2. Professor John Edmunds
PROFESSOR JOHN EDMUNDS (affirmed).
Questions From Lead Counsel to the Inquiry
Mr Keith: Professor, could you commence your evidence, please, by giving the Inquiry your full name.
Professor John Edmunds: Professor John Edmunds.
Lead Inquiry: Professor Edmunds, you have kindly provided a substantial witness statement, INQ000273553, we have it there on the screen. We can see from the bottom of the first page that that page is page 1 of 115, in fact, and it’s a statement that you signed, certified as being true on 30 August 2023; is that correct?
Professor John Edmunds: Yes.
Lead Inquiry: You are an expert in infectious disease modelling, in pandemic planning, by extension, and also, by virtue of your particular expertise, a de facto expert in epidemiology.
You are the chair in infectious disease modelling at the London School of Hygiene and Tropical Medicine?
Professor John Edmunds: I am.
Lead Inquiry: Have you been involved in pandemic planning at the United Kingdom level for many years?
Professor John Edmunds: Yes.
Lead Inquiry: Were you the head of the Modelling and Economics Unit at the Health Protection Agency? Is that the body now known as the UKHSA?
Professor John Edmunds: Yes, and it is, yes.
Lead Inquiry: Were you therefore, in fact, one of the first members of SPI-M –
Professor John Edmunds: I was, yes.
Lead Inquiry: – of which we’ve heard much? It’s the Scientific Pandemic Infections Group on Modelling, of course.
You left the Health Protection Agency in June 2008 when you took up your chair at the London School of Hygiene and Tropical Medicine, but did you carry on working on, in particular, pandemic influenza –
Professor John Edmunds: I did, yes.
Lead Inquiry: – influenza pandemics, over the years, whilst you were still serving on SPI-M? And were you at the forefront of the expert field of modelling in epidemiology in relation to epidemics both in the United Kingdom and abroad?
Professor John Edmunds: Yes, I suppose you want me to say, but yes.
Lead Inquiry: All right. You were also a member of NERVTAG, and you I think joined NERVTAG in 2014, and you served on that committee from 2014 through to 2022. So when we confronted the pandemic in the United Kingdom, you continued to serve on all those committees. I think you attended 97 SAGE meetings, 99 SPI-M-O meetings, and 91 other subgroup or related meetings?
Professor John Edmunds: As far as I could ascertain, yes.
Lead Inquiry: And I think, in addition, 74 NERVTAG meetings?
Professor John Edmunds: It was busy.
Lead Inquiry: It was indeed busy.
You participated in a number of other groups, of which we’ve heard mention, for example EMG, the Environmental Modelling Group, the Children’s Task and Finish Group, the Moonshot Scientific Advisory Group and a number of other bodies or committees set up by the public agencies in the United Kingdom –
Professor John Edmunds: Yes.
Lead Inquiry: – including Public Health England and government departments such as the DHSC.
Professor John Edmunds: I did, yes.
Lead Inquiry: To add to your burdens, throughout the pandemic, because of course you are the chair in infectious disease modelling at the London School of Hygiene and Tropical Medicine, you were intimately concerned with the work that continued to be done by the Centre for Mathematical Modelling of Infectious Diseases, which is an integral part of the London School of Hygiene and Tropical Medicine?
Professor John Edmunds: Yes, correct.
Lead Inquiry: I think throughout the pandemic, the CMMID, which is what I’m going to call the Centre for Mathematical Modelling of Infectious Diseases, produced a vast amount of learning and reports and advice for the United Kingdom as well as a host of other low and medium-income countries around the world.
Professor John Edmunds: Yes. It was an amazing effort.
Lead Inquiry: I’ll turn in a moment to asking you just to give us a flavour of the work that the CMMID did, but before I do, I want to ask you to put your mind back and give the Inquiry, please, a sense of what your understanding was in the middle of January 2020 as to the threat that was by then plainly emerging from China.
You say in your statement it was clear by early to mid-January 2020 that the novel coronavirus outbreak in China was a major public health threat. Did you mean – do you mean – by that that it was a major public health threat to the world, to the countries around China, just to China, or to the United Kingdom?
Professor John Edmunds: At that very time, at the middle of January, it wasn’t clear whether that was a threat just to China or whether it was a threat to everyone. I think all of us thought it might well be a threat to everyone across the world, but it wasn’t clear at that time, because of – it’s a technical issue, but there was – the way that the data were being reported from China, it looked at the time – there was only 41 cases that had been reported, they’d all been – they’d all attended the seafood wet market in Wuhan, and no other cases were being reported. So it could have been just some odd event, quite a large event, where people got exposed to something in that market. But it might not have been. And when we started to see cases outside China, then it was – it was very hard to believe that it was just a limited event.
Lead Inquiry: Whilst you give your evidence, Professor, could I invite you just to go a little bit slower as well.
Professor John Edmunds: Sorry. Yeah.
Lead Inquiry: Just to get our chronological bearings, the knowledge that there were cases outside China, of course, emerged at the end of January –
Professor John Edmunds: No, before then, the first case outside China I think was about the 13th, it may have been 11th or 13th January.
Lead Inquiry: But by the end of January, it was clear that it wasn’t just one or two cases sporadically in a country outside China, there were multiple cases in multiple countries?
Professor John Edmunds: There were. And by then the Chinese had changed the way that they were reporting their cases, and there were thousands of cases in China.
Lead Inquiry: We’ll come to this issue later of how it was that the early data grossly underestimated the spread of the outbreak in China.
But you’ve used the words major public “health threat”.
Professor John Edmunds: Yeah.
Lead Inquiry: It was clear by mid-January that what China was grappling with was a viral outbreak, a viral pathogen, a disease outbreak based upon a virus?
Professor John Edmunds: Yes, absolutely.
Lead Inquiry: And viruses have a tendency, it’s what they do, to spread exponentially –
Professor John Edmunds: Not all of them.
Lead Inquiry: Not all, but they may do so.
It was clear in mid-January, although nobody knew the extent of the spread in China, that this virus had the capacity to kill, to seriously harm, to hospitalise, and that people weren’t becoming infected just because they’d had contact zoonotically with an animal –
Professor John Edmunds: Correct.
Lead Inquiry: – they were becoming infected from human-to-human transmission?
Professor John Edmunds: That was then very clear by – certainly by the end of, you know, the third, fourth week of January, that was very clear, yes.
Lead Inquiry: So if human-to-human transmission was clear, and it was clear that it was spreading, although nobody knew to what extent, was that why you, as you say, appreciated there was a major public health threat?
Professor John Edmunds: Yes.
Lead Inquiry: Because if the virus continued to spread, and its reproduction number was more than 1, that is to say every single infected person would infect more than one other person in an unimmunised population, subject to control measures being applied, the virus would continue to spread forever, until herd immunity?
Professor John Edmunds: Yes. Even after herd immunity of course you get spread, like we have now.
Lead Inquiry: So the basic nature of the threat was clear: it was an issue, wasn’t it, of seeing whether it would spread significantly beyond China and the countries around China, and therefore, by extension, whether there was a need to apply control measures to stop it?
Professor John Edmunds: Yes, I would agree with that.
Lead Inquiry: All right.
If a virus spreads at a rate greater than R larger than 1, then it will spread, we’ve heard, exponentially, it will grow faster and faster and faster?
Professor John Edmunds: If you don’t take measures to stop that, yeah.
Lead Inquiry: If you don’t take measures. So is that why, in your field of expertise, there is this notion that when dealing with viral epidemics which may become a viral pandemic – which is just a difference of scale, is it not, a pandemic is a worldwide epidemic – a sensible and wise approach is to apply a precautionary approach, that is to say get on top of the problem before it beats you?
Professor John Edmunds: Correct.
Lead Inquiry: And in your statement, you refer on multiple occasions to the need for the precautionary principles to be applied; it is at the very heart of epidemiology, is it not, it’s how you deal with epidemics?
Professor John Edmunds: Yes, when you’re talking about response epidemiology, how to respond, then yes, you do – it is wise to apply that precautionary principle, because we – our surveillance systems are never likely to pick up every case, and they’re always a bit delayed, and so the epidemic is likely to be more widely spread than you think it is.
Lead Inquiry: Was that why you say in your statement that even in the early days or mid-days of January, it was essential for the United Kingdom, as with every other country, to assemble significant data in terms of the epidemiological nature of the virus that had by then already spread outside China, and the modelling data, in order to be able to work out precisely how the virus would spread and how to deal with it?
Professor John Edmunds: That’s right. So first of all you try to characterise what you’re dealing with, in terms of – you mentioned the reproduction number, so what – if you could try to estimate the reproduction number. And then other critical parameters related to the virus, for instance obviously how – the infection fatality rate or case fatality rate, which is the fraction of those – of the infections that might die, for instance. These are sort of absolutely critical numbers that you try to get an early estimate of, as best you can.
Of course you don’t stop there, throughout the epidemic you might refine those estimates and they might change a bit, but you spend a lot of your time trying to characterise – especially with a new disease like this, trying to understand it, how fast it might spread, and then you can start to put together models to play – you know, to look at different scenarios, as it were, to see whether – to see how you could, you know, what measures might be effective or most effective against this new threat.
Lead Inquiry: In relation to the coronavirus pandemic, that basic data, the reproduction rate, whether the virus killed, whether it hospitalised people, whether it was capable of being transmitted and was being transmitted human to human, and whether or not it was possible to become infected but not show symptoms, asymptomatic infection –
Professor John Edmunds: Yes.
Lead Inquiry: – whether or not it was possible to become infected and have a period of time during which you showed no symptoms, pre-symptomatic; all that in general outline was known fairly early on, was it not?
Professor John Edmunds: It was. Certainly by early February, or mid-February, I’d have thought, then we had probably reasonable estimates of most of these things. Some of them – some of these things take longer to estimate. For instance, the infection fatality rate takes longer, because sadly it takes time for people to die if they’re infected, and so you have to sort of wait for that. I know it’s a dreadful thing to talk about, but you have to wait for that to happen, so you don’t know how many people might die until people are dying.
Lead Inquiry: Could you keep your voice up a bit more, please, Professor.
Professor John Edmunds: Sorry, yes.
Lead Inquiry: So the infection fatality rate is vital, is it not –
Professor John Edmunds: Yeah.
Lead Inquiry: – in terms of assessing what might happen to any particular country’s healthcare system? You need to know what proportion of those infected in your population will die in order to know whether you’ve got enough beds, whether you’ve got enough healthcare facilities?
Professor John Edmunds: There’s two aspects. So one is the reproduction number, the basic reproduction number, and that gives you an indication of how many people might become infected – if you do nothing. So if you allow the epidemic just to sweep over the population – and the population does nothing. So they don’t change their behaviour. And that gives you – so that tells you how many people might become infected. And then, of course, you would need to know, of those who become infected, how many might die, how many might be hospitalised.
And it’s not just those crude numbers, you’d like to know it by different groups, like different age groups, which, for Covid, that was – there was enormous differences in risk by age, for instance.
Lead Inquiry: But the reproduction rate was estimated to be between 2 and 3 at a relatively early stage, in fact in late January. The infection fatality rate, in a very broad sense, how many people will die in an unimmunised population that takes no steps to protect itself, was assessed in mid-February preliminarily –
Professor John Edmunds: Yeah.
Lead Inquiry: – to be 1% overall. It subsequently transpired that if you were over 70 – or for the over 70-year olds the infection fatality rate, the proportion of over 70-year olds who would die once they become infected was much higher?
Professor John Edmunds: Yeah.
Lead Inquiry: Around 7% of them?
Professor John Edmunds: Correct.
Lead Inquiry: But the point, Professor, is this: plainly epidemiologists and modellers, to use your words, like to know the precise nature of the virus –
Professor John Edmunds: Yeah.
Lead Inquiry: – the detail of how it will behave, how it transmits, what the particular features are in terms of the impact on segments of the population, how the population might behave, how the virus might respond to self-imposed behavioural changes.
And the models, to use your word, because you used it, can be used to play at the figures, to demonstrate these more nuanced conclusions.
But the basic information about the threat of this virus and its potential fatal impact and the impact upon the healthcare systems of this country were known, was known, relatively early on?
Professor John Edmunds: Correct, yes.
Lead Inquiry: It was known, putting together the reproduction number, the infection fatality rate, the knowledge of the size of the population in this country, the knowledge of –
Professor John Edmunds: Demography, yes.
Lead Inquiry: – how big the NHS is –
Professor John Edmunds: Yeah.
Lead Inquiry: – that was all apparent to those in the know, to the experts, certainly by the end of February?
Professor John Edmunds: Oh, yeah. I mean, earlier than that, really.
Lead Inquiry: When earlier than that, do you assess?
Professor John Edmunds: Sort of mid-February, I think, where we had probably a pretty good – pretty good idea. You get an initial sketch even earlier than that, perhaps, but then – which might give you, you know, an initial impression, but of course then you improve on that and then you understand some of the nuances, like the – how risk varies with age and how risk varies perhaps with other – with other sorts of variables, ethnicity – obviously those sorts of things came later.
Lead Inquiry: So would it be fair to say that when that realisation dawned, perhaps in mid-February, the absolute core consideration then became: how do we control it? How do we stop it? How do we suppress it? How do we mitigate it? How do we do anything –
Professor John Edmunds: I think that had been a core consideration from before then, certainly from January when the alarm first came up: how do we stop this?
Lead Inquiry: And unsurprisingly, experts, government officials, scientists, epidemiologists, cast their minds back to what sort of control measures we had utilised in the past?
Professor John Edmunds: Yeah.
Lead Inquiry: And of course because of the flu pandemic of 2018, because of swine flu, because of –
Professor John Edmunds: I think 1918.
Lead Inquiry: Sorry, what did I say?
Professor John Edmunds: 2018.
Lead Inquiry: Thank you very much, Professor.
Professor John Edmunds: Sorry. I didn’t mean to put you off.
Lead Inquiry: No, no, no, it’s quite all right. 1918.
Because of swine flu, because of SARS and MERS and other – those two particular coronavirus –
Professor John Edmunds: Yeah.
Lead Inquiry: – epidemics, or pandemics perhaps, in the Middle East and Far East, there was a basic understanding of what sort of control measures might work?
Professor John Edmunds: Some, yeah. Almost as well a bit the other way around, what kind of control measures are unlikely to work as well. You know, there’s two aspects to that.
Lead Inquiry: Thank you.
For flu, there had been quite a prolonged debate about whether school closures, for example –
Professor John Edmunds: Yeah.
Lead Inquiry: – would work, and strategically the government and its advisers thrashed this issue around for a very long time indeed: is it a good idea to close schools in the face of a flu pandemic?
There had been a long running debate, again resolved in the context of flu, whether or not shutting borders would help?
Professor John Edmunds: Yeah.
Lead Inquiry: And it was generally understood that it wouldn’t?
Professor John Edmunds: There’s a difference between absolutely shutting your border, letting no one in –
Lead Inquiry: And restrictions?
Professor John Edmunds: And restrictions, yeah.
Lead Inquiry: But generally –
Professor John Edmunds: Restrictions were unlikely to buy much time.
Lead Inquiry: But we had never – at least –
Professor John Edmunds: We had never shut our border.
Lead Inquiry: We had never shut our borders to deal with flu. And we had never had a sophisticated or put into place a sophisticated system for test, trace, contact, to deal –
Professor John Edmunds: We had at the beginning of the swine flu pandemic, but mostly to understand its transmission characteristics here in the UK rather than as a concerted effort to try to actually stop it, because there was, you know, widespread recognition that it would be extremely difficult and extremely resource-intensive to actually try to stop a flu pandemic via contact tracing, because it – it moves so fast that the virus moves between one generation of cases and the next so quickly that it’s really impossible to keep up with it with contact tracing.
Lead Inquiry: And the contact tracing that was used for swine flu, and is used actually for any new or emerging –
Professor John Edmunds: Oh, and things that have been around forever. You do it for TB and – well, HIV’s not been around forever, but yes, you do it.
Lead Inquiry: It’s relatively limited. You pick up travellers, you test them, you test and trace, contact, trace index cases, and whether or not you’re focusing on people coming in with the infection or you focus on the first few hundred cases or you focus on the first few cases in the hospitals, it doesn’t really matter, the system was only designed to deal with the first relatively few cases?
Professor John Edmunds: Yes. So for flu the system was always a first few hundred system and the idea, as I said, is really to understand and characterise the virus here in the UK more than trying to stop it with the recognition that it was very, very unlikely to stop a flu pandemic.
Lead Inquiry: So drawing those threads together, and I should say, can you tell us whether or not there was in January 2020 any system at all, whether by utilisation of past control measures or anything drawn up on paper, any system of quarantining whole segments of society or whole-society, of self-isolation of the whole society or social distancing the whole society?
Professor John Edmunds: In January/February, no, there was no consideration of that. It was concentrating on contact tracing.
Lead Inquiry: And you knew that?
Professor John Edmunds: Knew?
Lead Inquiry: You knew that there was in place no system at all for social distancing –
Professor John Edmunds: Yeah.
Lead Inquiry: – quarantining –
Professor John Edmunds: Yeah.
Lead Inquiry: – for whole-society response?
Professor John Edmunds: Yeah. I mean, of course at that time, if we’re talking about, say, February, there would have been very few cases – even, you know, looking back at it now, and realising how many cases there were, there were still very few, so you’ve got to sort of have some sort of proportionate response. You know, do you put the entire country under some sort of restrictions when there’s, you know, perhaps a handful of cases? So the idea is to really try to target it around those cases. I think the issue was we always knew that it was likely that cases would not – some cases would not be picked up. We were targeting our contact tracing around cases who came in from high risk areas, China being the most obvious, but other places where there was – cases had been picked up, which were mostly in the Far East. But of course people could come indirectly into the UK via other routes, and of course they did, and so that contact tracing effort, it had – you know, it had to go really well everywhere in the world for it to be – for it to stop –
Lead Inquiry: For it to work?
Professor John Edmunds: Yeah, exactly.
Lead Inquiry: And you knew that?
Professor John Edmunds: Yeah.
Lead Inquiry: So you – and I make it absolutely plain, you are but one of a number of brilliant scientists and advisers who assisted the government and the country in the remarkable way that you did, but there must have been a general awareness, therefore, by February this viral, severe pandemic, this viral pathogenic outbreak is coming, and it can’t be stopped, and the measures which could stop it once it reaches the United Kingdom have either never been dreamt up or never been applied or won’t work?
Professor John Edmunds: I mean, you said can’t be stopped, I mean, it was worth trying to stop it in those ways. You know, there was a hope but maybe not an expectation that it would be stopped like that. But yes, we knew that there was a very high likelihood – I mean, you know, I’m a scientist, I’m not going to say there’s a – you know, there was an extremely high likelihood that we would – that we would face a very, very major pandemic, yes, we knew that.
Lead Inquiry: And when you say “we would face a very, very major pandemic”, you mean, so that we are clear?
Professor John Edmunds: Something like 1918. That was always – you know, that would have been – and of course that’s the great – it was the great influenza pandemic of more than 100 years ago. You know, it’s sort of etched in people’s – especially my field, of course, the sort of collective memory has been a horrendous event, and this looked, there was – it was, you know, every time a new bit of data came in they just sort of confirmed that this was going to be something like that, you know, a once in a hundred years event, horrific.
Lead Inquiry: And because there was no sophisticated test, trace, contact, isolate system in place, because such things weren’t generally used for flu, for which we’d been preparing, although this coronavirus had a latency period, a gap between when you become infected and when you can pass on the infection to somebody else, in which gap you can be tested and seen whether you are positive for the disease, until such a system could be developed, designed and put into place, it would be of little practical assistance?
Professor John Edmunds: So by late January, early – late January, let’s say early February, we knew something about the characteristics, you quite rightly say, so there was quite a long period between infection and you becoming ill of sort of five or six days, which is very different to flu, which is sort of one or two days, and so there was a possibility that gave you a bit more time, if you were trying to contact trace – I mean, if you’re trying to contact trace, it gave you a bit more time to be able to do it. In terms of are you infectious before you become symptomatic, with SARS-1 that didn’t look like that was the case. So with SARS-1 that time period was a bit longer, it was more like eight days, and it looked like you became infectious when you became symptomatic. And you were very ill with SARS-1 and so most people were in hospital very quickly. And so it was easier to contact trace with SARS-1 and that’s how it was stamped out globally. Flu you just wouldn’t be able to do it because of the speed. SARS-2, Covid, was somewhere in between. It gave you a glimpse of maybe that might be possible, but everything had to go really well for it to work.
Lead Inquiry: But in practice, whether epidemiologically a test system was possible, it didn’t matter, did it, because in January, February, March, beyond the first few hundred cases, before the first few index cases, there was no whole-society test, trace, contact system?
Professor John Edmunds: No. Strictly speaking you don’t need to test people, you can isolate them anyway, you know, on symptoms and things like that, so – obviously it’s much better to test them because then they know they have it or they know they don’t have it, but strictly speaking you don’t need to test people.
Lead Inquiry: So, to come back to your earlier answer, by mid-February there was an understanding that there was a major pandemic coming?
Professor John Edmunds: Yes.
Lead Inquiry: And so again so that we are clear, a major pandemic means tens of thousands of hospitalisation cases?
Professor John Edmunds: And more.
Lead Inquiry: And more. Hundreds of thousands perhaps. It means tens of thousands, perhaps more, of deaths?
Professor John Edmunds: Oh, yes, and again more.
Lead Inquiry: It means the country being overwhelmed by disease?
Professor John Edmunds: Yes. It’s more than that. You know, once – the reason why the flu pandemic was at the top of the National Risk Register, it was always known that an event like that would affect every aspect of society, every aspect of government. So it wasn’t just that it would overwhelm the health service and cause, you know, a huge amount of disease, but also it would affect people’s lives in other ways – and society quite fundamentally in other ways. That was always known for these major, major events.
Lead Inquiry: As you’ve said, by mid-February there was only the hope, not the expectation, that it might be stopped?
Professor John Edmunds: Yes.
Lead Inquiry: Why, then, as a country, did we not apply the precautionary principle to which you have already referred and do something about it then?
Professor John Edmunds: I think the risk then was still low to a person –
Lead Inquiry: Sorry, please speak more slowly. It’s very important that we record your answer.
Professor John Edmunds: I apologise.
So I think the risk for an individual in this country in February was very, very low – of Covid was very, very low. So could you take national restrictive measures, would people come along with that? You know, I think – I think that would be difficult. I think it would be a hard sell.
Lead Inquiry: But that, Professor, was surely a matter for our politicians and our decision-makers? That was for them to decide, was it not?
Professor John Edmunds: Yeah, it was, of course. I think there are other things in between. You’re going to – you’re kind of jumping to the nuclear option, I think there are other things in between that perhaps could have been done. I’ve thought about it later, I thought, you know: what could we have done? What would be more proportionate? I think things like advice to work at home we could have perhaps done that. Yes, it would have had an impact on the economy, but – and, you know, I regret that we didn’t look at that at that time.
And there are things – there are other things like we could have given – we gave public health advice, that was being given, to wash your hands and things like that, which are sensible, but we could have perhaps made it really clear that people should stay at home if they had any sort of symptoms. Despite the fact that almost all of them wouldn’t have had Covid. Almost all of them would have had flu or coughs and colds, whatever. You know, because Covid was vanishingly rare even at that time. So I think maybe there are things that we could have perhaps emphasised in February that might have slowed things a little bit. They weren’t going to stop it, but they might have slowed things a little bit more than they did.
Lead Inquiry: We’re going to come, of course, to the detail of the advice that you and SPI-M-O and SAGE gave to the government, but the nature of the response was, you accept, a matter for government.
What I’m asking you, though, is why was that terrible conclusion, that dawning realisation that the virus was coming, it was a fatal pathogenic disease, and there was in practice, you understood, not much more than a hope that it could be controlled, why was that warning, why was that realisation not made more apparent to government in the middle of February, to the public –
Professor John Edmunds: Yeah.
Lead Inquiry: – to the United Kingdom –
Professor John Edmunds: Yeah.
Lead Inquiry: – that this pathogenic tsunami was coming?
Professor John Edmunds: So I distinctly remember my feeling at the time. I assumed that the government did know all of this. I mean, you know, I can’t believe that they didn’t, quite honestly. I still can’t believe that they didn’t. So I assumed that they did know all of this, and that actions were being taken.
I – the messaging at the time was very reassuring, and I assumed that there was a plan: let’s not concern people and bother people now, because we’ll have to – we’ll have to get people prepared, and do it in the right way. That was my assumption at the time.
Afterwards, I look back on it and think: actually, really, you know, was there a plan? I’m not sure. But I’d assumed that there was. I assumed that the messaging being quite reassuring was there for a reason.
Lead Inquiry: I’m not asking you to speak for the government, and we’ll come later to how much the government responded to the advice you actually did give. I’m asking you and, through you, vicariously SAGE and SPI-M-O and SPI-B and all the august, brilliant advisory committees, the epidemiologists, the modellers, the virologists, why was that warning not being shouted out from all of you –
Professor John Edmunds: Yeah.
Lead Inquiry: – from mid-February?
Professor John Edmunds: Yeah. So I didn’t think we had to shout it. You know, in terms of the government, I – you know, something of this magnitude you’d have thought the government should have all its attention paid to it, you’d think. So there’s that.
Secondly, yeah, I kind of just assumed that there was some reason for not shouting it out. I remember quite distinctly – I remember Neil Ferguson gave a – did say something on Radio 4 and I remember Chris Whitty also saying something. There was this kind of funny period where people would talk about, as you’re talking about, the – you know, the reproduction number and the implications that would mean for how many people might get infected in an unmitigated wave, and there was talk about the infection fatality rate, and so, you know, you could easily just multiply those two numbers together and get a very big number for deaths. But people didn’t. I was … you know, people avoided multiplying, you know, in public utterances.
And I felt that – I honestly thought – I mean, it sounds really naive and silly, I think, but I honestly thought there was a plan. I didn’t want to be the person who multiplied those two numbers together and – I thought that should come from someone central in a kind of organised – in an organised comms plan way to prepare the country for what was going to happen. And I didn’t want to get – I didn’t want to mess that up in any way.
Lady Hallett: I appreciate you’re mid-flow, Mr Keith.
Mr Keith: May I ask one more question and there will be a very natural break?
Lady Hallett: Very well.
Mr Keith: Your statement makes plain, Professor, how much work was done by the CMMID working group at the London School of Hygiene and Tropical Medicine, you describe it as brilliantly led and organised by your colleagues, in particular a doctor Rosalind Eggo.
You describe how over those three months, January, February and March, you undertook – or rather the London School of Hygiene and Tropical Medicine undertook a huge range of work, right from the early days –
Professor John Edmunds: Yes.
Lead Inquiry: – assessing the nature of the initial outbreak, accumulating data, analysing the spread of the virus, looking at the reporting delays from China, how difficult it was to get a handle on the nature of the spread. You looked at airport screening, methods of transmission, rates of testing, contact tracing, isolation, the case fatality ratio, then latterly, in March, the effect of non-pharmaceutical interventions which, let’s just speak it out, how to control the virus, whether it would be a wave or a second wave, what was herd immunity, whether we should be suppressing or mitigating, whether we should have an episodic lockdown process.
But this vast learning nowhere says, at least until March, there is a pathogenic tsunami coming and it can’t be stopped.
Professor John Edmunds: You know, I think that was clear to all of us. Yes, it wasn’t me who raised that alarm to the public. I deliberately didn’t. As I’ve explained to you, I didn’t want to. I didn’t think – I didn’t think it was for me to do that, I thought it was for someone with more authority to do that, and to prepare people for what was likely to come.
Mr Keith: Thank you.
My Lady.
Lady Hallett: Thank you very much.
I hope you were warned, Professor, that we take regular breaks, so I shall return at 1.40.
(11.25 am)
(A short break)
(11.40 am)
Lady Hallett: Mr Keith.
Mr Keith: Continuing, Professor, with the theme of the generic understanding in the scientific community, the scientific advisory community in January, it is absolutely vital, I make plain and put to you, that you of course, Professor Edmunds, had absolutely no personal responsibility for having to stand up and tell the government what it should be doing, what was going to happen, because you were part of SAGE, SPI-M-O, all the many bodies, and it was those bodies which had been constituted in order to give government advice; that’s a fair summary, is it not?
Professor John Edmunds: Yes, but it doesn’t stop me feeling that I had some responsibility.
Lead Inquiry: Well, if I may say so, that is very much to your credit.
And the way in which the structure worked was that these many august and brilliant bodies were constituted to assemble information, assemble data, give advice, and then that advice – and it was very clear how it could be done and should be done – was routed to government through the CMO, the Chief Medical Officer, the Government Chief Scientific Adviser –
Professor John Edmunds: Yeah.
Lead Inquiry: – through the minutes, through the papers which were given to the committees, through the documents that you produced –
Professor John Edmunds: And can I say I’m absolutely sure that the CMO and the Government Chief Scientific Adviser both raised this. There is no way that they didn’t.
Lead Inquiry: Yes. And we’ll come to it, in a moment, your own emails, personal emails to Professor Ferguson, Professor Sir Chris Whitty, Sir Patrick Vallance, raise the issue of urgency and the need to act. We’ll come to those in a moment.
But the point, we’ll also look at SAGE, though is this, isn’t it, that systemically or systematically, there was a structure in place to give the government advice, to warn it, to tell it what might happen and to give it the information to enable it to decide to respond rapidly, proportionately, effectively, but that system doesn’t appear to have worked?
Professor John Edmunds: Clearly not. I mean, if you think about it, though, SAGE is – only sits in an emergency, and it was called to sit in – somewhere around the 20th, you’ll know the date exactly, but, you know, the 20-something of January. So somewhere someone in government thought that it was sufficient – you know, it was sufficiently – there was a sufficient emergency to call SAGE. SAGE doesn’t – only sits very seldom in these kind of situations. So someone thought that it was worthy of calling SAGE together.
Lead Inquiry: Before we leave the subject entirely of the working group at the London School of Hygiene and Tropical Medicine, and the issue of the vast amounts of work that were done, can I ask you to look at one particular paper dated from 7 February 2020, which is INQ000092645.
Professor John Edmunds: You can carry on, I know which paper it is, yeah. Yeah.
Lead Inquiry: Yes, we need to get it up on the screen, Professor, for everybody else.
So this is a paper dated 7 February. It’s called “Feasibility of controlling 2019-nCov outbreaks by isolation of cases and contacts”.
So at a relatively early stage, 7 February, the London School of Hygiene – and this isn’t a SAGE paper, it’s a paper done by your research institute’s working group, was on to the issue of how easy or difficult or effective controlling the virus by isolation of contacts and cases would be.
Professor John Edmunds: Yeah.
Lead Inquiry: Hence your evidence earlier about the very early understanding of how difficult it would be to control the virus by isolation and contact trace.
The summary of the findings in the bottom half of the page are these, or the summary is this:
“The percentage of contacts traced is critical to achieving control in all scenarios.
“Higher transmission (higher R0) makes outbreaks more difficult to control.”
By this time you did have some basic understanding of the likely –
Professor John Edmunds: Yeah.
Lead Inquiry: – reproduction number. What was it?
Professor John Edmunds: You know, it – there was still – estimates varied between about 2.5 and 3.5 at the time.
Lead Inquiry: So not as high as some other or some high-consequence infectious diseases, but –
Professor John Edmunds: Higher than most high-consequence infectious diseases. That – 2.5 or 3 doesn’t sound bad, but it’s bad.
Lead Inquiry: Yes. Not as high as some, but higher than many.
“30% transmission before symptoms makes control less likely in all scenarios.”
By that were you saying, was your working group saying: if you’ve got a high number of people who are asymptomatic, who –
Professor John Edmunds: Pre-symptomatic, that’s about pre-symptomatic –
Lead Inquiry: Okay.
Professor John Edmunds: So if you are infectious before you become symptomatic and we had different scenarios for that, so different assumptions – because we didn’t know that very well at that time, although that was becoming clearer –
Lead Inquiry: My mistake, the asymptomatic bullet point –
Professor John Edmunds: Is a bit lower down, yeah.
Lead Inquiry: Let’s have a look at that.
“Presence of subclinical (asymptomatic) cases has an outsize and negative impact on probability to achieve control.”
By that were you saying if a large proportion of infected people are asymptomatic, that is to say they don’t show symptoms, then your ability to achieve control is hindered and the probability that you will be able to achieve control goes down?
Professor John Edmunds: Correct.
Lead Inquiry: You also say:
“60-80% of contacts must be traced (and transmission stopped) in order to achieve control in most scenarios, and more for some characteristics.”
So you’ve got to, practically, be able to stop a very large number, a very large percentage of contacts for transmission chains to be broken?
Professor John Edmunds: Correct, so you have to – you have to quickly isolate – contact trace a large fraction of the contacts, and effectively quarantine them.
Lead Inquiry: Was it these findings in early February which led you to conclude that, as you began to appreciate, the asymptomatic, pre-symptomatic nature of the viral epidemic and the transmission rates, that effectively contact trace control was going to be extremely difficult?
Professor John Edmunds: I think it’s a little bit more nuanced than that. This paper was a little bit of a – one of those – the results here are a little bit of one of those – is the glass half full or is a glass half empty? It said it was possible to do it, potentially, to – but things had to go very well for that. Yeah, that’s really a summary.
Lead Inquiry: All right.
I want to ask you now about SAGE and functionally how SAGE operated vis-à-vis the government. You had attended earlier forms, emanations of SAGE, because I think you’d been on SAGE during the Ebola crisis?
Professor John Edmunds: Correct.
Lead Inquiry: So you were very familiar with the workings of SAGE?
Professor John Edmunds: Familiar. I wouldn’t say “very familiar”, yeah.
Lead Inquiry: When the virus began to emerge from China, SPI-M – of which we’ve heard a great deal – alongside SAGE being brought together was also put into place, was brought together, and changed its focus to looking specifically at Covid-19?
Professor John Edmunds: Yeah.
Lead Inquiry: NERVTAG, we’ve heard, continued to operate, it was a standing statutory committee to the DHSC, it deals with new and emerging viral threats, but it also looked at Covid-19, of course.
When you were on SAGE, were you attending as a representative of the London School of Hygiene and Tropical Medicine, or do you and all your colleagues attend in a personal capacity?
Professor John Edmunds: I was just there in a personal capacity.
Lead Inquiry: It’s self-evident, there were a very great number of experts on SAGE. You describe the level of advice and the level of understanding on the part of the attendees at SAGE, as being very high. SAGE was very, very well informed, was it not?
Professor John Edmunds: Absolutely.
Lead Inquiry: All of you were experts in your own fields, but you were obviously capable of opining on related subjects, and the evidence is that a great deal of information was culled by members of SAGE from their contacts and their professional colleagues abroad?
Professor John Edmunds: Correct.
Lead Inquiry: So in summary, do you agree that SAGE, in terms of its ability to locate, consider and report on data and on information and on this field of expertise, was very high indeed?
Professor John Edmunds: Yes, absolutely.
Lead Inquiry: The papers produced by SAGE, in particular the minutes, weren’t really minutes, though, were they, they were more of a consensus document bringing together a final concluded position?
Do you think that worked? Do you think having a consensus document was a good thing, because it gave the government a clear understanding of a final position, or perhaps was undermined by or flawed by the tendency of such an approach to conceal nuance, to conceal the width of debate?
Professor John Edmunds: I think that – you know, I think you could probably have done both, have a consensus statement and then have maybe fuller minutes or something, so if you were interested you could see the – how the debate went. But as it was, it was just this very terse, short document with a consensus.
Lead Inquiry: Was the information flow with government one-way or two-ways?
Professor John Edmunds: No, it was one-way. It came from us, through Patrick and Chris – sorry, Patrick –
Lead Inquiry: Sir Chris Whitty and Sir Patrick Vallance.
Professor John Edmunds: Yeah, Sir Chris Whitty and Sir Patrick Vallance to – to central government. We didn’t have any – we didn’t play any role in that.
Lead Inquiry: So that there is absolutely no question about it whatsoever, there is nothing to suggest that they conveyed the information from SAGE to the government other than properly, faithfully, and –
Professor John Edmunds: Oh, I’m absolutely sure they would have done. And it didn’t come back. I mean, they’re consummate professionals, of course, and so they – we didn’t know what the government was discussing – you know, they didn’t report on that, of course they didn’t. So it went one way. That’s how it was.
Lead Inquiry: Did you understand on SAGE that they were conveying the consensus position which SAGE had reached or that they were conveying the whole range of debate, the issues which had been explored, and perhaps the divergence of views which had been apparent in argument?
Professor John Edmunds: I don’t know, of course, because I wasn’t there. But we did used to try to include a statement about certainty or uncertainty in everything – I say everything; I would hope just about everything – so when there was a statement made then it was – there would be a very broad indication of how certain that statement was.
Lead Inquiry: You, or rather SAGE, is a scientific advisory committee. Did you see the role of SAGE as properly extending to giving the government policy advice or making specific recommendations as to what it should do?
Professor John Edmunds: I didn’t. I viewed the process in sort of three steps. I thought that there was the sort of evidence synthesis step, which was SAGE – and obviously there could have been evidence syntheses in other aspects, economic aspects, social aspects, that we weren’t covering, but I felt that we were involved in evidence synthesis, trying to summarise the evidence, and then that went forward to central government somehow, to the policymakers, who I – in my view are the senior civil servants who weigh up those – put that aspect of the evidence together along with the other, because of course any policy would have huge implications for society, you know, beyond the epidemiology or the health implications and so –
Lead Inquiry: Could you just slow down a little bit, Professor.
Professor John Edmunds: Apologies.
Lead Inquiry: You’re running away from us.
Professor John Edmunds: So I felt that then that second step was being done by the policymakers, the senior – the civil service. And then the final step, you know, they would come up with – this is my mental model, I don’t know whether it’s accurate, but – and then the final sign-off on which of the preferred options would of course be made by our elected representatives.
Lead Inquiry: Was it the role, do you think, of individual members of SAGE to publicly advocate for particular measures to be taken or for policy, to go to the press and say, “I think this should be done, why isn’t the government doing that?” or “We, SAGE, aren’t doing enough”?
Professor John Edmunds: I think it was difficult. So my – I think the answer to that is – should we have that sort of thing – probably no, because that didn’t necessarily help the government make its – I thought that – and we were – you know, Chris and Patrick both made this clear to us, that it didn’t necessarily help the government consider the evidence in a cool and calm way, if they were getting pressure from senior – from senior advisers, I have to say, so I tried to stick to that in the early part of the epidemic. Later in the epidemic, at times I struggled with trying to stick to that, and I don’t think I always did. I – I did – yeah.
Lead Inquiry: Professor, it’s fair to say that you gave a number of interviews to the press, you spoke to Reuters in April, on 8 April, I think, The Sunday Times in May, the Andrew Marr programme in May, you went on the Robert Peston programme I think at a later stage, perhaps Andrew Marr as well?
Professor John Edmunds: Yeah.
Lead Inquiry: Was the tendency of some members of SAGE to speak to the press and to talk about the guts of what had to be done and what was being done or not being done, do you think that helped this process of giving scientific advice to the government?
Professor John Edmunds: So possibly not. I tried not to give – to make statements about what the government should or shouldn’t do, in any of those interviews. Sometimes it’s – they’re very eloquent, they’re very clever at their art and they get things out of you that perhaps you didn’t want to divulge. So I tried not to.
What I tried to do, because I did think it was – well, I always thought that it’s important, that we should explain to the public – you know, science generally I think – you know, outside of a pandemic I think we should explain our work to the public, who are ultimately funding it in most instances. And in this particular case, of course, they were being directly affected by the measures that were being put in place or not being put in place, and I felt that it was – there was a responsibility on us to try to explain the science. And also I tried to explain – I mean, if you saw my interviews on wherever, I tried to explain that this was not easy, that there was never an easy solution to any of this, and this was difficult, and the government were having to make really difficult decisions, having to trade off different aspects of, you know, health and wealth and whatever. I tried to explain that this was a very, very difficult thing. Because it was. They were dreadful decisions that they were having to make.
Lead Inquiry: Indeed.
More on SAGE. The Inquiry has heard evidence from a number of attendees on SAGE that because the government never told SAGE what its strategies were, what its overall objectives might be or, in essence, what it wanted to achieve, when providing advice SAGE was to some extent shooting in the dark, would you agree?
Professor John Edmunds: Yes, I think – I think I said in my statement it’s very – it’s very difficult to plot a course when you don’t know what the destination is.
Lead Inquiry: In terms of the membership of SAGE, the membership of SAGE grew enormously, not least because it was able to go online and did go online –
Professor John Edmunds: Yes.
Lead Inquiry: – at the onset of the pandemic. It was obviously a scientific committee, and it had a number of august biomedics, epidemiologists, modellers, public health experts. It was attended also, wasn’t it, by representatives of NHS England, Public Health England, and of course the CMO and the Government Chief Scientific Adviser, who are well renowned experts in their own right?
Professor John Edmunds: Yeah.
Lead Inquiry: Would it have benefited from a greater input from frontline organisations?
Professor John Edmunds: I thought – I personally felt that that would have helped at times. I thought that there were times, particularly at the beginning, when our data were terrible, that our situational awareness of what was really happening wasn’t as good as it could have been. And so I would have – I would have preferred to have – yes, I thought – I would have liked to hear a little bit more from the frontline.
In fact, with NERVTAG, I knew that PHE, for instance, used – had started to do somewhat they called a sitrep, and this was a large number of slides, you know, there was – it was huge, it was like 50, 60 slides, that they were putting together every week which gave a summary, of – well, a situation report. I sort of – I asked on NERVTAG whether we could see that at the start of NERVTAG meetings, so that we could get a little bit better, a bit more holistic understanding of what was really happening. And that did happen, so that was accepted, and PHE used to start NERVTAG with a brief sitrep.
Lead Inquiry: What did SAGE make of the government’s mantra that it was, at crucial times, “following the science”?
Professor John Edmunds: Well, you know, the government couldn’t and shouldn’t ever have just followed the science. That was only one aspect of the – it’s only one aspect of the epidemic. And so they had to weigh advice or – you know, on various aspects, whether it was economic or social or, of course, operational, as well as the scientific aspect.
So I thought that that was always, I could see why they were doing it, they were doing it so they could hide behind us, I think, so when difficult decisions had to be made, they could hide behind us.
Lead Inquiry: Is science ever certain?
Professor John Edmunds: No.
Lead Inquiry: Can it ever be?
Professor John Edmunds: No.
Lead Inquiry: Is there ever one piece of science which can be followed?
Professor John Edmunds: No. That’s the – so that was – exactly – so that’s why we tried to represent the level of uncertainty in the statements we were making at these sorts of meetings. Because, of course, especially at the beginning of a pandemic, of a completely novel disease, I mean, uncertainty is huge.
Lead Inquiry: Why did SAGE, or perhaps you, feel the government was trying to hide behind you?
Professor John Edmunds: It’s what they do. It’s convenient, isn’t it?
Lead Inquiry: Was SAGE enormously assisted by, well, a great deal many other unsung heroes? I think a secretariat, you received enormous assistance from something called the Department of Health and Social Care Health Protection Analytical Team?
Professor John Edmunds: Yeah, they were amazing. The secretariat for – it’s hard to describe the – how much work was being done. And to bring that together, you know, and to make sense in – say if we think of the SPI-M work, enormous amount of work that was being done every week, technical, difficult, not something that lay people would necessarily be able to get a grasp of, and the secretariat, importantly, with SPI-M, included modellers. There’s a Health Protection Analytical Team within – it’s a small team, but within the Department of Health and Social Care. And they formed part of the secretariat for SPI-M, and – so then the discussions that we were having, they were following them, they were understanding them, so they could – because these discussions were technical, far ranging, difficult. And to summarise that in these consensus statements that they did was an amazing piece of work. And similar work was being done by civil servants, GO-Science and others.
The secretariat support was spectacular.
Lead Inquiry: To be clear, SAGE and SPI-M and NERVTAG weren’t just responding to particular commissioned requests from government, every week or perhaps every meeting these committees would have presented before them, because they had been prepared since the last meeting, round-up of information, updated projections, rolling charts, voluminous papers on what the position was –
Professor John Edmunds: Correct.
Lead Inquiry: – that you could consider as part of your – then your analysis?
Professor John Edmunds: Yeah, correct. So it’s probably worth – I don’t know whether you want to get into the details, but there was different ways of working on the different committees. SPI-M – or SPI-M-O more correctly at the time was a little different from the others, in that it had some routine tasks it did every week, which was short-term projections, medium-term projections, estimation of the reproduction number, and so on, and they were done by many groups contributing to that every week. So there was a kind of routine piece of work. There were the commissions that came to us from central government, asking us to do some work on a particular aspect. And they came most weeks, from recollection.
Then on top of that there was work that we did off our own bat, because we felt that it was important. Like, for instance, the work that you just highlighted earlier, nobody asked us to do that, we got on with that in January and then brought it to SPI-M, you know, at the appropriate time.
Lead Inquiry: Now can I turn, please, to modelling, which is, of course, your speciality.
Shortly, can you explain the difference, please, between scenario modelling and forecasts?
Professor John Edmunds: So forecasts are what we think will happen, and scenarios are what might happen under certain circumstances, and they’re usually run, those scenarios, over a longer period of time, so you could see the impact of those different circumstances.
So if I could give an analogy –
Lead Inquiry: Please.
Professor John Edmunds: – from the … so we have a weather forecast, and that tells us – that tells us – it gives a probabilistic statement about what the weather might be tomorrow or the day after or whatever. So it might say there’s a 80% chance of rain tomorrow.
There’s nothing you can do about that. It’s going to rain probably, there’s an 80% chance, or not. The only thing you can do is take an umbrella or a mac or something. Yeah?
So a scenario is something quite different and it runs over a much longer period. So the scenario models for looking at climate change, for instance, so looking at what might happen over the long term, over, you know, 10, 20, 30 years if we do something: if we take certain action to, say, reduce our CO2 emissions, for instance, this might happen to the climate.
Now, those are obviously very certain, they’re run over a very long time period, but you have the decision-makers, and in this case it’s sort of the – all of us, I guess, have some ability to change the future. So on the basis of these scenarios you could say, well, really we ought to be doing this to, say, reduce our carbon dioxide output, for instance, which then might change the future, we might have less of an increase in global temperatures.
It’s the same sort of thing for epidemiological forecasts, which are very short term and just say things like how many beds might there be required next week or perhaps the week after. They’re very short term, just like the weather forecast is very short term. Versus these longer-term scenarios: okay, if we put this policy in place, what might happen? If we put that policy in place, what might happen?
Now, they’re, of course, played out over a much longer period. They’re much more – because they’re going over a much longer period they’re not going to be right. The actual – the actual – “The epidemic will really be exactly like this in two or three months” is – the chances of that are very low, of course.
Lead Inquiry: Right.
Professor John Edmunds: They sketch out possibilities, just like the climate change modelling –
Lead Inquiry: Versus –
Professor John Edmunds: – sketching out possibilities.
Lead Inquiry: All right.
In the context of Covid, the forecasts therefore focused, did they not, on fairly – it’s no less important, but fairly basic information like how many people will die if you do nothing, how many beds will need to be occupied, how many hospital cases are there likely to be, and so on. Those are examples, fairly basic –
Professor John Edmunds: Yeah, and they were very short term, so it’s sort of looking ahead just one or two weeks.
Lead Inquiry: In order to be able to forecast in that way, as you’ve explained, a modeller needs to have an understanding of the reproduction number, the infection fatality rate, the hospitalisation rate, that sort of basic data?
Professor John Edmunds: Strictly speaking – yes, you certainly need the data, of course you need the data. But strictly speaking, you don’t necessarily need to know the reproduction number to forecast how many hospital beds you might need the next week. You need to look at the trends and you could just – so there are simple ways you could do it just looking at trends and projecting forward.
Lead Inquiry: All right.
Professor John Edmunds: And what happened was that there were a large range of different methods that were used by the different groups around the country, and brought together in a – and then combined in a statistical way to come up with a – what’s called an ensemble forecast.
Lead Inquiry: Even a forecast of a fairly basic type, perhaps based on fairly basic information like taking a percentage of how many people in the population might die or how many might be hospitalised, requires the modeller to have a good understanding of the underlying data. So if there is a delay in people being tested, or there’s a delay in getting the results of those tests to the modeller, or if there is an unwillingness on people who are infected to be tested at all, or if there aren’t any sophisticated surveys or blood tests which have been carried out in order to see how many people are infected if they’re not prepared to be tested, a lack of data of that type makes the modellers’ life very difficult indeed?
Professor John Edmunds: Of course. In fact, actually one of the things that we are – one of the roles in the – is to understand those delays. And so it’s not just – it’s not just a matter of forecasting into the future, but there’s this dreadful term “nowcasting”, which is how many cases there actually are now, because that’s not – because the reported cases won’t be reflecting the actual infections occurring on that day, they’re reflecting something that happened perhaps weeks earlier. So we can take – with understanding of these delays, then we can actually get a better idea about what’s actually happening now. It’s a dreadful term, but it’s quite explanatory, “nowcasting”.
So that was one of the roles that we were of course doing.
Lead Inquiry: So for SAGE and the modelling experts on it, there was a very real problem in February and early March, was there not, because you couldn’t be sufficiently precise in even these basic forecasts until you had the right data and you were receiving the data in good time?
Professor John Edmunds: We weren’t doing forecasts in February, there wasn’t really sufficient data to do it. We started doing it in March.
Lead Inquiry: Right. In terms of the scenario modelling, that is to say “what might happen if we do this”, do you think that that distinction between forecasting and the contingent possibility, “what might happen if we do or don’t do this”, do you think that distinction was properly understood by the government and the public?
Professor John Edmunds: No. I think sometimes at times it may have been deliberately misunderstood. So we were – so very frequently our scenarios about what might happen were afterwards treated as a forecast, when we’d changed the – the government had taken action to avoid that scenario. A classic example would be, I mean, the work on looking at the first wave and how many deaths there might happen in a first wave and a scenario that – that – you know, we were working on and Neil Ferguson’s group at Imperial was working on, would be – you know, there were many scenarios but one of which would be: what happens if we take no action and nobody changes their behaviour? That would be the kind of absolute worst-case scenario. And of course we took action, you know. And both my group and Neil’s group, the work suggested that that scenario would be devastating, there would be hundreds of thousands of deaths, hospitalisations way above what the NHS could cope –
Lead Inquiry: All right.
Professor John Edmunds: But we took action to avoid that, so the government took action to avoid that. So to compare then what happened with that scenario is actually meaningless really.
Lead Inquiry: I want to ask you about two particular examples. You’ve mentioned one of them indirectly already.
The Report 9, so-called, by Imperial College on 16 March I think –
Professor John Edmunds: Yeah.
Lead Inquiry: – was actually part of a wider body of material. You had drawn up, I think on 3 March, learning from a meeting on 1 March that also looked at how many deaths might occur or would occur if there was a failure to take control measures and what the impact would be on the NHS. And Professor Steven Riley, from whom my Lady heard, also gave evidence about his own work, a series of papers between 3 and 10 March.
Professor Ferguson’s work, or rather the work of Imperial College London, that Report 9, was met with a storm, really, of reaction and, in some places, criticism, and he was accused of being outrageously alarmist.
Were these scenario modellings, particularly of March, which set out what would happen if steps weren’t taken, in fact unduly alarmist?
Professor John Edmunds: I don’t think so. You know, we were, as you – we said before, from early on you could see that this had the – this was the – you know, this had all the characteristics of being a nightmare.
In terms of epidemiologically, it was a respiratory infection, so very easy to spread. Clearly very transmissible in the community. And although an infection fatality ratio of 1% doesn’t sound like a lot, when of course you match that with, if no action is taken, a large fraction of the potential will become infected very rapidly, that then – that then leads to a huge number of deaths.
Lead Inquiry: A second example, so moving forward, in fact, to the autumn, the government gave a press conference where some particular documents were used to – not directly used, I think, to justify the lockdown but they were certainly put into play, and they were documents which had been produced some weeks before by a number of modelling groups, so your own London School of Hygiene, I think Imperial, Warwick –
Professor John Edmunds: PHE in Cambridge.
Lead Inquiry: PHE, Cambridge, thank you. And they were work done at the request of the Cabinet Office to point out what the very worst or one of the worst or maybe even the worst, the reasonably worst-case scenario might be.
Professor John Edmunds: Yeah, there were – it was an early step to try to work up a new reasonable worst-case scenario. These reasonable worst-case scenarios were used for government planning. And it was an early step, actually at the request of SPI-M-O secretariat initially –
Lead Inquiry: All right.
Professor John Edmunds: – to come up with some … so to come up with some scenarios what might happen over the next few months.
Lead Inquiry: All right.
Weeks later –
Professor John Edmunds: Yes.
Lead Inquiry: – they were relied upon.
Professor John Edmunds: They were.
Lead Inquiry: The extent to which they were relied upon needn’t detain us, but there was a massive reaction in the press, was there not, because the press were saying: well, look, these documents appeared to show X number of deaths but they haven’t happened, or they won’t happen.
Professor John Edmunds: Yeah.
Lead Inquiry: The short answer was they were only scenario models, and they were reasonable worst-case scenario models to boot, and they were draft documents –
Professor John Edmunds: Correct.
Lead Inquiry: – and they were being prepared for a different purpose?
Professor John Edmunds: Correct. And it was worse than that, in fact, because every week we were doing medium-term projections, so, again, the various groups contributing to SPI-M-O were doing medium-term projections over a period of six weeks, I think is – four to six weeks is what we were doing, and every – and each of those groups were contributing to an ensemble estimate of what we thought would happen if nothing changed, and then every week we would look at how well we did last week and learn from it. So we would look at each individual model, how well that had projected what had happened in the coming week, and also the ensemble estimate, how well that had done, how well that had performed in the coming week, and the whole process would move on.
So since the date when those reasonable worst-case scenarios were generated at the beginning of October, there were three weeks or more of these more – what we think are more likely to happen, you know, and that had – those estimates had been validated by looking at what actually did happen. And they were doing – and they were actually capturing the trends really rather well.
So the government could have used that much more accurate – those much more accurate scenarios, medium-term projections, to – it didn’t matter, in a way. They were all still saying: unless action is taken, the NHS will come under severe – will come under severe stress very shortly.
But the way it was done and the way it was – to use the reasonable worst-case scenario, it reflected very badly on us, it made us look like we were, well, we were called doom –
Lead Inquiry: That you were being alarmist?
Professor John Edmunds: Yeah.
Lead Inquiry: All right. We’ll just have a quick look at some of the reaction, INQ000212171.
Professor John Edmunds: Yeah, yeah.
Lead Inquiry: “Apocalyptic forecast of 4,000 coronavirus deaths a day could be FIVE TIMES too high and had already been proved wrong when government revealed it at the weekend.”
Professor John Edmunds: Yeah, well, we would have said the same thing.
And the – of course the whole point of getting this ensemble estimate together is that it would downplay, downweight the more extreme estimates. Just the same way with sort of climate change, you know, some models might give a higher estimate of what the impact might be and some lower, and it’s the same thing here. And then by bringing many, many different models together, you’d get a consensus. And so what was done here was pick the worst – the worst – the most alarmist bit of the – of that – so of those four reasonable worst-case scenarios, the Daily Mail here is picking the worst one, and we would never have presented – we would never have presented it like that. We were presenting these consensus estimates, which of course would downplay the extremes and focus on the most – you know, where there’s most support from the different statistical – the different models.
Lead Inquiry: All right.
Professor John Edmunds: So it’s very ironic, really.
Lead Inquiry: You say in your statement that –
Professor John Edmunds: It’s them being alarmist, not us.
Lead Inquiry: All right.
You say in your statement, Professor, there are some lessons which can be drawn.
Professor John Edmunds: Yeah.
Lead Inquiry: Firstly, the limitations of models needs to be more clearly, widely understood?
Professor John Edmunds: Yeah.
Lead Inquiry: These are scenario models, they are all contingent, what might happen if we don’t do something. Secondly, government in future needs to be much clearer and more straightforward in the way in which it will rely upon such models and use them and –
Professor John Edmunds: It needs them, of course, it needs to have those forward looks, and – but it needs to be treated with some care.
Lead Inquiry: And also, thirdly, I think you would suggest that the way in which this valuable work was treated in some parts of the press was very unpleasant –
Professor John Edmunds: It was indeed, yeah.
Lead Inquiry: – as well as being wrong?
Professor John Edmunds: Exactly.
Lead Inquiry: All right. I now want to come, please, to discuss some of the particular measures that SAGE debated during the course of February and early March.
On 29 January, you were party to an email string with Professor Chris Whitty.
Could we have that up, please, INQ000212194.
We can see at the top of the page that the final email is from Chris Whitty, “Thanks that lot …”
Further down the page, on 29 January you’ve written to him saying:
“We are going to have a go at looking at the potential impact of mass school closure over the next few days.”
Obviously closing of schools was an important issue that was being looked at?
Professor John Edmunds: Yeah.
Lead Inquiry: But if we go further down to the – nearer the origin of – the beginning of the string and over the page, we can see that you’ve written a fairly lengthy email to Sir Chris Whitty:
“My comments are:
“1. Given the apparent speed of spread, it seems unlikely that contact tracing and isolation is going to be effective at buying us much time.”
Is that a reflection of the debate in fact – or the evidence you gave earlier, which is it was apparent this was the –
Professor John Edmunds: Yeah, so that work was being finalised at the time. I mean, this is 29th, I think we put it on our website –
Lead Inquiry: Yes.
Professor John Edmunds: – one week later.
Lead Inquiry: So you were clear and you told obviously the recipients of this email that your view was that contact tracing and isolation would be unlikely to be effective at buying much time?
Professor John Edmunds: I was taking the glass half empty view of it, of the results.
Lead Inquiry: But you were right.
Professor John Edmunds: Yeah.
Lead Inquiry: In relation to –
Professor John Edmunds: Unfortunately.
Lead Inquiry: In relation to travel advice, and exit screening, you’ve already given some evidence about that, was the position that the World Health Organisation had beforehand generally advised that screening and restrictions short of complete closure of a border were unlikely to be efficient or effective?
Professor John Edmunds: Correct, yes. They’d done a review of all of the global literature on it and come to that conclusion.
Lead Inquiry: And so –
Professor John Edmunds: That was for flu, of course, it was concentrating on flu, but it wouldn’t be very – very – different, but we did look at it.
Lead Inquiry: If we could have INQ000212206, did you enter into, again, another email string with, I think this time, Sir Patrick Vallance, Sir Chris Whitty, Professor Sir Jonathan Van-Tam, Dame Jenny Harries and Charlotte Watts at the Home Office? You say in this email at the top of the page:
“A concerted travel ban with our closest neighbours, from whom indirect travel from China would be expected, is going to be far more effective than us going it alone. However, even that is likely to have relatively limited impact, buying a few weeks at best. The question is what you could you achieve in this time? Very little in terms of … vaccines … [but it might] give the Chinese enough time to bring the epidemic under control.”
Professor John Edmunds: Yeah, well, I thought if they could bring it under control, they were under lockdown then, then maybe we might get away with this.
Lead Inquiry: And SAGE around the same time, the next day in fact, 3 February, INQ000212208, concluded, based upon a paper with which it was provided, if we could go to page – I think the second page, please, of this document, 3 February:
“1. On the expected impact of travel restrictions, SAGE estimates – with limited data – that if the UK reduces imported infections by 50%, this would maybe delay the onset of any epidemic in the UK by about 5 days; 75% would maybe buy 10 … days; 90% maybe … 15 additional days …”
Professor John Edmunds: Yeah.
Lead Inquiry: SAGE considered a report, we won’t need to get it up, in which I think the London School of Hygiene perhaps, rather than ICL, had concluded that tests or modelling had shown that 46% of infected persons would never be detected by screening at a border?
Professor John Edmunds: This was looking at temperature screening, symptom screening, which is usually done with temperatures. The problem is, of course, if you – it takes a few days for you to develop a temperature, you know, five or six days, so if you travel on day 0, day 1, day 2, day 3, day 4, you don’t get picked up.
Lead Inquiry: Contact tracing.
Professor Sir Chris Whitty asked in January for an investigation to be carried out into whether or not that would be effective. The London School of Hygiene produced a number of papers which they put online and then they published, I think, in The Lancet.
Professor John Edmunds: Yeah.
Lead Inquiry: Let’s have a look at that Lancet health article, INQ000212222.
Professor John Edmunds: I think this is the same one as before, with Joel Hellewell, is it?
Lead Inquiry: It’s the one to which you were a contributor. 212222. “Feasibility of controlling Covid-19 outbreaks by isolation of cases and contacts”. The findings, if we could scroll in – thank you very much.
There is a description of the consequences or the analysis of simulated outbreaks, but essentially, without going into the detail of that paragraph, what that data or what that analysis showed is that in order to be effective, contact tracing has to pick up a very large percentage, an overwhelming percentage of the people who are the contacts in order to work?
Professor John Edmunds: Correct.
Lead Inquiry: Was the fraction of contacts which have to be picked up to make it work as high as 70% to 80%?
Professor John Edmunds: It’s very difficult to tell, because of course, almost by definition, you don’t pick up the contacts you didn’t pick up. Yeah?
But there are clever ways that you can get to that, and actually later in the epidemic, when Test and Trace was launched, it was one of the things we – I myself kept raising with Test and Trace was to try to put these measures in, to see whether – to see what fraction of the contacts were being missed. But at that time it was impossible to tell.
Lead Inquiry: Sporting events. This was analysed by CMMID, the London School of Hygiene research institute or centre, as well as SAGE. Could we have INQ000212210.
This is dated 11 March, on page 1, “The impact of banning sporting events and other leisure activities on the COVID-19 epidemic”, prepared on behalf of the CMMID Covid modelling team.
Did it essentially conclude that banning mass gatherings would be unlikely to have a great deal of impact?
Professor John Edmunds: When looked at in the whole of the epidemic. So we were – we were – I think this is a kind of example of where there was a kind of over-reliance on modelling. So, yes, attending a sporting event would be, you know, more risky than staying at home, of course, but actually if it’s outside that risk was probably quite low, although we didn’t know it.
But at the population level, stopping sporting events is not really going to do very much, because we’d found, or I’d found on a Google, I’d found some attendance data, how many – how many times – what the global attendance – or the entire attendance of sporting events in the UK in, I think, 2018 or 2019, I can’t remember. And it was something like 75 million ticket holders, as it were, 75 million attendances at sporting events of every type, whether it’s the cricket or the football or Wimbledon or whatever it might be. And if you think about it, there’s 67 million of us, roughly, so that means on average – on average – we attend about one sporting event per year. And so if you stop the sporting events, is that going to stop the virus? Well, no, because it’s going to make a tiny impact on the total number of impacts that we make. So that –
Lead Inquiry: Throwing, it’s been described as, a lit match on a raging fire?
Professor John Edmunds: Yeah. So – but that’s looking at it at the population level, so – and of course that’s what we do, we are modelling things at the population level. Whereas actually at the individual level maybe it’s not a good idea to go to a sporting event in a pandemic. So for an individual, you know, sensible public health advice might be to say, well, “Don’t go”. But that doesn’t mean to say it’s going to have a big impact on the epidemic. It wouldn’t.
Lead Inquiry: So if answer to a question that in fact my Lady put to an earlier witness, if you attend a mass gathering event, there is a risk you will become infected and it’s a risk that you wouldn’t otherwise have run?
Professor John Edmunds: Yes and no, depends what you would do if you hadn’t have gone to the event. So if you’d have gone to the pub instead, then maybe the risk in the pub was greater than being at the event, if the event is outside.
Lead Inquiry: All right. But at a micro level there is obviously a risk for the individual?
Professor John Edmunds: Yes.
Lead Inquiry: But if you look at it on a population modelling level, there is a tendency, isn’t there, to overlook the significance of that risk?
Professor John Edmunds: Because of the population level, it’s tiny. It makes a tiny contribution to the entire – yeah, your analogy is a very good one.
Lead Inquiry: So, in truth, by relying upon modelling in order to answer the question, should we ban mass gatherings –
Professor John Edmunds: It’s the wrong – you’re really – you’re asking the wrong group of people, you should just take a decision about it. And, you know, there’s a lot – this is – you know, there’s lots of reasons why you might – might – why you might want to do it even though it might not have an effect or a very small effect at a population level. One we just talked about, an individual risk. Two is the optics, it doesn’t necessarily look good. You know, imagine the situation if we’d had our schools closed and the football was still going on. I don’t think anybody would have accepted that. It would have looked a bit strange.
Lead Inquiry: And in terms of the precautionary principle to which you referred earlier, there was obviously a good argument for banning mass gathering events?
Professor John Edmunds: Yes, even though I think, and we did work on it later, actually, it’s something we did some work on later in the epidemic, and it did show that actually the risk is really quite small.
Lead Inquiry: You’ve referred to the fact that modellers were handicapped to some extent by the delays in, originally or initially, receiving data from China, and understanding that data, and then towards the end of February and the beginning of March the delays of which you spoke in relation to the delays between testing and getting the data to you in terms of delays in people getting tested or testing the right number of people or getting an understanding of who was infected.
You raised with SAGE, didn’t you, on 13 March your concerns about how the significant delays were impacting your ability to model efficiently?
We’ll just have a look at that. INQ000212212.
Page 1 shows 13 March, the second page, paragraph 1, this is the date on which –
Professor John Edmunds: Yeah –
Lead Inquiry: – SAGE says: we now believe there are more cases in the United Kingdom than SAGE currently expected.
Paragraph 7:
“… we may be further ahead on the epidemic curve …”
Professor John Edmunds: Yeah.
Lead Inquiry: “The change in numbers is due to the 5-7 day lag phase in data availability for modelling.”
So you in essence said to SAGE, “We’ve been undone, there has been a delay in getting data to us, but now that we’ve got a better understanding, our situational awareness is better, we can now see we’re further ahead on the curve than we thought we were”?
Professor John Edmunds: We always thought there would be a delay, because of course there is, nothing’s – you know, it takes time for the data to come in, of course. But that was the first time we’d been able to estimate it. And that was the average delay. Some individuals on the database, the delays between them was up to three weeks. And so, yes, with having estimate – it was two bits of work that we were doing that week. I don’t know if you want the details or not.
Lead Inquiry: I don’t think we need trouble you for the detail of the work. The main point is you were working very hard on the modelling, but the output –
Professor John Edmunds: It was much worse. There was two bits, there was this – and we used to start SAGE meetings with a quick update, like a one-minute update from Chris Whitty or Jonathan Van-Tam on – just on the numbers of cases that had been reported. And of course those cases – because of this delay, those cases hadn’t actually been – become a case on that day that we were getting reported. They’d actually become a case a week earlier. So what, you know, Chris was reporting on was what was happening a week earlier.
And it’s worse than that, if you think about it, because it takes about five or six days between getting infected and becoming a case, and so actually we were being – you know, I thought that we were being lulled into a bit of a false sense of security here, in that actually the numbers of cases – because what was being reported on was infections that had happened perhaps two weeks earlier.
Lead Inquiry: All right.
Professor John Edmunds: And that’s just the ones we knew about.
Lead Inquiry: Because for all the asymptomatic infections –
Professor John Edmunds: Or even just the cases that had come through different routes, because we were still – to be tested you had to – there had to be a reason for you to be tested and to become a case, as it were, and that was you had to have symptoms but you also had to have come from a high-risk area, China, Singapore, mostly other places in the Far East initially. And so we weren’t testing people who had symptoms that hadn’t come from there initially.
That did – we did put systems in place at the end of February that would give an idea of infections in the community, infections in the community, and they immediately picked up a case – cases. So that –
Lead Inquiry: Sporadic?
Professor John Edmunds: These were the sporadic ones.
Lead Inquiry: All right, I’m going to pause you there, because we’ve got to move on to other topics.
So in summary, Professor, by this time, the beginning of February –
Professor John Edmunds: No, this is March.
Lead Inquiry: Well, sorry, I meant to say March. In fact 13 March.
Professor John Edmunds: Yeah.
Lead Inquiry: You’ve told us that by the end of January, the broad nature of the threat was known. By February, mid-February, the broad nature of the possible fatalities and hospitalisations and infections were known. The modelling process and the enormous amount of work dedicated to trying to bottom out the figures and get a proper handle on the nature of this pandemic continued. And then at the beginning of March SAGE was blindsided by the discovery that not only, as you’ve described, was there no effective means of containing the virus, and not only the virus was as deadly as it was, but that it had spread through the United Kingdom far further than anyone had realised?
Professor John Edmunds: Yeah. By picking up these sporadic cases they were not linked to importations or anything like that. So hopefully we’d have seen no – none of them. And this – by no means were we picking up every sporadic case. It was – this was like a sieve with huge holes in it. But there was two systems, if you think of two sieves, mostly holes and very little …
Lead Inquiry: Professor, I’m –
Professor John Edmunds: But at this point, so we should have seen none and of course we did start to see them, so – and we were trying to work out from the growth of those – so at this point it really was apparent that there was far, far more – not just had the – was the infection spreading but it was spreading much more widespread than all of us hoped.
Lead Inquiry: All right.
Professor John Edmunds: So it was – we were in big trouble.
Lead Inquiry: Would the full application of the precautionary principle in February, based upon the understanding of how fatal or damaging the virus was, have allowed the government, the country as a whole, not to have to wait to find out how far the epidemic had spread before realising that action and severe action was absolutely necessary?
Professor John Edmunds: Yeah, I’m not sure exactly what we – what would have been a proportionate response in February. That’s –
Lead Inquiry: All right.
Professor John Edmunds: Of course I wish we had taken more action in February, but I’m not sure – I’m not sure what would have been proportionate when the cases would have been very, very, very low.
Lead Inquiry: But in reality, nothing was done in February, other than a fairly low level surveillance on travellers –
Professor John Edmunds: So we were concentrating on kind of trying to pick cases up coming from overseas and we were concentrating of course on the places where we knew there was transmission, and there was always a risk that transmission was happening somewhere else, which indeed it was. In fact we imported most of our cases from Italy and France and Spain in the early parts and we were not looking there, initially at least.
Lead Inquiry: Was that predominantly also the half term –
Professor John Edmunds: This was after –
Lead Inquiry: – break –
Professor John Edmunds: Exactly, skiing holidays and the like. And just because of the – just the travel, how much travel there is between – between our countries.
Lead Inquiry: On 3 March, a report was prepared for SAGE, INQ000212223 by the LSHTM CMMID team.
Professor John Edmunds: Yeah.
Lead Inquiry: Which set out in very clear terms what the likely deaths would be?
Professor John Edmunds: Possible. I mean, this is – these are possibles and these are scenarios.
Lead Inquiry: Professor, forgive me, I hadn’t finished – I’m afraid I was just taking my time in formulating the question. What the likely deaths would be if social distancing measures were not applied. It was a classic scenario model: what might happen if something is not done or only something else is done.
The report showed to SAGE, did it not, we can see from the results in the middle of the page:
“The unmitigated epidemic is expected to result in 570,000 deaths … in England and result in a peak demand of [almost a million] non-ICU beds … 130,000 ICU beds … at peak. Closure of schools is estimated to be the least effective of these policies … Cocooning of the elderly, general social distances, and case isolation are all estimated to reduce deaths by about 25% … social distancing reduces peak demand on hospital services more than the other strategies. The combination of school closure and social distancing … [a reduction] of about 75% [in beds] … 32% [in deaths]. The combination [that’s to say all of them] would reduce demand by about 75% and reduces death by about half.
So, again, this was not an alarmist production, was it?
Professor John Edmunds: No, this was just what you would get from those scenarios. I mean, obviously the worst case – the unmitigated one, I can’t imagine it would ever have happened, we must have – we must have taken action at some point, but … and of course it doesn’t take into account – and this is important – it doesn’t take into account spontaneous behaviour change, because we had no way of estimating what that might be, what that might – we’d never done anything like that before. And in previous epidemics, because I did measure contact patterns in the 2009 pandemic, people didn’t change their behaviour at all. Obviously it was low risk. So it doesn’t take into account spontaneous behaviour change, which would have probably happened, but there’s no way we could predict that.
Lead Inquiry: Estimate that. And it didn’t take into account, of course – well, it didn’t say – it projected one outcome of what might happen if these steps were taken individually or in combination?
Professor John Edmunds: And, you know, I regard these as – as I said before, I think these are broad sketches of what might happen rather than precise … but they were huge numbers, you know, that was the –
Lead Inquiry: Huge numbers. And this report set out in clear terms, did it not, that the NHS would be overwhelmed if certain measures –
Professor John Edmunds: There’s no way you can cope with that sort of level of demand, you know.
Lead Inquiry: This was plainly brought to the attention of SAGE, of course, it was consistent, wasn’t it, with the outcome of Professor Riley’s reports and also Professor Neil Ferguson’s reports of a few days later?
Professor John Edmunds: Yeah, Professor Riley’s were less detailed than this, so I would say that it’s more consistent with Neil Ferguson’s estimates. And if you compare the two, there are differences, but broadly they’re kind of in the same ballpark.
Lead Inquiry: Your report notably, or rather the CMMID report notably doesn’t get into the conceptual debate of suppression or mitigation. Professor Riley’s and Professor Ferguson’s do. There are references to that –
Professor John Edmunds: They were done later – they were done later, those. This was very early March as opposed to sort of mid-March.
Lead Inquiry: To what extent did any debate about reasonable worst-case scenario, whether a response was mitigation or suppression, whether or not herd immunity was good or bad, assist in understanding these basic thoughts, which is unless practical measures are taken, the deaths are going to be huge?
Professor John Edmunds: Yeah. That’s the simple message.
Lead Inquiry: Do you –
Professor John Edmunds: And we were not alone in this. So Neil and Steve, their work was similar. And other people were doing similar things elsewhere, not just in the UK, but we were all – it all pointed to extremely – you know, the sort of situation that I don’t think anybody could possibly just let happen.
Lead Inquiry: Did those debates about what was a reasonable worst-case scenario, was it going to happen, are we suppressing or mitigating, need to be resolved in order for SAGE to be able to say to the government, “There is this massive problem coming”?
Professor John Edmunds: I guess if there is one thing saying there is a problem, it is better to come with a solution. And I don’t think we had the solution – I don’t think we had the solution at that time, so we were looking at these sort of measures – you can see, I mean, even with these measures and combinations of these measures, it still looked horrendous. The –
Lead Inquiry: I’m sorry to interrupt you, but you were looking at measures, you weren’t engaging here in a polemic about whether it’s suppression or mitigation or a reasonable worst-case scenario; you were focusing on what practically needs to be done?
Professor John Edmunds: Yeah.
Lead Inquiry: All right.
Professor John Edmunds: And it was more than this, in my view.
Lead Inquiry: As proved to be the case.
The government had already produced a report on 3 March, a Coronavirus: action plan, of which a major part, the first stage, was contain.
Was that a publication of which you had become aware prior to its publication?
Professor John Edmunds: No, I mean, those sorts of strategy documents that the government published periodically over the course of the epidemic, of which that was the first one, we didn’t see those before they came out.
Lead Inquiry: What was your reaction on seeing that the government’s future strategy, because it was a document produced for what should happen going forward, what was your reaction on seeing that an element, the first element was containment?
Professor John Edmunds: Yeah, I … I mean, it would have made more sense for that to have come out a month earlier. At that time, I know we were officially still in the containment phase, I think, but, you know, the – as I say, from these sporadic cases you could see, there was – we hadn’t contained the virus, you know, at that point. So there was that.
There isn’t a lot of detail in that document as well, so it is very general, it doesn’t really say what really we would do. And maybe that was fine, because I don’t think that had been worked out, but it was a very kind of high level document.
Lead Inquiry: Bluntly, Professor, the ship had sailed. There was and could be no containment, the virus was rife in the population?
Professor John Edmunds: Rife I don’t think is right, yet. I mean, are we talking about 3 March? It was certainly here, it was certainly spreading, and this was the work that we were trying to do. Actually later than this, it was around 8 March when we were looking at these sporadic cases and trying to work out how many – what was the scale of the epidemic, because the reported cases was not reflecting that by any means.
Lead Inquiry: All right.
Professor John Edmunds: So we didn’t really know the scale of it although the very fact that we picked up these sporadic cases was an alarm bell.
Lead Inquiry: In your statement you say, recognising that some observers have indicated that SAGE appeared to be too slow to recommend action during the early weeks of the epidemic, that you have some sympathy with this view and that you had become increasingly anxious yourself?
Professor John Edmunds: Absolutely.
Lead Inquiry: Is that because you say that this understanding of the sheer number of deaths and hospitalisations and the impact on the healthcare system in the United Kingdom should have been understood earlier or –
Professor John Edmunds: I mean, it was, I mean, everybody, I mean, I saw you inter – well, Mark Woolhouse’s evidence from a few days ago, and, you know, he did this sort of simple back-of-the-envelope calculation based on the reproduction number – and he had done it back in January, based on the reproduction number and guestimates of case fatality ratio and come with very big numbers. We’d all done the same calculations back in January and early February. This was the – this and when – and Neil’s was – and Neil was – Ferguson was doing the same, you know, in parallel doing – looking at similar things. This was when we had kind of – it had gone through the formal modelling kind of process and those numbers were coming out and they were – they were truly horrendous.
Lead Inquiry: And that should have been understood earlier, is what you’re saying in your statement?
Professor John Edmunds: So I think our broad – if you want – so we didn’t know – the last piece of the jigsaw was related to hospitalisation. It was difficult to understand exactly what fraction of people would be hospitalised, because in the early days, particularly in East Asia, and even here as well, early cases were hospitalised whether they needed it or not. So they were hospitalised for public health measures – reasons, so that they wouldn’t spread. This was in the containment phrase. Yeah?
Lead Inquiry: All right.
Professor John Edmunds: So it was difficult then to know exactly what fraction would need to be hospitalised for clinical – on clinical grounds, and how long they would have to be hospitalised for and what fraction might need intensive care. So that was the last bit of the jigsaw. I mean, you could get a guesstimate at it, and a reasonable – and as February moved on that became more clear, but we didn’t have a – I would say we didn’t have a solid estimate of it until really that meeting on 1 March, on the Sunday, 1 March, when we really – we had a meeting with colleagues at Oxford, Imperial obviously, and the NHS, and then we got a much clearer idea. So that was the final bit of that jigsaw. But you didn’t really need the whole jigsaw, I mean, you could see the picture was pretty obvious from – from, you know, much before then.
Mr Keith: It’s the perfect moment.
Lady Hallett: I’m afraid going to complete you today – sorry, today we’ll complete you, but this morning. I hope you were warned that you might go over lunch.
The Witness: Yeah, that’s okay.
Lady Hallett: Thank you very much indeed, I will return at 2 o’clock.
(1.00 pm)
(The short adjournment)
(2.00 pm)