Prescribing Centre) out of a total diabetes prescribing cost of GBP725,000,000 (NHS Digital); in 2018, the figure was GBP157,000,000 (National Prescribing Dataset) out of a total of GBP1,012,000,000. During that period, the number of patients with type 2 diabetes increased from 2,300,000 to 3,800,000 (Public Health England). The East London study, estimated a reduction in diabetes prescribing costs in 2013 by GBP21,800,000 for a UK-wide population [C]. A similar estimate of potential saving was made in a review by Clar et. al. (2010), which reported that GBP17,000,000 per year spent on self-monitoring blood glucose in patients with type 2 diabetes in the UK would be saved by applying the findings of DiGEM to clinical practice [H]. Scaled to the period from the date of publication of the NICE guidelines in 2015 to the end of the REF2021 period, these estimates suggest that savings equivalent to between GBP85,000,000 to GBP100,000,000 may have arisen from implementation of the NICE guidance.

The American Diabetes Association cite the DiGEM trial in their 2018 standards of care for diabetes, which was updated in 2019 with new wording of the recommendations, again citing DiGEM, to recognise the strength of accumulated evidence: “ In people with type 2 diabetes not using insulin, routine glucose monitoring may be of limited additional clinical benefit” [I]. Continuing education programs in the United States are incorporating this advice with explicit reference to the 2007 and 2008 DiGEM publications [2,3] and suggesting that pharmacists should “ challenge prescriptions … for frequent … SMBG testing in patients with prediabetes or diabetes on monotherapy with agents that have minimal hypoglycemic risks” [J].

Similarly, in Canada, local clinical practice recommendations issued by RxFiles, a well-established Saskatoon academic detailing programme providing objective data on drug use, has issued explicit guidance advising against using self-monitoring of blood glucose in non-insulin treated type 2 diabetes [K], citing the 2008 and 2009 DiGEM papers [3,4] and the 2012 meta-analysis [5].

The Royal Australian College of General Practitioners and Diabetes Australia issued guidance in 2016 including the advice that “ routine self-monitoring of blood glucose for people with type 2 diabetes who are considered low risk and using oral glucose lowering drugs (with the exception of sulphonylureas) is not recommended” [Li], referencing the 2012 meta-analysis [5].

The Ministry of Health in Singapore issued guidelines in 2014 advising that it did not support considering self-monitoring of blood glucose for people with non-insulin treated type 2 diabetes. The DiGEM study was cited in the included evidence statement [Lii].