Written evidence submitted by Gregory Lewis


Executive summary


The UK has many lessons to learn from COVID-19. From a national security perspective, I would highlight two:


  1. To protect its population from biological threats, the UK sorely needs technical countermeasures that are both effective and fast. Current and horizon biotechnology promises such capabilities, and the UK is well-positioned to develop them. It should.
  2. The national risk register suggested an emerging infectious disease could potentially cause “up to 100 fatalities”. Hundreds of times that number have died from COVID-19. This error suggests a need better account for extreme events, and to better integrate existing best practice for forecasting.


Major events can also provoke mistakes in government response. I also caution against two ‘lessons not to learn’:


  1. Avoid fighting the last war by preparing for the previous pandemic.
  2. Do not de-prioritize accidental and deliberate biological threats.




0.1 The COVID-19 pandemic has inflicted vast damage on the UK population, whether put in health, economic, or wellbeing terms. It demonstrates our vulnerability to biological threats, and underlines the importance of strengthening our health security for the future.


0.2 I work at the Future of Humanity Institute at Oxford University, where I investigate the risk of extremely high consequence biological threats, and how these dangers are best addressed. (I write in a personal capacity.)


0.3 The pandemic is far from over, and there will surely be many security lessons to learn. I highlight two I think are important, and caution against two strategic mistakes the UK could make in response to COVID-19.


Lessons to learn


1 New technical capabilities need to be developed to meet 21st century threats.


1.2 The current portfolio of countermeasures to fight a pandemic can be divided into two categories. ‘Pathogen blind’ countermeasures, which do not require knowledge specific to the pathogen (e.g. supportive hospital care; quarantine, contact tracing, and other NPIs); and ‘pathogen specific’ counter-measures (e.g. vaccines and therapeutics), which do.


1.3 Pathogen blind countermeasures can be pre-positioned and deployed rapidly against an emerging pandemic. Yet they possess limited efficacy: across the world, they have only been effective at suppressing COVID-19 if implemented to a degree which causes substantial hardship to the population in their own right.


1.4 Pathogen tailored countermeasures can be extremely effective, but these can only be developed after the pathogen and its disease are known. This takes too long: despite unprecedented efforts worldwide, a vaccine may not be available this calendar year; effective therapies (i.e. Dexamethasone, Remdesivir), were only found several months after the pandemic began.


1.5 ‘21st century biodefense’ needs capabilities which are both effective and fast. Present and horizon biotechnology promises such capabilities. Of many, I highlight two:

       Sequence-based detection: Genetic sequencing has become roughly 100 000 times cheaper since 2001,[1] making it a viable tool in outbreak monitoring.[2, 3] It could also be a valuable tool for early detection: ‘metagenomics’ - sequencing all genetic material in a sample - could allow novel pathogens to be detected before they cause symptoms.[4, 5]    

       Platform vaccine technology: Vaccines often need to be developed ‘from scratch’, both lengthening development and complicating rapid manufacture, as vaccine production capacity can neither be easily repurposed or quickly expanded.[6] Platforms, shared backbones which can be adapted for a given pathogen by ‘plugging in’ the relevant antigen, could greatly ease both challenges.[7,8]


1.6 Recommendation: The present pandemic demonstrates these future capabilities are already overdue. The UK is better placed than many countries to pioneer developing and deploying these due its advanced bioscience and biotech corpus, alongside initiatives to better utilize the same (e.g. UKARPA). It should.


2 Develop excellence in forecasting to aid decision-making.


2.1 The UK national risk register assessed the danger of an emerging infectious disease as “several thousand people experiencing symptoms, potentially leading up to 100 fatalities.”[9] There have been over 42 000 COVID-related deaths so far.[10] The risk assessment underestimated the potential danger by over 400-fold.


2.2 Although stark, an error of this magnitude is regrettably easy to explain. Many emerging infectious diseases are discovered each year, and even the most notorious recent examples (e.g. Zika, Ebola, SARS, MERS) have had relatively mild impacts on the UK. Anchored on recent memory, ‘up to 100 fatalities’ is a reasonable assessment for the typical event of this type.


2.3 The benefit of hindsight is not needed to see this was a very poor assessment of a reasonable worst case scenario. Emerging infectious diseases (which are not pandemic influenza) killing more than 100 people in the UK is hardly unprecedented: HIV was also an emerging infectious disease, and kills more than 100 in the UK each year even now.


2.4 The impact of disease outbreaks have vary extremely widely: atypically bad outbreaks can be hundreds or thousands of times worse than what is usually observed. This means that most of the danger arises from extreme ‘once in a generation’ (or ‘once in a century’) events rather than typical ones.[11] Risk assessment based on what typically happens will greatly underestimate the real danger.


2.5 Bitter experience also teaches us that people tend to be inaccurate and overconfident when forecasting what could happen, especially with rare events. Subject matter expertise does not guarantee much greater forecasting skill:[12] prediction platforms have outperformed experts in forecasting various aspects of the current pandemic.[13] Yet good forecasters can be identified, and we already know of practices that improve corporate performance such as using precise probability estimates rather than vague terms (e.g. ‘possible’, ‘likely’) and reviewing the track record of previous predictions.

2.6 Recommendation: With risks that can vary greatly in severity, the UK should explicitly consider reasonable worst case events alongside typical events when weighting risk. Forecasting skills in government should be identified and cultivated, and known best practice in forecasting should be deployed for national security strategy.


Lessons not to learn


3. Preparing for the previous pandemic


3.1 Health security has an infamous cycle of ‘panic and neglect’. Rare major events prompt a surge in interest and activity, which wanes over subsequent years as the major event fades from recent memory. This is compounded as rare extreme events provide limited data which is easy to misinterpret, e.g.:


       Prevention paradox: It is hard to observe a ‘successfully prevented pandemic’, and still harder to attribute any such success to any particular attempt at risk mitigation. Thus effective efforts that reduce a real risk can look identical to pointless efforts addressed to an illusory one.

       ‘False alarms’ and crying wolf: Early stage outbreaks tend to be extremely hard to forecast. Optimal policy would err on the side of overreaction, as over-reacting to a ‘false alarm’ is much less costly than underreacting in the early stages of a pandemic. Yet these false alarms and overreactions can erode public confidence.  


3.2 One risk of panic in policy-making is developing a response narrowly tailored to ensure ‘this never happens again’. The next serious outbreak the UK faces will not be a carbon copy of COVID-19: it may be a flu pandemic, but it may also be something which, like COVID-19, is to some degree unprecedented.


3.3 Recommendation: Do not ‘fight the last war’ in biosecurity. The UK should instead aim for its response to enhance preparedness across the entire threat landscape. It can accomplish this by prioritising interventions which have broad benefit rather than those which apply to specific narrow scenarios, and by continuing to balance its mitigation portfolio.


4. De-prioritising accidental and deliberate biological threats


4.1 Historically, naturally arising biological threats have caused vastly more damage to the UK than accidentally or deliberately arising ones. COVID-19 as a single example of a naturally arising threat, has probably caused more harm than all accidental and deliberate threats over the last several decades.


4.2 I believe learning from this that (naturally emerging) pandemics and emerging infectious diseases should be given overwhelming priority over accidental or deliberate events would be a mistake. There are significant threats in the present day, are anticipated to increase, and I believe they will overtake natural events as the principal biosecurity risk over this century.


4.3 The evidence for the historical pattern changing cannot be found in the track record. I point instead to the following.


4.4 Accidental: The last known cases of smallpox and SARS were caused by laboratory exposures, and both cases involved secondary transmission.[14] The 2007 foot and mouth disease outbreak in the UK had a laboratory origin.[15] The 1977 Influenza pandemic was caused by a strain closely related to those isolated in the 1950s, suggestive of an anthropogenic origin.[16]


4.5 Deliberate: Biological weapons may be attractive additions to a state’s pre-existing portfolio of violence. Not all countries are signatories to the Biological Weapons Convention, and multiple signatory states have probably violated the treaty in the past.[17] Non-state actors have also sought biological weapons.[18-19]


4.6 These dangers increase in step with the rapid march of biotechnological progress. The prospects of what 20th century state biological weapon programs (some of which employed thousands and spent millions) could have accomplished with 21st century technology is harrowing. They should remain an urgent and extremely important priority.


22 June 2020



  1. Human Genome Research Institute. DNA Sequencing Costs: Data [Internet] [Cited 22 June 2020] Available from: https://www.genome.gov/about-genomics/fact-sheets/DNA-Sequencing-Costs-Data
  2. Hoenen T, Groseth A, Rosenke K, et al. Nanopore Sequencing as a Rapidly Deployable Ebola Outbreak Tool. Emerg Infect Dis. 2016. 22(2): 331-334. doi:10.3201/eid2202.151796
  3. Nexstrain. SARS-COV-2/COVID-19 [Internet] [Cited 22 June 2020]. Available from: https://nextstrain.org/help/coronavirus/SARS-CoV-2
  4. Moore M, Landree E, Hottes AK, Shelton SR. Environmental Biodetection and Human Biosurveillance Research and Development for National Security. [Internet] Rand Coporation. 2018. [Cited 22 June 2020] Available from: https://www.rand.org/content/dam/rand/pubs/research_reports/RR2300/RR2398/RAND_RR2398.pdf
  5. Miller RR, Montoya V, Gardy JL, et al. Metagenomics for pathogen detection in public health. Genome Med. 2013. 5:81. https://doi.org/10.1186/gm485
  6. Gastfriend D. Daniel Gastfriend: Manufacturing medical countermeasures against catastrophic biothreats [Internet] [Cited 22 June 2020] Available from: https://www.effectivealtruism.org/articles/daniel-gastfriend-manufacturing-medical-countermeasures-against/
  7. World Health Organization. WHO Research and Development Blueprint: Evaluation of Ideas for Potential Platforms to Support Development and Production of Health Technologies for Priority Infectious Diseases with Epidemic Potential. [Internet] August 2016. [Cited 22 June 2020] Available from: https://www.who.int/medicines/ebola-treatment/R-D-Blueprint_Evaluation-of-platform-technologies-for-priority-patho.pdf?ua=1
  8. Adalja AA, Watson M, Cicero A, Inglesby T. Vaccine Platforms: State of the Field and Looming Challenges. [Internet] Baltimore: Johns Hopkins Center for Health Security; 2017. [Cited 22 June 2020] Available from: https://www.centerforhealthsecurity.org/our-work/pubs_archive/pubs-pdfs/2019/190423-OPP-platform-report.pdf
  9. UK Government. National Risk Register of Civil Emergencies: 2017 Edition. [Internet] [Cited 22 June 2020] Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/644968/UK_National_Risk_Register_2017.pdf
  10. UK Government. Coronavirus (COVID-19) in the UK. [Internet] [Cited 22 June 2020]. Available from: https://coronavirus.data.gov.uk/
  11. Fan VY, Jamison DT, Summers LH. The inclusive cost of pandemic influenza risk. National Bureau of Economic Research; 2016 Mar 31.
  12. Tetlock PE. Expert political judgment: How good is it? How can we know?-New edition. Princeton University Press; 2017 Aug 29.
  13. MacAndrew T. Covid-19-survey-8 [Internet] [Cited 22 June 2020] Available from: https://works.bepress.com/mcandrew/5/download/
  14. Heyman, DL, Aylward RB, Wolff C. Dangerous pathogens in the laboratory: from smallpox to today’s SARS setbacks and tomorrow’s polio-free world. Lancet. 2004 363, 1566-1568.
  15. Pennington TH. Biosecurity 101: Pirbright’s lessons in laboratory security. BioSocieties. 2007 2, 449-453.
  16. Rozo M, Gronvall GK. The reemergent 1977 H1N1 strain and the gain-of-function debate. mBio. 2015. 6, e01013-15.
  17. Carus WS. A century of biological-weapons programs (1915–2015): reviewing the evidence. The Nonproliferation Review. 2017 Jan 2;24(1-2):129-53.
  18. Carus, WS. Bioterrorism and Biocrimes: The Illicit Use of Biological Agents Since 1990 (Fredonia Books, Amsterdam 2002).
  19. Meulenbelt SE, Nieuwenhuizen MS. Non-State actors’ pursuit of CBRN weapons: From motivation to potential humanitarian Consequences. Int Rev. Red Cross. 2015. 97 (899), 831-858.