2. Professor Graham Medley
PROFESSOR GRAHAM MEDLEY (affirmed).
Questions From Counsel to the Inquiry
Mr O’Connor: Thank you, Professor, do take a seat.
Can you give us your full name, please?
Professor Graham Medley: Yes, Graham Francis Hassell Medley.
Counsel Inquiry: Professor Medley, you have, at our request, prepared a statement for the Inquiry. We have it up on screen. It weighs in at 123 pages, and I think on the last of those pages – yes, don’t worry, we don’t need to call it up, but you have signed the statement under the statement of truth indicating that you believe the facts stated in the statement to be true; is that right?
Professor Graham Medley: That’s correct, yes.
Counsel Inquiry: That signature was dated 4 September 2023; yes?
Professor Graham Medley: Yep.
Counsel Inquiry: Thank you.
Professor, your witness statement gives us some detail of your career to date. You are currently, and you have been since 2015, professor of infectious disease modelling at the London School of Hygiene and Tropical Medicine?
Professor Graham Medley: Correct, yeah.
Counsel Inquiry: That institution in London is one of the main centres of epidemiological research in this country?
Professor Graham Medley: Yes.
Counsel Inquiry: And we’ll be hearing also from one of your colleagues, Professor Edmunds, who is also there.
Before moving to, do we call it LSHTM or do we say London School of –
Professor Graham Medley: The “London School”, with apologies to the London School of Economics.
Counsel Inquiry: That abbreviation may end up saving us hours of time in the next couple of weeks, Professor.
So before moving to the London School, we see that you were based, first of all, at Imperial College London for ten years, from 1983 onwards.
Professor Graham Medley: Yep.
Counsel Inquiry: Then you spent some time at the University of Warwick before moving to the London School?
Professor Graham Medley: Yes, correct.
Counsel Inquiry: I think what we will discover is that, in fact, those other two institutions, Imperial College and also Warwick University, are two of the other leading epidemiological centres in this country?
Professor Graham Medley: Yes.
Counsel Inquiry: Also relevantly, Professor, you were co-chair of the committee SPI-M – which we’ve heard something about from Dr Wainwright, and of course we will be hearing much more from you about – from October 2017 until, first of all, the start of the pandemic, then, as we’ll hear, SPI-M was, as it were, replaced by SPI-M-O during the pandemic –
Professor Graham Medley: Yep.
Counsel Inquiry: – and you chaired that committee during that time. Then in February last year, when SPI-M-O stood down, you returned to being co-chair of SPI-M?
Professor Graham Medley: Yes.
Counsel Inquiry: You have, though, resigned from that position in June of this year?
Professor Graham Medley: That’s right, correct, yeah. And I was co-chair of SPI-M-O as well.
Counsel Inquiry: Yes.
Professor Graham Medley: So the co-chair is important.
Counsel Inquiry: Yes, and I’m going to come back to it, but thank you, thank you.
Just before we get into the detail, though, Professor, just help us, give us an idea of just how much of your time you spent on these matters during the pandemic. You’ve already told us that you were chair of SPI-M-O for that two years or so; we’ll hear you were also someone who attended SAGE meetings; you were also, I think we’ll hear, a member of the Welsh Technical Advisory Group. That must have taken up a lot of your time?
Professor Graham Medley: Yeah, so SPI-M, the pre-pandemic, is about six meetings a year, although we didn’t actually meet, I don’t think, in 2019 because of Brexit preparations, and then from January 2020 my involvement ramped up so that from mid-February 2020 until March 2022 I was essentially full-time working with the members of the committee and chairing the meetings, and working with the secretariat.
Counsel Inquiry: So really very little or no time left for your research or teaching commitments –
Professor Graham Medley: No.
Counsel Inquiry: – that you would otherwise have had?
Professor Graham Medley: No. The institution was extremely kind and my colleagues at the London School were extremely good at filling in for me, but no, I effectively gave up independent research for that period.
Counsel Inquiry: And I think, from having read your statement, it’s fair to say not just a full-time, as in Monday to Friday, 9 to 5, but rather more than that?
Professor Graham Medley: Yes, it was –
Counsel Inquiry: As –
Professor Graham Medley: Yes, I mean, I was – I still had commitments that I had previously, so I was filling in those, but – but anyway, the demands of what I was trying to do. And I took on other things. I’m involved, for example, in the Academy of Medical Sciences’ reports, and that was partly to kind of help things be joined up and to function to – for Wales, as you’ve mentioned. I mean, those were things that technically I suppose I didn’t have to do but actually I felt were important to be doing.
Counsel Inquiry: Professor, thank you. Let’s take then, shortly, a step back, I just want to ask you a few questions about SPI-M before the pandemic.
Let’s look, if we can, at page 9 of your statement, paragraph 3.1, that’s where you set out the dates that we’ve already discussed, you refer there to having been co-chair of SPI-M from October 2017.
Could I ask you what you might think is a surprising question: what does the I in SPI-M stand for?
Professor Graham Medley: It did stand for influenza, so the committee was originally I think set up in sort of 2005 and – or 2009 to cope – to deal with pandemic influenza. It has since changed to infections, pandemic infections, and one of the – I noticed three – two things when I took over as chair in 2017, one was that we really shouldn’t be just thinking about infections and so it was agreed – sorry, influenza, and it was agreed we would switch to infections more broadly, and we were in the process of going through what we had in terms of preparedness document and modelling capacity, so in 2018, 2019, but obviously we didn’t – because of the delays in 2019, we didn’t get through that process.
And the other thing I noticed was to do with devolved administrations, which we might come back to it, but that was noted before the pandemic.
Counsel Inquiry: I see. So just to be clear, I won’t call it up, because you’ve explained it all very well, but we were looking at an organogram with Dr Wainwright’s evidence and SPI-M on that organogram had a yet different meaning of I, for insights, but that’s wrong, isn’t it? It was influenza, and, as you’ve explained, it’s now infections?
Professor Graham Medley: Yes. Well, I’m not sure. That definition was created earlier. I’m pretty sure it was pandemic influenza, but I might … I might be wrong. I only ever called it SPI-M.
Counsel Inquiry: Well, it’s quite important, isn’t it, Professor, to know what the committee is called?
Professor Graham Medley: The committee was focused on influenza and that’s why I challenged it when I became chair, and the then Deputy CMO, Professor Jonathan Van-Tam, agreed with me, and so it was changed to infections at that point.
Counsel Inquiry: Perhaps we can just call the organogram up.
It’s INQ000303289, page 1, please.
So if we look there, do you see the top and bottom row?
Professor Graham Medley: Yes.
Counsel Inquiry: Do you think, then, Professor, this actually might represent a misunderstanding within GO-Science as to what the I stands for?
Professor Graham Medley: It clearly represents a misunderstanding between myself –
Counsel Inquiry: Well, it could just be a mistake, because the I, for example, in SPI-B certainly stands for insights?
Professor Graham Medley: Yes.
Counsel Inquiry: But with your experience of working on this committee, I’m asking you whether you think that it may be that other people on the committee and who support the committee might actually think that the I stands for something different?
Professor Graham Medley: My understanding is that everybody on the committee was focused on influenza, and that is why I challenged it when I became chair, because I said that the next pandemic may well not be influenza.
Counsel Inquiry: All right. Let’s leave it there. Thank you, we can take that down.
So you’ve explained then that when you took over in 2017 it was an influenza committee, if I can use that shorthand, but that before the pandemic you were already trying to make that shift to broaden the scope of the committee so that it looked more broadly at infection –
Professor Graham Medley: Yep.
Counsel Inquiry: – rather than simply at influenza, and I think you said a moment ago that something in the nature of six meetings a year would have been –
Professor Graham Medley: Yes.
Counsel Inquiry: – the norm?
Professor Graham Medley: Yes.
Counsel Inquiry: Although you also said that in fact the year before the pandemic it hadn’t met at all?
Professor Graham Medley: I think it met once at the beginning of 2019. We – the role of that committee was really to address policy questions that were raised to us, and we relied upon people coming from the centre of government with particular questions, and if they were not available, then in most – a lot of the rest of the work could be done by email and so there was no point in us gathering if policy wasn’t there to ask us questions and interrogate us.
Counsel Inquiry: Does that explain what might otherwise have been a bit – it explains your remark that perhaps the reason you didn’t meet in 2019 was because of Brexit?
Professor Graham Medley: We were told that was why.
Counsel Inquiry: There’s no obvious connection between the risk of a pandemic and Brexit, but is your explanation that the officials in government were so busy concentrating on Brexit they didn’t have time to ask you –
Professor Graham Medley: Yes.
Counsel Inquiry: – about pandemics?
Professor Graham Medley: Yes.
Counsel Inquiry: But I think you’re telling us that the fact you didn’t meet didn’t mean that, as a committee, you weren’t doing any work; is that fair?
Professor Graham Medley: Yes, I think so. We had – we have – the main output from the government was the preparedness document, and this is a short summary of what modelling tells us about the start of epidemics, and it’s designed – was written originally to fill the gap, that kind of knowledge gap between the start of the epidemic and before we have sufficient data within the UK to be able to make more focused – produce more focused evidence, and it’s there to, as I say, fill that gap, to explain to policymakers what we understand generically about controlling the start of epidemics. And we had just refreshed that document and we were going through to refresh it, and I can’t – I can’t remember at the moment what else was on our agenda, but I think we were being asked about the – the Department of Health had conducted an analysis of pre-buying vaccines, influenza vaccines, and investing that, and we were being asked to discuss that work.
Counsel Inquiry: Right. So that was the sort of background activity –
Professor Graham Medley: Yep.
Counsel Inquiry: – in the year or so before the pandemic?
Professor Graham Medley: Yes.
Counsel Inquiry: Just for completeness, you mentioned the Department of Health, that was the body, the department, to which SPI-M reported?
Professor Graham Medley: Yes.
Counsel Inquiry: Although, as we have seen from the organogram, it’s accurate in this respect, once the committee became SPI-M-O, it reported to SAGE?
Professor Graham Medley: Correct. We had a secretariat associated with SPI-M, and we retained that same secretariat as SPI-M-O, so the people we were working with day to day didn’t change, but the position in the organogram changed.
Counsel Inquiry: Let me ask you about the switch from SPI-M to SPI-M-O. We can look briefly, if we may, at page 23 of your statement, paragraph 3.43. You explain here, Professor, that SPI-M, so the old committee, was emailed in late January, 21 January, about what you describe as the novel coronavirus.
Professor Graham Medley: Mm-hm.
Counsel Inquiry: You spoke to the secretariat, and in fact there was then a meeting of SPI-M on 27 January to discuss preparedness and so on, and it appears to have been shortly after that that the change was made from SPI-M to SPI-M-O, because you then mention the first meeting of SPI-M-O on 3 February, so a week or so later.
On the paper, it’s simply a change of name from one committee to another. What about the membership of the committee, did it remain the same as between the old and the new?
Professor Graham Medley: No, it didn’t. I advised – I mean, it’s not my committee, it is run by the secretariat, but I gave a lot of thought and we had a lot of discussion about the membership, and so the – I think all the members of SPI-M became members of SPI-M-O, but we also involved a much wider range of people as well. And my principal – well, I had several concerns, but at that point my main concern was that we had sufficient expertise to be able to answer any – all the questions that we were – we might be asked, but I also gave thought to the composition of the committee in terms of the people, their relationships to each other, and the relationships to me.
I had no thought at that point that this would go on for two years, meeting every week, and I did have qualms at the beginning that I had involved too many people, but in fact that turned out to be very good, because in the end SPI-M-O was essentially co-created by the members and the secretariat, and we needed to have that community to be able to carry through, as I say, meeting weekly for two years.
Counsel Inquiry: Yes. Just give us an idea, Professor, we don’t need exact numbers, but the move from two committees, roughly how many academic members of the first committee and then how many – how much larger was SMI-M-O?
Professor Graham Medley: That’s a good question. Off the top of my head I’m going to say three times bigger, but I would need to count them.
Counsel Inquiry: A substantial increase.
Professor Graham Medley: Yes. And the workload meant that not every member could attend every meeting, so having that, if you like, redundancy in the numbers, so meant that there was always a – always a quorate in the meeting.
Counsel Inquiry: Just before we get much further, I want to take you to a part of your witness statement where you crystallise what the function of this organisation was. So if we can look at page 86, please, and it’s the paragraph at the bottom of that page, 8.27, you say:
“The role of SPI-M-O [so the new committee created in early February 2020] is to generate scientific evidence based on transmission dynamic modelling of the epidemic. The key questions for SPI-M-O are what drives the tram (epidemiological parameters, core groups …) and what determines disease given infection.”
And you refer to the models as being “intended to inform policy”, so “guided by what policy options are being considered at the time”.
And you go on to say:
“The age-dependent risk of severe outcomes [that’s of the infection Covid] … were well established by the end of February.”
So we’ll come back, I don’t want to unpack all of that now, Professor, because in a sense that’s what we’ll do for the rest of the morning, but it was a modelling committee, and those key aspects of the disease, what drives it and what determines the disease, given infection, those were at the heart of the modelling exercise?
Professor Graham Medley: Yes, I mean, the – just to caveat that, we are not clinicians, we’re not medics, and we’re not immunologists, although all of those disciplines are pulled into modelling at one point or another, so we are not looking at the process of disease, but we are looking, at a population level, at the consequences of widespread infection.
Counsel Inquiry: Moving on to a couple of more just practical points about the committee, you mentioned that you were co-chair of SPI-M-O, as you had been of SPI-M, and more precisely you were the academic co-chair, and we know from your statement that there was also a policy co-chair. Certainly as from March of 2020 the co-chair, the policy co-chair, was Angela McLean?
Professor Graham Medley: Yes, correct.
Counsel Inquiry: Who was then the chief scientific adviser in the Ministry of Defence?
Professor Graham Medley: Yes.
Counsel Inquiry: Tell us in a few sentences what the significance of there being two chairs and their different functions and responsibilities.
Professor Graham Medley: So, in order to kind of make a lot of sense of the rest of my evidence, I just need to kind of point out that government – what happens in government is very different to what happens outside. The code of conduct, the employment of civil servants, the conditions under which they’re employed, is very different from outside, and particularly from academia, which is where I’m coming from. And so crossing that fence, the things that we were allowed to be – to know and the things that we were allowed to say, in some senses, but not – it’s more about what we could know and what we couldn’t know – is not simple.
And so it was decided, I think as I became chair, the co-chair situation was developed. So there was somebody inside government, and their task is to talk to people in government about the modelling, to bring the appropriate questions to us, and to take the answers back. And then I’m the academic co-chair and my role is to try and get – or to get the members to provide the evidence in answer to the questions.
But because I am not a member of government, I can’t go into government and talk to civil servants openly, you know, about what they might be thinking or what they’re worried about.
Counsel Inquiry: Yes, and I think you explain in your statement that you felt that Angela McLean did a very effective job of fulfilling that function –
Professor Graham Medley: Yes.
Counsel Inquiry: – of the policy chair that you just described?
Professor Graham Medley: It was an essential – well, essentially it was an extremely important step change in the way that SPI-M-O worked in March 2020.
Counsel Inquiry: Moving on, we’ll see from some of the documents we look at that the routine during the pandemic was for there to be three SPI-M-O meetings a week, two subcommittee meetings, and then, if you like, a plenary meeting which considered what had happened at the two earlier subcommittee meetings; is that right?
Professor Graham Medley: That was the normal pattern, but there were many other ad hoc meetings as issues arose, and of course the meetings between members, you know, to discuss particular technical issues.
Counsel Inquiry: Yes, but what we may see is that the subcommittees then fed into, and other discussions, the main meeting?
Professor Graham Medley: Yes.
Counsel Inquiry: Then that main meeting produced material, which we’ll have a look at in due course, which then went to SAGE?
Professor Graham Medley: Correct, yeah.
Counsel Inquiry: We’ve heard from Dr Wainwright that that pyramid, if you like, of advice being generated by SAGE and its subcommittees, at the top were Chris Whitty and Patrick Vallance?
Professor Graham Medley: Yes.
Counsel Inquiry: Did they ever attend SPI-M-O meetings?
Professor Graham Medley: They didn’t in person, except very occasionally, but the meetings, especially when they went online, so at the end of March, there would be typically sort of somewhere between 20 and 40 members of SAGE, there would be sort of the order of ten members of the SPI-M-O – sorry, members of SPI-M-O, and then about ten members. SPI-M-O secretariat, and then a whole host of other people, sort of numbering more than the other attendants combined, of people from across government listening in. And so I’m pretty sure that there will have been people from SAGE secretariats and the CMO’s office also listening in to that meeting, but they came only very – I can only remember twice where Patrick Vallance came to SPI-M-O.
Counsel Inquiry: And you’ve described, I was going to mention it, but from the end of March, so the first lockdown, your meetings were all done remotely?
Professor Graham Medley: Yes.
Counsel Inquiry: And the output, again we’ll come back to it, but the output from these, the larger SPI-M-O meetings, would it be fair to say that, first of all, there was something called a consensus statement, and we’ll have a look at one or two of them in due course, which was the work, or the outcome of the meeting which was going to SAGE, and you might also send papers, academic papers, that had been prepared by members of SPI-M-O and discussed as well?
Professor Graham Medley: Yeah, they were technical papers rather than – I mean, they were written by academics but they hadn’t been through the peer review process.
Counsel Inquiry: No, that’s an important qualification. Because of the speed with which these matters with being dealt with, they were being turned around very quickly, but they look like short academic-style papers?
Professor Graham Medley: Yes, they often were the beginning of bigger amounts of work.
Counsel Inquiry: One other practical aspect of SPI-M-O, and for this can I ask that we go to page 35 of your statement, and paragraph 3.84, it’s the question of tasking, Professor.
I think you were watching Dr Wainwright’s evidence, is that right, and so he was of course talking about SAGE, not SPI-M-O, but he explained that, yes, SAGE received questions from policymakers that they were asked to consider, but there was also at least a degree of self-tasking going on on SAGE, so proactivity as well as reactivity.
What you describe in this paragraph is a bit different. What you seem to be suggesting is really you were purely reactive. You say that although you had some influence with the secretariat, the agenda was set by policymakers and decision-makers and you weren’t able to determine the scope of the agenda.
So are we right to see a bit of a difference between what you say about SPI-M-O and what Dr Wainwright said about SAGE?
Professor Graham Medley: Well, it might just be a matter of language and degree, but I felt that we were – myself especially, but members were able to raise issues, and one that was raised a lot was the relationship between – with economic analysis, and so we were – felt quite happy to be able to tell the secretariat, SAGE and SPI-M-O secretariat, that this ought to be considered. But in the end it’s not our meeting, we are independent academics who are coming along to a meeting organised and held by government in order to get our input into the questions that they wish to address.
Now, if we spot a gap and say, “No, you’re asking the wrong question, and this is a whole area that you need to look at”, then they have the decision about whether to take that advice or not.
Counsel Inquiry: But you could be proactive at least in the sense of expressing a concern that there was a gap?
Professor Graham Medley: Oh, yes, and do it in the strongest, most vigorous possible terms. And because, of course, we’re outside government, we could take – you know, take that to the public, as it were. So we did have quite a lot of power, soft power, if you like –
Counsel Inquiry: Yes.
Professor Graham Medley: – in terms of determining the agenda. But in the end they weren’t our meetings.
Counsel Inquiry: Understood. I’m going come to come back to one or two of those points in due course but let me just ask you one or two more quite practical questions.
The first is: we’ve seen that pyramid before, it’s right, isn’t it, that SPI-M-O itself had a number of subject-based subcommittees, not just the two that met before the plenary session, but various that were little committees that were focusing on particular issues?
Professor Graham Medley: Mm-hm.
Counsel Inquiry: You’ve listed them in your statement, I’m not going to go through them. But can you help us with whether there was something called a behavioural and social interventions subgroup of SPI-M-O?
Professor Graham Medley: I think very early on there was, but that got moved, got I think turned into a child – a school-specific –
Counsel Inquiry: Right.
Professor Graham Medley: I think it was called BSI at that point, but yes, as you’ve described it, but then it changed its name.
So the acronyms for what we now call NPIs changed a lot in that period. Whether that subgroup started, I think, and I think Professor Julia Gog was the –
Counsel Inquiry: Yes.
Professor Graham Medley: – leading that at one point, but I don’t remember it meeting beyond the end of March 2020.
Counsel Inquiry: Right, so early on.
On a similar theme, Professor, you’ve mentioned your involvement in the Welsh Technical Advisory Group. I’ll ask you about that in a moment, but before I do, much more generally, as far as SPI-M-O was concerned, did you think it was important that the devolved nations themselves were represented through membership on SPI-M-O?
Professor Graham Medley: Yes. So the devolved administrations or devolved nations question is something which worried me a lot over the two years. As I said in 2017, when I took over as chair, one of the questions – the other question I looked – that concerned me was: who were we talking to in the event of a pandemic? And clearly because health is a devolved responsibility, it meant that the different nations could make different decisions and go in different directions, and there’s a danger, in epidemiological terms, if that happens that nations start gaming against each other. In other words, the best thing for Scotland to do depends on what England does, and the best thing for England to do depends on what Scotland does. So – and that almost never produces an optimum outcome. It’s much better if you have both administrations agreeing a common goal and co-operating closely.
So I raised that then with the secretariat and I don’t know who they spoke to but the message came back that, no, that would not be a problem, that the United Kingdom in the event of a pandemic would respond as one unit of administration. I didn’t take it any further, it’s not – my role is to raise problems not solve them. But clearly as the epidemic – well, let me start at the beginning, as it were.
The epidemics, when it first started, as you have heard and you may well ask me in the future, the data flowing for analysis and modelling was really very poor, and it wasn’t until late in April 2020 that we actually had data from across all four nations.
So there was clearly not a seamless navigation of the four nations issue, and nations data definitions changed differently during the epidemic, which caused us some analytical headaches. But clearly, as policy started to diverge, it became, I think, you know, important that not only did we have a central group of SPI-M-O, I needed to – or it needed to be done to make sure that there was modelling capacity within each of the nations to help support their decisions, and so we co-opted a member from the University of Swansea, who essentially formed the – Mike Gravenor, who formed the kind of modelling unit for Wales, but with input from other SPI-M-O members, so I think Matt Keeling worked quite close, and John Edmunds worked quite closely with Mike Gravenor, and Wales invited me to their TAG meetings and I went to ensure that there was some coherence in terms of capacity and capability and understanding of the modelling.
Scotland had a much bigger internal capacity, so I really wasn’t involved that much at all with Scotland. They came to our meetings in SPI-M-O, and some of the submeetings that you’ve talked about the devolved administrations brought their own analysis to those. The nation I didn’t really have much involvement with at all is Northern Ireland. I think right at the beginning or early in the epidemic it had been suggested that I have a call with the Chief Medical Officer for Northern Ireland, but that I don’t think ever transpired. So, yeah, I’m – unfortunately, and I don’t – I’m not very proud of that, it didn’t happen.
Counsel Inquiry: Your discussions didn’t happen?
Professor Graham Medley: The discussions didn’t happen, and I don’t – I didn’t have sight of what Northern Ireland were doing in terms of modelling.
Counsel Inquiry: So you’ve, I think, explained that both in Wales and Scotland there was, as it were –
Professor Graham Medley: Yes.
Counsel Inquiry: – a freestanding, albeit linked, modelling capacity that was being developed?
Professor Graham Medley: Yes.
Counsel Inquiry: Is the position that you can’t really help us with what, if anything, similar took place in Northern Ireland?
Professor Graham Medley: I don’t – I don’t know what took place in Northern Ireland.
Counsel Inquiry: Yes, thank you, Professor, that’s very helpful.
I want to move on to a slightly different topic, albeit one that’s at the heart of your committee, which is about modelling. And I do so with some trepidation, given the complex detail on models which is to be found not only in your statement but in the statements of your colleagues which have been prepared for this Inquiry.
I want, if I may, to at least start with some very basic principles, and it may be that we can build our understanding with some of your colleagues who are giving evidence next week.
I’m looking for these purposes at page 31 of your report, and it starts at paragraph 3.69.
You describe there, Professor, a very basic distinction in modelling between, on the one hand, what you describe as “statistical models”, which are “data-driven”, which have few, if any, assumptions built into them, and which generate what you call a prediction or, in longer terms, a quantitative forecast.
So that’s one thing, the statistical model.
On the other hand, you describe something called a mathematical model, which is driven to a much greater degree by assumptions, what you describe as a series of what ifs, and those models you say generate not predictions or forecasts, by which you mean the same thing, but scenarios, which is I think a term of art and means something rather different.
So far, so good?
Professor Graham Medley: Yes. No, absolutely. I mean – but virtually all the models, the ones we use, are somewhere between the two.
Counsel Inquiry: Exactly. So I wanted to say, although they’re capable of being described as alternatives, if you like, they are in fact on a spectrum, they’re at either end of a spectrum –
Professor Graham Medley: But they are a mixture of data and assumption.
Counsel Inquiry: The more you go towards a mathematical model, the more assumptions that the model builds in?
Professor Graham Medley: I’m a biologist by background, so a mathematician I’m sure – I think your next witness might well argue about some of the wordings and the definitions, but essentially once you replace, start to replace data with assumption, then you have moved towards a more mechanistic description, set of assumptions about the processes that determine transmission.
Counsel Inquiry: As you do that, what – the output from that model –
Professor Graham Medley: Yeah.
Counsel Inquiry: – becomes not a prediction but a scenario?
Professor Graham Medley: Yes, at the statistical end – I mean statistical models still have models in them, they still have mathematics in them, but they are much more data driven, and from those you can create formal statistical predictions, and we did that, on SPI-M-O, early in the epidemic. But it was quite short lived because it’s the sort of thing that can be done relatively easily, to –
Counsel Inquiry: I want to move from the general to the specific, or the practical. Because, as you say, when one looks at the SPI-M-O papers, and we’ll look at them, there are various things that are being done. Some of the – it appears, anyway – product of your meetings was what you describe in your statement as “nowcasts”, in other words estimating the position of the disease at the time of the meeting, and that I take it would be performed by a statistical exercise; yes?
Professor Graham Medley: Yes, a more statistical –
Counsel Inquiry: Yes, and then, perhaps a little bit further along the range, we will see in the papers “Medium term projections”, which appear to be a prediction or a projection of how the pandemic is likely to develop in a matter of weeks, three weeks, four weeks, those are the sorts of periods we see in the papers?
Professor Graham Medley: But that’s – yes, but that is only possible with the relatively strong assumption about transmission in the coming –
Counsel Inquiry: Yes, so that – but just to help us sort of place these things in the –
Professor Graham Medley: Yes.
Counsel Inquiry: – hierarchy, that is clearly less –
Professor Graham Medley: You are correct and I’m highlighting the fact that the big difference is a strong assumption that you have to put in for something that hasn’t happened yet.
Counsel Inquiry: But it is still something in the nature of a prediction of what, assuming that that – with that strong assumption that things are going to carry on as they are, how you are able to give a projection of how you think the pandemic is going to develop in the next few weeks?
Professor Graham Medley: Yes. Although I don’t want it to be thought of as a statistical prediction –
Counsel Inquiry: No?
Professor Graham Medley: – in the same way –
Counsel Inquiry: As the nowcast?
Professor Graham Medley: As a nowcast.
Counsel Inquiry: Then something different again which we also see in the papers is a reasonable worst-case scenario. That is something completely different, it’s not a prediction of any sort, it’s a very assumption-driven model which doesn’t try to predict the way the pandemic is likely to develop at all, it is simply one course that the pandemic might take –
Professor Graham Medley: Yes.
Counsel Inquiry: – depending on the assumptions that are used?
Professor Graham Medley: Yes, although we have to be a bit careful with the language, because whilst it’s not a formal statistical prediction, and – it’s a set of scenarios of things that might happen, but the circumstances under which they might happen are really quite rare. On the other hand, they do have elements of prediction in them, so we’re not drawing random graphs. You know, they have –
Counsel Inquiry: Yes.
Professor Graham Medley: – the salient features that we think of a particular disease within them. You know? And it’s that nuance which I think has, you know, caused lots of problems, that – on the one hand if we say, “Well, they’re not predictions”, which they’re not, then the temptation is to say, “Well, they’re valueless then”. You know?
Counsel Inquiry: One way, you make this point in your statement Professor, as I understand it anyway, is that when we are dealing with the nowcast and also the medium-term projection, what is generated is, as it were, a single projection, with all the caveats you’ve described built in: the R number today is 1.5. Or: over the next three weeks we expect or we project the pandemic will do this or that.
So there is a single product which one is looking at, albeit understanding it with the appropriate caveats.
When one has a scenario, there is no value in looking at a single scenario, the whole purpose of scenarios is to develop a number of scenarios and look at them all together. Is that one – I think that may be really what you’re saying by saying if one just looks at the reasonable worst-case scenario, it’s valueless because you need to look at it with all the other scenarios that have been created and look at them together?
Professor Graham Medley: Generally with scenarios you’re absolutely right, the more that you can look – or looking at one has less value than looking at several. The reasonable worst-case scenario, though, is a slightly different thing, it’s more of a planning tool than it is an epidemiological modelling –
Counsel Inquiry: And I think you, certainly someone refers to the fact that, for example, some risk registers require there to be a reasonable worst-case scenario –
Professor Graham Medley: That is my understanding.
Counsel Inquiry: – for planning purposes.
Professor Graham Medley: Yes.
Counsel Inquiry: And perhaps that’s where some of the confusion arises.
Just a couple of sample documents produced by SPI-M-O. I mentioned earlier, just by way of example – well, first of all, let’s look at the minutes of a meeting, if we may, and I’m going to look at two documents, they’re not from the same meeting, but I don’t think that matters for these purposes.
Can we go to INQ000233688, please. This, we see immediately, is a meeting on – it’s quite late, it’s February 21, it’s on a Wednesday, I think it’s one of these plenary meetings we’ve discussed, would that be right?
Professor Graham Medley: Mm. We call it the main meeting.
Counsel Inquiry: The main meeting, I’ll remember that.
We don’t need to look at this in any detail, although I will come back to it in due course for one point. Just casting our eyes down, we can see the topic headings: forward look, R, growth rate, incidence, medium-term projections, restrictions and so on. These were not considered every week, although I think the R rate was something you looked at very regularly, but these were the type of issues that were addressed at your meetings regularly, Professor?
Professor Graham Medley: Yes, yes, and that agenda was, as I said, decided by myself, the co-chair and the secretariat.
Counsel Inquiry: If we look at the second topic, “R/Growth rate/Incidence”, we see there is reference there to the committee having discussed the issue and a consensus view having been agreed?
Professor Graham Medley: Yes, so the main meeting was every Wednesday morning. On the Tuesday afternoon we met as a group to discuss the reproduction numbers and the consensus, and essentially came to an agreement at that – that was a very technical meeting. I mean, typically the SPI-M-O meeting was three hours, the previous day’s meeting would often be two hours. Very technical discussion about the reproduction numbers.
That was set because we got a data drop on the Monday evening, so members had, you know, 24, 12 hours to do that analysis. And then the SAGE meeting would follow – the Wednesday, on the Thursday, then we would – I would meet with the secretariat and co-chair on Thursday evening and we would discuss the issues that had come up in that SAGE to go to the next, the meeting the following week.
Counsel Inquiry: Yes, so that was the pattern of things?
Professor Graham Medley: That was, yeah, the –
Counsel Inquiry: Then, just lastly on this, if we look at the bottom we can see what’s being sent up to SAGE, the consensus statement we’ve already discussed, and then the medium-term projections and some papers –
Professor Graham Medley: Yes.
Counsel Inquiry: – of the type we mentioned.
Let’s, if we may, look at a sample consensus statement. As I say, I’m afraid it’s not the precise one that’s referred to here, but I don’t think that matters.
Thank you very much.
So this is in fact earlier in the pandemic, it’s September of the year before, Professor, but we see here the first paragraph:
“SPI-M-O’s best estimate for R in the UK is between 1.2 and 1.5 …”
That’s the type of consensus that you reached week by week –
Professor Graham Medley: Yes.
Counsel Inquiry: – about the R number?
Professor Graham Medley: So the consensus – well, it differed on – for what it was we were talking about. For this reproduction number, which we ended up doing every week – so even when SAGE didn’t meet, SPI-M-O had to meet because we were producing this official government statistic, and was – actually came from a statistical combination. So many groups contributed estimates of the reproduction numbers in the different devolved administrations and the different regions of England. They were then combined statistically and the technical meeting was really to discuss why that was wrong, and so there may well be individual reproduction numbers where it hadn’t worked or it was completely out of kilter with the rest and we’d have long discussion about it and whether it should be included or not, and so having decided what was in and what was out, then these, this consensus is essentially a statistical combination.
Counsel Inquiry: Yes.
Professor Graham Medley: Other things, where you can’t do a statistical combination, the point of the consensus was to ensure – sort of cover the variability and cover the uncertainty in the views on the committee. So we didn’t – it’s not consensus as in agreeing a single thing, it’s consensus as in agreeing what the uncertainty was.
Counsel Inquiry: Well, I wanted to ask you about consensus, Professor, and let me come on to that now. Can we just look, for these purposes, at your statement at paragraph 3.9 on page 13. Perhaps you say there something similar to what you’ve just told us, Professor, I’m starting three lines down, you said:
“SAGE and its sub-groups developed a consensus in the sense that what was written was agreed by all not to be wrong.”
Just pausing there, that sounds a rather, if you like, reductionist way of producing a piece of advice.
Professor Graham Medley: So we weren’t producing advice. I mean, perhaps I’m wrong to include SAGE in this description, but SPI-M-O was very much about creating evidence, not advice, and the documents were written by the secretariat, the SPI-M-O secretariat, and myself and the co-chair then went through them, corrected them, changed them, discussed them, and, as I say, the point was that the documents emphasised that the – uncertainty. So we didn’t come to an agreement about what the evidence was, we came to an agreement about, if you like, what the evidence wasn’t, so that the statements that were included had to be something that captured the disagreement.
Counsel Inquiry: It sounds like the R number is actually quite a good example of this, because if amongst the committee there is a disagreement about whether, for example, the R number is in one place or another, the consensus statement would, as it were, sort of find the lowest common denominator, it would capture the range of views?
Professor Graham Medley: Lowest common denominator I think is wrong. It’s not the lowest common denominator, because at the time you don’t know what is correct. So you can have everybody agreeing with one piece of evidence except for one person. It would be wrong to dismiss the wrong person – the one person. You have to try to capture everything. So the reproduction number, I think that was a particularly narrow range that you just showed, of between 1.1 – was it 1.2 and 1.5? In other circumstances we have much wider ranges.
Counsel Inquiry: If on your committee you had a few people, expert, who took the view that the R number was quite low, say below 1 –
Professor Graham Medley: Yes.
Counsel Inquiry: And you had other people who disagreed with them, and thought that the R number was above 1 –
Professor Graham Medley: Yes.
Counsel Inquiry: – from what you’re saying, that would translate into a consensus statement that we are agreed the R number is between 0.8 and 1.4?
Professor Graham Medley: So what would happen on the Tuesday afternoon is that we would have a vigorous discussion about it, and because we had multiple groups, then – and because it was a relatively technical issue, then the person who is out, if there’s one person whose estimate is very different from the rest, then they would be challenged, and say: well, why are you getting that? And often it would be because they were looking at a different data stream. You know. Or they had made a different set of assumptions. And so if there was no reason to exclude it, then we would include it, and say: yes, there is uncertainty here, we – the estimate would come out as being, saying: well, it’s likely or highly likely the reproduction number is above 1, if that was, you know, the majority, but we can’t be 100% sure.
Counsel Inquiry: What this process doesn’t seem to generate, Professor, is a statement saying something like: we had a debate about it, some people thought it was 0.8, some others, perhaps more, thought it was 1.2, these are the reasons why each group took the view they did. Because of the need to reduce what is being –
Professor Graham Medley: Yes.
Counsel Inquiry: – said into a consensus. And you know that that is an outcome that has been criticised?
Professor Graham Medley: Well, I’m not sure what the criticism –
Counsel Inquiry: Well, let me show you. If we can look perhaps at Professor Woolhouse’s statement, which is INQ000250231 at page 7, and if we can look at paragraphs 35 and 36. You can see – it’s really perhaps 36, where he’s talking about consensus, he says:
“Another weakness is that on many issues there was not complete consensus at the time and reporting what was effectively the majority view might have given an impression of groupthink. In my experience, minority views were not always communicated to officials and ministers.”
So that’s the point. If there is a minority view, why not explain that there are different views on the committee, explain the pros and cons of the different arguments, so that those who are receiving the benefit of your expertise can understand and, if necessary, take up with those involved what the difference of view is, rather than reducing it to what Professor Woolhouse suggests is a sort of bland statement which doesn’t actually capture the disagreement on the committee?
Professor Graham Medley: Yeah, so I completely respect what Mark Woolhouse is saying, but in my view I didn’t feel that at all, otherwise I would have changed it. But I think the question then of how – so we’ll take a step back for a moment. Uncertainty is, I think, the – one of the core issues in terms of how science and policy work, because as scientists, you know, we have, I think, an understanding of what we mean by uncertainty, and I think that doesn’t always map directly on to policymakers, of what they’d understand by uncertainty. And of course uncertainty within decision-making is absolutely critical. And within the pandemic, you know, uncertainty was almost, you know, the biggest issue in the decision-making.
How we communicate that uncertainty I think is a very valid – as scientific subgroups, is a very valid question, and I think I was content with this consensus approach, but of course I think the people who need to be asked are those who received it: did it work for them? The consensus, the – for example, the scale we had in terms of SAGE papers, there’s kind of a likelihood scale ranging from almost certain not to happen, through unlikely, likely – plausible, likely, highly likely, to almost certain to happen. And again I found that very useful as a sort of extended traffic light system, but it’s, you know – the system wasn’t designed for my benefit, it was designed for somebody else to read.
Counsel Inquiry: Certainly not, and you do explain in your statement, exactly as you’ve said, that you felt that this idea of uncertainty within the scientific analysis may not have been properly understood by ministers, and do you think that it may be this consensus approach may be one of the reasons why it wasn’t properly understood?
Professor Graham Medley: Again, you’ll have to ask the people to whom we were communicating. A lot of my understanding about uncertainty and consensus – you know, and how we reached – how we defined it and used it, have come on in reflection. I think during the pandemic I was quite – it appeared to me to be a very sensible way of working, to capturing that uncertainty, in a statement rather than in a series of bullet points.
Counsel Inquiry: We certainly will, in due course, ask the people who received that, your materials, Professor, but even then, with hindsight, are you able to say that you think perhaps it might have been done a better way?
Professor Graham Medley: But I think an advantage of the consensus approach that we had, which I don’t think you can – should be minimised, is that the way in which it enabled the committee to function. I think if – given that it’s a group of academics, that if we had said, “Right, everybody write down your views”, then we would have ended up with 20 bullet points, because people would have found all sorts of nuances, we would have been emphasising the differences in the actual meetings. And part of my concern with a group of academics who are always competing with each other for grant funding and publications and data, was to get a cohesion, and so having a consensus approach whereby we had to agree on something was actually extremely useful.
I – the approach – running a committee or trying to get within those – these timeframes, capturing all the possible nuances I think would have been much more difficult.
Counsel Inquiry: Yes.
Professor Graham Medley: But of course I’m saying that’s running it from my point of view, not from the people who need it.
Counsel Inquiry: It’s a useful insight, Professor, and of course one of the things which Professor Woolhouse mentions there and which you mention, I haven’t touched on so much, is the delay aspect.
Professor Graham Medley: Mm.
Counsel Inquiry: And one of the concerns about consensus that’s been expressed is that it takes time to create a consensus and therefore it delays the evidence, but I think what you’re saying is that actually trying to capture a range of different views may be even more time consuming, because all the members of the committee would want their views to be represented?
Professor Graham Medley: I think that’s correct. And the other point to remember is that we are independent academics and we are having our meeting in front of the secretariat, who then produced these documents. So – and I as co-chair have input into the SPI-M-O document, but we are doing it for their purposes rather than for ours.
Counsel Inquiry: Let me move on.
Lady Hallett: Mr O’Connor, before you do, it looks as though we’re not going to finish the professor before the break.
Mr O’Connor: We’re not, no.
Lady Hallett: Can you be back this afternoon, Professor?
The Witness: I can.
Lady Hallett: In which case I think best, probably, to break now?
Mr O’Connor: Certainly.
Lady Hallett: Very well, I’ll be back at 1.50, please.
(12.48 pm)
(The short adjournment)
(1.50 pm)
Lady Hallett: Mr O’Connor.
Mr O’Connor: Professor Medley, I want to take you to a slightly different topic from what which we were discussing before lunch, and that is a more general consideration of the balance of the make-up of SPI-M-O, and in particular the question of whether it might have included experts from wider disciplines.
Let me start, if I may, by asking you to look at two paragraphs in Professor Woolhouse’s statement, and it’s page 4 of his statement, please, and paragraphs 18 and 19. We see here that Professor Woolhouse, in the first paragraph, makes the point that SPI-M-O was exceptionally well qualified in epidemiology and so on, and he says the group was fit for purpose with regard to quantifying the direct health harms caused by Covid-19, but it didn’t have appropriate expertise to assess, using models or any other form of analysis, the harms being done to the economy, nor harms to education, mental health and societal wellbeing.
He goes on to say that this “lack of attention”, as he puts it, to the wider harms was repeatedly raised by him and others during SPI-M-O meetings. He understood that it wasn’t the committee’s role but no one, DHSC officials or Cabinet Office observers, was able to say whose role it was.
Thank you.
Now, I think that you broadly agree with those comments, first of all in the sense that clearly those wider disciplines weren’t on your committee, and secondly you were told that it wasn’t your job to consider those wider harms?
Professor Graham Medley: Yes, not just broadly, I completely agree, and it was raised many times and particularly in regard to the strategy that the Government wished to pursue across the whole epidemic, which was an area in which modelling was underused, I think.
Counsel Inquiry: Yes. Can I just take you to one passage in your statement, please, and that’s paragraph 34 – sorry, page 34, and we see at the end of paragraph 3.80 you say:
“It is explicitly states in [minutes in] SAGE … [in] September 2020 … that the economic harms of interventions were being addressed outside of the SAGE structures.”
Then further down, if we can look at the next paragraph, you say something rather similar to Professor Woolhouse, about four or five lines up from the bottom:
“I was assured that the quantitative exploration of the impact of measures on the economy, education, mental health and societal well-being was being done.”
Professor Woolhouse said he never found out who was supposed to be doing that modelling or exploration. Did you ever receive an answer to that?
Professor Graham Medley: No, I didn’t. We raised it – as I say, we raised as a gap, or us not doing it, and our concern about who was doing it, several times. It’s included in some of our consensus statements. I notice that the one we produced in February regarding school closures, we state in that that school closures causes harms, but we are no experts in that area.
And … yeah.
Counsel Inquiry: Let’s just look at one more paragraph on this in your statement, and we need to go forward, please, to page 119 and paragraph 12.27. You are responding there to a question about the development of epidemiology economic models, so this is the broader type of modelling.
You say they do exist, they’re used to evaluate cost efficiency of things like vaccines and new medicines. And dropping down a couple of lines, you mention the fact that the impact of the epidemic was exacerbated by disparities in, for example, housing, access to healthcare and occupation, and that the epidemic increased those disparities. You then say this:
“This is a major gap in modelling and should be addressed.”
How, by whom?
Professor Graham Medley: So it’s a major gap in infectious disease modelling, not just in this epidemic but generally. We – at the moment decisions are made on the basis of averages, so – and I’m not aware that we were asked during the epidemic, but indeed have never been asked to model the outcomes within different groups. So if something is deemed to be cost-effective on average, you know, if it prevents – let’s say it prevents a thousand – there is a very simple example. Suppose you have two interventions and one saves a thousand lives, the other saves 500, then clearly the thousand is better, but if those thousand are all from one particular group, and the 500 are a much better cross-section of society, then maybe the thousand isn’t better. That’s not a question we can answer directly, that’s, you know, then a policy interpretation of the models, but we were never asked those questions, and as a consequence we’ve never developed the models. Now, the policy side, of course, can say that: we don’t ask those questions because you haven’t got the models.
So we need to address that, and I think it’s a subject-specific issue that we need to go to the research councils to make sure for the next pandemic we have those models available.
It’s not straightforward, but it’s perfectly possible to do, if the data are available, and we know what type of policy questions we might be asked.
Counsel Inquiry: Was this gap something that was discussed on SPI-M –
Professor Graham Medley: Yes.
Counsel Inquiry: – before 2020?
Professor Graham Medley: Yes, and I had actually published about it previously. I had a piece of work which was trying to address this.
We as modellers know of this gap, and we have been trying to address it, but we hadn’t solved it by the time the epidemic started.
Counsel Inquiry: Is it closer to being solved now?
Professor Graham Medley: A good question. I’m – not that I am aware of, no, but it might be.
Counsel Inquiry: It sounds as though it’s not being treated as an urgent –
Professor Graham Medley: Well, I know people have written grant applications, but they have to be funded. It’s not something that we can just do, it’s something that we as a community can try to address.
Counsel Inquiry: Going back to the point you referred to, you certainly say in your statement, and it makes sense, that it’s something that should be addressed between pandemics rather than during one?
Professor Graham Medley: Absolutely. And many of the issues that I raise in my statement are of that nature.
Counsel Inquiry: Thank you.
I want to move on to another related topic. It’s a theme of your statement, Professor, and in fact it’s something that Dr Wainwright gave us some evidence about at the beginning of the day, that, at least at various stages during the pandemic, you, as a committee, lacked sufficient understanding of government policy to be able to give them as much help as you would otherwise have been able to do.
Can you expand on that for us?
Professor Graham Medley: Yes. So I think there are two – two answers to that., the first of which is really a between-epidemic problem, and that is governments have – well, what the policy needs to do is to give some indication of what it regards as a better outcome. And by an outcome I mean over the whole epidemic. So from beginning to end.
So some idea in 2020 – we didn’t have vaccines and there was no guarantee that they would have arrived, in which case the epidemic might have lasted three, four, five years, we might still be in the epidemic now without the vaccines, but what would the decision-makers think of as being success over that period of time. Because that then provides the framework, the strategy for understanding what government is trying to achieve. It’s not up to us to determine what that is, because they’re value-based judgements as to whether the United Kingdom did badly or well, and personal perspectives in terms of whether the country did badly or well, and we have elected representatives to make those very difficult decisions for us.
Counsel Inquiry: Sorry, can I just add a thought, which is that it was apparent that the government had some policies. I mean, for example, we knew they wanted to save the NHS. Was that too broad a policy objective for you to work with?
Professor Graham Medley: Yes. And that was – only became – that was only stated early in the epidemic, but the pre-pandemic strategy was to have the epidemic in one wave, which has – the only advantage it has is that it’s over more quickly. Yeah? So that was the strategy. There were no other strategies.
That changed, then, as soon as the epidemic – or very quickly into the epidemic, and the strategies were much more short term, and modelling has a particular problem with a lack of discussion and sort of a lack of understanding what the policies might be, because we cannot make a policy-neutral model. So in the types of model that you talked about at the beginning, particularly for the scenario models but also for the medium-term projections, if we don’t know what government might do when or why they might do it, then we have to second-guess, we have to make it up. We can’t put nothing into the model.
So ideally what happens is that you have a discussion with the policymakers to come – to ensure that there’s shared understanding of what evidence is going to be most useful for them to make their decisions.
Counsel Inquiry: What I wanted to understand, and I think you’re helping us in this regard, is: is it a question of detail? Is it a question of temporal scope, long-term policies? Or perhaps it’s both.
Professor Graham Medley: I think it’s a combination of both. And the best way that I can describe it, I think, is the comparison in the epidemic. So up until January 2021 we didn’t really know what – what the government would have thought of as being a good 2020, you know, or – sorry, “good” is the wrong word, less worse. Epidemics are bad things, bad things happen in them. But what would the government have regarded as being a less worse outcome?
After January, from January 2021 afterwards, the Cabinet Office started to send people to the SPI-M-O meetings and we started to have much more of a dialogue. Now, this fence between within and outside of government still exists, but it’s possible to talk and signal over the fence, and we clearly understood, although we might not have been directly told, that the government would have regarded a success in the next six months by not having a major surge of infection. In which case we can say: well, do it slowly, take your – you know, don’t rush to open up immediately, remove restrictions slowly, and remove them in an ordered way so that the data analysis and the modelling can inform you what the impact of the previous decision was. And that was the roadmap: the data, not dates process.
Counsel Inquiry: I was about to say, that’s data not dates?
Professor Graham Medley: Yes.
Counsel Inquiry: And we can see that that was a process where there was no doubt a dialogue between the government and you which –
Professor Graham Medley: Yes.
Counsel Inquiry: – informed when the various steps –
Professor Graham Medley: And critically was that time period. So we had raised very early, and it’s in one of our consensus statements, that we – it’s not just us as modellers but data analysis needs a period of several weeks between making a change, a decision, and seeing it reflected in the data.
So there are several instances of measures being introduced by the government too late, so that you couldn’t evaluate them properly before the next decision had to be made.
So I think one of our – I think one of SPI-M-O’s sort of contributions to that roadmap was to say: spread out the decisions and tell us when you’re going to make it, because then when we do the modelling we can include that date in the modelling as the date at which things might change.
And I think the evidence that we produced for that period was much improved in terms of its relevance to the decision-making, and supported the decisions much more. And it wasn’t – much of a relief to me – it wasn’t then put front and centre to explain why the decisions were being made.
The amount of work done in that period was absolutely huge in terms of the members of SPI-M-O and their teams. You know, not just the people coming to the meetings but the teams of people behind them. But it was not presented as being: this is why we have to make this decision. And it was a much more ordered, rational period of time in which hopefully the decisions were better.
Counsel Inquiry: Last thought on this subject, Professor, we’re obviously at least partly thinking about recommendations for the future and how the next pandemic might be handled. The period that you’re referring to as being a much better period was, of course, a year into the Covid pandemic. How realistic is it, do you think, that were there to be another pandemic, or when there is another pandemic, the initial period, which will be one of inevitably great uncertainty, whether it would be possible for the government to provide the modellers and the scientists with that type of sort of calm, thought-through strategic information at the very start of a pandemic as opposed to well into it?
Professor Graham Medley: Well, the time to do it is now. I was disappointed that as soon – you know, when we reached February 2022 the policymakers who we had been interacting with were disbanded and – you know, almost immediately. I would very much like to have gone with them, with that group, gone back to February 2020 and re-gone through that process.
I mean, critical in that was Dame Angela McLean, the policy co-chair, who I suspect had brokered that interaction. Because it’s a risk for government officials to come and talk to people, so – even though it wasn’t direct talking – to come and meet with people outside government and discuss policy options. And so – but that is absolutely critical, and I’ve mentioned in the report we had sort of – Dame Angela McLean also managed to get two people into Cabinet Office, so people from SPI-M-O who worked in Cabinet Office, just to keep the – that dialogue flowing. When I say dialogue, I don’t necessarily – I mean information rather than words. So that we are able, as modellers, to understand what government wants to try and achieve, so that we can support their decisions to achieve that. Or, of course, we can tell them: no, that’s impossible.
Counsel Inquiry: Yes.
Professor Graham Medley: Yeah.
Counsel Inquiry: We heard that you had resigned the chairmanship of SPI-M earlier this year?
Professor Graham Medley: Yep.
Counsel Inquiry: Are you in fact still a member of the committee at all?
Professor Graham Medley: No, I’m not.
Counsel Inquiry: Regardless, the type of work that you’re describing, the sort of between pandemic preparation work, is that something that SPI-M ought to be involved with?
Professor Graham Medley: We ought to be, but, you know, much to my disappointment it’s not happening, or it wasn’t when I was there. And, you know, we – actually, the first piece of work that we came back to on SPI-M was that piece of work that we hadn’t finished before, which is this pre-purchase of vaccines. So the Department of Health is asking whether it should invest a large amount of money to buy vaccines for influenza for the next pandemic.
Well, the answer to that question very much depends on what government would do. Because if the government was prepared to stop the epidemic with a lockdown at the next time, then it changes, you know, the amount of investment now that’s worthy of putting in. But we don’t have any idea of what that is, and that policy landscape is still developing. So UKHSA, Department of Health, how SPI-M feeds into it is all still there, but it needs to have – or we’re not quite clear what those relationships are, but it needs to have input from the centre of government, because if it gets bad enough that you need Cabinet Office policymaking, then Cabinet Office should be involved now to discuss what that’s going to look like.
Counsel Inquiry: Thank you, Professor.
I want to change focus again and in fact take us back to the beginning of the Covid pandemic, early 2020. We’ve heard your evidence about SPI-M, the I standing for influenza but transporting itself across to infection at that time. But, as you’ve said, a committee that, historically at any rate, had very much focused on influenza planning.
Professor Graham Medley: Yeah.
Counsel Inquiry: Let’s look, if we may, at another passage from Professor Woolhouse’s statement.
It’s on page 44, paragraphs 243 and 244, I think. Yes, thank you.
So 243, yes, he says:
“An important question in those early stages was whether a Covid-19 epidemic would be more influenza like or more SARS-like. The UK response initially assumed … an influenza-like event. This was not unreasonable but, in my view, it took looking to recognise that Covid-19 had many similarities to SARS: we knew from early January that [it] was closely related to SARS. Compared with influenza, Covid-19 had a higher R number and had little impact on children while being highly dangerous to the elderly, frail and infirm.”
Do you agree with that so far?
Professor Graham Medley: Well, I’m not sure what he means by “early stages”. I mean, the timeframe is – generically I agree that pre-pandemic preparedness was influenza, and I have – if you look at the cover of the SPI-M modelling summary, it spells out influenza as the I. So – and SARS – so this was transmissible SARS, but I think I first used “transmissible SARS” as a phrase in February, so I think we realised quite quickly that this was not influenza. That is important in public health considerations, but in terms of the modelling was not critical, I don’t think.
Counsel Inquiry: You’ve already stated in the course of your evidence today, and as I think we know, that in early 2020 there really was only one policy that the Government had, it’s the delay, contain, mitigate. You’ve also referred to the idea of just letting the epidemic take its course. And you say quite straightforwardly in your statement that that turned out to be outdated.
Two questions –
Professor Graham Medley: Sorry, it would have been outdated for influenza as well, because influenza still would have had the same kind of health impact.
Counsel Inquiry: Let’s just focus on Covid, if we may. The two high level questions that arise, Professor, is firstly, and that’s really the point raised by Professor Woolhouse, whether the UK pivoted away from its pre-existing influenza-based plans too slowly in early 2020, and the related question is whether SPI-M-O or SAGE should have done more to alert policymakers that there was a problem with the influenza plan, or was it that the policymakers themselves weren’t tasking SPI-M-O and SAGE appropriately?
Let’s take them in stages. Do you think that the pivot from the earlier contain, delay, mitigate strategy towards suppression happened too late?
Professor Graham Medley: So that question of – which is why I pointed out that the – that plan would have been inadequate or, sorry, wouldn’t have … given the preferences that the government subsequently showed, that they didn’t want to have the epidemic in one wave, that would have been true for influenza as well. So the realisation that there would be a large number of hospitalisations and deaths, would mean that they were going to stop the epidemic, would have to have occurred with influenza. Yeah?
Counsel Inquiry: I don’t want to get into a debate with you about a pandemic that didn’t happen, Professor, but vaccines might have played a part in influenza –
Professor Graham Medley: But the virus that was causing the epidemic wasn’t the issue. The issue was that in the plan these large numbers of – you know, hundreds of thousands of deaths were just going to happen. It was going to be a natural event that the country would get through.
Counsel Inquiry: All right.
Professor Graham Medley: Yeah?
Counsel Inquiry: Yes.
Professor Graham Medley: The question of suppression, of stopping the epidemic, yeah, is – was clearly made, but that wasn’t because it was coronavirus, presumably the government would have made that same decision had it been pandemic influenza.
Counsel Inquiry: So I think what you’re saying is that the reason why there was a delay in changing, in pivoting from one strategy to the other was really the fact that policymakers themselves had previously signed up to the earlier policy, and it was only when they decided on a change of course that a new policy was adopted?
Professor Graham Medley: Yes. And that policy, I’m not sure where it originates from, but, you know, clearly the world and society has changed a lot in the intervening period. And the ability to stop it. You know, I mean, I don’t think ten years ago it would have been possible to have the kind of lockdowns that we had.
Counsel Inquiry: Can we look at another passage of your statement, please, Professor, it’s page 18, and I want to look at paragraphs 3.27 and following. Picking it up at the end of paragraph 3.27, please, you say, last sentence:
“My view is that the reality of the epidemic, via the SAGE process, was not given sufficient weight initially.
“3.28. There was a sense that [the] government strategy was being created ‘on the hoof’ during February and March …”
Just pausing there, are you here referring to some similar points to those we were just discussing? What do you mean by policy being created “on the hoof”, Professor?
Professor Graham Medley: I mean the lack of a plan. It wasn’t clear what the plan was going to be. If they were not – government were not going to follow the pre-pandemic plan, and clearly at that point, I don’t think – it depends when I was writing this, but I never got a sense that not having a single wave was something that the government was thinking about. I had a conversation on 4 March which convinced me, you know, very quickly that even if that was the government plan they would change their mind. And so it was really, at that period of time, was a sense of, from my point of view, the lack of strategy, of the lack of decision-making under uncertainty, the things that might happen in the future and what we would do at that point.
Counsel Inquiry: I suppose not lack of strategy, because I think what you’re telling us is there was the strategy, there was the single wave strategy?
Professor Graham Medley: I realise the inconsistency in what I’m saying, yes.
Counsel Inquiry: But were you expecting them to change course but that call didn’t come or –
Professor Graham Medley: I thought they would, and that was partly, of course, from the international perspective, in the sense that we had already seen, you know, the first country, China, close the economy rather than face the epidemic. And even from that perspective you thought, well, that might well be the response.
Counsel Inquiry: So, and this is what you mention at the bottom of the page we’re looking at, you talk about China’s response being to “close their economy rather than face the epidemic and its consequences”. You say you thought it was “likely that generalised NPI would be a widely used intervention in the pandemic”, and that what happened in Italy confirmed your view.
So looking at – focusing in on Professor Medley and SPI-M-O in, let’s say, February 2020, it sounds almost as though you were waiting for the government to ask you about lockdowns or to issue you with some tasking saying, “We don’t think we can bear to take this wave, what else can we do?” But that call didn’t come. Is that what you’re saying?
Professor Graham Medley: Yeah, we were doing things, we were looking at the impact, and we were particularly asked about schools. You know, much to some people’s frustration, because – some members’ frustration, because, as Professor Woolhouse has already pointed out, that relatively early, in February, having realised that this was more SARS than influenza, then care homes would be more of a problem than schools. So – nonetheless the government was asking us about the impact of closing schools, and, you know, that is the only NPI, really, for which we had prior information. We were not being asked about alternatives to the strategy.
Counsel Inquiry: No. Were you, either on SPI-M-O or on SAGE, encouraging the government to think about alternative strategies?
Professor Graham Medley: Regardless of the – or under most strategies that you would think of, then putting in place NPIs and, I mean, restrictions such as – or guidance such as asking people with clinical disease to stay at home, so individual isolation, would have to be done, and my frustration at that point was that those were not being put in place early enough, because we didn’t know whether they would work.
So coming back to this point about sufficient time and data to enable – see interventions working, the doing nothing and then suddenly changing your mind, which is to some extent what happened, was perhaps where I got the – you know, some of that confusion from, that if – regardless of what the outcome you want to achieve is, putting those kind of interventions, individual isolation, in place sooner would have seemed to be more sensible.
Sorry, does that answer your question?
Counsel Inquiry: I want to stay on this subject but move on to a slightly different issue, which is the NHS, and which – as we know, the cause of the NHS, whether it would be overwhelmed or not, became a key issue.
If we look at paragraph 4.9 of your statement at page 43 –
Professor Graham Medley: Just to point out whilst that is coming up, in the pre-pandemic preparedness plan, hospitalisations are not mentioned.
Counsel Inquiry: No. Let’s just look at this paragraph, Professor. It’s the second sentence:
“Throughout February 2020 it became increasingly clear …”
And I take it you mean – well, is that clear to you, clear to SPI-M-O?
Professor Graham Medley: Yes, to SPI-M-O.
Counsel Inquiry: “… that NHS capacity in the UK would be overwhelmed.”
Professor Graham Medley: Yes.
Counsel Inquiry: And you say that SAGE asked a working group be set up to discuss the extent of the overwhelm.
Now, we may hear from those, for example, in Number 10 that this prospect of the NHS being overwhelmed wasn’t something that at the very least they adverted to until a couple of weeks later than the end of February, in mid-March.
But are we to take it from this that – not the possibility but the certainty of NHS capacity being overwhelmed was something that was clear to you and to SPI-M-O during February?
Professor Graham Medley: So the extent of the epidemic became very clear during February and so I think at that point we were, you know, the – we’d essentially established the infection fatality rates, so that’s the proportion of people who are – die following infection, at about 1%, big variation with – especially with age, but – and if 80% of the population becomes infected in a single wave, then you can calculate the numbers of people who would die.
The question, then, of hospitalisation is slightly more complicated, because of course it depends on who goes to hospital under those kind of circumstances.
So it needed a meeting with the NHS and with clinical colleagues for them to say what symptoms would justify admission into hospital, because that is a variable, it’s a clinical variable that if somebody goes to hospital then they might not be admitted, and that – those kind of clinical questions determine the extent to which the epidemic results in hospitalisations, for example.
Counsel Inquiry: Well, Professor, just looking at this statement, and maybe you want to qualify or change it, looking at the statement, what you say is it became increasingly clear that capacity would be overwhelmed, and all you were doing with the NHS was talking about the extent of the overwhelm?
Professor Graham Medley: I think that’s right. I mean, throughout – I mean, yes, I stand by the statement, throughout February. What I mean is that, in terms of being able to make a clear statement, it was absolutely clear the expected deaths. Hospitalisations are slightly, you know, more uncertain because it requires more clinical input.
Counsel Inquiry: I want to have a look – sorry, Professor.
Professor Graham Medley: I don’t think it’s recorded in the minutes, but just to make it clear, so Dominic Cummings’ iPhone X, for example, attended all of the SPI-M-O meetings, even those, I think, pre-pandemic. Whether it’s recorded in the minutes I’m not sure, but, we – as I said, there was a whole host of people phone – people phoning in from across government. So even if it might not be in the paperwork, but I’d argue about that, it was known.
Counsel Inquiry: Let’s have a look at what is in the paperwork, Professor, and I want to look at the minutes of two SAGE meetings at the end of February.
So first of all for 25 February – thank you – we see a SAGE meeting. Though this is pre-lockdown, so it looks as though it was held in person, although there may well have been people dialling in. We see that you were there.
If we can go over to the next page, please, we can see that the measures to limit spread are referred to. And if we look at paragraph 13, it says:
“Any combination of measures would slow but not halt an epidemic.”
Then this:
“NHS needs must be considered in any decisions to alter the epidemic curve.”
Is that consistent with a view that you’d reached that NHS overwhelm was at that stage inevitable?
Professor Graham Medley: I think so, yes. Remember these documents are written by civil servants for civil servants –
Counsel Inquiry: Well, just pausing there, Professor, we may hear that SAGE minutes are written for policymakers.
Professor Graham Medley: Who are civil servants.
Counsel Inquiry: Well, they are politicians, Professor.
Professor Graham Medley: Well, I always think of them as decision-makers.
Counsel Inquiry: I’m not going to quibble with you about language, Professor, but I want to press you on the rather more important point, which is whether the sentence “NHS needs must be considered in any decisions to alter the epidemic curve”, does that really mean, does that sentence signal a conclusion that it is inevitable that the NHS will be overwhelmed in the approaching epidemic?
Professor Graham Medley: It certainly indicates that the NHS is a major factor.
Counsel Inquiry: Well, that’s not the same thing, Professor, is it?
Professor Graham Medley: No. No. I mean, I don’t remember that – as I said, I went to many meetings over the epidemic, I don’t remember that particular meeting.
Counsel Inquiry: Let’s have a look at the next one, Professor. It’s a meeting in fact only two or three days later, two days later, so it’s the 11th meeting of SAGE.
It’s a different document. INQ000213175.
Professor Graham Medley: No, this is a paper I wrote.
Counsel Inquiry: I may have the wrong …
(Pause)
Counsel Inquiry: We may have to come back to this. Just bear with me one moment.
(Pause)
Counsel Inquiry: We can take that down, that’s the wrong reference. Yes, could we have INQ000106129 on the screen, please.
There may be a problem with the document. I tell you what I’m going to do, Professor, I’m going to read out – there was a SAGE meeting a couple of days later – I think there is a problem with getting the document on the screen, but I will read out the entry. In fact it’s the same entry that you refer to in your statement, but what it says is:
“UK academic modelling groups, Imperial Oxford, London School of Hygiene and NHS planners to organise a working group in the week starting …”
I think we’ve got it on screen – yes.
So if we go over to the bottom of the second page, please, thank you. Yes, this was the passage I was reading.
If we see the very bottom:
“UK academic modelling groups … and NHS planners to organise a working group in the week starting 2 March [so the next week] to analyse key clinical variables for reasonable worst-case scenario planning for the NHS, for review by SPI-M and then discussion at SAGE.”
So on the one hand, Professor, we have your statement which says that during February it became clear that the NHS would be overwhelmed, and then we have these two sets of minutes at the end of February, neither of which refer to an overwhelming of the NHS. Here what is referred to is reasonable worst-case planning. So, as we’ve already discussed, it’s a scenario, it’s not a prediction. How can we reconcile those two things, Professor?
Professor Graham Medley: So at the beginning of February it wasn’t clear at all. By the end of February SAGE was asking for quantitative details. So that meeting happened on 1 March, I think. It was asked for in SAGE, so by 1 March we had the – a good understanding of what the impact would likely be on the NHS. So that’s, you know, that’s throughout February.
Counsel Inquiry: Professor, from a lay perspective, and of course with hindsight, but if you and your colleagues had already realised during February that the NHS was going to be overwhelmed by the pandemic that was developing, why didn’t you just say so?
Professor Graham Medley: Well, we – I mean, we – the secretariat that we were talking to completely understood. The SAGE secretariat were taking the minutes. I didn’t write these documents, the documents were written by somebody else for somebody else, they’re not written for my benefit, and it’s really the writers and the readers that you need to question. I was – at the end – by the end of February it was clear, I think, that the NHS would be overwhelmed, and I don’t think that was being kept a secret.
Counsel Inquiry: So the situation you’re describing is a meeting, a SAGE meeting, where everyone round the table is discussing the fact that the NHS is going to be overwhelmed –
Professor Graham Medley: Yes.
Counsel Inquiry: – but the minutes record simply that there needs to be some reasonable worst-case planning for the NHS?
Professor Graham Medley: Yes.
Counsel Inquiry: I see, thank you.
I want to move to a related document, and that is, please, INQ000129093. I hope this one is right. Thank you.
This is – Professor, I think you’ve seen this document before, it’s a report or a note written by someone called David Halpern, who was a Number 10 civil servant, a behavioural expert who will be giving evidence to the Inquiry in two or three weeks’ time.
We can see it’s dated at the top September 2020, and he’s looking back, is he not, at the early days of the pandemic? It’s a sort of a – well, it says, it’s a lessons learned document.
If we can turn over to the second page, it’s the section which is headed “The Early misstep” that we’re interested in, and it’s in fact the next page where we – the particular point, but I’ll just introduce it, if I may.
You can see he says:
“Arguably the most fundamental misstep in the UK response was the presumption that Covid would be an unstoppable flu-like wave.”
He says, the next paragraph:
“It is important to see that this presumption was not based on ignorance, but on a century of prior knowledge and assumptions.”
Then the next paragraph, he refers to the mass expertise which he has described converging on the conclusion that once early containment had failed, a flu-like wave was inevitable.
If we can go to the next page, he carries on referring to this, as it were, received wisdom, which he perceived in any event, and he describes it as being a block to any suggestion that there might be an alternative policy of suppression.
If we can pick it up at the bottom of this page, please, so the paragraph – he says:
“If I had to choose a single issue – and moment – that embodied this failure, I’d point to the unshakeable conviction of the SPI-M modellers that suppression – the sustained holding down of Covid prevalence – was not a viable strategy. For example, Graham Medley, chair of SPI-M, was asked in SAGE in the second week of March, along with the other modellers: ‘how certain they were that major second waves would arise in China and other Asian nations?’ (ie that suppression was not a viable strategy). Medley answered immediately, and with total conviction: ‘as close as 100% as possible…yes, 100%’. His colleagues echoed the conclusion ‘yes, 100%’. They were totally convinced that as soon as the harsh lockdowns in Wuhan, South Korea and elsewhere were lifted, cases would immediately surge again.”
And he observes:
“Nothing in science, and certainly nothing in statistical modelling, is 100%. Let alone in the face of the data already emerged from the Asian experience. This was doctrine, not science.”
Professor, is it right that you believed, as it were, as a matter of doctrine, that suppression wasn’t viable in February and March 2020?
Professor Graham Medley: So epidemics have happened throughout history, and the study of them through models is a science, scientific discipline, so it’s absolutely not doctrine. And I think that we were correct that – there were surges. You know, the virus did come back in every country in the world. It was a question of how long they suppressed it for rather than whether they suppressed it. And in terms of health outcomes, clearly suppressing it until a vaccine was available and getting the vaccine into the people was critical in determining the outcome of that. But every country has had a subsequent epidemic, and to some extent, you know, the – so what is said here is correct, but it’s out of context, because I think the – if I remember this particular discussion, it was, you know: will having a lockdown solve the problem? And the answer to that is: no, it will not solve the problem, because it will continue to be a problem. And if you go back to what you were doing previously, you – it will re-occur. Which of course it did in the United Kingdom.
Counsel Inquiry: It did, but, you see, the debate we’ve been having for the last ten minutes or so is trying to understand what was happening in that time.
Professor Graham Medley: Yeah.
Counsel Inquiry: And why it was that lockdown doesn’t seem to have been considered during February or even in early March.
Professor Graham Medley: Yeah.
Counsel Inquiry: And what you were telling us was that you were, as it were, waiting for the government to ask you about it, but they didn’t. But I wonder whether actually it’s possible that you and your colleagues didn’t think it would work and so weren’t pressing for it and that was at least one of the reasons why it wasn’t on the agenda?
Professor Graham Medley: So – well, we didn’t know if it would work, but even if it worked, it doesn’t resolve the problem in terms of the longer term. You know, because unless you continue that suppression, then it will come back.
Counsel Inquiry: Of course if, as you say, you had – it was very clear to you that the NHS would be overwhelmed –
Professor Graham Medley: Yes.
Counsel Inquiry: – then that puts a rather different perspective on whether there needs to be a lockdown, doesn’t it?
Professor Graham Medley: Potentially. I mean, you’re asking me about perspectives. My role and the role of the models is not to provide those kind of decisions. Our job is to give the evidence, or was to give the evidence.
Counsel Inquiry: Yes.
Professor Graham Medley: Now, I completely understand what you’re saying, is that had it become – had we made it clear or had policymakers known what the potential for a single wave epidemic was, that they would have been asking much earlier. I don’t think that they could not have known. But I might be wrong.
Counsel Inquiry: What was, as you have said, known to be certainly going to happen to the NHS, wasn’t a question of modelling at all, by that stage, was it? It was a statistical prediction. It wasn’t a scenario, to use that binary approach we were discussing. On your analysis, you knew that was going to happen.
Professor Graham Medley: Yeah, and we had already seen the policy response to it internationally.
Counsel Inquiry: But it wasn’t flagged up in the SAGE minutes?
Professor Graham Medley: No, but then nor was it – we asked about, you know, what – tell us about lockdowns in terms of policymakers.
Counsel Inquiry: Professor, that –
Professor Graham Medley: There wasn’t that dialogue going on, as I pointed out before.
Counsel Inquiry: Thank you. I’m going to move on. I’ve got a few more slightly shorter topics to deal with, with you, and the first of them is about – is to do with nosocomial infection after the first lockdown.
If we can start, for these purposes, with paragraph 6.6 of your report, which is on page 55.
So here, Professor, you address the question of transmission within LTCFs, which we know is long-term care facilities, I think.
Professor Graham Medley: Yes.
Counsel Inquiry: We can call them care homes, nursing homes.
You refer to an email that you sent to Patrick Vallance in April 2020, do you not? Perhaps we can call that up on screen, and I think the best number to do that with is INQ000 – well, no, well, shall we try INQ000260625.
What we see here is an email exchange. Let’s look first at your email to Sir Patrick Vallance, or Patrick Vallance as he was then. So we see it’s dated 17 April of 2020, and looking at the second paragraph, you talk about two conclusions being relatively robust. The first is that the number of deaths in hospital is less than 50% of all deaths. Secondly, that deaths at home and in care homes are not plateauing.
Then looking at the next paragraph, second sentence:
“Consequently my reading of the situation is that we have widespread ongoing transmission in the health and social care systems. Hospital and community health and social care appear to be driving transmission, and potentially at an increasing rate.”
Then you make this rather striking observation:
“In effect, this is the opposite of shielding – vulnerable are being preferentially infected.”
Can you expand on the observation you were making?
Professor Graham Medley: So this is 17 April, so lockdown has been going for three weeks, and we’ve seen the hospitalisation rate and the death rates turn over, much to everybody’s relief, that government could introduce measures that would stop the epidemic. So this is a relative – this is a dataset looking at deaths. Now, I’m not sure that the data, when they were fully unwound and looked at, actually supported the conclusion, the worry that I had in this email.
Counsel Inquiry: Right.
Professor Graham Medley: But clearly we had stopped transmission – my feeling at this time was that transmission within the community had been stopped, or reduced greatly, so that this reproduction number was less than 1, so the prevalence in the community is falling. But the deaths in care homes and in the community were continuing to rise. And the care population, which it turns out we knew very little about, so the people receiving care – there are more people in residential and informal care than there are in care facilities, and others will know more about that than I do, but this rise in deaths of people at home and in care homes was concerning to me because it meant that we were effectively closing schools and locking down on one hand, and being successful, but there were epidemics in the care sector and the health sector which were continuing to rise. And that’s, of course, exactly what you don’t want to happen.
Counsel Inquiry: It’s what you don’t want to happen and it’s what we now know at around this time was happening?
Professor Graham Medley: Yes.
Counsel Inquiry: So here you are sounding the alarm to Patrick Vallance. If we can go back to –
Professor Graham Medley: Just to come back to another point, I mean, this is part of the difficulty – you know, the situation I was in as an independent – so if somebody from within government has sent me a document that he’s described as not for sharing, I’ve then shared it potentially with somebody else inside government, I mean there are all sorts of boundaries going on partly because of this problem that I was outside but …
Counsel Inquiry: Well, you obviously thought it was the right thing to do, to –
Professor Graham Medley: I did.
Counsel Inquiry: – to share your concern with –
Professor Graham Medley: I did.
Counsel Inquiry: – the Chief Scientific Adviser.
Professor Graham Medley: Clearly.
Counsel Inquiry: And I want to ask you about his response, please. So if we look at the very top of the page, it’s a short response, he says:
“Thank you for sharing [it]. I think that’s what we’ve been driving at in SAGE and I will reinforce again with the accountable departments that this is a very, very key area that they need to get on top of.”
Do you know, can you help us any further with what –
Professor Graham Medley: No, I –
Counsel Inquiry: – Patrick Vallance did about that?
Professor Graham Medley: Patrick Vallance was the most senior email I had, so I had told the person the most – the highest up the chain, we had been talking about it at SAGE and SPI-M-O, we’d discussed this possibility previously, and, you know, he presumably did something about it, but I didn’t hear any more.
Counsel Inquiry: Right. That was on 17 April. If we could go to another document, please, it’s INQ000213298.
This, when it comes up, we will see is a consensus statement from your committee, SPI-M, SPI-M-O, three days later on 20 April.
If we look at the second paragraph, the first paragraph talks about hospital transmission, and then the second paragraph says:
“This estimate does not include people who acquire infection in hospital, leave (either because they are discharged, or because they are outpatients) and are then readmitted with COVID-19. This requires urgent investigation.”
So again, the nosocomial infection. Was an urgent investigation undertaken?
Professor Graham Medley: So I contacted a scientist who I knew working in PHE, who had – was – had some expertise in nosocomial infections. I was also – I think about this time the SAGE nosocomial working group was set up, so there was urgent investigation of it, and work has subsequently been published looking at the impact of this particular phenomenon.
Counsel Inquiry: And this issue was exposed, over time at least.
Yes, thank you, we can take that down.
Finally, Professor, I want to ask you, I hope quite briefly, just about a few unrelated matters. The first of them is the resignation of Neil Ferguson from SAGE, something we all remember, and if we can perhaps call up on screen INQ000267746. If we hadn’t remembered, I hope this is the document which will remind us of that event.
We can see it’s a BBC News article from 6 May 2020, so only a couple of weeks, in fact, after those emails we were just looking at. It refers to him quitting what’s described as his “government role” after “‘undermining’ the lockdown”. We see there’s a reference to him having quit after admitting “an error of judgement”.
If we can go over to the next page, please, we see the Health Secretary, Matt Hancock, quoted as saying it was extraordinary that Professor Ferguson – but he had taken the right decision to resign. Then he says, this is Mr Hancock saying it was “just not possible” for Professor Ferguson to continue advising the government.
We read on, possibly with some degree of irony, that Mr Hancock said that the social distancing rules were there for everyone and were deadly serious, and there is a reference to Scotland Yard having made a statement as well.
If we could go over to the next page, please, page 3, the BBC correspondent refers to Professor Ferguson’s resignation as being “a really big deal”, says he is “the most influential scientist” in the virus outbreak apart from Chris Whitty and Patrick Vallance.
Then further down the page there is a Sir Robert Lechler, said he didn’t think that Professor/’Ferguson’s resignation would “have any material impact” on the work of SAGE. He says that although Professor Ferguson made an important contribution, he is sure the group would continue to provide valuable support.
The point I want to ask you about, Professor, is that although, as we see here, Professor Ferguson on this occasion resigned from SAGE, he in fact remained a member of SPI-M-O, did he not?
Professor Graham Medley: He did.
Counsel Inquiry: We can look at an email from you to Paul Allen and Angela McLean, in fact I think it was the day before that BBC report, but obviously very much at the same time. If we can look at the bottom half, please, we can see you in the second paragraph saying that you are quite keen to keep him on SPI-M, he is a modelling expert, you don’t want to lose his expertise. You say that his input into forming SPI-M consensus is greatly valued but then you say his presence might damage the science and SPI-M and there is a reputational risk, and then you refer at the end to the fact it’s a DHSC decision.
So we know he did stay on SPI-M. We can see in this email that you understood the tension perhaps between the statements that were being made publicly about Neil Ferguson not being an appropriate person to advise government, but your desire to keep him on the committee. Presumably that’s what you mean by the reputational risk?
Professor Graham Medley: Yes, well, and I saw SPI-M’s role as mainly providing evidence rather than advice. You know, this process of creating evidence that then gets transformed into advice to be given to decision-makers.
Counsel Inquiry: As a matter of –
Professor Graham Medley: SPI-M-O is one step further away from decision-makers.
Counsel Inquiry: As a matter of transparency and public confidence, Professor, did you think that perhaps it was important that the public should understand that, notwithstanding the public statements that had been made, although he was resigning from SAGE he was in fact continuing to be a regular member of SPI-M-O?
Professor Graham Medley: So the reason why we were all doing this is for the public good, because we wanted – I firmly believe that modelling is essential and important within an epidemic, and we wanted the government to make the best decisions it could, because, you know, we’re living through it as well.
I greatly value Neil Ferguson’s professional expertise. I think he is, you know, undoubtedly one of the leading modellers internationally, and wanted SPI-M-O to produce the best evidence it could, and therefore was making clear to my policy co-chair and the head of – Paul Allen’s the head of the SPI-M-O secretariat, that that was what I felt. But on the other hand, you know, it wasn’t – it’s their committee, not mine.
Counsel Inquiry: Exactly. And do we therefore assume from what you say that in the end the decision that he should stay on SPI-M-O was not your decision, but do you know whether it was –
Professor Graham Medley: He didn’t come for many meetings, I can’t remember how many, but then he reappeared.
Counsel Inquiry: Do you know whether it was a decision taken by Mr Hancock or not?
Professor Graham Medley: I have no idea how far up that chain it went, but I was very glad for two reasons. One was because the – it’s not only Neil Ferguson himself, but, I mean, he has a large group with a national investment in it to provide that – this kind of evidence, and I was also worried about the kind of morale on the committee at the time, and, you know, felt it important that his expertise was recognised in order to keep the committee functioning.
Counsel Inquiry: Yes. Thank you.
I want to ask you about a separate matter now, and in fact this involves going back to the consensus – sorry, no, the agenda, the SPI-M-O agenda that I showed you before lunch, so that’s INQ000233688, please.
You’ll remember looking at this earlier. One of the agenda items I didn’t refer you to earlier because I knew we were coming back to it, it’s the fifth one down, paragraph 8, Long Covid.
Now, we noted earlier that this is a meeting quite late in the pandemic, it’s February 2021. Can you tell us how it came to be that you were discussing Long Covid on that occasion?
Professor Graham Medley: Thank you. So you alerted me to this. I looked it up yesterday. I was trying to remember or trying to find out whether this was members suggesting we talk about it or the secretariat asking us to talk about it, and I couldn’t find out which, so presumably it was done in a phone conversation rather than by emails. But we knew from the outset that there were likely to be post-infection sequelae, I think clinically they’re known as, the consequence of infection. Principally I remember discussion about whether or not Covid increased susceptibility for bacterial infections. That turned out not to be true. But without formal case definitions and good data we can’t include it explicitly in the models, which is really the conclusion of this discussion. Clearly we can and we do include infections, so we were always talking about infections, but there is little we can do in terms of modelling to enhance the data and the information that’s available.
Counsel Inquiry: Does it follow from what you’ve said that following this discussion there wasn’t any modelling of Long Covid, at least at that stage?
Professor Graham Medley: Yes.
Counsel Inquiry: Has there in fact to your knowledge ever been any SPI-M, SPI-M-O modelling of Long Covid?
Professor Graham Medley: Not to my knowledge. And until we, for example, have case data and good data then it wouldn’t be possible.
Counsel Inquiry: You may not be able to answer this, in which case say so, but from your understanding about what is known about Long Covid now, is there enough data, enough understanding to model it if one wanted to?
Professor Graham Medley: Well, because, as where we started, that models are a combination of data and assumptions, so if you wanted to put some assumptions in you can model anything. If you want to be able to provide policy-specific evidence from a model then the data do not exist to be able to include Long Covid in a model that actually produces realistic numbers.
Counsel Inquiry: Thank you.
Just one or two more short topics. The first of those is to do with interactions with the media by members of your committee. We will all, I’m sure, recall during the pandemic hearing about the latest announcement by the government or the latest decision by SAGE or one of its committees, and then perhaps hearing one of – a scientist who may or may not have been on that committee, but may have been, expressing their own view, and the broadcasters were always very careful to introduce the person as expressing his or her own view rather than the view of SAGE or SPI-M or whichever committee it was.
Do you have a view about whether – well, perhaps I’ll ask the question in this way: can we have a look at a document, please.
It’s INQ000102129.
This is a series of WhatsApp messages, Professor. So they’re WhatsApp messages from Matt Hancock’s phone, and we can see it’s June 2020. It starts with Boris Johnson saying:
“These sage geezers now saying we should have gone into lockdown earlier … can we gently ask them why they did not make their anxieties public at the time???”
Then Patrick Vallance, relevantly for our purposes, says:
“I think there is too much enthusiasm for the camera at the moment and will speak to them again. All the minutes of SAGE are published and so dates of recommendations are clear.”
Then Matt Hancock says:
“It is exceptionally unhelpful having individual members of SAGE making comments like this. It undermines us all.”
So we see apparently a view being expressed by Patrick Vallance and Matt Hancock which is disapproving of members of SAGE and other committees making media statements.
We know that you did talk to the media during the pandemic. Help us with that issue as to whether, looking forward, people like you who attend these committee meetings and feed into these consensus statements ought or ought not to express contrasting views publicly?
Professor Graham Medley: As I said, I think this is a difficult area in terms of kind of the inside/outside government and independence. Clearly the government values independence, and so wishes to have independent people giving advice or providing evidence, and of course if we’re independent we can say what we like.
Slight irony of saying, well – in here, a paradox of, you know “Well, why weren’t they speaking out publicly at the beginning?” as a defence or a criticism for why they’re speaking out publicly now.
So – and it is a difficult area, because in an epidemic, you know, one of the key things that determines outcome is the coherence of the population, and we’re very well aware of that. So being on message, as it were, supporting government communications, even if you might think that they are personally wrong, you know, is – puts you in a difficult position. I’m not – whereas – as other scientific groups were quite happy to advocate for things that the government weren’t supporting.
So it just creates a minefield.
Counsel Inquiry: One way of looking at it is it undermines the consensus statements we were talking about earlier, because of course the SAGE minutes were being published –
Professor Graham Medley: I think we didn’t have it as a consensus statement. If we had a series of, you know, “Five people think this and three think that”, then potentially you get arguments in public about which of the – which is right, the three or the five.
And so having a consensus statement I think helps because that does give people a clear guideline of what we were, what our position was as a group. But we don’t have any – we weren’t asked to follow that in public, so people quite happily go out and – quite within their rights to go out and disagree with their own consensus, which might sound incoherent, but we are independent academics and, you know, it’s that – that’s the nature of the beast.
I mean, in some ways it would have been much easier for me and for others if they had agreed to pay my salary and co-opted me into the civil service and taken me into government. Then that would have made my life a lot easier. But then I wouldn’t have been independent. So that independence question and how you use it across the barrier, across into government, I do think is a critical one for understanding how SAGE works.
Counsel Inquiry: It is, Professor, and it actually brings me to the very last point I wanted to mention, because as you say, as we’ve heard, certainly you and your academic colleagues were independent, you weren’t being paid for all of the work you did on SAGE and the like committees.
If we could finally then look back at your statement, please, and go to page 25, and look at paragraphs 3.48 and 3.49, you make the point which you refer to in various places in your statement, Professor, that – picking it up towards the end of 3.48, you say:
“[You] expected that SPI-M-O would be mostly involved in supporting the technical functions of government and boosting the modelling capacity and creating the breadth of models required to generate ensembles and provide SAGE with scientific advice.
“3.49. In the event, [you] were solely responsible for the majority of the formal government modelling during the epidemic.”
And you describe then the vast amount of work that was done.
Then finally – if we can go, please, to page 101 – you pick up on very much the same point. At paragraph 9.9 you say:
“… personally [you] were not comfortable that SPI-M-O, a group of volunteers, was producing a government statistic …”
And you’re referring to the R number here, aren’t you?
“… which had been given prominence in government strategy, as our role was providing evidence rather than fulfilling operational functions.”
You’re making very much the same point in those two paragraphs, are you not, which is that you had expected to be, as it were, providing an additional later to government –
Professor Graham Medley: Yes.
Counsel Inquiry: – work, and in fact you found yourself week in, week out calculating the R number?
Professor Graham Medley: Yeah, so even when SAGE didn’t have to meet, SPI-M-O had to meet. And the setting up of the Joint Biosecurity Centre, JBC, in June, I think, 2020, I was hoping that they would take up this function but it took them over a year to get their act together to do that. And, you know, a lot of what we did would have been done by PHE had PHE had the capacity to do it. But, you know, a lot of the people we knew in PHE on SPI-M had left PHE and there was a loss of capacity. And so we did, to some extent, step in and fill that gap.
Counsel Inquiry: Looking forward, on your understanding of how things work now and how they might work in the future, were there to be another pandemic, would the government be reliant on, as it were, the volunteer academics to provide that basic modelling function, or –
Professor Graham Medley: That’s under discussion, and clearly UKHSA which, you know, is a new body put in, but has many of the same functions that PHE has, I think now, at this point in time, has much more capacity. How that would fit in with an equivalent SPI-M-O, I don’t know. But of course the danger is not now, the danger is in ten years’ time, that that capacity being reduced again.
Mr O’Connor: Professor, thank you very much. I’ve taken you through some quite lengthy points.
My Lady, those are all the matters I have for this witness.
Lady Hallett: Thank you very much indeed, Professor. I’m very grateful to you, and of course for all the work that you and your colleagues did. Thank you.
The Witness: Thank you.
(The witness withdrew)
Lady Hallett: Right. Well, as you know, Mr O’Connor, but others may not, I have to finish by 4 o’clock at the latest, so I shall return at 20 past, and I’m afraid anything you can’t elicit by way of oral evidence this afternoon, I’m going to have to rely on the written statement.
Mr O’Connor: My Lady, yes.
Lady Hallett: Thank you.
(3.08 pm)
(A short break)
(3.20 pm)
Lady Hallett: Mr O’Connor.
Mr O’Connor: My Lady, our last witness of the day is Professor Matthew Keeling.