1 00:00:03,880 --> 00:00:07,360 Okay. Can you start by saying your name and what your current title is? 2 00:00:08,340 --> 00:00:11,880 Hello, I'm David Eyre and I'm professor of Infectious Diseases. 3 00:00:12,420 --> 00:00:16,139 Okay, that's great. At can you give you a department? 4 00:00:16,140 --> 00:00:23,280 And so I'm, I'm based at the Big Data Institute, which is within the Department of Population Health. 5 00:00:23,700 --> 00:00:30,660 Okay, that's great. And without going into enormous detail with your whole life story, can you just tell me, 6 00:00:31,200 --> 00:00:37,590 starting from when you first got interested in medicine, the main milestones on your way to your current position? 7 00:00:39,050 --> 00:00:46,010 So when you get to the end of 2019, because we're going to talk a little bit about the area of work generally before we get into COVID. 8 00:00:46,820 --> 00:00:51,860 So I originally went to medicine in Cambridge in 2000, 9 00:00:51,860 --> 00:01:00,739 and then I came over to Oxford to do the clinical part of my degree in 2003 and then since then have worked in an era in around Oxford. 10 00:01:00,740 --> 00:01:07,729 So I, I started off doing my initial medical training split between the John Ratcliffe and 11 00:01:07,730 --> 00:01:15,350 the Horton Hospital and then spent some time in other hospitals in the region as well, 12 00:01:15,350 --> 00:01:20,110 and and became interested in becoming an infectious disease specialist because 13 00:01:20,120 --> 00:01:24,769 it offered a really nice combination of being able to look after patients with 14 00:01:24,770 --> 00:01:29,209 a range of problems and problems that often we could diagnose and offer treatment 15 00:01:29,210 --> 00:01:34,550 for and also some really exciting opportunities to be involved in research. 16 00:01:34,550 --> 00:01:42,020 So before starting my specialist training, I did a PhD and basically that was around the area of pathogen sequencing and particularly 17 00:01:42,020 --> 00:01:46,940 looking at an important problem in hospitals called Clostridium difficile or C2, 18 00:01:47,510 --> 00:01:56,810 and thinking about whether we could use information from sequencing of the pathogens DNA to then I'm sorry, 19 00:01:58,400 --> 00:02:01,580 you're getting the background noise through. No, I'm not hearing anything. 20 00:02:02,030 --> 00:02:06,740 I'm sorry about that. I've got I'm just going to stick to the do not disturb thing. 21 00:02:06,920 --> 00:02:10,069 So hopefully we get no no pinging in the audio. 22 00:02:10,070 --> 00:02:18,800 So if I just go back to it, say so starting off with thinking about sequencing it, see if this pathogen seeded for sale, 23 00:02:19,010 --> 00:02:21,709 which is an important problem in hospitals, 24 00:02:21,710 --> 00:02:26,480 thinking about whether we could work out how infection was spreading and what could we could be done to stop it. 25 00:02:26,480 --> 00:02:34,490 And what was really interesting there was that we we found that actually much less of it was acquired from other patients than we previously thought. 26 00:02:34,490 --> 00:02:38,660 And and that's really changed how we how we think about this infection and how to prevent it. 27 00:02:39,080 --> 00:02:44,570 And then following on from that, I did my specialist training alongside doing some further academic work, 28 00:02:44,900 --> 00:02:50,990 thinking about what we could learn about sequencing other pathogens and was involved in work, 29 00:02:52,040 --> 00:02:57,709 thinking about very antibiotic resistant gonorrhoea or super gonorrhoea in the in the media. 30 00:02:57,710 --> 00:03:04,400 And we had some involvement in cases where we had infections that were resistant to nearly all 31 00:03:04,400 --> 00:03:09,500 antibiotics that have been produced and worked on trying to understand how they'd occurred, 32 00:03:09,500 --> 00:03:12,530 but also sort of some of the messaging around how to stop spreading. 33 00:03:13,010 --> 00:03:20,510 And then also involved in an outbreak of a new emerging super fungus, something called Candida Auris here in Oxford, 34 00:03:20,960 --> 00:03:26,360 where we successfully ended a big outbreak and contributed to guidelines to stopping that. 35 00:03:26,360 --> 00:03:30,439 So and then in in around 2018, 36 00:03:30,440 --> 00:03:34,610 I finished my specialist training and became a consultant in infection and 37 00:03:34,610 --> 00:03:38,860 around the same time also took up a fellowship of the big data institute here. 38 00:03:38,870 --> 00:03:47,899 And so going into 2019 and the start of 2020, I'd got all sorts of plans for what I was going to do. 39 00:03:47,900 --> 00:03:51,830 But that has ended up changing with the with the pandemic. But we'll get to that. 40 00:03:51,830 --> 00:04:03,770 But I'd like to get go into a little bit more detail into how you use genomic sequencing to deal with the problem of infection prevention and control. 41 00:04:04,070 --> 00:04:08,180 So, I mean, it seems to me that there's there's two things you can do with a sequence. 42 00:04:08,180 --> 00:04:20,240 You can use it as a kind of barcode to label to, to finally identify a particular bug, or you can use it to try and find out how that works. 43 00:04:21,290 --> 00:04:25,549 Is that is that correct? Is that the reason? So I think that's right. 44 00:04:25,550 --> 00:04:30,890 So we can do a number of things with sequencing information. 45 00:04:30,890 --> 00:04:37,219 And one thing that we can do is to use it as a way of trying to see whether transmission is likely or whether we can rule it out. 46 00:04:37,220 --> 00:04:43,490 So as an infection spreads, if infection is spread from one person to another person or from one place to another place, 47 00:04:43,880 --> 00:04:50,120 then that infection looks really quite similar and actually in some cases completely identical when we sequence it. 48 00:04:50,120 --> 00:04:54,049 And so we can use that as a way of identifying when transmission might have occurred. 49 00:04:54,050 --> 00:04:59,209 And of course, if two things look very different and that's quite a good rule out to saying transmission hasn't 50 00:04:59,210 --> 00:05:03,740 occurred and actually we need to look somewhere else for the source of the second infection. 51 00:05:03,740 --> 00:05:08,450 So that's, that's one piece of work and and that can be used in a number of ways. 52 00:05:08,450 --> 00:05:14,660 So we can use that to study outbreaks where we've got an unusually unusual cluster of infections. 53 00:05:15,260 --> 00:05:21,229 We can also use it to monitor an infection that we're interested in over time. 54 00:05:21,230 --> 00:05:24,800 And we can look to see whether we can pick up clusters of infection that way. 55 00:05:25,160 --> 00:05:33,680 And that turns out to be quite a sensitive way of doing surveillance that you wouldn't necessarily you would have to 56 00:05:33,980 --> 00:05:38,450 you would have to expend quite a lot more effort if you were just monitoring the number of cases you had because. 57 00:05:38,700 --> 00:05:47,339 It's harder to spot sort of rises in a specific type of case if you don't have that sort of more detailed information that we get from sequencing. 58 00:05:47,340 --> 00:05:57,360 So that's sort of one area. And then that another area where there's interest in sort of looking at specific traits of the of the pathogen. 59 00:05:57,360 --> 00:06:04,010 So it might be, for example, can we predict whether an antibiotic is going to work to treat an infection using sequencing data? 60 00:06:04,020 --> 00:06:12,030 And I'm actually the big success story there has been with TB where now actually the treatment of TB in the UK is 61 00:06:12,030 --> 00:06:19,559 based on where if we find a pathogen to be completely sensitive to an oral antibiotics or based on sequencing, 62 00:06:19,560 --> 00:06:24,000 then we actually use the sequencing and don't do any further diagnostic work. 63 00:06:25,200 --> 00:06:28,590 We rely on the sequencing data to sort of make decisions about treatment. 64 00:06:29,790 --> 00:06:34,800 It's it's not quite so advanced in other areas and it's still a sort of research technique, 65 00:06:34,950 --> 00:06:39,420 but largely but but TB is certainly one one success story there, 66 00:06:39,780 --> 00:06:46,440 and that can be particularly helpful where something takes a long time to grow in the lab inside a traditional sort of lab workflow is slow. 67 00:06:47,040 --> 00:06:53,880 And then we might sort of with TB previously, it may take several weeks to get an answer and now we can get an answer within a week or two. 68 00:06:54,180 --> 00:06:57,450 And there are also times where we can't grow a pathogen. 69 00:06:57,450 --> 00:07:03,360 So some things that simply it's just very difficult to grow in the lab and we can sort of make a diagnosis using sequencing. 70 00:07:03,840 --> 00:07:08,820 And then increasingly it is interesting, can we just detect things faster using sequencing? 71 00:07:08,860 --> 00:07:15,659 So there's a whole area of research called better genomic sequencing where we try and without using the sort of 72 00:07:15,660 --> 00:07:21,330 traditional culture based methods that we've been using in microbiology for the last last hundred years or so, 73 00:07:21,330 --> 00:07:26,670 we actually just try and go straight from the original sample and see what's in it using sequencing. 74 00:07:27,480 --> 00:07:32,660 And you mentioned that you're now based in the Big Data Institute. So where does where does big data come into all this web? 75 00:07:32,700 --> 00:07:37,320 Where does computer analysis extend what you do? 76 00:07:37,980 --> 00:07:47,340 So a typical typical virus is has a sort of genome sequence that's maybe a few thousand acres jus inches long, 77 00:07:47,340 --> 00:07:49,530 whereas a bacteria, it might be a few million. 78 00:07:50,040 --> 00:07:58,589 And so and when we do sequencing, typically we can't get that entire sequence all in one, get it, we get it in fragments. 79 00:07:58,590 --> 00:08:07,799 And so actually what we need to do is to to reconstruct the the sequence data that was there in the original pathogen from lots of small pieces. 80 00:08:07,800 --> 00:08:10,140 And so this is sort of rather like putting together a jigsaw puzzle. 81 00:08:10,440 --> 00:08:16,530 And you can do this through a number of ways, some of which involve sort of knowing what it looks like, 82 00:08:16,530 --> 00:08:20,909 sort of analogous to having the picture on of what the jigsaw supposed to look like. 83 00:08:20,910 --> 00:08:25,260 And then the other way is really just looking at overlap of bits of the sequencing data 84 00:08:25,260 --> 00:08:29,520 and trying to piece it together just from sort of what looks similar around the edges. 85 00:08:29,520 --> 00:08:42,270 But doing that involves a degree of computational computational processing and then trying to make inferences from the sequencing data that we get. 86 00:08:42,270 --> 00:08:48,390 So trying to piece together how related the different sequences are or to look for specific 87 00:08:49,440 --> 00:08:54,180 pieces of sequence within the overall sequence also requires a degree of computation. 88 00:08:54,900 --> 00:09:01,920 The the other thing that and one of the big reasons that I'm based at the BDA is actually we're not just looking at sequencing data in isolation. 89 00:09:01,920 --> 00:09:05,460 Often we're looking at it in conjunction with other health care data. 90 00:09:05,880 --> 00:09:09,280 So we're thinking about maybe how patients move around the hospital. 91 00:09:09,300 --> 00:09:16,300 So piecing together data from electronic patient records on when patients are admitted, which wards they go to, 92 00:09:16,300 --> 00:09:21,600 who they share time and space with while they're in hospital and and sort of coming out of that. 93 00:09:21,600 --> 00:09:27,719 I've also become increasingly interested in how we can use electronic health care data actually as a tool in itself. 94 00:09:27,720 --> 00:09:29,820 So not necessarily just with sequencing data, 95 00:09:29,820 --> 00:09:38,460 but can we use it to better understand infection in terms of the epidemiology of infection and the transmission of infection, 96 00:09:39,030 --> 00:09:42,359 but also thinking about diagnosis, prognosis and treatment? 97 00:09:42,360 --> 00:09:48,450 So can we use information on what happens in routine care within hospitals to then try and make 98 00:09:48,630 --> 00:09:54,630 try and give advice on how to better manage manage patients and and also not just patients, 99 00:09:54,630 --> 00:09:57,360 but sort of thinking about hospitals as systems as a whole. 100 00:09:57,660 --> 00:10:04,130 And can we can we control infection, spread of infection in hospitals better, but can we also just manage hospitals better? 101 00:10:04,140 --> 00:10:12,240 Can we cope with peaks of seasonal infection and be able to sort of help help patients journeys through hospitals be more efficient? 102 00:10:12,240 --> 00:10:16,500 And and so that so that is really not just sequencing, 103 00:10:16,500 --> 00:10:21,060 but that sort of come out as a sort of increasing, increasing interest in the practice of works of. 104 00:10:22,560 --> 00:10:25,890 Are a bit before COVID, but certainly the afterwards as well. 105 00:10:27,100 --> 00:10:31,120 So let's just one final question, because not everybody necessarily knows. 106 00:10:31,630 --> 00:10:36,160 Can you tell me a bit about the genome sequencing technology that's available to you now? 107 00:10:36,310 --> 00:10:42,880 Because I think people may have heard the story of how the human genome sequence was originally 108 00:10:43,570 --> 00:10:50,410 deciphered and it cost $3 billion and took is clearly we're not still in that in that world. 109 00:10:51,220 --> 00:11:02,680 Yes. And so when when I first started thinking about what I was going to do for my Ph.D. and sort of towards the end of 2008, 110 00:11:02,680 --> 00:11:05,350 2009, sequencing was really an emerging technology. 111 00:11:05,350 --> 00:11:10,830 And we thought if we're lucky, we might just be able to sequence a few bacteria and that would be a big, big success. 112 00:11:10,840 --> 00:11:14,440 And what really happened is over over the few years following from that, 113 00:11:14,440 --> 00:11:23,799 there was really a kind of explosion in the availability of sequencing data such that actually we could see we could go from thinking we'd only built 114 00:11:23,800 --> 00:11:31,870 sequence a handful of things to actually sequencing thousands of bacterial genomes and probably doing so for around about the sort of £50 genome, 115 00:11:32,170 --> 00:11:41,200 which was really unthinkable previously. And what drove a lot of that was this sort of expansion in what we call short read sequencing technology, 116 00:11:41,200 --> 00:11:47,470 where we could sequence up to several hundred base pairs of several hundred 117 00:11:47,890 --> 00:11:53,260 genes and tease of the bacterial genome in one go and generate many thousands or 118 00:11:53,260 --> 00:11:58,629 millions of these these bacterial reads that would then put together what's what's 119 00:11:58,630 --> 00:12:03,640 happened more recently is there've been different different approaches taken, 120 00:12:03,640 --> 00:12:06,310 including approaches where you can generate much longer. 121 00:12:06,580 --> 00:12:11,560 So rather than just fragments of sequencing of a few hundred long getting up to several thousand long 122 00:12:11,570 --> 00:12:16,300 and this has the property because it makes it it much easier to reconstruct these bacterial genomes. 123 00:12:16,310 --> 00:12:22,780 And and also we've seen changes in the way that technology is deployed from being something where you'd 124 00:12:22,780 --> 00:12:27,969 have to generate big batches of sequences to where you can actually it's not quite random access, 125 00:12:27,970 --> 00:12:32,410 but that sequence sort of a relatively small number of things together, 126 00:12:32,410 --> 00:12:36,040 which has meant that we're able to to offer sequencing and sort of more real time, 127 00:12:36,580 --> 00:12:40,210 real time service when we're thinking about how we want to think about wanting to deployed. 128 00:12:40,600 --> 00:12:45,900 Mm hmm. So, I mean, do hospitals already have sequencing available? 129 00:12:45,910 --> 00:12:49,120 It's something it's done relatively routinely. Yes. 130 00:12:50,050 --> 00:12:55,360 So the it's largely been a research tool or something that reference laboratories have used. 131 00:12:56,620 --> 00:13:05,919 One of the things that has happened, and we may come on to talk to is about is that because of everything that's happened in Cambridge, 132 00:13:05,920 --> 00:13:09,670 the capacity of routine laboratories in hospitals to. 133 00:13:11,520 --> 00:13:18,810 Undertake the sorts of molecular microbiology work that are involved in sequencing has gone up substantially. 134 00:13:19,230 --> 00:13:24,870 And so with that, there's been a clear focus on sequencing COVID as well. 135 00:13:25,140 --> 00:13:28,920 And that's and that's meant that both the sequencing machines, 136 00:13:28,920 --> 00:13:36,500 but also the kind of familiarity and competence of a stopwatch in hospitals is massively increased. 137 00:13:36,510 --> 00:13:43,830 And so, for example, in our own hospital, we would now have been able to routinely sequence COVID. 138 00:13:44,610 --> 00:13:49,040 And I'd also thinking about routinely sequencing a number of a number of other things, 139 00:13:49,050 --> 00:13:56,560 But there's still challenges around maybe demonstrating that the sequences cost effective and that it may be a nice to have things. 140 00:13:57,120 --> 00:14:01,139 But does it actually does it actually improve outcomes for patients or outcomes hospitals? 141 00:14:01,140 --> 00:14:05,930 And I think this is probably still a bit more work to be done there. Right. 142 00:14:05,940 --> 00:14:15,839 Well, I think we've arrived at coded. So can you remember when you first heard that there was an outbreak happening in China and how soon it was 143 00:14:15,840 --> 00:14:21,840 before you realised that this was something that was going to have a major effect on you and your colleagues. 144 00:14:22,800 --> 00:14:32,910 So I think we still would remember the exact dates, but we certainly became increasingly aware as reports came out of China, 145 00:14:32,910 --> 00:14:39,300 that there was this unexplained respiratory problem and, and then actually had. 146 00:14:40,280 --> 00:14:45,880 A number of visitors to Oxfordshire from hand in the in the run up to it. 147 00:14:45,890 --> 00:14:51,800 And I was involved as the infectious diseases consultant on COOL and we had a number of queries about sort of what. 148 00:14:52,730 --> 00:14:56,750 What should be you know, what should be happening and what would happen if people became unwell. 149 00:14:57,020 --> 00:15:02,599 And then we gradually got more and more calls about what to do with people who come 150 00:15:02,600 --> 00:15:06,980 back from various parts of the world where there was concern about about COVID. 151 00:15:08,090 --> 00:15:14,870 And we had some quite initially sent quite some quite clear but sort of restricted guidelines about who 152 00:15:14,870 --> 00:15:19,330 needed testing and then gradually became sort of more and more aware that this was a sort of growing, 153 00:15:19,940 --> 00:15:20,990 growing issue. 154 00:15:21,350 --> 00:15:29,850 And and I think from a from a clinical point of view, this initially was something that sort of the infectious diseases team dealt with, 155 00:15:29,850 --> 00:15:36,110 but it became increasingly apparent that this was really going to be something that impacted the whole the whole hospital. 156 00:15:36,680 --> 00:15:45,649 And it's I think although although we hoped it wouldn't become a big problem, it's quite interesting, actually, 157 00:15:45,650 --> 00:15:50,240 even looking back at a sort of WhatsApp message sent between me, my wife and her colleagues at the time, 158 00:15:50,240 --> 00:15:54,379 sort of in February, in February, saying, you know, this could be a really big problem, 159 00:15:54,380 --> 00:16:00,140 it could have a substantial economic impact and really being aware that it could be a big problem. 160 00:16:00,140 --> 00:16:06,780 But then. I interestingly myself reflecting that I then sort of took minimal personal action to sort of, 161 00:16:06,780 --> 00:16:12,190 you know, I still run out of food along with everyone else when we ran out of food or or, you know, 162 00:16:12,220 --> 00:16:18,570 sort of failed to sort of fail to sort of take all sorts of actions that you maybe could have done with this prior knowledge, 163 00:16:18,630 --> 00:16:23,280 because maybe deep down sort of hoping that it just wouldn't become such an issue. 164 00:16:23,860 --> 00:16:26,550 And. That said, I was lucky. 165 00:16:26,950 --> 00:16:35,170 I managed to go to my cousin's wedding on the 29th of February, and in 2020, a big, big wedding with over 250 people in London. 166 00:16:35,710 --> 00:16:42,130 And so, so obviously wasn't so worried that I because I was still prepared to go to that. 167 00:16:42,490 --> 00:16:49,870 And and then and then actually subsequent to that personally became unwell and developed a fever and 168 00:16:49,870 --> 00:16:53,559 cough and was subsequently antibody positive where I thought at the wedding or from somewhere else, 169 00:16:53,560 --> 00:17:00,010 I don't know, but spent much of that sort of first the sort of middle of March actually feeling ill 170 00:17:00,010 --> 00:17:06,790 and being off work and sort of not feeling necessarily a part of what was going on. 171 00:17:06,790 --> 00:17:13,569 But then sort of towards the end of March and coming back into the hospital and and really in that first part of the pandemic, 172 00:17:13,570 --> 00:17:17,950 actually just being a part of a medical team looking after after patients in the hospital. 173 00:17:18,370 --> 00:17:26,290 Yes. I was going to ask you about that, because obviously you clearly had a need to deal with patients and work on your clinical side. 174 00:17:26,980 --> 00:17:33,730 But that was also, I mean, across the university, a very quick realisation that this was a massive research opportunity. 175 00:17:33,820 --> 00:17:38,410 Indeed, research is going to be necessary in order to tackle the pandemic, 176 00:17:38,920 --> 00:17:43,870 and it had to be done while there were cases you had to do it while you had the pandemic going on. 177 00:17:44,890 --> 00:17:48,460 So how soon did you know? Let's do the clinical side first. 178 00:17:48,730 --> 00:17:54,520 So what what how were you deployed? And as patients began to come into the hospital with COVID, 179 00:17:55,270 --> 00:18:03,040 So we we had several wards which were sort of given over to looking after patients with COVID. 180 00:18:03,040 --> 00:18:11,529 And I was involved in being the consultant running one of those, and you sort of organised the wards into the level of support that people needed. 181 00:18:11,530 --> 00:18:15,010 So there were some wards where people needed to go to the intensive care unit, 182 00:18:15,020 --> 00:18:21,280 other units where people still needed oxygen, but maybe less, less support with the with their breathing. 183 00:18:21,280 --> 00:18:24,580 And so they were on the the ward that I was helping run. 184 00:18:26,470 --> 00:18:35,299 And. I have to say, initially it was really quite a frustrating experience because as I said at the beginning, 185 00:18:35,300 --> 00:18:44,150 I think in infectious disease, often we we can we there's both often a bit of a challenge about working out what's wrong with people. 186 00:18:44,330 --> 00:18:51,139 And then secondly, we then usually able to treat things, whereas here it was fairly obvious what was wrong with people, 187 00:18:51,140 --> 00:18:54,410 but we were not able really to offer much meaningful treatment. 188 00:18:54,410 --> 00:19:00,380 And so, so both those things really quite, quite different and difficult to what we were used to. 189 00:19:01,490 --> 00:19:09,940 And, and I think really a very challenging time in that actually, you know, lots, lots of people have pushed me to either nothing I, 190 00:19:10,520 --> 00:19:16,129 I certainly remember times where you'd come into the hospital and actually I just 191 00:19:16,130 --> 00:19:20,180 really vividly remember coming in one Saturday morning and there was all this backed 192 00:19:20,180 --> 00:19:26,719 up possessions that had been left from people who died the night before or the day 193 00:19:26,720 --> 00:19:32,150 before that were just waiting on the ward and electronic bed ward of patients. 194 00:19:32,150 --> 00:19:37,580 And the hospital was both full of patients who had been there and patients who'd been admitted. 195 00:19:37,970 --> 00:19:42,140 And there was just no one necessarily available to update the system. 196 00:19:42,140 --> 00:19:50,190 And so it sort of this sort of rather rather chaotic, but everyone trying their best to try and help people as best they could. 197 00:19:51,560 --> 00:19:56,150 So I think I think that was a very difficult time and I think people were also quite. 198 00:19:57,150 --> 00:20:03,030 Quite worried and lots of uncertainty and we've not really experienced anything like this before. 199 00:20:03,030 --> 00:20:09,329 And in infectious diseases we have a relatively good amount of training around using personal protective equipment. 200 00:20:09,330 --> 00:20:19,559 But actually even having had the people stuck in it, people still feeling personally worried about about their own health and safety as well. 201 00:20:19,560 --> 00:20:23,640 And certainly that colleagues around you not had so much training. 202 00:20:23,740 --> 00:20:26,910 So, yeah, I think a really challenging time. 203 00:20:26,910 --> 00:20:33,690 And I think you mentioned research and I think there was this sort of tension where I was off feeling unwell and then I was working in the 204 00:20:33,690 --> 00:20:40,710 hospital and sort of being aware peripherally that there were a number of people then with various bits of research work starting up and. 205 00:20:41,770 --> 00:20:45,670 I think I had quite mixed feelings about that because from a personal perspective that, 206 00:20:45,760 --> 00:20:51,280 you know, that actually there opportunity to contribute that I was potentially missing out on, 207 00:20:51,310 --> 00:20:55,330 but also recognising that there was an important, 208 00:20:55,570 --> 00:21:04,070 important role to do in the hospital to say this is a bit of a sort of conflict in terms of how I felt about so. 209 00:21:04,350 --> 00:21:10,329 So that was really the the sort of first first few weeks of the pandemic. 210 00:21:10,330 --> 00:21:15,700 But then I then ended up getting involved in a number of different things, 211 00:21:15,700 --> 00:21:25,420 mostly around the area of diagnostics and also around a programme testing healthcare workers in Oxfordshire. 212 00:21:26,590 --> 00:21:31,780 And then following that, actually quite a lot of work to do with antibody testing and using antibodies as well, 213 00:21:31,780 --> 00:21:38,960 tracking response to to vaccination, infection, sex. Yeah, happy to sort of pick up, pick up on each of those. 214 00:21:38,990 --> 00:21:47,380 Yes. Yeah. That's all of that. So, so one of the first things I ended up doing was there was a hope that. 215 00:21:48,560 --> 00:21:55,340 The sort of lateral flow tests that we become very used to using for detecting whether people would have infection initially, 216 00:21:55,340 --> 00:22:00,470 that you would be able to use these for detecting antibodies and seeing who had been infected. 217 00:22:00,770 --> 00:22:05,720 And then maybe if you knew who'd been infected, then you might be able to relax restrictions for those people. 218 00:22:06,670 --> 00:22:15,560 And there were a number of tests that were procured from various suppliers, which we were then involved in evaluating in Oxford. 219 00:22:16,580 --> 00:22:23,300 And so. Quite a few people were involved in going around taking blood from people who had been infected. 220 00:22:23,930 --> 00:22:31,430 In fact, some of the earliest cases of people who'd been infected sort of driving around the countries that are going and seeing those people and 221 00:22:31,430 --> 00:22:40,339 asking if we take blood from them so that we we had samples from people where we knew we sort of had that what happened to their antibodies. 222 00:22:40,340 --> 00:22:43,310 And then we had a number of blood samples that were, say, from before the pandemic, 223 00:22:43,670 --> 00:22:49,700 which we could use to work out sort of whether whether we got negative results when we expected to get negative results. 224 00:22:50,930 --> 00:22:57,170 And we tested a number of these different devices, and it turned out that none of them really worked very well. 225 00:22:57,560 --> 00:23:03,070 And. This is sort of frustrating on a number of levels. 226 00:23:03,080 --> 00:23:05,899 It would have been obviously really helpful to have had that information. 227 00:23:05,900 --> 00:23:13,100 But it also meant that a number of the tests that had been had been bought then then could be couldn't be used. 228 00:23:13,100 --> 00:23:19,250 And it was sort of quite, quite an interesting initial experience at the sort of interface between 229 00:23:19,880 --> 00:23:26,030 research and then subsequent policy and that sort of quite a bit of discussion 230 00:23:26,030 --> 00:23:33,980 at the time about what we could and couldn't publish in the publication that we eventually released didn't actually name the the device manufacturers, 231 00:23:35,570 --> 00:23:40,250 but there was sort of quite, quite clear contractual reasons why that had to be the case, 232 00:23:40,250 --> 00:23:43,520 but quite different to how we might normally normally do science. 233 00:23:43,520 --> 00:23:53,300 But I think that the main message that came out of that was that we knew that these tested didn't work. 234 00:23:53,840 --> 00:23:57,720 So what? Well, you tell what were you testing them against? What was your gold standard? 235 00:23:57,950 --> 00:24:00,859 The gold standard was people we knew who had an infection. 236 00:24:00,860 --> 00:24:07,099 So people who tested positive on a PCR test for infection and these were some of them were were health care workers, 237 00:24:07,100 --> 00:24:08,720 some of them were people who were sort of early, 238 00:24:08,960 --> 00:24:17,840 early case early cases from from yeah, from around the country and had blood samples taken at different time intervals after they'd become infected. 239 00:24:18,500 --> 00:24:25,100 And then the sort of negative controls with the blood samples which we had stored from from before the pandemic. 240 00:24:25,100 --> 00:24:31,280 So going back. Oh, I see. Did you not have any method at that stage that could detect antibodies? 241 00:24:31,850 --> 00:24:39,319 So So we were yes. So we were relying on the fact that we were essentially saying, well, how many times in people have been infected? 242 00:24:39,320 --> 00:24:41,240 Do they did these devices test positive? 243 00:24:41,630 --> 00:24:47,870 And then if people who we knew hadn't been infected because of the time window, how many times did they test negative? 244 00:24:48,350 --> 00:24:52,669 And what we found was simply that we got some of the positive when they shouldn't do. 245 00:24:52,670 --> 00:24:56,420 And then we also got quite a few that didn't test positive when we expected them to test positive. 246 00:24:56,780 --> 00:25:06,680 And as you say, we didn't have a good antibody test, sort of a lab antibody test at the time to to compare performance against. 247 00:25:07,850 --> 00:25:15,259 And so that really comes on quite nicely to to sort of second second area of work that I was involved in, 248 00:25:15,260 --> 00:25:19,579 which was actually around developing an antibody, a better antibody test. 249 00:25:19,580 --> 00:25:28,010 So we'd seen these devices didn't work very well. So it was quite clear we needed something that worked better and colleagues here in the university 250 00:25:28,010 --> 00:25:35,270 were involved in sort of rapidly trying to trying to try to develop synthetic antibodies, 251 00:25:35,270 --> 00:25:41,690 monoclonal antibodies that could mimic the kinds of antibodies that were produced. 252 00:25:41,690 --> 00:25:47,420 And then also the the coating of the SARS-CoV-2 virus, the spike protein, 253 00:25:47,660 --> 00:25:54,590 also creating synthetic versions of that so that you could use the synthetic version of that to capture antibodies, 254 00:25:54,590 --> 00:25:56,660 which you could then could then measure. 255 00:25:57,350 --> 00:26:08,059 And what I was involved in was taking some of the initial synthetic antigen, a synthetic viral spike protein and synthetic antibodies, 256 00:26:08,060 --> 00:26:12,800 and turning that into a sort of more high throughput antibody testing platform. 257 00:26:13,430 --> 00:26:20,630 And probably to summarise what was a lot of work quite quickly, I mean, it ended up being successful, 258 00:26:21,920 --> 00:26:28,130 but it was clear that in order to have an impact, it would need to be delivered at really quite, quite a quite wide scale. 259 00:26:28,490 --> 00:26:37,280 And so we initially went to colleagues at the the Target Discovery Institute here in Oxford who worked on screening a large number 260 00:26:38,820 --> 00:26:47,450 of drug compounds and other compounds very rapidly using robotics and said could they help running antibody tests at scale? 261 00:26:47,810 --> 00:26:54,799 And so we initially got something set up there and a subject this was this was Daniel Hepner and Stephanie Hatch, was it? 262 00:26:54,800 --> 00:27:03,950 Yes, that's yes, right. Yeah. I got something set up with Donald Leiber and Stephanie Hatch and and then subsequently. 263 00:27:05,070 --> 00:27:10,140 Actually, then we're able to port that to an even higher throughput robotic robotic platform. 264 00:27:10,900 --> 00:27:17,070 That and that and that sense has delivered over a over sort of to 2 million antibody 265 00:27:17,310 --> 00:27:23,370 tests that supported the National Office for National Statistics COVID Infection survey. 266 00:27:23,410 --> 00:27:30,660 So that's really been given us an enormous understanding of what happens to people after they get infected, 267 00:27:30,660 --> 00:27:37,560 what happens to their antibodies, how they rise and then fall, also how antibodies respond to vaccination. 268 00:27:37,860 --> 00:27:44,700 And then putting that together, thinking about sort of what levels of antibodies are associated with protection against infection. 269 00:27:45,180 --> 00:27:49,919 And we've been able to sort of forward then project as to how long immunity from 270 00:27:49,920 --> 00:27:55,170 vaccination might last or how long immunity from previous infection levels as well. 271 00:27:55,850 --> 00:28:00,980 So we might we might just do the those conclusions. Well, while we're here, Ross isn't going to do it. 272 00:28:01,780 --> 00:28:09,060 So what what were your main discoveries? I mean, yeah, over the next 18 months or so, probably. 273 00:28:10,380 --> 00:28:16,290 So I think we what was so what, what came out was that. 274 00:28:17,670 --> 00:28:25,200 Antibodies were when you got infected. We did see in most people antibodies increase and rise. 275 00:28:25,200 --> 00:28:32,309 And actually it depended a little bit on how new or how you detected the infection, 276 00:28:32,310 --> 00:28:42,540 but probably in at least three quarters and possibly more people who were exposed and tested positive at some time for the virus that causes COVID. 277 00:28:42,540 --> 00:28:48,359 SARS-CoV-2 They then developed an antibody response and those antibodies rose 278 00:28:48,360 --> 00:28:53,070 and then and then lasted for for at least several for at least several months, 279 00:28:54,090 --> 00:28:59,430 possibly even possibly even sort of several years after that after people were infected. 280 00:29:00,540 --> 00:29:04,349 And we initially thought that this might offer quite a long standing protection against reinfection. 281 00:29:04,350 --> 00:29:10,579 But what happened was, is that the virus changed, and each time that the virus changed and evolved, 282 00:29:10,580 --> 00:29:19,830 then actually you tended to need more and on higher levels of antibodies to try and offer the same level of protection. 283 00:29:20,610 --> 00:29:26,070 And and so we also found that the vast majority of people, when they were given the vaccines, 284 00:29:26,460 --> 00:29:32,790 responded in probably as many as 95% or more people responded to an initial course of two 285 00:29:32,790 --> 00:29:39,329 vaccines and produced antibody levels that were as high or even higher than we possibly saw. 286 00:29:39,330 --> 00:29:45,569 So when in fact after infection and but those antibodies were possibly less, 287 00:29:45,570 --> 00:29:50,040 less longer lived, they lasted a few months rather than sort of a couple of years. 288 00:29:50,160 --> 00:29:55,879 And they were predicted to last a few months rather than a couple of years at levels that often offered protection. 289 00:29:55,880 --> 00:30:02,490 And so this really fed into some of the some of the booster campaigns which came along to both reflecting the fact 290 00:30:02,490 --> 00:30:08,069 that we saw waning of protection when we could follow sort of whether people were infected and after vaccination. 291 00:30:08,070 --> 00:30:12,270 But also we could see this sort of waning antibody levels at the same time. 292 00:30:13,560 --> 00:30:16,410 And and then as people were boosted, 293 00:30:16,740 --> 00:30:25,590 we then saw even more people respond and get response and get responses to vaccination in terms of the antibody levels that we saw. 294 00:30:25,770 --> 00:30:31,590 But again, trailing off after they were they were boosted and then as each new successive variant came out. 295 00:30:31,600 --> 00:30:39,509 So then the Delta and Omega from that, the the amount of antibody that you needed against the original spike protein and the original virus 296 00:30:39,510 --> 00:30:47,970 needed to be sort of progressively higher in order to get similar levels of protection and and. 297 00:30:49,750 --> 00:30:52,500 And then. This sort of most recently, 298 00:30:52,500 --> 00:31:02,070 we've been thinking about the difference between the protection that you get when you have a have a reinfection or get 299 00:31:02,080 --> 00:31:08,969 infected after you've been vaccinated versus the sort of ongoing protection that you get from various booster campaigns. 300 00:31:08,970 --> 00:31:15,090 Because what we've what we've seen is that although we've continued to try and get regular 301 00:31:15,090 --> 00:31:19,890 boosters to those who are most vulnerable to to sort of poor outcomes from infection, 302 00:31:21,000 --> 00:31:24,240 there are a number of people now who aren't being routinely offered vaccination. 303 00:31:24,270 --> 00:31:27,000 So we wanted to understand what was what might happen to them. 304 00:31:27,480 --> 00:31:34,740 And it it turns out that in people who have been vaccinated and then go on to get infected, 305 00:31:35,040 --> 00:31:43,530 that they they then tend to generate relatively sustained antibody responses suggesting that they may be protected from infection, 306 00:31:43,770 --> 00:31:52,120 getting infected again for some time. Whereas the people we need to distinguish, don't we, between getting really ill and getting infected. 307 00:31:52,140 --> 00:32:01,410 So yes, and I think that's been the. The other success story really with vaccination has been that although maybe people 308 00:32:01,410 --> 00:32:04,590 haven't been protected from reinfection as much as we might have hoped initially, 309 00:32:04,590 --> 00:32:11,700 actually, the degree of severe illness has gone down and and possibly also that's been that's 310 00:32:11,700 --> 00:32:16,160 been a consequence of changes that have happened in the SARS-CoV-2 virus as well. 311 00:32:16,170 --> 00:32:22,530 But as time has gone on, the infection that's been caused has been less severe as well, particularly for the combination. 312 00:32:24,940 --> 00:32:28,000 So that's the antibody side of things. And. 313 00:32:29,620 --> 00:32:35,440 Which we are back to. I've got a note here that says Web based application to support trust staff and collect data. 314 00:32:36,070 --> 00:32:42,370 Yes. So. So we. The other thing that sort of working in the hospital motivated us to think about was, well, 315 00:32:42,370 --> 00:32:47,649 what what's happening to stop, you know, our staff more more at risk if they don't know or not. 316 00:32:47,650 --> 00:32:51,580 And if they are, well, why and what can we do to to protect them? 317 00:32:52,690 --> 00:33:01,329 And what we did back in April 2020 was we were fortunate to sort of make it make a bid directly to the Department Health and Social Care to say, 318 00:33:01,330 --> 00:33:09,160 could we set up large scale testing of health care workers in Oxford as a way of trying to understand risk to health care workers? 319 00:33:09,340 --> 00:33:14,590 And some of our really key questions about COVID at the time, You know, did people get reinfected? 320 00:33:15,340 --> 00:33:16,660 How often did that happen? 321 00:33:17,900 --> 00:33:28,280 And so what we what we did was to start by offering staff regular antibody testing and also regular testing for infection with PCR testing and. 322 00:33:29,290 --> 00:33:34,570 In order to do that, we had there was an enormous amount of logistics involved in that, actually. 323 00:33:36,280 --> 00:33:40,300 The hospital employs over 13 and a half thousand staff. 324 00:33:40,600 --> 00:33:52,509 And so being able to offer them regular PCR testing, all of which was performed by by a team we put together and antibody testing, 325 00:33:52,510 --> 00:33:57,190 which was done through blood tests that were also done by the same team, was a really big undertaking. 326 00:33:57,320 --> 00:34:05,620 And part of that was it became quite clear that we needed a system for for people to be able to book into appointments. 327 00:34:05,620 --> 00:34:11,590 We needed a system for people to be able to receive their results because we felt that feeding back to people what the test results were, 328 00:34:12,070 --> 00:34:15,820 was sort of a really important way of getting them engaged and getting them to participate. 329 00:34:16,090 --> 00:34:20,080 People particularly wanted to know what their antibodies were, you know, had they been infected in the past. 330 00:34:22,060 --> 00:34:28,389 And also we wanted a system that was safe because part of the reason for doing this was that we wanted to know when people were healthcare workers 331 00:34:28,390 --> 00:34:36,520 were infected so that we could then get them home to isolate them and protect other patients in hospital and also other staff from getting infected. 332 00:34:37,570 --> 00:34:43,540 And so I said I set up what I thought initially was just going to be a sort of short term fix. 333 00:34:43,540 --> 00:34:54,339 Was it which was a sort of web based application that enables people to to register their interest and to and to look into slots. 334 00:34:54,340 --> 00:34:56,379 And I think when we first made it live, 335 00:34:56,380 --> 00:35:02,830 we had thousands of appointments made within the first few minutes of it being available and it performed and it worked, 336 00:35:02,840 --> 00:35:06,520 didn't fall over, which is certainly sort of a nice, nice feeling. 337 00:35:07,600 --> 00:35:12,250 And then we then got people regularly coming, coming to complete the testing and, 338 00:35:12,460 --> 00:35:18,430 and that was possible because we had an amazing group of research nurses who switched from doing all sorts of things, 339 00:35:18,430 --> 00:35:24,489 including some of the sort of hepatology research nurses supported by lots and lots of medical students, 340 00:35:24,490 --> 00:35:35,260 other people from around the hospital who were redeployed, helping us both with sort of supporting the as supporting the actual testing of people, 341 00:35:35,260 --> 00:35:43,120 swabbing people, and then also some of the sort of additional cooling outs of results, the sort of follow up of people who tested positive. 342 00:35:44,680 --> 00:35:53,350 And and then from that, we we collected the sort of first thing we did was to work out who'd had COVID, and we showed that. 343 00:35:54,400 --> 00:35:58,020 More health care workers have had COVID than in the general population. 344 00:35:58,690 --> 00:36:03,280 So I think we showed our sort of initial result was towards the end of April, early May, 345 00:36:03,280 --> 00:36:09,189 that around ten or 11% of our health care workers had had COVID compared to estimates from antibody tests. 346 00:36:09,190 --> 00:36:12,400 And another test in the community of around 5%, 347 00:36:13,540 --> 00:36:18,129 And we also showed that staff looking after patients with COVID were even greater risk and 348 00:36:18,130 --> 00:36:24,130 actually that was sort of up around the 20% mark or so in areas of the hospital where people, 349 00:36:24,130 --> 00:36:32,380 particularly patients with COVID. We also showed that actually different staff groups were differentially at risk. 350 00:36:33,340 --> 00:36:36,970 So different medical specialities have different degrees of risk. 351 00:36:37,000 --> 00:36:45,500 People particularly working in acute medicine, looking after unwell patients coming first, coming into hospital. 352 00:36:46,900 --> 00:36:51,180 Then also different different sort of staff groups within the hospital. 353 00:36:51,190 --> 00:36:57,459 So nurses and health care assistants, porters and other support staff at greater risk. 354 00:36:57,460 --> 00:37:03,290 And then also staff of Asian and black ethnicity also were at greater, greater risk, 355 00:37:03,290 --> 00:37:07,719 even even an adjusted analysis, adjusting for other factors that we could measure. 356 00:37:07,720 --> 00:37:15,310 Six Coming out of that, we it really became clear that we needed to do more to protect staff. 357 00:37:15,310 --> 00:37:23,390 And we also had, unfortunately, several staff who died of COVID as well and those in the early months of the pandemic. 358 00:37:23,410 --> 00:37:26,820 And so there is a sort of multi, multi sort of action, 359 00:37:27,280 --> 00:37:30,909 sort of multi-point action plan that went to the trust board and was largely 360 00:37:30,910 --> 00:37:37,600 implemented in full as a result of some of the findings that we had from that. 361 00:37:37,600 --> 00:37:46,149 And following on from that, actually, we had some of the lower rates of hospital associated COVID compared to other other hospitals in the country. 362 00:37:46,150 --> 00:37:52,870 And people were sort of coming to us as a sort of source of best practice for how to help try and prevent transmission in hospitals. 363 00:37:52,880 --> 00:37:57,310 So so that was a really, really positive initial outputs of that. 364 00:37:57,580 --> 00:38:02,200 But having tested all these people in the first wave of the pandemic, 365 00:38:02,200 --> 00:38:08,350 then when the second wave of the pandemic came along in the autumn of 2020, we were in a position to ask, well, 366 00:38:08,650 --> 00:38:15,340 given we had a group of people who've been infected before with a protected from getting infected again, 367 00:38:15,640 --> 00:38:23,660 and that allowed us to actually quantify for the first time anywhere in the world what your chances were being infected, infected again. 368 00:38:23,680 --> 00:38:31,509 So towards the end of October into November 2020, it became quite clear that there was actually quite good protection against against reinfection, 369 00:38:31,510 --> 00:38:39,070 that your risk might be only around sort of 10% of what it was if you hadn't been previously infected. 370 00:38:40,660 --> 00:38:48,150 And. And we then we then did some work with our health care workers as well, thinking about sort of response to vaccination. 371 00:38:48,960 --> 00:38:53,520 But that sort of actually that that initial testing program really, 372 00:38:54,150 --> 00:39:00,930 really helped really sort of helped help both establish protection from reinfection and also sort of have a better shot of the staff. 373 00:39:00,930 --> 00:39:05,489 And then we were then when lateral flow testing was rolled out as a routine thing to start, 374 00:39:05,490 --> 00:39:14,280 we were then able to use the same systems to track lateral flow test results and and continue to offer testing to staff first, 375 00:39:15,210 --> 00:39:18,990 first first half of the pandemic. Really using that is in our system. 376 00:39:19,680 --> 00:39:24,330 It's now been retired, which I'm very pleased about because it was a lot of work keeping it going better. 377 00:39:24,840 --> 00:39:33,570 But I think we in the end certainly supported hundreds of thousands of tests, the staff, the staff through it, 378 00:39:34,410 --> 00:39:41,489 and several thousand positive results notified staff and their staff so that that sort of be able to 379 00:39:41,490 --> 00:39:47,430 be able to sort of be able to go home and isolate and prevented from transmitting on in hospital. 380 00:39:48,310 --> 00:39:55,600 Mm hmm. I think you listed three things earlier and I now can't remember what the other one was. 381 00:39:59,280 --> 00:40:06,120 So we. Yes. So so I think and sort of developing the antibody tests, 382 00:40:06,120 --> 00:40:14,310 the and then the sort of using them through the COVID infection survey and start testing other things. 383 00:40:14,340 --> 00:40:17,130 Other things we have have been involved in to. 384 00:40:17,150 --> 00:40:23,970 To some extent was thinking about whether we could use sequencing as a tool for tracking infection in hospital. 385 00:40:24,510 --> 00:40:28,919 And certainly others have done larger scale studies than what we did. 386 00:40:28,920 --> 00:40:35,790 But we did do some work in Oxford and also the experience was that we found that 387 00:40:36,360 --> 00:40:40,679 it was actually quite a bit of the transmission from hospital in hospitals. 388 00:40:40,680 --> 00:40:45,830 That happened happened from a relatively small percentage of the patients and staff who had COVID. 389 00:40:47,280 --> 00:40:54,900 So and it tended to be sort of particularly patients with newly acquired infection in hospital who were most. 390 00:40:55,640 --> 00:41:04,500 It was the sort of most infectious and who sort of posed the greatest sort of risk of onward onward transmission to it to other patients. 391 00:41:04,500 --> 00:41:15,360 And so you almost got this sort of this sort of ever increasing feedback loop where patients that came in from the community with COVID, 392 00:41:15,810 --> 00:41:21,240 they'd been ill for a few days at home and most infectious in that period, and then once they came into hospital were less infectious. 393 00:41:21,630 --> 00:41:26,040 But if that then that then happened to trigger off one additional infection in hospital. 394 00:41:26,340 --> 00:41:31,110 That person was then much more infectious. A number of people then became infected who were also much more infectious. 395 00:41:31,110 --> 00:41:38,220 And so we saw these sort of clusters of cases where you got this progressive amplification of transmission that we could see. 396 00:41:38,940 --> 00:41:44,249 And so that really prompted us to particularly focus on making sure our infection control 397 00:41:44,250 --> 00:41:49,860 was really was really tighter around patients who had been newly diagnosed with COVID. 398 00:41:50,160 --> 00:41:55,020 And it also underlined the importance of making sure that if staff became infected, 399 00:41:55,020 --> 00:42:01,110 that we were aware of as soon as possible, because actually they were likely most infectious at the start of their infection. 400 00:42:01,110 --> 00:42:09,090 And so making sure that that people followed advice and testing and rapidly isolated was also also really important. 401 00:42:09,940 --> 00:42:15,190 And did you continue your regular surveillance of other pathogens in the hospital? 402 00:42:16,550 --> 00:42:21,129 Yeah, and. Know, from a research point of view, 403 00:42:21,130 --> 00:42:27,459 I think people continue to be aware of some of the pathogens that we might routinely monitor for infection control purposes. 404 00:42:27,460 --> 00:42:32,200 And so one of the one of the challenges was with that increased PPE, 405 00:42:32,200 --> 00:42:35,259 personal protective equipment that people were wearing in the intensive care unit, 406 00:42:35,260 --> 00:42:44,110 that actually we had an increase in infections due to a bacteria MRSA, which actually we hadn't had a particular issue with in the hospital. 407 00:42:45,010 --> 00:42:55,450 But but it sort of came back to being more of an issue because actually some of the sort of washing of hands and changing of gloves and 408 00:42:55,450 --> 00:43:03,520 changing and sort of changing equipment became much more difficult where people were having to sort of continuously where we're keeping at. 409 00:43:03,800 --> 00:43:05,140 And also actually, 410 00:43:05,140 --> 00:43:12,459 it is just difficult to be as sort of physically precise about how you might clean things and maintain things when you've got all the barriers, 411 00:43:12,460 --> 00:43:14,450 the extra, extra people as well. 412 00:43:14,450 --> 00:43:20,829 So so that was an example of something we saw rise and then had to then think again about sort of how we how we addressed that. 413 00:43:20,830 --> 00:43:30,060 Did we need to modify the way we use people need needed. We need to change what people we use it to try and control and and set up that. 414 00:43:30,460 --> 00:43:39,060 That was one thing we did see happen. But otherwise actually that we did, as people would imagine, see quite a lot of sort of the routine activity. 415 00:43:39,080 --> 00:43:46,150 The hospitals slow down or stop. And so some of the other things that we would normally monitor or have challenges with actually went went down. 416 00:43:46,540 --> 00:43:53,680 And I think that's what I was thinking, that you had fewer patients from the general community coming in with all kinds of things up their noses. 417 00:43:53,980 --> 00:44:00,120 Yes. Yes. So. 418 00:44:03,370 --> 00:44:07,900 What kind of hours we work. I mean, I've just I've just been trying to tot up your workload. 419 00:44:08,170 --> 00:44:12,640 I mean, one of the things I looked at was your list, the list of publications that have got your name on. 420 00:44:13,150 --> 00:44:16,530 And it seemed to massively increase over the past three years. 421 00:44:16,540 --> 00:44:22,509 I don't know how much actual writing you had to do yourself because they're mostly papers with huge numbers of of authors. 422 00:44:22,510 --> 00:44:28,600 But I mean, that's just kind of the end of the pipeline, isn't it? You've got to work the research work do you've got all the clinical work to do. 423 00:44:29,620 --> 00:44:33,100 How much pressure were you under in terms of the hours you put in? 424 00:44:34,180 --> 00:44:40,540 Yeah, I, I think you clearly can't work the kind of hours that people worked. 425 00:44:40,540 --> 00:44:46,029 And, you know, I think lots of people worked enormous numbers of hours in all sorts of capacities. 426 00:44:46,030 --> 00:44:50,319 So, you know, I don't think it was by any means unique, but you clearly can't sustain that forever. 427 00:44:50,320 --> 00:44:59,650 But what I think what a number of people did was to work enormously hard and much, much longer hours than they would they would do normally. 428 00:45:00,250 --> 00:45:09,040 I sort of came in in several forms, I think previously that no one would have dreamt of having organised meetings at eight 429 00:45:09,040 --> 00:45:13,359 or 8:00 in the morning or 730 in the morning or late on a Friday evening or whatever. 430 00:45:13,360 --> 00:45:18,969 And those things then almost became normalised and people sort of routinely did those things or spoke 431 00:45:18,970 --> 00:45:22,870 to each other at the weekend in a way in which people wouldn't have done necessarily previously. 432 00:45:22,870 --> 00:45:27,699 But there was this real sense that it was it was really important to get answers quickly, 433 00:45:27,700 --> 00:45:34,570 that actually we need to do everything possible to to to get answers. 434 00:45:34,990 --> 00:45:40,510 And I think that was largely motivated by sort of. Altruistic reasons for public good. 435 00:45:40,690 --> 00:45:43,390 But but there's also this sense of enormous competition as well, 436 00:45:43,420 --> 00:45:49,450 because actually everyone was working very hard and we're all trying to answer similar questions. 437 00:45:49,450 --> 00:45:54,460 And so I think there was also this external pressure that you wanted to be first. 438 00:45:55,240 --> 00:46:01,810 For example, we talked about the health care workers study. Of course, we wanted to report our results before anyone else reported that. 439 00:46:01,820 --> 00:46:07,479 So we we knew we were kind of likely to have enough results to have a paper. 440 00:46:07,480 --> 00:46:12,790 And so we were almost sort of writing the paper and had it ready to go so that once we had enough data, 441 00:46:12,790 --> 00:46:19,749 we could sort of we were sort of slightly second guessing what we'd found so you could top and tail it with the sort of relevant discussion. 442 00:46:19,750 --> 00:46:23,860 And then once you got the results, sort of slop them in and send it off. 443 00:46:26,230 --> 00:46:33,370 And so I think, yeah, there were, there were times where definitely sort of people were working from 730 in the morning through to, 444 00:46:33,370 --> 00:46:37,329 you know, even 1130 at night on things. I think that was every day. 445 00:46:37,330 --> 00:46:40,610 But there were times when it was it was like that. 446 00:46:41,080 --> 00:46:44,139 And I'm interested what you say about competition, 447 00:46:44,140 --> 00:46:47,959 because I've one question I've been asking people is whether they felt that working 448 00:46:47,960 --> 00:46:52,150 with the pandemic led to it an increased level of openness and collaboration. 449 00:46:52,450 --> 00:46:59,799 And most people say, yes, I think you're the first person who's frankly said that they were driven by competition. 450 00:46:59,800 --> 00:47:04,330 But did you also experience that sense of collaborating with a wider group? 451 00:47:04,900 --> 00:47:14,200 I think I think, yes, there was definitely an immense amount of collaborative working both within the university and sort of across institutions. 452 00:47:14,200 --> 00:47:22,059 And yes, so I think it would be it would be wrong to say everyone was sort of motivated by sort of their own their own selfish sort of motivations. 453 00:47:22,060 --> 00:47:25,810 But I do think that there were times where. 454 00:47:27,080 --> 00:47:33,649 We were aware that other people were doing similar studies and and and yes, everyone. 455 00:47:33,650 --> 00:47:37,160 So wanting to. Wanting to share information quickly. 456 00:47:37,160 --> 00:47:43,400 But also I guess the ultimate ultimately that there are rewards for being for being first and I think that kind of. 457 00:47:44,470 --> 00:47:52,670 Even amongst everything else. I think that's still kind of place some some degree of a role in and what people did and. 458 00:47:54,080 --> 00:48:01,549 But but yeah, I mean, I think as I said, as I said, I think the main thing it was that people wanted to get an answer because it really mattered. 459 00:48:01,550 --> 00:48:10,710 And so to give you another example of something I was involved in as the pandemic went on, there was a a a child on a lateral flow testing in schools. 460 00:48:10,730 --> 00:48:12,140 And I'm thinking about whether. 461 00:48:13,920 --> 00:48:20,190 Rather than sending students home to isolate at home for ten days when they've been a contact in the case where they're actually 462 00:48:20,190 --> 00:48:28,110 they could test regularly and then stay in school because it's a real sort of real issue with with people missing out on school. 463 00:48:28,110 --> 00:48:32,069 And I think unavoidably so when they had coded and were infected. 464 00:48:32,070 --> 00:48:37,050 But actually if they were just a contact and then weren't going to go on to be infected, that they really need to go home. 465 00:48:37,440 --> 00:48:41,640 And so that the trial that was done was to take 200 schools and roughly 100 of them 466 00:48:41,790 --> 00:48:46,949 sort of followed the standard where where when people were a contact of the case, 467 00:48:46,950 --> 00:48:48,780 they were at home to self-isolate for ten days. 468 00:48:49,170 --> 00:48:56,280 And then in the other one, they were offered regular lateral flow testing until one they were tested negative. 469 00:48:56,290 --> 00:49:04,349 They were then able to come into school and and they the outcome was that actually we saw that in the 470 00:49:04,350 --> 00:49:10,980 schools using the lateral flow testing strategy that we didn't see evidence of any large increase in cases. 471 00:49:12,030 --> 00:49:19,319 And actually, although people were absent for quite a lot of other reasons other than other than being contacts of cases, 472 00:49:19,320 --> 00:49:25,200 but still potentially that that strategy could get could get children back into back into school, 473 00:49:25,200 --> 00:49:28,470 and also that teachers and other staff in school as well. 474 00:49:29,100 --> 00:49:33,629 But as we came to finish that, there was this sort of the background to this. 475 00:49:33,630 --> 00:49:36,780 Was this sort of the what got caught at the time in the press, 476 00:49:36,780 --> 00:49:43,709 the pandemic where so many people were off work because they'd been contacts that actually 477 00:49:43,710 --> 00:49:49,890 there was sort of there was a lack of a lack of people actually being able to do things. 478 00:49:49,890 --> 00:49:54,720 And and that was a problem for us in health care, but it wasn't a problem for all sorts of other things as well. 479 00:49:55,770 --> 00:50:01,610 And so. Although it wasn't the subject of the trial, we were mostly looking at children. 480 00:50:01,630 --> 00:50:06,760 There was a lot of pressure that actually what we found may be very relevant to the population more widely. 481 00:50:06,760 --> 00:50:09,310 And so we were under quite a lot of pressure. 482 00:50:09,520 --> 00:50:14,380 It was a trial that was sort of joint sponsored by the Department of Health and Social Care and also education. 483 00:50:14,860 --> 00:50:19,780 And there's quite a lot of pressure to get really quite a rapid answer. 484 00:50:20,620 --> 00:50:25,359 And so unlike a sort of normal study where we might spend months going over things from when we 485 00:50:25,360 --> 00:50:30,940 had the final sort of fixed dataset to actually producing the paper was just a number of things. 486 00:50:31,300 --> 00:50:36,390 And again, we knew what the data that we were going to get. 487 00:50:36,400 --> 00:50:42,580 We knew what they looked like. And so we sort of had already coded up all the sort of analysis as much as 488 00:50:42,580 --> 00:50:47,020 we could based off sort of what we thought the data were going to look like. 489 00:50:47,020 --> 00:50:51,790 So that actually once we got it, we could run it really, really quite rapidly. 490 00:50:51,790 --> 00:50:55,210 So we hadn't seen the results unblinded until the end of the study. 491 00:50:55,480 --> 00:50:59,049 But once we saw them, then actually we could generate lots of that in a few days. 492 00:50:59,050 --> 00:51:05,750 And yes, and that was an example of a paper where some students wrote and wrote and wrote and it was on an incredibly hot ticket to the summer. 493 00:51:05,750 --> 00:51:09,129 And I remember sort of the computer sort of whirring with all the analysis, 494 00:51:09,130 --> 00:51:15,910 sort of creating loads and loads of heat in the room and then being really hot and you sort of frantically writing, trying to do the analysis, check. 495 00:51:15,920 --> 00:51:22,190 You got things right, sort of getting in contact with people, making sure that you were doing the right thing and and that. 496 00:51:22,210 --> 00:51:27,070 Yeah, so getting getting that paper and that result out was sort of under enormous time pressure. 497 00:51:27,110 --> 00:51:32,200 But then also also really important because it changed policy overnight. 498 00:51:33,790 --> 00:51:41,710 I mean, you did you mentioned just in passing that obviously people can't can't work that that kind of pitch the whole time. 499 00:51:42,310 --> 00:51:46,900 Did you actively take any steps to support your own well-being? 500 00:51:47,290 --> 00:51:52,210 Did were you able to say, I must give myself some space? 501 00:51:52,750 --> 00:51:56,170 Yeah, I think you have to. 502 00:51:56,230 --> 00:52:04,209 I would, yes. I sort of put our sort of professional portfolio for kind of the medical side of things. 503 00:52:04,210 --> 00:52:06,460 Part of it was sort of reflecting on things. 504 00:52:06,460 --> 00:52:12,220 And actually one of my reflections from that time is I've got a whole load of lovely pictures of butterflies in our garden, 505 00:52:12,220 --> 00:52:14,020 and I don't know why they were so many that year, 506 00:52:14,020 --> 00:52:19,329 but there were just loads and loads and loads and actually just having the kind of opportunity just to kind of decompress, 507 00:52:19,330 --> 00:52:23,170 go and do something completely different was definitely very helpful. 508 00:52:24,130 --> 00:52:30,340 It's also been very interesting talking to people about kind of how you unlearn the very time 509 00:52:30,340 --> 00:52:34,860 pressured sort of way of working and particularly talking to people from other industries. 510 00:52:34,860 --> 00:52:40,959 So I was fortunate to do a leadership course at the business school sort of towards the end of the pandemic. 511 00:52:40,960 --> 00:52:47,710 We spent two, two weeks getting to know and work with a number of people from all sorts of different sectors and talking to people, 512 00:52:47,710 --> 00:52:52,330 particularly from banking who'd been involved in the financial crisis in the early 2000s, 513 00:52:52,330 --> 00:52:56,620 and how they then had worked in a very pressured environment then, 514 00:52:56,620 --> 00:53:06,249 and then kind of over time had to unlearn that and actually had to be very kind of active and deliberative in saying I was coming to work, 515 00:53:06,250 --> 00:53:10,360 then I'm actually going to make a conscious decision to not do that or I was having a meeting. 516 00:53:10,360 --> 00:53:16,930 Now I'm going to make a conscious decision to actually say I'm not going to have meetings at that time of day so I can possibly avoid it. 517 00:53:19,300 --> 00:53:23,379 And I think, yeah, by being aware of that for yourself, but also for other people. 518 00:53:23,380 --> 00:53:27,850 So the people on the team and being able to say, look, we done an amazing job, 519 00:53:27,850 --> 00:53:36,820 but actually it is okay now to get back to working in a way that's more sensible or more sustainable, maybe. 520 00:53:37,960 --> 00:53:41,110 Do you think that the fact that you knew all the time you were working on something, 521 00:53:41,110 --> 00:53:44,979 it was really important actually, that helped to support your wellbeing or what? 522 00:53:44,980 --> 00:53:50,379 You too stressed that that I think I think it did. 523 00:53:50,380 --> 00:54:00,190 I'm not sure how much or how much perspective know the busyness dominates rather than having that much time to pause and reflect. 524 00:54:00,310 --> 00:54:08,049 But yeah, I how we were, you know, how I was I think there was you know, there were definitely times where you feel exhausted, 525 00:54:08,050 --> 00:54:15,190 you feel really tired and particularly sort of some of the clinical work you've directly confronted with some really challenging things. 526 00:54:15,190 --> 00:54:21,940 And I, I might pick out the sort of kindness of people originally at the beginning of the pandemic as being a real encouragement. 527 00:54:21,940 --> 00:54:28,479 I think there was sort of one particularly miserable weekend where it all felt pretty horrible in the hospital 528 00:54:28,480 --> 00:54:34,780 and someone brought in this massive tray of Danish pastries from their business or whatever and gave them to us. 529 00:54:34,780 --> 00:54:43,030 And it was amazing. And I kind of felt actually they were the people who needed help because their businesses were essentially like sort of, 530 00:54:43,600 --> 00:54:47,829 you know, that their customer disappeared overnight. And yet there were people being really generous, 531 00:54:47,830 --> 00:54:51,489 kind of bringing in all sorts of stuff to us who actually were were busy and 532 00:54:51,490 --> 00:54:56,560 getting very well paid while they work because they're actually struggling. But I think that was a real encouragement, I think. 533 00:54:58,790 --> 00:55:05,579 I certainly found the whole sort of the whole the whole public support for the NHS, also a great deal, a great encouragement. 534 00:55:05,580 --> 00:55:09,920 And so I think that that definitely contributed to to being supportive of that. 535 00:55:10,830 --> 00:55:17,620 And then I think having sort of family at home and people to talk to also also really helped them on. 536 00:55:17,720 --> 00:55:21,270 Yeah. My wife's also a doctor, she's a urologist, 537 00:55:21,270 --> 00:55:26,370 so I actually spent some of her time being redeployed back to doing things that she hadn't been doing for some time. 538 00:55:26,370 --> 00:55:30,600 But having that that shared experience was definitely also also helpful. 539 00:55:32,250 --> 00:55:40,139 So what how do you think the work that you did during the pandemic has changed what you do now? 540 00:55:40,140 --> 00:55:43,880 Has it has it at all? What's your plans now? 541 00:55:44,520 --> 00:55:50,579 It's a really interesting question because I think definitely the the pandemic 542 00:55:50,580 --> 00:55:54,420 there was sort of a real sense that what you were doing really mattered. 543 00:55:54,420 --> 00:56:02,100 And that manifested in a number of ways, you know, whether it was contact with decision making sort of organisations, 544 00:56:02,100 --> 00:56:06,149 whether that's within government or other advisory bodies or directly being in contact and 545 00:56:06,150 --> 00:56:11,340 participating in meetings with people there and and seeing the direct consequences of your work. 546 00:56:11,550 --> 00:56:14,730 Also the opportunity for sort of various media coverage and, you know, 547 00:56:14,730 --> 00:56:22,740 actually being on the television or being on the radio or being in the print media and and previously sort of, 548 00:56:23,130 --> 00:56:30,740 you know, having had some involvement with the media around some of the other work we've done, whether it was sort of, you know, 549 00:56:30,760 --> 00:56:34,559 sort of some of the work around multi-drug resistant gonorrhoea, which, as you can imagine, 550 00:56:34,560 --> 00:56:40,620 got sort of couched in terms of the world's worst super gonorrhoea or some of the work around Candida Auris, 551 00:56:40,620 --> 00:56:49,110 which we found was linked to transmission to these these axillary thermometers which the sun characterised as deadly armpit super fungus, 552 00:56:49,520 --> 00:56:56,879 which is, which is fantastic. So I had, I had had some of that involvement, but, but actually then through COVID, 553 00:56:56,880 --> 00:57:02,430 actually the kind of media coverage and press involvement was sort of significantly greater. 554 00:57:02,440 --> 00:57:05,480 So so you have this sort of sense of urgency in the work, 555 00:57:05,500 --> 00:57:11,579 you have this sort of connection with sort of official sort of impact of what you do and this contact with the 556 00:57:11,580 --> 00:57:18,510 media and sort of indirectly also through social media and people's interest in your work sort of skyrockets. 557 00:57:18,510 --> 00:57:22,770 And then and then suddenly that's all there anymore and you're sort of left reflecting? 558 00:57:22,770 --> 00:57:29,280 Well, actually, you know, I had I had this list of things I maybe wanted to achieve academically in terms of papers. 559 00:57:29,280 --> 00:57:33,170 I'd like to get published in particular places or numbers of publications or, 560 00:57:33,490 --> 00:57:39,780 or collaborations with people or and previously it might have taken you ten years to accrue those. 561 00:57:39,780 --> 00:57:47,310 And then all of a sudden, within two years, you've racked up an enormous number of publications and, and lots of you know that. 562 00:57:47,580 --> 00:57:50,610 And it had the opportunity to publish in lots of journals that you might not 563 00:57:50,730 --> 00:57:54,629 have otherwise done and has your work covered in a way that you might not. 564 00:57:54,630 --> 00:57:58,080 And so it does make you think, well, what do I want to do next? 565 00:57:58,380 --> 00:58:07,240 And maybe the plans that you had in 2019 kind of don't quite feel the same now compared to that and that. 566 00:58:08,130 --> 00:58:11,280 And so, yeah, I'm not sure it completely got to the answer, 567 00:58:11,280 --> 00:58:17,820 but I think there's there's definitely a sense that you have to sort of pull back and saying, well, what am I what am I good at? 568 00:58:19,290 --> 00:58:26,909 How can I continue to sort of use that to to actually still be benefiting other people? 569 00:58:26,910 --> 00:58:34,170 And so I think coming back to that, yes, it may it may not there may not be the same level of reward, 570 00:58:34,170 --> 00:58:38,250 but it may still be important to continue to do those things and to do them well. 571 00:58:40,050 --> 00:58:46,620 That said, it it also makes you think that it would be quite nice to do things a little bit differently, 572 00:58:46,620 --> 00:58:53,189 whether that's working with different groups of people or working in slightly different areas or doing academic work a little bit differently. 573 00:58:53,190 --> 00:58:59,339 And I think I'm still thinking about that, still exploring still, but it does make you think, 574 00:58:59,340 --> 00:59:05,610 okay, I've done quite a lot and quite a lot of work academically and I'd like to, 575 00:59:05,940 --> 00:59:10,290 yeah, like to make sure that I sort of take the opportunities to do as many different things as I can. 576 00:59:11,880 --> 00:59:15,930 So. So yes, I think it definitely changes your perspective. 577 00:59:15,930 --> 00:59:20,610 I'm not sure quite. I've quite got to where I find me realising where I'm going to be. 578 00:59:20,610 --> 00:59:27,419 But I think yeah, I think I think that but that kind of in the meantime, having the encouragement to go back to doing what you're doing, you know, 579 00:59:27,420 --> 00:59:34,200 you can do well and to do it well and to make sure you're answering meaningful questions, I think has been sort of really where I got to so from. 580 00:59:35,360 --> 00:59:37,310 That's great. Thank you very much indeed.