6. Mr Scott Heald

MR SCOTT HEALD (sworn).

Questions From Lead Counsel to the Inquiry for Module 2A

Mr Dawson: You are Scott Heald?

Mr Heald: I am.

Mr Dawson: And you are David – Roger Halliday?

Mr Halliday: Roger Halliday.

Mr Dawson: Could I just ask you, first of all, to try to speak into the microphones, but as you’re giving evidence together, we’ll try to avoid you speaking over each other – I know you’ve got a lot of interesting things to tell us – I’ll try to direct my questions to each of you individually, but if you both have things to contribute in certain areas, I’d very much like to hear from both of you on those matters to the extent appropriate, thank you.

Mr Heald, you have provided a witness statement dated 11 October of this year to the Inquiry; is that right?

Mr Heald: That’s right.

Mr Dawson: The Inquiry reference is INQ000335154. A copy has just come up.

If we could just go to the page 28. It’s a couple of pages before that, I think. Yeah, that one there.

There you have signed the statement; is that correct?

Mr Heald: That’s correct.

Mr Dawson: As far as you’re concerned does the content of that statement remain true and accurate?

Mr Heald: It does.

Mr Dawson: Mr Halliday, similarly you have provided a statement to the Inquiry dated 15 November of this year; is that correct?

Mr Halliday: Correct.

Mr Dawson: That’s the statement there, it’s under reference INQ000274011.

And again if we could go to the final page?

You see there you’ve signed the statement, is that correct?

Mr Halliday: Absolutely.

Mr Dawson: Does this statement remain true and accurate as far as you’re concerned?

Mr Halliday: It does.

Mr Dawson: May I also ask my Lady simply to read into the record, we’ll come later to some slides which have been put together relating to statistical matters, the reference for that being INQ000274150, and that’s simply so that others can look at that if they consider it appropriate.

Could I start with you, Mr Halliday, just to get some details. You were the Chief Statistician for Scotland from 2011 to 25 April 2022 when you left the Scottish Government to become the chief executive of Research Data Scotland; is that correct?

Mr Halliday: Correct.

Mr Dawson: During the pandemic you held a number of roles in addition to being Chief Statistician. Perhaps most relevantly for our Inquiry, you were the joint head of the Covid analytical team. Is that correct?

Mr Halliday: Correct.

Mr Dawson: I understand that the joint head of that team was a lady named Audrey MacDougall?

Mr Halliday: That’s right.

Mr Dawson: And you were responsible for the quality and accuracy of the data that was published by the Scottish Government; is that correct?

Mr Halliday: That’s absolutely correct.

Mr Dawson: And as part of your role, I understand that you were a member of a group that we’ve heard a little bit about already, the Scottish Government Covid Advisory Group; is that correct?

Mr Halliday: Yeah, from, until January 2021.

Mr Dawson: And you were a member of some other groups, something called the Scottish Covid chiefs group, until April 2022?

Mr Halliday: That’s right.

Mr Dawson: And something called the Scottish Covid Data and Intelligence Network delivery group; is that correct?

Mr Halliday: That’s right.

Mr Dawson: You also attended meetings of a group that we’ve heard of called SGoRR?

Mr Halliday: From time to time. I wasn’t a standing member.

Mr Dawson: Thank you. And you also attended Cabinet meetings.

Mr Halliday: Again, from time to time.

Mr Dawson: How was it determined when you would attend SGoRR or Cabinet meetings?

Mr Halliday: When there was relevant evidence or data that I or my team collated that was relevant to the agenda of those meetings.

Mr Dawson: Thank you.

Mr Heald, if I could just run through a similar background to you. Over the course of the period on which this module is focused, you were the associate director and head of profession for statistics at the Information Statistics (sic) Division, which was incorporated into PHS when it became operational in April 2020; is that correct?

Mr Heald: That’s correct.

Mr Dawson: You continued when PHS was formed and became the interim contact tracing director from May 2020 to January 2021?

Mr Heald: That’s correct.

Mr Dawson: You were the chief officer from January 2021 to May 2021?

Mr Heald: That’s correct.

Mr Dawson: Since June 2021 your title has been director for data and digital information?

Mr Heald: Data and digital innovation.

Mr Dawson: Innovation?

Mr Heald: Yeah.

Mr Dawson: Also, between January 2020 and April 2022 you were the head of profession for statistics at PHS?

Mr Heald: That’s correct.

Mr Dawson: You were accountable, as I understand it, for the statistical methods, standards and timing of statistical release from PHS?

Mr Heald: That’s correct.

Mr Dawson: And you say in your statement that whilst part of your role involved advising Scottish Government officials, the final decision regarding the publication of PHS statistical material lay with you?

Mr Halliday: That’s correct.

Mr Dawson: And that was the case throughout the pandemic?

Mr Heald: It was.

Mr Dawson: Thank you very much.

Could I just ask you, I’ll direct the question to Mr Halliday first and then Mr Heald will have something to say about this, some questions broadly about, as far as the Scottish Government response to the pandemic was concerned, the purposes for which the various datasets that you were involved in collating and analysing and presenting, what the purposes of those might be.

Could you tell me whether the purposes for which data was being collected during the course of the pandemic changed as the pandemic progressed and if so in what ways? Mr Halliday.

Mr Halliday: Well, I would say yes, that did happen. So to – I would say initially it was – the data that we had around infections, hospitalisations and deaths were used partly to communicate to the public. They were partly used for decision-making as part of modelling. And I would say that – and other reasons for, in terms of managing the business and decisions of the government.

I would say as we went on, the nature of those decisions would need to change. So, for example, some of those datasets formed part of the decision-making or the evidence for decision-making as part of the levels approach, for example.

So I guess, yeah, I would start off by saying that.

Mr Dawson: So as far as the levels approach was concerned, about which we heard a little yesterday, would it be fair to presume that the data that you required became more localised, given the fact that the levels approach involved local area levels being applied?

Mr Halliday: Indeed, and I guess the interest from members of the public as the Covid pandemic sort of went on, again, became more intense, and the demand for local area data by the public certainly increased during that time as well.

Mr Dawson: In the very early stages of the pandemic, would it be fair to say that there was a limited amount of data that was available?

Mr Halliday: It certainly developed the amount of data we had. You know, in a large area, particularly for Public Health Scotland, they had existing data systems that served us well, but in many areas what we did was we adjusted either the data collected so that it was looking, for example, at schools and looking at the impact on staff and students at schools, and attendance and absence, or at the frequency of the data that was collected. So the nature of the data collection changed in response to the need for government information and to support decisions.

Mr Dawson: Mr Heald, was there any perspective you have to add to that?

Mr Heald: Yeah, I would agree with the points that Roger has made. I think what I would reflect is that the data that we held, and the data that we published – so Public Health Scotland had a role to make the data public – changed over the course of the pandemic. So, as Roger says, infections and hospitalisations, deaths, very much the focus at the start, but as the pandemic and the approach to the pandemic changed, so things like vaccinations became really important, that we released data on vaccinations into the public domain. We also released data on aspects of the Test & Protect system, just so people could understand how that was operating. So I would say that we adapted what we published as the pandemic progressed.

Your point about data at local level being important, so one of the key differences between the data that was published by Scottish Government, which tended to be headline Scotland numbers, and the data published by Public Health Scotland each day was that we provided more granular data at a more local level, and I think one of the successes for us was the Public Health Scotland Covid dashboard, which had data to, I guess, locality levels or very low levels of geography, that allowed users to log in and see kind of how the pandemic was affecting their local areas.

Mr Dawson: We will come back to it in a bit more detail, but could you remind us, because I’m sure everyone at one stage was aware, of what the Covid-19 dashboard was?

Mr Heald: Yeah, so the Covid-19 dashboard was basically a tool that Public Health Scotland updated every day that contained data statistics about the pandemic, so updating for the most up-to-date figures. It presented data at Scotland level, so similar to what was published by Scottish Government, but also published data at the more granular, local level. And that was the key difference between what Public Health Scotland published each day and what government did.

Just to add alongside that, we also published the data in what we call open data format, which was we released the data so that others could pick up the data and use it. And that open data also fed into the UK Covid dashboard, which the UK Health Security Agency also published, so there was a real stream of data going out each day.

Mr Dawson: Do I take it from that then that you were feeding the Scottish data into the UK dashboard as well as publishing it separately as a Scottish entity?

Mr Heald: Yes.

Mr Dawson: Thank you.

You have anticipated the area I wanted to go to next, which was the interplay really between both of your roles, one within the Scottish Government and one within PHS.

Our understanding from the material is that both Scottish Government and PHS published daily statistical updates throughout the pandemic, and the Scottish Government-published data included some data provided by PHS, as you’ve already said, Mr Heald, and other sources as well, which we understand to include things like the National Records of Scotland.

Is it correct to say that overall the Scottish Government published a daily update on the internet from March 2020 until April 2022, and that PHS produced a daily dashboard; is that correct?

Mr Halliday: That’s correct, and what we took was the judgement that actually we wanted to make it as easy as possible for people to access the headline statistics that were of significant interest, and so by bringing that together in a single place, we hoped to achieve that.

Mr Dawson: Thank you.

Do I take it to be the case, then, that the data that was produced by PHS was available to Scottish Government and formed a subset of the overall material that was published by the Scottish Government?

Mr Halliday: That’s pretty much correct. We didn’t – the distinction is in the local area data, that the Scottish – the data that was published on the Scottish Government website was national data and then Public Health Scotland produced the dashboard which showed that local area data.

Mr Dawson: So just to understand that, the position is that Scottish Government were draw on PHS data, it would extract from that for publication purposes certain elements of it but not necessarily with the granularity that Mr Heald referred to; is that correct?

Mr Halliday: That’s correct.

Mr Dawson: I would just like to separate out two concepts here. One is the question of publication of the data for public information, which you’ve both referred to, and one is the data that would be available for various people within Scottish Government to be able to process, analyse and ultimately inform high-level decisions with which this module is concerned. So the position is that all of the PHS data would be part of a wider suite of data available for to the Scottish Government for that decision-making purpose; is that correct?

Mr Halliday: That’s correct.

Mr Dawson: But what you’ve both spoken about, I think, is that publication was a separate matter because thought was put by both the Scottish Government and PHS into what would be appropriate to release into the public domain, which might not be everything that would be compiled?

Mr Halliday: Correct.

Mr Dawson: Thank you.

One of the decision-making bodies with which we are concerned, or bodies which is connected to decision-making, is one we have mentioned already, the Scottish Government Resilience Room. Mr Halliday, you have told us that you would on occasion be asked to provide information to that.

I understand that information, statistical and data related information was fed into that body by a series of documents which were known as the SGoRR sitreps; is that correct?

Mr Halliday: That’s right.

Mr Dawson: I understand that these were documents which were provided in connection with SGoRR meetings where decisions might at least be discussed and that the data that was provided in the sitrep was assimilated and put together to try and assist with that decision-making process?

Mr Halliday: The data was provided on a daily basis, or updated on a daily basis – well, some of the elements of the report were updated on a daily basis, some of it was weekly or less frequent, but updates were given every day to make sure that the information that – was available to the meetings, and more broadly that there was a clear definitive set of data for government to make decisions upon.

Mr Dawson: If I could have up, please, a document under INQ000214776, thank you very much, this is an example of one of the SGoRR sitreps; is that correct, Mr Halliday?

Mr Halliday: That’s right.

Mr Dawson: And we see from the top corner that this is from June 2020. Could we – I wonder whether we might look through this document to a certain extent, and you might be able to tell us a little bit – for example, if we were able to go to page 3 of the document, there is a colourful arrangement there with a lot of information on a single page under the title “Key indicators”, which appears to be split into four separate boxes; is that correct, Mr Halliday?

Mr Halliday: That’s right.

Mr Dawson: Could you explain to us broadly what the information is that’s contained within that, not looking at the detail but the sort of thing that you were trying to convey when putting these things together.

Mr Halliday: Yeah, so in April, if I recall, the Scottish Government published the – a paper about the handling of the pandemic under the theme of the four harms, which are, here: the – Covid direct, directly from Covid; harm because of the effect of Covid on the health service; on society; and on the economy.

And what our role was as analysts in government was to bring the range of evidence together under each one of those harms, and what the picture shows here are some – five key indicators for each of the – the harms, with a picture of what – the value of the indicator and how that’s changed – how that compares to the status for the pandemic. And it’s red, amber and green to mark – to highlight areas of potential concern.

Mr Dawson: Would it be fair to say in this four harms strategy that where a box was marked red, which would be the highest category, that would be an indicator for the fact that there was a particular strain in that area that was increasing that harm potentially?

Mr Halliday: It’s an indicator of that, yes.

Mr Dawson: Yes. And the colour-coding is in order to try to catch the reader’s eye and attract them to the things that are perhaps more stable and things that are perhaps less stable, based on the statistics?

Mr Halliday: Indeed so. And later on in the report, as then – a lot of the detail that goes behind these headline numbers.

Mr Dawson: And as you say they’re split into four harms. One of the questions I wondered if you might help us with, Mr Heald, is obviously the four harms are – as we’ve heard already: the first relates to the direct threat to health from Covid; the second, broader health harms; the third, society; and the fourth, economic.

Which of the harms would data be fed into this machine which would emanate from PHS?

Mr Heald: Yeah, so looking at that report, data from harm 1, so Covid direct health, there are a number of indicators there that would have come from Public Health Scotland, and broader health, harm 2, would be the other area where data from Public Health Scotland would have fed in.

Mr Dawson: Because those are the two health-related harms?

Mr Heald: Yeah.

Mr Dawson: And who fixed what the indicators were in each box?

Mr Halliday: So this was a decision of the analytical and also sort of the – the leads within Scottish Government on each harm, so we’d be – my team would work with the Chief Medical Officer on harm 1 and harm 2. On harm 3 the Chief Social Policy Adviser would take the decisions on which indicators are, and on harm 4, on the economy, it’s the Chief Economist at Scottish Government.

Mr Dawson: So they would fix what the indicators were they wanted information about, and it would be provided, it may emanate from PHS or some other source. Were you involved in the actual fixing of the indicators to any extent?

Mr Halliday: So I would just clarify –

Mr Dawson: Thank you.

Mr Halliday: – the fact that what would happen is a discussion between the analysts and those people that I’ve mentioned, because not necessarily – the data might not necessarily be available for the exact concept that they’d be looking for, and I guess the role of my team was to collate this information – so the information would be put together by different statistical and analytical teams from around Scottish Government or from other places, and we’ve mentioned Public Health Scotland and National Records of Scotland as well, and it would be up to my team to commission updates from the various statisticians and to put it together and put it into the format that we can see presented here.

Mr Dawson: And what was done to try to make the information not just contained on this page but throughout this quite lengthy document, which you say was produced regularly, to try to make the information digestible and comprehensible to those who would need to take decisions on the basis of it?

Mr Halliday: Yes, well, I guess that what we have is a group of quite senior analytical staff with significant experience of doing exactly as you’ve described, presenting complex information, multifactorial information in ways that can be digested by politicians and by senior officials, and so the people putting this together are well trained in exactly that task.

Mr Dawson: Is the risk by using, for example, a single page like this, and by using the colour-coding, that someone might, looking at this, simply look at which area is more or less red, think “That’s the thing we need to deal with now”, and not interrogate the detail?

Mr Halliday: I think that would be up to them, but as – we’ve used similar kind of presentations when we’re looking at overall performance of government in the past, so I think it’s something that ministers and other senior officials are relatively used to, the risks that you presented there.

Mr Dawson: Did you, from your perspective, get feedback from ministers or senior officials about the comprehensibility of this obviously very significant and broad database?

Mr Halliday: We certainly got questions about some of the detail.

Mr Dawson: Yes, but as far as the overall system of presentation was concerned in these quite lengthy documents with this key indicator element to it, was that something that they fed back saying, yes, they had a good handle on it, or was that something that they struggled with?

Mr Halliday: I certainly can’t recall any feedback about them struggling with the presentation of the information. It was more that we would get questions that looked very much like they had understood and were reflecting upon and asking for further detail on some of the evidence that’s provided.

Mr Dawson: And this may, I have to be clear, be indicative of the timing of this, but we’re not going to go through every page of the document, but having done that myself, I wondered whether it might be fair to say that the majority of the document contains information, much of which I suspect may have emanated from PHS, about what’s described as the first harm, tracking the ebb and flow of the pandemic; is that correct?

Mr Halliday: Yeah, I suppose we wanted to make sure that it wasn’t a document that was focused just on the first harm, that it was – reflected indicators across all four. What was unique about the first harm was that data was updated for at least a couple of the indicators on a daily basis, so it was much more frequent and it changed much more frequently than – in terms of what the numbers were actually saying, than some of the other harms. So it was unique in that respect.

Mr Dawson: And as far as – there is, I should say, some information about the economic side which is contained later, but was it difficult to try to either find data or find data that would assist in particular with harms 2 and 3, which might be slightly more difficult to encapsulate in a format such as this?

Mr Halliday: I think that it actually was – I thought it would be more difficult than it actually turned out, in that some of the data available for the economy was actually available on a fortnightly basis where previously it had been available on a less frequent basis.

Mr Dawson: Because one might say, for example, in harm 3 there’s information about vulnerable children attending school, people describing themselves as lonely, people who trust the Scottish Government to work in Scotland’s best interests, applications to the Scottish Welfare Fund and the total coronavirus interventions by Police Scotland. One might say that there are a very large number of categories that aren’t taken into account which would fall into the area of societal harm.

Mr Halliday: That’s absolutely right. Some of the – as you’ll have read later on in the document, go into that, but we had to take a judgement on what information to present to make it digestible.

Mr Dawson: Thank you.

Perhaps I might ask you probably my final question, Mr Heald, just in relation to harm 2, where PHS had a significant input, if we could just have a look at that.

Again, there are a number of criteria that are there relating to hospitalisations and in particular cancer.

Mr Heald: Yeah.

Mr Dawson: There are, I think, a number – one might quite reasonably say there are a number of non-Covid harms that aren’t reflected there. Were similar issues experienced as Mr Halliday has described it in that regard?

Mr Heald: Yeah, I mean, I think again, as Mr Halliday’s said, a judgement call about what’s available, and I guess this is a snapshot of the support at a particular point of time, so again I’m unfamiliar with whether indicators changed throughout the course of the pandemic.

I think the other thing I would say is while this is a document that was shared within government, Public Health Scotland did still publish, continue to publish data on a whole range of health and care statistics that we had in place prior to the pandemic that continued beyond that, so other data about other areas of health were still available throughout the pandemic.

Mr Dawson: So, for example, mental health obviously –

Mr Heald: Yeah.

Mr Dawson: – as we know and we’ve heard was a very significant non-Covid-related harm. It doesn’t feature there, but aren’t you suggesting that that would be something that PHS would have been compiling throughout the pandemic?

Mr Heald: And we still published statistics on mental health throughout the pandemic, yes.

Mr Dawson: Thank you very much indeed.

If that’s a convenient moment, my Lady.

Lady Hallett: It is, certainly. I shall return at 1.45.

Mr Dawson: Thank you very much.

(12.46 pm)

(The short adjournment)

(1.45 pm)

Lady Hallett: Mr Dawson.

Mr Dawson: Thank you, my Lady.

I’d like to return to a subject that we touched on briefly near the beginning of your evidence, and that’s to do with publication of data.

I understand from the statements that both the Scottish Government and PHS published data, so there were two sources from which data came.

Perhaps Mr Halliday first, why was it that it was thought useful for data to be published by both sources?

Mr Halliday: Well, I would say that it was important to have a very clear place to have data brought together, and the data that we did bring together in Scottish Government, yes, it included Public Health Scotland data, but it also included data from other sources, and I guess that we had that central role of co-ordinating sources of data and Public Health Scotland could focus on the excellent publication of its own data.

Mr Dawson: In terms of what I think you accepted earlier was the ultimate aim of the publication of the data, which was try to keep the public informed in a way that was effective, was the publication of data from both sources potentially confusing, given that the PHS data was a subset of the Scottish Government data?

Mr Halliday: I’d like to suggest that it wasn’t, I mean, and the Office for Statistics Regulations in fact, who were the organisation that comment on the quality, trustworthiness and value of statistics said exactly that, that these two things worked well together.

Mr Dawson: And you mentioned a moment ago that the Scottish Government data included data obviously over and above the PHS data. What were the other sources, the main other sources that were included within that extra data?

Mr Halliday: Are you referring specifically to the direct effects of Covid or –

Mr Dawson: Really –

Mr Halliday: – is it much wider? Because I would say that in Scottish Government we had portfolios of around – well, more than 100 regular statistical publications that adapted themselves to describing the effects of society, economy and environment during the pandemic.

Mr Dawson: That’s what I was interested in, really, the broad range of sources which you called upon. As far as the Covid-related information is concerned, you also published, I think, National Records of Scotland data.

Mr Halliday: Indeed.

Mr Dawson: As far as that data was concerned, to what extent did that differ from the PHS data? If that’s the right way of putting it.

Mr Halliday: Yeah, well, there’s a difference in the definition. There we’re talking about mortality data from Covid, so there was a different definition that was used for the National Records of Scotland data and the Public Health Scotland data, and in broad terms the National Records of Scotland data referred to situations where a death was – Covid was recorded on somebody’s death certificate, and Public Health Scotland where somebody had died within 28 days of a positive test. And after the spring of 2020 those two things were very, very similar indeed but during the early part of the pandemic the death certificate data was higher than the Public Health Scotland data, and that’s – I guess reflected the development of testing during that time, because the Public Health Scotland data required a link between a positive test and somebody dying.

Mr Dawson: Right. And as far as the mortality data was concerned, was there a possibility that the discrepancies in those two data sources might be confusing as regards the level of mortality?

Mr Halliday: There is – there’s the potential of that, and what us statisticians did to avoid that was to have very clear descriptions of what each statistic was representing, and the differences between the two, and when to use one set of data versus when to use another set of data.

Mr Dawson: Okay, thank you.

I’d like to ask you a few questions about accessibility of data, please. How did you – I think this is for both of you – both factor communication needs and the issue of digital exclusion for members of the public into your decision-making about how you would go about publishing data?

Mr Halliday: I guess this is – you know, we had established processes that are under the – a code of practice for official statistics and which we were working, which essentially …

Essentially communication and making sure that people could access and understand was an important part of how statistics are compiled and how they’re made available, yeah, to – as part of our standard processes.

Mr Dawson: Would that be the same for PHS?

Mr Heald: Yeah. I think another important point, particularly as we developed our Covid dashboard, that we got a lot of feedback from users about what was helpful and what was not helpful so that we could adapt the outputs based on the feedback we were getting.

I mean, I think an important point to stress is at the time the data and the outputs was being produced at great pace, and therefore it was really important that we got the data out into the public domain, but I would say we learnt over the course of the pandemic the most effective ways of getting that into the public domain so that people could understand what was happening.

Mr Dawson: What about consideration being given to people with particular needs, in the sense of perhaps disabled people who would have difficulty accessing the information, was that something that featured in the thinking behind publication in either the Scottish Government or PHS?

Mr Halliday: I would say that the thing that comes to mind when you’ve posed that question is about the accessibility of data via the Scottish Government website, but all the presentation of our information was specifically designed to be as accessible as possible, to high accessibility standards.

Mr Dawson: Would that apply to the PHS –

Mr Heald: (overspeaking) – Scotland, yeah.

Mr Dawson: One of the themes that we’ve heard from evidence that’s been collated by the module and indeed in other parts of the United Kingdom is the theme of digital exclusion. I think it’s the case that the data was all simply published through the internet, the dashboard, for example, that we’ve discussed.

Was any consideration given to the fact that there were sectors of society who, for various reasons, suffered from digital exclusion and how that might be addressed?

Mr Halliday: I would say perhaps not directly but I was – you know, an important route for making a lot of the information available was not just directly through the statistical publications but in things like the First Minister’s daily address where the statistics featured heavily as part of that.

Mr Dawson: I think we’d said earlier that the daily address was headline figures.

Mr Halliday: Indeed.

Mr Dawson: So that would be a means of communicating that. There would, of course, be people who would struggle to be able to understand that information. Was any consideration given in that regard about how information would be communicated better through that forum, that involved you?

Mr Halliday: That involved me? Not directly.

Mr Dawson: Thank you.

Mr Heald: Likewise for me: not directly. But as far as I recall people could contact Public Health Scotland. If they, for example, were requesting particular pages, we could print them out, make them more accessible. So that was certainly an option. But you are correct the majority of the outputs that came out from Public Health Scotland were in digital means.

Mr Dawson: What the evidence tends to show, and you may be aware of this from a lot of the statistical material you’ve looked at yourselves, is that the most vulnerable in society were the most likely to be the most vulnerable to Covid, or the most likely to suffer from digital exclusion, the most likely to have particular difficulties accessing the information. So would it be fair to say that efforts were necessary in order to get the information to the people that were most affected and those efforts might have been done better?

Mr Heald: I think that that would be fair. I think there’s always learning with these things. I think the key thing was that we were doing our utmost best to get the data out to the public in as easy accessible formats as possible on a daily basis, and this was running every day with data asked adapting to different stages of the pandemic. So there’s always learning from these approaches, but I think we did our utmost best to present the data in a way that people could access it and use it and understand it.

Mr Dawson: Thank you.

From a Scottish Government perspective, Mr Halliday?

Mr Halliday: I think I would agree with Mr Heald’s assessment of the situation.

Mr Dawson: Thank you.

I’d like to ask you a few questions, as you’ve helpfully included information about this in your statement – it’s particularly you, Mr Halliday, on this topic – it’s about data sharing on a UK basis.

One of the general questions I was interested in asking you was the extent to which, as far as informing key decision-making is concerned, local data is preferable, important, part of the picture? What would your view be on that?

Mr Halliday: I think that that depended on the stage of the pandemic that was being discussed, and the kind of decisions that were being taken.

As I alluded to earlier on, that when the levels approach was being developed and operated, that used local area data to a much more overt kind of – yeah, used that data much more overtly than at other stages, for example, when national – at the start of the pandemic, national modelling was what was a vital piece – vital piece of data rather than a lot of the local effects.

Mr Dawson: So would it be fair to say you’ve made a time distinction there at the beginning of the pandemic: because there was a limitation on information one had to try to use whatever information one could get, and therefore a more national reliance was prevalent and perhaps that became more local as the pandemic went on?

Mr Halliday: I would say that it was more that the scale at which the – you know, the numbers that were involved weren’t of a scale where there was a local – a significant local dimension to it. So the national – and therefore the focus on the national impact.

Mr Dawson: A scale that was statistically apparent based on what testing was available, for example?

Mr Halliday: Indeed.

Mr Dawson: So the flip side of that, I suppose, is to ask the question: what – beyond what we’ve discussed about the early stage of the pandemic, what was the use to the Scottish system, ultimately Scottish decision-making, of data which came from other parts of the United Kingdom? What sort of data was helpful or important?

Mr Halliday: So I put together – you’ll have seen in the situation report and – I put together a report called the “state of the epidemic”, and what that did was put – we used data from other nations of the UK or other nations internationally to put Scotland’s position in context, and that kind of helped frame some of the Scottish data.

Mr Dawson: When you say you used data from other places, to what extent was that data available to you? Did you have problems accessing data, in particular from the UK or more widely from these international sources?

Mr Halliday: So I would say that there’s two parts to this. So in terms of the aggregate data that might be about the number of cases in a particular country or a region of a country, that flowed very well through the central Cabinet, Cabinet Office arrangements that we had, and we fed data into that, and that was a reciprocal arrangement, so that worked really quite nicely. I think that it’s fair to say that data at an individual – around an individual person that could be used for research was more difficult.

Mr Dawson: So to set the Cabinet Office to one side, you mention in your statement at paragraph 12(d) that there were issues in obtaining data from UK Government departments and you cite at least an example of getting data from the DWP in June 2020. First of all, what was the significance of that data, why were you interested in that data?

Mr Halliday: So we were interested in that data to try to understand the effect of the pandemic on people’s finances and welfare, and the data I’m talking about is of individual case – individual people level data rather than aggregate data, and we found it quite difficult to come to an arrangement with the Department for Work and Pensions for sharing that data, which is a bit disappointing.

Mr Dawson: And did that continue throughout the pandemic or was that ever resolved? You mentioned the date, June 2020 –

Mr Halliday: That’s not yet been resolved.

Mr Dawson: That would be something, would it not, where you’ve identified, and you’ve explained why, that data which comes from the DWP for particular purposes would be useful, and this is the sort of thing that an inquiry might look into as suggesting would make a pandemic response more effective in future?

Mr Halliday: I think it would.

Mr Dawson: Thank you.

Were there any other such UK Government departments with which you had difficulty?

Mr Halliday: Not that I can recall, but I don’t think I particularly asked for the similar – made a similar ask to the – that I did to DWP to other UK Government –

Mr Dawson: That’s the one that sticks in your mind?

Mr Halliday: Indeed.

Mr Dawson: You mentioned also preparing some analyses of data on an international basis. Could you tell us a little bit more about that and how that was used to try to assist Scotland’s pandemic response?

Mr Halliday: Yeah, I can point to a couple of – a couple of times. So the first one was in our modelling, and our model inputted data from a range of different countries where there was some easing of restrictions – this was during the first wave of the pandemic, that we were looking at different options for easing restrictions, and we were able to get some evidence about the effect of different interventions in different countries through our modelling. So we were able to use data from other nations in order to be able to estimate what the effect of different policy interventions might be on things like the R number, or the number of people infected with Covid. So that was the first one.

The second one was –

Mr Dawson: Sorry, just on that one, was that something that continued throughout the pandemic or was specific to a particular time period?

Mr Halliday: That was predominantly during the first wave of the pandemic, but to some extent that did continue.

The second one was around foreign travel, and at that time we used a combination of modelling that was done by the London School of Hygiene and Tropical Medicine, and some data from internationally – internationally comparative data from a European agency to look at the incidence of Covid and the point – prevalence of Covid, ie the number of infective people or the rates of infective people in different nations, in order to make decisions about travel corridors to different nations.

Mr Dawson: On the very subject that you mentioned there, the international information, first of all, was that sourced then through the UK Government or did you have independent sources of that information?

Mr Halliday: For international travel, there are two routes for that. The first is that the UK Government did some analytical work to bring a set of data together for us, data and modelling together, and that we also looked at some websites that had comparative data on them, where we needed additional granularity in that data or we wanted to make sure that we understood that data properly.

Mr Dawson: To what extent was the data compiled at the end of the day which would have been used by ministers or other advisers to make decisions about border controls as you discussed, would that have been different for Scotland than, for example, at the UK Government level, or would the figures have been –

Mr Halliday: The figures were from the same source.

Mr Dawson: Thank you.

I had some other questions about something quite specific from your statement this time. You noted that you worked closely with the ONS, I think you’ve mentioned them already, and were briefed by them in respect of various nationwide surveys they were undertaking.

The ONS, we know from other evidence, commenced the Covid Infection Survey in England in May 2020, and is it correct to say that Scotland was the last of the four nations to be admitted to that?

Mr Halliday: It’s the last – the last nation for the survey to start in, yes.

Mr Dawson: Yes, and that was around 3 October 2020.

Mr Halliday: It started recruiting participants at the beginning of September and the first report was for 3 October, yeah.

Mr Dawson: And my understanding is that that is rather looked at as the gold standard of statistical evidence in certain areas of mapping the pandemic; is that correct?

Mr Halliday: It is for particular – particular things. For understanding the level of prevalence across a nation of the UK – a region of England level, that’s absolutely right. For understanding the pandemic within Scotland, then it’s not appropriate because it’s of a – while it went to quite a lot of people, there’s still a lot of uncertainty in the estimates for Scotland and for other nations.

Mr Dawson: So my understanding broadly, please correct me if I’ve got this wrong, was that the approach taken by the ONS was that they looked at sectors of society, looked at prevalence and therefore extrapolated out numbers that would tell you things about infection and mortality; is that broadly right?

Mr Halliday: About infection rates, certainly.

Mr Dawson: Yes.

Mr Halliday: Not about mortality.

Mr Dawson: Okay, so would that approach generally be deemed to be the gold standard?

Mr Halliday: Having a survey that was – that used consistent methodology in questions across the whole of the UK is the gold standard in terms of the ability to compare data between nations.

Mr Dawson: Is that because, at least in part, the data you might otherwise arrive at or based on, for example, positive tests, would not necessarily reflect the number of people who were actually infected?

Mr Halliday: There’s the potential for that to be the case. I think when you actually look at the charts, they track each other very, very closely and so actually that’s – the data on positive tests is a good proxy and therefore we were confident to use the data at a local level as well as a national level, which the Covid Infection Survey couldn’t do.

Mr Dawson: I think, Mr Heald, as I understand it, the data on positive tests that you were providing, that was the PHS data –

Mr Heald: It was.

Mr Dawson: – test positivity?

Mr Heald: Yeah.

Mr Dawson: And you mentioned a moment ago that the potential problem of using what you describe as English prevalence data but you needed to apply to that Scottish local data, is the idea that it would have been better for Scotland to have been involved in that type of approach at an earlier stage, to provide this additional source of information?

Mr Halliday: I think that would be useful. I think what we did was we took the time to make sure that that methodology would give us the most useful data from making decisions in Scotland, and once we were confident that the survey would go to enough people here to provide that estimate, then that’s when we adopted the survey.

Mr Dawson: Okay.

I understand that in your statement you talk about having requested case level survey responses from the ONS in the summer of 2020. What were they about, what did they tell you?

Mr Halliday: So this recognised that the individual survey responses can be particularly useful for research by enabling the linkage of the results from the survey to other routinely collected health data, say, for example, on vaccinations, and that that would be particularly welcome from the research community, which – and it was – we were aware that this was particularly helpful already in the UK context and what we were looking for was the Scottish data so that we could conduct some of that useful research for Scotland too.

Mr Dawson: Okay.

One of the themes again that emerges from evidence that we’ve heard from a number of groups, in particular representing vulnerable or at-risk individuals, is that they found, when they tried to influence or plead their case for different decisions being taken within Scottish Government, that there was a lack of base data relating to these groups. One particular group, for example, which I referred to in the opening yesterday, had complained about the fact that what this meant was that they had to plead their case more anecdotally and it was difficult to be able to prove the effects that they asserted, in their case in the ethnic communities of Scotland, using statistics or data.

Is it correct to say that there was a lack of data at the beginning and during the pandemic relating to at-risk and vulnerable groups such that this was the result?

It’s really for both of you.

Mr Heald: So I think it’s definitely fair to say that at the start of the pandemic, but as the pandemic progressed, and recognising the importance of data for those groups that you’ve talked about, we did take steps to address that.

I would still say that’s work in progress, so there’s currently a – for example, a data group that I’m chairing that’s looking at, you know, how do we improve the recording of ethnicity and the datasets that we have for health and care. So that’s important. But where we could we did publish data.

And one source that we’ve not touched on during the hearing so far is that Public Health Scotland as well as having the daily dashboard also had a weekly report which allowed us to kind of deep dive into more detail into particular topics. So we did throughout the pandemic have particular chapters that majored on the impacts of different aspects of it on ethnic minorities as the group that you’ve – highlight in particular the impact of vaccinations. So it definitely was recognised and we took steps as best we could to address it, and we’re still working now to make sure we’ve got systems in place to improve that going forward.

Mr Dawson: So that’s a work in progress. I’ve focused on ethnicity because it’s one particular example, but across what one might call “protected characteristics” generally would you say, Mr Heald, that that is a work in progress?

Mr Heald: I would say that’s a work in progress. And when I, you know, give an example of ethnic groups, you know, we’re doing work at the moment on all the protected characteristics, and that’s a really important aspect of what we do and it’s really, really important to the work that Public Health Scotland does more broadly, so not just for the work we do in Covid.

Mr Dawson: I think actually we heard yesterday that it was one of the purposes of the formation of Public Health Scotland to try to address health inequalities more effectively.

Mr Heald: Yes.

Mr Dawson: And this would be an example of trying to do –

Mr Heald: Yeah.

Mr Dawson: I’d like to ask you a few questions about some other data areas.

You gave evidence, I think, Mr Heald, to the Scottish Parliament Health, Social Care and Sport Committee at a hearing on 23 November 2021.

Mr Heald: I did.

Mr Dawson: There are a number of aspects. For the sake of the transcript is reference is INQ000286854.

If we could just have that up.

There are a number of reflections, I think, in this similar to the one that you’ve just made, Mr Heald, which are very interesting to us, about issues that were experienced with data access within PHS and efforts, indeed, that are being made to try to look at that issue.

If I could look, for example, at page 2, I think these are four pages – oh, no, that’s not quite the same as the version I have.

I think here you say that it is in your response – this is a few lines down in the first paragraph, you refer to – yes, you refer to:

“We have a lot of data that we can use to good effect, and we have the ability to link the data in order to understand pathways of care. It is important to recognise that we are building on strong foundations. There are a couple of areas that we need to focus – and are focusing – on: social care in particular, and primary care. Those are the two big areas to which we need to direct our attention.”

Now, just to be clear, that was you speaking in November 2021, so we were still in the pandemic, but the later stages of the pandemic. Roughly, in context, about the time that Omicron was about to strike or had just struck.

So you were saying at that time that you had identified these problems based on the prior experience with the pandemic.

Mr Heald: Yeah.

Mr Dawson: I’d be interested in particular in understanding more about the difficulties you faced accessing data from social care, and I’ll ask you about primary care in a moment.

Mr Heald: Okay, so, yes – so as I’ve outlined there, we have got good, well established data systems and processes around collecting a flow of what I would call health service data. One area where there is a gap is social care, as you’ve alluded to, and that is an area that’s still work in progress. So although identifying it back in November 2021, it’s still work that we’re doing at the moment. And in fact across Scotland, I can’t remember the date of the strategy being launched, but we did have a health and care data strategy joint between Scottish Government and local government, and one of the key aspects of that is the desire to address issues with social care data.

So Public Health Scotland does collect data from social care, we have a system called Source which collects data about individuals who are receiving care at home. One of the challenges though is that the frequency of that data is currently collected on a kind of ongoing basis but is made available quarterly and is used in annual reporting. So we didn’t have the set-up that we had for the other health service data that would have allowed us to do more granular reporting on a more regular basis.

And part of that is, you know, we need to have things in place within Public Health Scotland to receive the data, but there is also – investment is required in infrastructure or locally, in local government, around being able to collect or maintain that data in the first place. So we do need to be thoughtful about the burden on the data providers but it was recognised as a gap.

It’s work in progress, we do have some data and we’ve currently got, as I say through the data strategy, a group looking actively at this as an area of particular focus is data on care homes, in particular what’s happening –

Mr Dawson: That’s one of the areas I’d like to ask you about, Mr Heald, because obviously in this module we’re interested in infections in care homes across the pandemic but particularly during the first wave when a high proportion of deaths occurred in care homes, and we’ll look into the details of that with other witnesses.

But there would be a number – I think it would be fair to say there would be a number of datasets, if you like, that would be useful to have in analysing and strategising for the types of issues that might arise in a serious infectious disease which might affect predominantly older people, would that be fair?

Mr Heald: Correct.

Mr Dawson: So, for example, data about the number of people in care homes might well be a useful starting point.

Mr Heald: Yes.

Mr Dawson: It might be useful to know, in the context of the Covid pandemic, the number of people that would be likely to be transferred between hospitals and care homes, for example; would that be right?

Mr Heald: Yes.

Mr Dawson: It would be useful I think also to know in this sphere the numbers that might be transferring between the community and care homes.

Mr Heald: Yeah.

Mr Dawson: And it might be useful to know the number of people that are receiving care at home who are in that vulnerable group.

Mr Heald: Yes.

Mr Dawson: Were, in the early stages of the pandemic, these datasets available?

Mr Heald: So they weren’t. So that was a definite gap. Although one way – how would I best describe this – it is possible from other datasets to triangulate to inform some of those particular questions that you are asking about. So, for example, the Care Inspectorate maintain a register of care homes, that register has the address of the care home, much of the data that we get coming into Public Health Scotland’s at an individual level, so we’re able to map the postcode as best we can to care homes to understand where people are. But I would agree this is still an area that needs further development.

Mr Dawson: One of the factors you mentioned earlier, which is of course very pertinent to any data provision in the Covid pandemic, was the need for data to be provided quickly because decisions needed to be made quickly and therefore the data backing them up needed to be provided quickly. Even where the data you’ve referred to might have been available, can I take it from what you’re saying, because they would have had to have been sourced from other places, they wouldn’t have been available, certainly, quickly?

Mr Heald: Partly true, I guess it depends – depended on the analysis that we were doing. I think one thing we pride ourselves on in Scotland is our ability to link the data quickly. So the data on testing is available every day, we’ve touched on that. The register of care homes, for example, that’s held by the Care Inspectorate doesn’t change that frequently, but we would be able to link to those data on a regular basis, so it really depended on the analysis that we were doing.

Mr Dawson: You mention in the paper – it’s actually at page 6, I won’t go to the direct quote – but one concept that you mention as being relevant to this is the fragmentation of the system. I was interested in exploring that word, but perhaps you’ve already told us what that means in the way that you’ve explained things.

Mr Heald: Yes, so just – again, to just refresh my memory, which paragraph?

(Pause)

Mr Dawson: I’ve got the quote here –

Mr Heald: Okay – oh, I can see it now, it’s at the top –

Mr Dawson: Yes, it’s the fragmentation of the system that I was interested in exploring with you – yes, that’s it, thank you very much indeed.

Is that something – you’ve already, I think, alluded to something about that because you wanted to talk about the Care Inspectorate, is that why you’ve already told us or is there another aspect that –

Mr Heald: So I guess in the context of that, the other aspect of that is that – so data particularly – this is particularly in relation to local government – is held by – by large local authorities, and that landscape is quite – well, at the time and still is quite fragmented. So what I mean by that is that there is no kind of standard way of collecting or then extracting data. So one of the challenges you have then is that you’ve got different approaches in 32 different local authorities and standardising that would take time and then, therefore, affect that ability to get data more quickly.

Mr Dawson: You go on just after this to raise, in the same area, the issue of differing information governance procedures. Is that part of what you’ve just described?

Mr Heald: Partly. I mean, the information governance really is about that ability to share the data so that local authorities or the data controllers – so they have a say in what happens to the data.

Mr Dawson: Yes.

Mr Heald: So we have to have the dialogue about kind of what Public Health Scotland’s use would be. So that’s one aspect. But even if that was resolved and was straightforward, you would still have the issue that the data’s fragmented locally and would still need to be –

Mr Dawson: Yes.

Mr Heald: – addressed.

Mr Dawson: These are separate problems?

Mr Heald: Separate problems.

Mr Dawson: I understand.

Mr Heald: (inaudible).

Mr Dawson: Thank you.

One other aspect I wanted to just touch on, as you’ve mentioned it before, was difficulties in accessing primary care. What would the value have been of being able to access primary care better than it appeared actually happened?

Mr Heald: Yes, so one of the key values of primary care data is that it tells you a lot about what’s happening in the general population, so the reasons why people would go to a general practitioner can be quite different to the reasons why people ended up in hospital, so a lot of the established datasets that we’ve got are from the hospital acute sector. That data’s good, is robust. That ability to understand more locally what’s happening within general practice would have been a really helpful additional dataset to have.

Mr Dawson: Because of the difficulties in accessing hospitals during the pandemic, would that have made primary care data perhaps even more revealing?

Mr Heald: Yes, potentially. And I can talk about steps we took to address some of the shortfalls in primary care data, but –

Mr Dawson: I’d very much like to hear –

Mr Heald: Yeah. So, I mean, it’s worth saying. So I’ve mentioned the Scottish Government data strategy. A key aspects of that is also work we’re doing around primary care data. I’ll just explain what – some of the challenges with the primary care data. One is that each of the general practices or the GPs within the general practices are the data controllers, so they have a say in what happens to the data, and so we’ve been working closely with kind of partners across Scotland to talk through the types of uses we can make of the data.

So there’s a couple of things to highlight that we have done. One is around data at what I’m going to call aggregate level. So it’s at a reasonably high level, it’s not at an individual level. We had a lot of engagement with GP bodies about that and that enabled us to start reporting on activity, effectively face-to-face activity, or telephone calls in general practice.

Mr Dawson: At what stage was that something you were able to institute?

Mr Heald: So forgive me around the exact dates but my memory is from –

Mr Dawson: Broadly.

Mr Heald: From reading, it was broadly shortly after my appearance at the committee in 2021, so –

Mr Dawson: Pretty much the end of the period we’re interested in.

Mr Heald: The period that you’re interested in. But we were able to get the data for that.

The other important area which, again, has been touched on, and was touched on in the opening statement that Public Health Scotland gave, was the EAVE II study. You’ll forgive me, I can’t remember what EAVE II –

Mr Dawson: Well, could we come back –

Mr Heald: Oh you can’t –

Mr Dawson: – questions to go through – (overspeaking) – research access.

Mr Heald: But an important aspect of that, and for the surveillance work that was done by Public Health Scotland, we did get agreement to get data at a more individual level from primary care to assist with the surveillance, and we managed to achieve essentially using the emergency powers that Covid brought, and we’re currently in conversations again with the GP community about continuing with that essentially beyond this Covid period, because the emergency powers we had then are no longer in place.

Mr Dawson: Would one of the things that primary data would have been of assistance in would be informing you about what we called the second harm, the extent to which people are suffering other health harms that may not come to the attention of hospitals?

Mr Heald: I think that’s fair. I mean, I think it’s also worth highlighting. A bit like in Scottish Government, you know, we’ve got many other data sets within Public Health Scotland that address other harms. So, for example, we used the example earlier of mental health, you know, we’ve got a lot of other datasets that look at different aspects of mental health, so not having the primary care data didn’t completely exclude us from being able to look at other aspects, but it’s an important gap, I would say, in our data estate and it’s an important gap that I would say we’re making good steps with at the moment to address.

Mr Dawson: I had another PHS-specific question. Another thing that was mentioned in the opening about PHS yesterday was the reorganisation that went on within PHS.

Mr Heald: Yes.

Mr Dawson: That, as I understand it, was a pre-planned organisation – reorganisation, PHS having been formed as a corporate entity late in 2019 but was only going to become operational on 1 April 2020. And that of course happened at a time when we were in the middle of the first lockdown.

Counsel for PHS accepted that there had been a number of issues, including staff changes and the need institutionally to bed in the new organisation, which are understandable when any large organisation forms like that.

I was interested in the specific element of the extent to which that reorganisation caused difficulties in data provision such as, or perhaps others, the ones that you have frankly pointed out.

Mr Heald: So I think in terms of challenges around data provision, you know, I would say that it didn’t cause problems. I think the important point was, as counsel mentioned yesterday, there were three different bodies that came together essentially to form Public Health Scotland and in essence those bodies had, you know, the existing data streams already in place, and those carried on into Public Health Scotland, so data that we routinely collected and had access to prior to Public Health Scotland we still had that as part of Public Health Scotland.

One of the areas that I would say we did make good strides and there was a real benefit of Public Health Scotland being there was the fact that we have the expertise on the kind of data, the data capture aspects of it and the analytical work from one of the previous organisations, ISD, the Information Services Division, that I was part of that could work more closely with our health protection colleagues to make sure that we kind of had our processes as automated and streamlined as possible. So actually I would say a benefit was we were able to bring additional capacity into the Covid space than might otherwise have been more challenging had Public Health Scotland not –

Mr Dawson: Presumably that amalgamation was part of the (unclear) –

Mr Heald: That’s right.

Mr Dawson: – reason we discussed earlier, trying to improve public health delivery –

Mr Heald: Yeah –

Mr Dawson: – was bringing together these two organisations?

Mr Heald: Yeah, and important that we didn’t want each of the organisations to continue as the previous organisations; that would defeat the object of Public Health Scotland coming together, so that ability to work together. Obviously at the time of the formation of Public Health Scotland we weren’t anticipating the pandemic hitting on day one but I think, certainly from a data analytical perspective, we rose to the challenge well.

Mr Dawson: Another matter that was mentioned yesterday was, as I understood it, difficulties with getting access to data from the original source. A computer system called ECOSS was mentioned.

Mr Heald: Yes.

Mr Dawson: What were the issues around that?

Mr Heald: Yes, so not so much about getting – problems with getting access to ECOSS – so just to explain how that works. So essentially data about testing is run through the lab system until Scotland and, latterly, through some of the UK labs that were set up during the pandemic, and those data flow into a system in Public Health Scotland called ECOSS.

What I would highlight is that ECOSS is what we would call a legacy system, it’s old, serves its purpose and prior to the pandemic a lot of the surveillance work that we were doing would have been about instances of disease that were a lot smaller in case than what we ended up seeing in Covid, so one of the challenges quickly became the sheer volume of data that was coming through. Not so much in the very early stages of the pandemic because case numbers each day, although rising, were still relatively small, but we needed to kind of take steps to address that. And in essence what we did do was, without getting too technical, we established what we call a data warehouse, which is on newer technology, that allowed us to feed the testing data into that new platform on a daily basis, and that allowed us to run the analyses, automate what we were producing a lot quicker, which would ease the burden essentially on a lot of our staff in terms of what they were having to do more manually in the early stages of the pandemic.

Mr Dawson: Thank you.

If we could have a document up, please, INQ000366002.

Now, this, as I understand it, is a National Records of Scotland document from 24 March 2021, which sets out statistics in particular relating to various indicators during the previous week but it also includes an overall aggregate total of various things that have happened in the past. There is one particular aspect, which I think is page 10, that I’d like to ask you about.

So this sets out at that time a number of specific aspects based on a number of specific things, and at the bottom we see that there is reference to data being included about disability, and although a number of the other sources of evidence and types of evidence seem to have been introduced into this type of analysis at quite an early stage, the disability information was only introduced on 24 March 2021.

Is there a reason – I think it’s for you, Mr Halliday – is there a reason why the disability information hadn’t been factored into this very useful document earlier than that?

Mr Halliday: Yeah, I would say that in general it’s because the data on disability came from – wasn’t recorded as part of the standard information on some of these death certificates and as such we had to bring that information in from the 2011 population census. Now, getting those two sources of data together, the deaths data and the census data and the – developing a method in order to provide some useful statistics and ensure that we could explain that in a useful way so that people understood the strengths and weaknesses of that analysis, that took a bit of time.

Mr Dawson: Did that mean that information, important information about people with disability was not available for decision-makers as it might have been in the earlier part of the pandemic?

Mr Halliday: Certainly information on disability relating to mortality was not available before that time.

Mr Dawson: Thank you.

In paragraph 36 of your statement you refer to a project to use data linkage to pull data from various sources which may hold different datasets with a view to improving the available data on protected equality characteristics. What’s the current progress of that project?

Mr Halliday: That dataset’s now together and available for research in the public good, and it’s held very securely in a – in the Edinburgh – the University of Edinburgh National Data Safe Haven.

Mr Dawson: Does that project allow intersectional analysis to be carried out?

Mr Halliday: Indeed, that’s exactly what it will allow.

Mr Dawson: It will allow that?

Mr Halliday: Yes.

Mr Dawson: Thank you.

A general topic which I’d like to touch upon, about which there are a lot of documents that I won’t get into, but I’m sure it’s one we’ve touched upon already, it’s the extent to which data was made available to research, research organisations outwith the organisations where you were working, the PHS or the Scottish Government.

There are a number of places where we have indications that for some time there had been concerns raised by academics, for example, about access to research.

Just for the sake of a transcript I’ll give some examples. INQ000149111 is an exchange between Professor Mark Woolhouse and the then Chief Medical Officer, Catherine Calderwood, Dr Catherine Calderwood, from May 2018, where concerns are brought up about researchers like Professor Woolhouse being able to access information and data.

Similarly, in the statement of Professor Andrew Morris, who played a prominent role as you’ll recall, as chairman of the Scottish Covid Advisory Group, his statement being INQ000346264, at paragraph 16, they raise concerns about the way in which data was provided to researchers.

I’d be interested to hear your perspective on that, in particular whether you feel greater efforts could have been made, but my ultimate objective really is to ask you the extent to which that – had data been made available to these individuals and institutions, their work with that data would have better informed Scottish Government decision-making.

Mr Halliday: I’m happy to cover this.

Mr Dawson: Mr Halliday.

Mr Halliday: So this is, yeah, clearly a known problem before the pandemic, and in fact the Scottish Government decided to set up an organisation to deal with this, which is Research Data Scotland, which is the job that – I’m now leading that organisation, and that was announced in 2019 in the Scottish Government’s programme for government as an organisation to enable data access and data to be brought together around a person, place or business. That – this is a – quite a tricky problem and a problem that isn’t unique to Scotland, isn’t unique to the United Kingdom, is much more – broader than that, and I guess goes back to the sort of concerns, on balance, of the owners of data wanting to make sure that they protect the privacy of individuals with the fact that there’s a lot of utility in the data.

So to answer your question, if we’d made this data available, would this significantly have improved the research base and potentially the evidence base in Covid, I think absolutely it would. And I would say that the government, the Scottish Government, recognised that that was the case. That’s why they announced the set-up of Research Data Scotland in 2019, and that’s where, why they decided to fund the organisation up to £25 million in – from 2021, and I’m delighted to be able to make a contribution to addressing this particular challenge.

Mr Dawson: Thank you.

As we touched upon Professor Morris, just one aspect, we talked earlier about the way in which data was provided to SGoRR, and we looked at one of the sitreps. As we found out yesterday, the Scottish Government set up its own Covid Advisory Group set up in the end of March 2020, started really working in April 2020. As far as data provision to it was concerned, I was interested in exploring how that worked.

Was it possible for that group, for example, to commission or at least ask for specific data either from the Scottish Government or PHS to assist with its work?

Mr Halliday: The – so my team provided data and evidence to that group partly on modelling and partly on other things, and I recall quite a number of cases where members of the group would ask me analytical evidence-related questions that I was able to respond to.

Mr Dawson: And would they be able to do that, would they be able to request it specifically about things they were interested in or would you just provide it and they would have to use what you provided?

Mr Halliday: Well, both of those things. So …

Mr Dawson: Yes, okay.

I would just like to ask you a few questions about matters that have arisen elsewhere around specific incidents of data presentation, which you’ve addressed in your statements. One for you, Mr Halliday, and one for you, Mr Heald.

Could I go, please, to INQ000239682.

This is a witness statement, I’m hoping, of Ed Humpherson, the Director General for Regulation at the Office for Statistics Regulation, and if I could go to paragraph 35, please, in that it states that:

“In September 2020 concerns were raised with me about a claim made by the First Minister of Scotland that around 40% of care homes in Scotland allowed and enabled indoor visiting. An FOI published on 5 November set out the source of this statement and made clear that the 40% figure was a loose approximation based on incomplete data. We advised the Scottish Government’s Head of Covid-19 Analysis that the uncertainty in this data should have been more clearly reflected in the FOI response and the associated published material. We also stated that it should not have been necessary to wait for the information to be published as part an FOI. It would have been more appropriate to share the data publicly through an ad-hoc release shortly after the statement was made.”

I think, Mr Halliday, you will recall that there was some correspondence with you about this. There is a letter INQ000092824. This is a letter, I think, where Mr Humpherson is writing to you about this particular issue.

I’m more interested in the generality of rather than this specific incident, but what’s being highlighted here is that there was a piece of information, important statistical piece of information that was relied upon by the First Minister and then it turned out that there were concerns about its accuracy and reliability.

Can you explain the process about how information like the care home indoor visiting statistic would have been provided to the First Minister?

Mr Halliday: My suspicions, though I don’t know, I guess, in this specific instance, but in general I would say that there are two classes of information, there’s statistical information and management information that are collected by professional statisticians and there are also a range of other management information that are collected by other Scottish Government officials. And with specialist statisticians we all work to the code of practice and that is – has a proactive publication approach, and what we would do is ahead of that or at the time of publication we’d be – provide written briefing to the First Minister on the contents of that data, and I think that goes for official statistics and management information. And so I don’t know in this particular instance but I would have thought that the First Minister will have received this information by a written submission and made the decision to use this information. And I don’t know what particular advice was given at that time to the First Minister on its use.

Mr Dawson: As I said, I’m not necessarily focusing on this particular instance but just how it may illustrate the generality of the process which you’ve just set out.

It seems that this case illustrates that there will, I think, inevitably, be times, especially when you’re dealing with a lot of data that might be complex where data is presented simply in that way: this is a – here is a figure. And there might, as I think Mr Humpherson was pointing out, be nuance or approximation about it, that it might be misleading for someone to rely on a figure like that, where perhaps they know the figure but they don’t know about the loose approximation aspect to it or perhaps aspects of its reliability.

What facility was there in the system for ministers or their senior advisers to be able to understand more fully in particular statistical information upon which they intended to rely in this fashion?

Mr Halliday: Yes, we would be – by “we”, statisticians would be very clear about the status of that data, when it would be put into the public domain, because, as I said, our default position was to make all management information, statistical information available publicly. So I think that’s how they would be told.

Mr Dawson: Thank you.

Mr Heald, there was one similar thing coming out of this statement which I was going to ask you about, which is on page 3. I’m not going to go to this document but it concerns an October 2020 PHS report which was related to the discharge of patients from hospitals into care homes. It can be found at INQ000147514 but I am still sticking with the Humpherson statement, INQ000239682.

At page 3, please – excuse me one second.

(Pause)

Mr Dawson: In any event – I’m sorry, I can’t lay my hand on the exact document, but you’ll be aware, as you’ve addressed I think in your statement, Mr Heald, that there was a specific quite subtle, I think, observation made by Mr Humpherson in connection with this information. In particular there was an aspect of the part of the report which related to the extent to which the discharge of patients from hospitals to care homes had led to specific outbreaks.

Mr Heald: Yeah.

Mr Dawson: And what was being pointed out on that very important piece of information in this very important report, that there were certain confidence intervals that had been used, which – I don’t think ultimately there was a criticism about the fact that that had not been mentioned, but that it was mentioned by Mr Humpherson in his important capacity that that was something that was certainly relevant to a complete understanding of the data.

On that subject, again to try to use this as a means of understanding the generality but on this very important topic with which we are concerned, to what extent would information like those confidence levels have been communicated to Ms Freeman, for example, who was the recipient of that information at the time, who was the Cabinet Secretary?

Mr Heald: Yeah, so some important points, I was one of the authors of that report, so I was involved in doing it –

Mr Dawson: Absolutely.

Mr Heald: – so I know it well.

So that report was a stats report produced by Public Health Scotland that followed all the same processes that other reports had. It was pre-announced. Pre-release access to the report was given to Scottish Government, which is our standard practice. And then the report was published as you say.

I was involved personally in briefing Ms Freeman in the contents of the report but she did not have access to the report prior to that pre-release access period that I am referring to.

Mr Dawson: Okay.

So in that particular instance – the reference I was looking for earlier was INQ000286856. I think it’s actually a letter that Mr – or some form of contact between you and – from Mr Humpherson relating to this issue.

Mr Heald: Okay.

Mr Dawson: And he point out – I think it’s page 3 in this document that I’m looking in. Yes, at the bottom:

“When looking at the different types of discharge, we see adjusted hazard ratios of 1.00 for tested negative, 1.27 for untested and 1.45 for tested positive. Although the confidence intervals again suggest these findings are not significant, the observed ‘dose-response’ pattern in the adjusted hazard ratios is consistent with a causal relationship between positivity and outbreak. Given the sensitivity of the care home setting during this pandemic, and the likely uses of the evidence from this analysis, some users may have benefited from additional discussion of this in the report.”

So I think what he’s trying to say is along the lines I suggested earlier, that it may not have been a point that he raised at all, but for the fact that this was a very important matter, as you know as an author of the report. Did you – do you think that you would have explained these sorts of things to Ms Freeman at the time?

Mr Heald: No, so we did explain this type of thing to Ms Freeman at the time, and I think the important point about this particular report was, given the importance as you rightly highlight of this particular topic, this was being produced at pace to get the results out into the public domain, and what Mr Humpherson essentially was pointing out in his letter is that there are, I guess, some nuances in terms of the analysis that was undertaken and the results that we could have made clearer in the report in terms of that communication. So it was more about the communication of the results rather than the results per se. And we did take that on board and we did an update to the report the following April where we went into a lot more detail with a lot more visuals to help people understand what we were saying.

Mr Dawson: So am I correct in understanding your evidence that there is a distinction to be made between the criticism, if we can call it that, by Mr Humpherson, which is about communication of this aspect of the data to the public, whereas what you’re saying is that that aspect would have been communicated to the decision-maker?

Mr Heald: It would have been, yes.

Mr Dawson: Yes. Thank you.

I’d like to move on now to deal with a completely different part of your evidence with which you have very helpfully agreed to give us some assistance, which is to look at some of the slides that have been put together.

These are at INQ000274150, and you have very helpfully looked through these.

These were originally compiled, my Lady, by the Inquiry team but they were based on publicly available information, and Mr Halliday and Mr Heald have very helpfully looked through this in order to confirm that this is in fact accurate data, and there are certain graphs and statistics within them that emanate indeed from, for example, PHS or sources to which PHS have contributed. And the purpose of looking at this is to try to understand some of the overall features of the pandemic in Scotland and indeed try to understand some of the statistical basis, which at times gets a little tricky.

So if I might take you, first of all, to slide 6. Although, as I said at the beginning, I’d be very happy for either of you to contribute, I had a slight idea as to who might lead on each one. If I’m getting it wrong, please tell me.

I wonder whether, Mr Heald, you might lead on this one.

Mr Heald: I would.

Mr Dawson: This comes from the UKHSA Covid dashboard. I think you told us earlier that that would be a dashboard to which PHS would contribute Scottish data.

Mr Heald: Yes.

Mr Dawson: And so this reports daily number of reported Covid-19 cases by specimen date from March 2020 to April 2022, ie the period that we’re primarily interested in in this module.

The slide I think suggests possibly, if one were to look at it, it tells us about the number of cases plotted over a period of time. It might suggest at the beginning that there were relatively few cases. Would that be an accurate interpretation of it?

Mr Heald: It would be in terms of cases reported. And I think a really important context on looking at this graph though is understanding the volumes of tests that were going on at the same time in the wider community. So the volume of testing, which is covered in some of the earlier slides, did change dramatically over the course of the pandemic. More testing, you’re more likely to find more positive cases.

Mr Dawson: So if we were to look near the beginning of the period and it might suggest there was a low number of cases, that would be because there would be a low number of tests because this particular graph is based on testing?

Mr Heald: Yes.

Mr Dawson: Thank you.

As we go along, I think we can see that there are a number of peaks in the graph, and I think, for example, we can see that there are a number of different ones, for example, cases starting to rise around about the – around about October time; would that be right?

Mr Heald: October ‘20 you mean?

Mr Dawson: 2020.

Mr Heald: Yes.

Mr Dawson: And then there are various other peaks. The fact that we can see these peaks may be indicative of the fact that there’s more testing, would that be right?

Mr Heald: Yes.

Mr Dawson: Yes. And do we see that there is potentially a – so we see some level of rise around about that time, and then we see perhaps another peak, which seems to occur – would it be around December/January 2021?

Mr Heald: Yes.

Mr Dawson: And I think you have confirmed with us through other sources that in that peak, from PHS data we know that its peak was 29 December 2020, do you recall that from the PHS data?

Mr Heald: Yeah.

Mr Dawson: Then I think, although I won’t hold you to the exact number, that what you’ve told us is that there were 3,137 confirmed cases on that day, so that was the peak of that particular wave.

Mr Heald: Yeah.

Mr Dawson: Yes. Then I think there’s a further wave which one can see, which seems to start around about May 2021, as we go along the line. Would it be correct to say that that was thought to be primarily associated with the Delta variant?

Mr Heald: It was, that’s correct.

Mr Dawson: And again I think that you have provided us with helpful other data to suggest that the peak of that is 2 September 2021, when there were 7,622 cases.

Mr Heald: I think if I may, the other important point around those dates is the easing of restrictions at different points in Scotland, and corresponding with that there was also increased testing, so you’re absolutely correct the Delta wave – or Delta variant was present at that time, but there was also increased testing, which also then leads to an increase in overall numbers.

Mr Dawson: Indeed. So you have to take into account both of those figures.

Mr Heald: Yes.

Mr Dawson: The figures are very much higher, but that’s due to a combination of a greater number of infections and a –

Mr Heald: Absolutely.

Mr Dawson: Thank you.

I think as we go along the graph we see a very large peak in December of 2021, and I think you told us that that peak peaked on 29 December, which was 23,539 cases; is that what –

Mr Heald: That’s correct, yeah.

Mr Dawson: Then there is a fall in mid-January 2022. Cases remain at a level consistently above however even the September 2021 peak, and there’s a further peak which rises to about 15,000 cases I think in around about March 2022; is that right?

Mr Heald: That’s right.

Mr Dawson: Are those later peaks attributed to the Omicron –

Mr Heald: They are.

Mr Dawson: Is that correct?

Mr Heald: Yes.

Mr Dawson: Is it correct to say that lateral flow tests were used from December 2020 but until January 2022 positive lateral flow tests required a confirmatory PCR test?

Mr Heald: That’s correct.

Mr Dawson: Would that be another reason why earlier figures may appear lower than they actually were?

Mr Heald: No, because if somebody had a lateral flow test and was positive, they would –

Mr Dawson: It would appear.

Mr Heald: – have had then a confirmatory PCR test which would’ve appeared in the numbers, so –

Mr Dawson: I see, so that factor –

Mr Heald: – that factor doesn’t feature. It’s – primarily the main reason for the change in January is that you did not require to have a confirmatory PCR, and so we went with the LFD positive data from that point.

Mr Dawson: I see. Thank you very much indeed.

If I might then turn to slide 8, I had thought that this one might be for you, Mr Halliday, but again if you’re able to contribute, Mr Heald, please do so.

This is the ONS Infection Survey, is that correct?

Mr Halliday: That’s correct.

Mr Dawson: We touched upon that briefly earlier, it was the one that started in May 2020 and in Scotland was October 2020 and we discussed the reasons for that. And to what extent does – what does this illustrate? I think it’s fair to say that this plots the four nations of the United Kingdom against each other; is that correct?

Mr Halliday: That’s correct.

Mr Dawson: And just to reflect again on the basis upon which these figures are calculated, this is the prevalence basis, I think, that you described earlier; is that correct?

Mr Halliday: That’s right.

Mr Dawson: So that was a – it’s not a test based but based on proportions of the population and extrapolating out to a total?

Mr Halliday: Yeah, a random sample of the population. But it was using testing in the same – similar sort of PCR testing, but it’s not just on who comes forward, it’s a deliberately chosen random set of the population.

Mr Dawson: And can you – you helpfully have marked on the graph that it begins on – the Scottish line, the dark blue line, beginning in October for the reasons we’ve discussed, could you please just take us through the periods that represent particularly significant Scottish peaks, in particular how they sit against the UK, the position in the other UK nations.

Mr Halliday: I suppose the first thing to say is actually if you were to look at just the Scottish peak, the Scottish line by itself, then it would show something in broad terms similar to the chart that we were just describing before – that Mr Heald just described and took us through –

Mr Dawson: Could I just ask about that, that was for the reason you said earlier, which is the testing results broadly show the same thing as the prevalence type method; is that correct?

Mr Halliday: That’s –

Mr Dawson: Thank you.

And so you’re going to track the peaks for us against the UK –

Mr Halliday: Yeah. So we’ll – you’ll see the first area of interest I think is in the winter of 2020, the beginning of 2021, but there … there’s a peak in Scotland but it’s generally over the periods of December, January, February that Scotland had a lower prevalence than England and Wales, and a lower peak than Northern Ireland.

Mr Dawson: Just to put that into its context, what we’ve been looking at there – because this is a percentage of population basis, so we can plot them against each other, it’s not numbers where England would come up the highest number?

Mr Heald: Indeed.

Mr Dawson: So what we’re looking at here is that Scotland has a lower prevalence over that period, which is the Alpha variant period; would that be correct?

Mr Halliday: That’s right.

Mr Dawson: Thank you.

And if we were to go through it further, we might see that certainly there’s a – the lines in around March 2021 all seem to be roughly the same for a period; would that be right?

Mr Halliday: Yes, absolutely.

Mr Dawson: Then we see Scotland starting to rise perhaps in around about July, June 2020 to a certain extent, but then there is a peak maybe somewhere slightly before September, maybe August 2021, where Scotland is certainly representing the highest figure.

Mr Halliday: That’s right.

Mr Dawson: Okay. And then as we follow the line further across we see as we get into the latter part of – very latter part of 2020 and through into the end of the slide, would that period roughly be the Omicron period?

Mr Halliday: Yes.

Mr Dawson: Would it be fair to say that overall there Scotland plots certainly at its peak the highest but generally pretty high against the other nations?

Mr Halliday: I would say that in the peak in October – in autumn 2021 and in winter – early in 2022 that the peaks happened at different times for the different nations, with Northern Ireland followed by Scotland followed by England and Wales at broadly the same time. And while Scotland is a little bit higher in the winter – in the spring of early 2022, whether that’s a significant difference or not, certainly there’s a level of confidence in the statistics that comes from the uncertainty of not sampling everybody that – it was certainly higher – statistically higher than England but not so for Wales or for Northern Ireland.

Mr Dawson: Okay. And if we just look at the Scottish line alone, one thing which is potentially significant to our overall understanding is that whereas at around the time of the Alpha variant around 1% of the Scotland population appears to have been infected, by the time of the Omicron variant in early 2022 the peak reaches over 8% of the population infected.

Mr Halliday: Yes.

Mr Dawson: So that is linked to the fact that Omicron was a more transmissible variant.

Mr Halliday: Absolutely. And also that the restrictions that were in place at the time meant there was a lot more mixing between people than there was during the Alpha variant.

Mr Dawson: Okay. There were lesser restrictions at the period when the peak was, as compared to late – the late 2020 –

Mr Halliday: Correct.

Mr Dawson: Thank you. That’s – thank you.

Mr Halliday: Also worth just flagging at that period there’s also the impact of the vaccination, so obviously a key development in December 2020 was the start of the vaccination programme, so in terms of, yes, case numbers higher but I guess the sickness of people lower because of the impact of the vaccination.

Mr Dawson: That’s a very important observation, thank you.

Would that be how you would characterise really the impact of vaccination, in the sense that it doesn’t stop people becoming infected, but it does perhaps in some people stop the worse effects? Is that your interpretation?

Mr Halliday: My understanding is that there’s a limited impact upon transmission; the much bigger impact is upon the impact in terms of sickness and mortality, yeah.

Mr Dawson: Okay, thank you.

If that would be an appropriate point to break, that would be –

Lady Hallett: Yes, certainly.

Mr Dawson: Thank you very much, my Lady.

Lady Hallett: I shall return at 3.15.

(3.00 pm)

(A short break)

(3.15 pm)

Lady Hallett: Mr Dawson.

Mr Dawson: Thank you, my Lady.

If we could just move on to the next of the slides, thank you, I’m wanting to look at slide 15, please.

Does this slide, which is entitled “Per capita Rates of Covid-19 patients in hospital” – perhaps we could try you, Mr Heald, this time, and again if there’s anything that could be contributed – or would it be easier –

Mr Halliday: I’m happy to surrender, thank you.

Mr Dawson: This is entitled “Rates of Covid-19 patients in hospital”, March 2020 to April 2022, and shows I think a comparison between the UK and Scotland in respect of the proportion of patients in hospital with Covid-19 over a similar time to the one we looked at in the ONS study.

Mr Halliday: That’s correct, though the ONS study started in – had data from October 2020 on a consistent basis.

Mr Dawson: Right, that’s right, thank you.

And there is an indication on this one around about October 2020 that, as far as the Scottish blue line is concerned, there was some change in methodology. Could you just explain briefly what that is and in particular what effect that had on the data before and after it.

Mr Halliday: Yeah, of course, but we noticed that the pattern for Scotland and the United Kingdom was – or other parts of the UK was slightly different and did a clinical audit to investigate what the reason was, and that was done in July 2020, and there was – that found that there were actually quite a number of people in Scotland who were in hospital that had at some point been – tested positive for Covid but Covid had resolved itself and they were still in hospital, and that – that we made the decision to exclude those people once the – beyond two weeks from when they were admitted or when they tested positive for Covid – when they were in hospital, they were excluded from the figures, and that brought us onto a consistent basis with other parts of the –

Mr Dawson: So what that means, I think, is that up to a certain point there were a number of people that were being included as being in hospital with Covid who were in hospital who had had Covid but were no longer suffering from Covid.

Mr Halliday: Correct.

Mr Dawson: They were in for other –

Mr Halliday: For other reasons.

Mr Dawson: Yes.

So would that mean that before the change of methodology kicks in that we are getting perhaps a slightly inflated number for Scotland and that maybe the true line lies something nearer the UK number?

Mr Halliday: I think so.

Mr Dawson: Yes, okay, thank you.

So as we go on one can see again there are peaks, there’s a significant peak which I think represents the Alpha variant where the number of hospitalisations with Covid in the UK is above Scotland; does that reflect the figures we saw before related to the numbers of infections perhaps?

Mr Halliday: That’s exactly how I see it.

Mr Dawson: As we go on, I think there’s a slight peak around July but possibly around about August 2021 again there seems to be a significant rise of Scotland for a period above the UK average; is that right?

Mr Halliday: That’s right.

Mr Dawson: And again I think that roughly coincides with the period that we had identified as being one where Scotland’s infections went up significantly due to Delta; is that right?

Mr Halliday: That’s right.

Mr Dawson: And again beyond that, when we get in towards the end of 2021 and the Omicron wave, you see the lines mirroring each other almost exactly for a period, but then towards the end of the period that we’re interested in, this takes us up to April 2022, there’s a significant jump in Scotland as compared to the rest of the UK. Would that tend to suggest that at the very end of the period in which we are interested there is a statistically significantly higher number of people in Scottish hospitals with Covid than in the UK? Is that what it shows?

Mr Halliday: It shows that – so statistical significance with administrative data is a slightly more complicated phrase –

Mr Dawson: I wouldn’t want –

Mr Halliday: Certainly the Scotland number is of an order higher than the rest of the UK. Which is – I’m not quite sure as to the reason given that at the time – when we referred back to the Covid Infection Survey, whilst the Scottish peak was a little bit higher, it certainly wasn’t higher to the magnitudes that the hospital –

Mr Dawson: If one were to assume that one would be in hospital with Covid if one were iller with Covid, would that tend to suggest that there was a higher proportion of people in Scotland who were iller with Covid at that time at the severe end, that would make you go into hospital, than in the rest of the UK by way of average?

Mr Halliday: That’s right. And again from the vaccinations data, it just – the numbers are broadly comparable in Scotland and the rest of the United Kingdom, so it’s – I’m struggling to explain exactly what that is using the other data that’s available to us around Covid.

Lady Hallett: Mr Dawson, just going back to the Delta peak, autumn 2021, the whole of the UK suffered from the Delta variant, didn’t it, so is there any explanation for that peak or is that not statistically significant?

Mr Halliday: I think there is a noticeable difference between – I think that’s a – I’d put that in the same group as what happened in April 2022, that the levels of infection are slightly higher in Scotland, the levels of vaccination are broadly the same, and so I – beyond that it’s difficult to quite understand this. I guess one factor in this may be something to do with the underlying health conditions of people in Scotland relative to other parts of the United Kingdom.

Lady Hallett: I think that’s the point that Mr Dawson was then making.

Mr Dawson: Thank you, my Lady.

If I could take you on to the next slide, which is slide 27. This is one of the dashboards, so perhaps you, Mr Heald, on this one, “Cumulative Covid-19 deaths for Scotland”. The graph shows the cumulative number of deaths in which Covid-19 is mentioned on the death certificate rather than excess deaths; isn’t that right?

Mr Heald: That’s right.

Mr Dawson: So would these have been the figures that PHS would have been releasing or were the PHS figures based on some other statistic at the time?

Mr Heald: So my understanding is that these cumulative deaths are the deaths published by NRS, National Records of Scotland, and as we’ve talked about before, the figures from Public Health Scotland, once testing really ramped up, mirrored the figures that we saw in NRS. But these are the National Records of Scotland figures rather than the Public Health Scotland ones.

Mr Dawson: Just to be clear, we all remember that the headline statistics that would be given, for example, in daily briefings would include statistics for the previous 24 hours’ infections and mortality.

Mr Heald: Yeah.

Mr Dawson: Those would have been provided by PHS, isn’t that right?

Mr Halliday: Yeah.

Mr Dawson: And those PHS numbers for the most recent time period, I think it varied when it was a weekend for a slightly longer period or something, but would they – what would the source about the information about the mortality have been in that?

Mr Heald: So for the daily reporting that was done, the source of the data would have been Public Health Scotland. And what we did in essence was link the positive confirmed cases to very fast data we were getting from National Records of Scotland about death registrations, but those death registrations didn’t have the detail, so we only knew about people who died and we matched them to the Covid data that we had –

Mr Dawson: Because I was wondering whether I could explore with you the possibility that that, in real time, creates a statistic which says “this is the number of deaths that have been recorded” –

Mr Heald: Yeah.

Mr Dawson: – but it may be that, for example, certain circumstances of deaths would result in quicker certification, for example perhaps patients who are in hospital, than perhaps people who have died in the community where possibly whether it’s a Covid death might not be entirely clear, does that statistically change things very much or are you basically reporting the deaths –

Mr Heald: It’s the same – so it’s the same – it’s based on the registration, so it’s the same time period. What I’m highlighting is that the level of detail available when the registration first comes through –

Mr Dawson: Yes.

Mr Heald: – is not as detailed essentially as we were using the raw data about people who’d died.

Mr Dawson: Yes.

Mr Heald: NRS would then use the full record when it became available and –

Mr Dawson: I see, and I suppose overall it’s possible that the overall numbers we see here might suffer from the fact that death certification might not be entirely accurate and there might be some Covid deaths missed.

Mr Heald: I think if you look at slide 25, it’s in the pack, kind of shows the difference between what was recorded on the death certificate and what we found in Public Health Scotland. So we’ve already touched about on the kind of early period when there was a higher number –

Mr Dawson: Yes.

Mr Heald: – from NRS, and that’s down to the fact there was less testing at that time.

Mr Dawson: Yes.

Mr Heald: But you can see from the kind of July 2020 period all the way through, I would say that the Public Health Scotland and NRS figures matched pretty consistently –

Mr Dawson: I see.

Mr Heald: – so that would suggest then that the recording of Covid, when it became available on the death record, was pretty consistent with what we got by matching –

Mr Dawson: Thank you.

Mr Heald: – the deaths to the test data.

Mr Dawson: As far as 27 is concerned, we see a rise and then a plateau, if you like. Would that be roughly telling us the number of deaths that were in the first wave?

Mr Heald: Yes, so that March to –

Mr Dawson: It’s March. It plateaus at around about June 2020 and that lasts till around about November 2020.

Mr Halliday: That’s exactly right.

Mr Dawson: Yes, and I’m wondering, it might be overly simplistic, but it rather looks like you see a similar pattern in the second and what you might call the third waves, is that there’s roughly 5,000 deaths in the first wave maybe slightly more than that but around about the same in the second wave; is that right?

Mr Halliday: That’s right. It’s over – the second wave is over a slightly longer period of time but it’s broadly similar.

Mr Dawson: Yes. And then we see a rise again from around about July 2021 through to the end of the period. Again, that’s a more gradual line rather than a steep line that we saw before, but that would be a combination of Delta and Omicron, would that be correct, over that period? And again if one combines those two in that later period you see, broadly speaking, roughly the same number of deaths, it comes to somewhere around 15,000 deaths, around about slightly under 5,000 in the first part, slightly over 5,000 in the second and slightly under 5,000 in the third; is that right?

Mr Halliday: That’s right, but again it’s over a longer period of time –

Mr Dawson: Yes.

Mr Halliday: – the first to the second to the third.

Mr Dawson: Hence the lines are more or less steep.

Mr Halliday: That’s exactly it, yes.

Mr Dawson: Thank you very much.

Could I take you on now to slide 28, please. I won’t dwell too long on this one because it’s quite complicated this one, as far as I can make out.

This is reflecting something different from the previous slide, which is excess deaths rather than deaths on certificates; is that correct?

Mr Halliday: That’s correct.

Mr Dawson: And what this traces is the certification with the light blue line and excess deaths with the dark blue line, and very deliberately inviting you to try to keep your explanation as simple as possible I wondered if you could explain to us broadly why it is that the lines appear as they do.

Mr Halliday: Okay. So these two things are measuring related but different, distinctly different things.

So the deaths with Covid-19 on the death certificate are the figures from the National Records of Scotland that we’ve discussed before. Excess deaths is a measure of all deaths whether that’s related to Covid or not, and the chart here compares what happened during 2020 up to 2022 with what happened in the five years, month by month, and the numbers are above zero where there are more deaths than there would be expected at that time of the year, and it would be below zero, for example in spring 2021 and spring 2022, where there are fewer deaths than would be expected at the time of the year.

So what it shows is that there’s a high peak both in excess deaths and deaths from Covid in wave 1, and the two lines match up pretty well, and that there is then – the relationship between these – the deaths from Covid-19 and excess deaths is relatively sort of – there’s a – you know, those two things are relatively well aligned in wave 2 but then it becomes less clear cut the relationship between those two things.

Mr Dawson: Because excess deaths, as I understand it, doesn’t mean just Covid deaths, it’s the number of deaths more than would have been experienced – had been experienced in a previous time period at a given moment.

Mr Halliday: That’s right.

Mr Dawson: Would that give us some indication about deaths that occurred over this period that weren’t due to Covid?

Mr Halliday: Yes, it would do.

Mr Dawson: Thank you. There’s a couple more slides I’d like to take you to quickly. The next one is slide 34.

This is Covid mortality rates by self-reported disability category. These are, I think, age adjusted; is that correct?

Mr Halliday: That’s right.

Mr Dawson: So it would be wrong to say, for example, that disability can be equated to old age, these figures were designed to strip out that aspect of the analysis, is that right?

Mr Halliday: That’s right.

Mr Dawson: And this slide shows that, adjusted for age, those members of society who self-report as disabled had a significantly higher rate of Covid mortality when compared with non-disabled members of society; is that right?

Mr Halliday: That’s right.

Mr Dawson: Thank you.

One further slide that I wanted to go to, which is slide 35. What this slide, which is based on Scotland again, tells us, I think, are the ethnic minority communities, and those which suffered the greatest numbers of mortality in Scotland; is that correct?

Mr Halliday: Yeah.

Mr Dawson: From which we can see that the community that suffered the greatest number of deaths is the Pakistani community; is that right?

Mr Halliday: What this chart does show – I mean, that’s broadly correct. There is also a confidence interval here that says it was significantly higher also amongst the other Asian –

Mr Dawson: Yes, I see that. So the fact there’s a broad horizon line indicates the confidence interval, I think. So I suppose it’s possible that the second category is slightly more, but relatively speaking it seems to be that the Pakistani or other Asian communities suffered the greatest likelihood of death.

Mr Halliday: Absolutely.

Mr Dawson: Thank you.

Those are the questions that I have for you, thank you very much. If you just bear with me one moment.

There is nothing from the core participants, my Lady.

Lady Hallett: Thank you very much indeed. I followed nearly everything that you’ve said, which is a miracle, given my self-confessed difficulty with graphs.

I’m very grateful for your help and I hope that you found having the two of you together worked quite well. It did for us. So thank you for everything you’ve done.

Mr Halliday: Thank you.

Mr Heald: Thank you very much indeed.

(The witnesses withdrew)

Mr Dawson: I think that the next witness will be Dr Audrey MacDougall. My colleague Ms Arlidge will be dealing with her.

Ms Arlidge: My Lady, may I call Dr Audrey MacDougall.