Transcript

1
00:00:03.370 –> 00:00:10.419
Harry Malcolm: Hi everyone, I’m Harry from Rubica, where we help organizations unlock radical performance improvements that benefit customers, patients, and employees.

2
00:00:11.020 –> 00:00:18.500
Harry Malcolm: I’m very pleased to welcome you to the next webinar in our series on what HDPs value, Achieving Relevance and Results with Account Insights.

3
00:00:19.040 –> 00:00:29.739
Harry Malcolm: In the last webinar, we spoke with Fran Bannon from Grunenthal, and Ana Isabel Gonzalez-Gonzalez from the Madrid Healthcare System. You can watch the recording of that by following the link in the chat panel: https://www.rubica.co.uk/webinars/achieving-results-with-account-insights/

4
00:00:30.310 –> 00:00:33.780
Harry Malcolm: Please add your questions throughout this webinar for the Q&A at the end.

5
00:00:33.940 –> 00:00:37.059
Harry Malcolm: We also have the resources section, so take a look in there.

6
00:00:37.470 –> 00:00:50.839
Harry Malcolm: On with today’s event, where I have the pleasure of talking with Professor David Lowe, who brings a unique blend of clinical expertise, innovation leadership, and deep experience in transforming healthcare pathways. David, welcome.

7
00:00:51.460 –> 00:00:54.520
David Lowe: Well, hi, hi. Hi, Harry, how are you?

8
00:00:54.520 –> 00:00:56.589
Harry Malcolm: I’m good. Where are you today?

9
00:00:56.630 –> 00:01:16.220
David Lowe: So, I’m actually in the main campus of the university today. We’ve got a group visiting from Taiwan, from their engineering and medical schools, looking at how we create a corridor for innovation between Taiwan and the UK, which is sponsored by the Royal Society of Edinburgh. So that’s what I’m doing for the rest of today.

10
00:01:16.220 –> 00:01:35.559
Harry Malcolm: Wonderful. Collaboration, this is going to be at the heart of the discussion today. You wear a lot of hats. Can we start with David, the clinician? Even if you don’t wear that hat as much as you once did, perhaps, what is daily life like as a healthcare practitioner at a central hospital in a city like Glasgow?

11
00:01:36.060 –> 00:01:45.200
David Lowe: Perfect. So I’m a consultant in emergency medicine based at the Queen Elizabeth University Hospital in Glasgow, which is one of the last year’s health boards in Europe. And I suppose

12
00:01:45.340 –> 00:01:58.569
David Lowe: Across the healthcare system at the moment, we are challenged, this is well described in the 10-year plan and in other documents around about how do we respond to the rising tides of age, comorbidity, and expectation for delivery.

13
00:01:58.570 –> 00:02:08.229
David Lowe: And therefore, my clinical setting is perhaps one of the most media-friendly examples of that challenge we experienced by the number of ambulances outside.

14
00:02:08.229 –> 00:02:27.460
David Lowe: And there’s some really great work by my college really describing how we need to shift care back into the community. We see, for every patient who sits more than 8 hours in an ambulance or in a corridor, in a med department across the UK, we see one additional death for every 75 patients in that position, which is very much

15
00:02:27.460 –> 00:02:31.659
David Lowe: motivating around about some of the rest of the discussion we’re going to have around about how we

16
00:02:31.660 –> 00:02:46.959
David Lowe: change pathways of care, how we optimize the patient journey, recognizing that in my context in the emergency department, we are there as a safety net to respond to people with Thoman disease, or when they’re unable to respond in another method, but

17
00:02:46.960 –> 00:02:58.040
David Lowe: as it’s been described, we need to increasingly move towards preventative care, which means not ending up in an emergency department, or indeed, within a hospital environment. Thank you.

18
00:02:58.440 –> 00:03:08.969
Harry Malcolm: And when you are in that, you know, that busy, as you said, media-friendly, from the perspective of kind of waiting times and statistics, emergency department.

19
00:03:09.950 –> 00:03:16.750
Harry Malcolm: What’s, like, your capacity to even think about changing pathways and moving towards prevention and so on?

20
00:03:17.660 –> 00:03:42.459
David Lowe: So I think across the system, there’s a recognition that we do need to change pathways, and it’s how much do we kind of shift those diagnostic pathways into community sessions, and that’s where some of the discussion around about diagnostic hubs comes to bear, is how can we, in my world, use AI and machine learning to predict case find those with potentially undiagnosed disease, those that are potentially deteriorating.

21
00:03:42.460 –> 00:03:48.409
David Lowe: and need optimization of therapy or treatment, and equally those who need review. And this is

22
00:03:48.410 –> 00:03:53.710
David Lowe: Complex because of comorbidities, but ultimately, to your question.

23
00:03:54.030 –> 00:04:05.000
David Lowe: when we’re… when I’m consultant in charge of the medicine department, there’s very little time to think about pathway change, and often the conversation is, this is not the time to reimagine pathways of care.

24
00:04:05.000 –> 00:04:05.320
Harry Malcolm: Yeah.

25
00:04:05.320 –> 00:04:10.380
David Lowe: But it’s how do you create time to be able to do that? And certainly, within

26
00:04:10.380 –> 00:04:17.940
David Lowe: within Glasgow, we’ve created an innovation hub and do try to give innovators time to be able to think about how they

27
00:04:17.940 –> 00:04:34.409
David Lowe: Transform or potentially disrupt clinical pathways to actually realize some of those benefits around about optimizing patient outcomes, ultimately, which is around about early detection, optimization, and case management in the community.

28
00:04:34.710 –> 00:04:51.499
Harry Malcolm: Yeah. I mean, you… David, the innovator is a hat I think you’re probably always wearing, and certainly when we’ve had discussions, and even now when you’re talking about the day-to-day, you are constantly thinking about the opportunities for innovation and transforming pathways.

29
00:04:51.500 –> 00:04:56.099
Harry Malcolm: I know many of the projects that you’re involved in are with industry partners.

30
00:04:56.850 –> 00:05:04.060
Harry Malcolm: What are you looking for in a collaboration with industry, and kind of what helps those partnerships to be successful?

31
00:05:04.760 –> 00:05:20.660
David Lowe: Yes, it’s a great question. I suppose my group, which is the Digital Validation Lab that sits within the university, has one mantra that sits across most, which is, do we want them in our lives? And that’s, really can be broken down into

32
00:05:20.760 –> 00:05:42.359
David Lowe: when these collaborations work well, we’re able to have a very honest conversation around about priorities. Now, that might be pipeline of new and novel therapeutics that are coming through, it might be around about the next step in regulatory approval for a med tech device, but it’s understanding what that… what those drivers are for the industry partner.

33
00:05:42.360 –> 00:05:48.439
David Lowe: How we can best support it, and being obviously very honest about, is this a priority for our healthcare system?

34
00:05:48.440 –> 00:06:01.960
David Lowe: And equally, is that a priority? And do we have that capability within our university group to be able to do this? And we function as a triple helix, relatively, effectively, which means that we have to bring people together.

35
00:06:01.960 –> 00:06:14.650
David Lowe: understand priorities, have sometimes very honest conversations about saying, well, I know this is the top thing you really, really want to do, but actually that sits lower for us, or perhaps our capability’s not there.

36
00:06:14.650 –> 00:06:34.139
David Lowe: But typically, because of the nature of the problems we’re trying to solve, as you’ve described, around clinical care, pathways, case finding, we can, usually relatively easily find a kind of shared, shared goal, ultimately, around about transforming a pathway, creating the evidence.

37
00:06:34.240 –> 00:06:49.080
David Lowe: And really, where we sit in this space is trying to avoid really carefully and meaningfully pilots, but actually create evidence that means that you do not have to go from site to site doing individual optimization of.

38
00:06:49.080 –> 00:06:49.460
Harry Malcolm: Yep.

39
00:06:49.460 –> 00:07:03.079
David Lowe: a description of a pathway or decision-making process, but actually say, how do we create evidence to be able to shift the paradigm of care, be it using an AI or med tech tool, or simply a new way of working?

40
00:07:03.290 –> 00:07:05.639
Harry Malcolm: You talked there about,

41
00:07:05.790 –> 00:07:24.079
Harry Malcolm: you know, a joint goal between yourselves and the industry partner. We sometimes talk about common goals, you know, so where there is that benefit for patients, for the healthcare system, and for industry. Do you have any kind of good juicy examples of those goals that you’ve agreed upon?

42
00:07:24.400 –> 00:07:43.069
David Lowe: Yeah, so I think one of the great examples is a study called Opera that was run through the MOU between University of Glasgow, NHS, Greater Glasgow, and Clyde, our center for Sustainable Delivery, which is part of the, kind of, broader change transformation engine within Scotland and AstraZeneca.

43
00:07:43.100 –> 00:07:54.289
David Lowe: And that was very simply put, around about a heart failure diagnosis. So, 80% of people continue to get a diagnosis of heart failure within an emergency department, acute admission.

44
00:07:54.600 –> 00:08:02.480
David Lowe: We know that within 6 weeks of initiation of therapy, that we can actually reduce hospitalizations, improve outcomes, and reduce costs.

45
00:08:02.580 –> 00:08:17.050
David Lowe: But how do we do that? And within Glasgow at that time, and this was per year just before COVID, but we delivered the study during COVID, we had about an excess of a year to ECHO, which is obviously the definitive test for heart failure.

46
00:08:17.260 –> 00:08:27.989
David Lowe: So recognizing that that was a priority for our industry partner, and equally an absolute priority for us, both clinically and to strengthen our university with Professor Mark Petchy, to do this work with

47
00:08:27.990 –> 00:08:41.240
David Lowe: other collaborators, we set upon transforming that pathway, so we worked with a company called Behring to validate their approach to case finding in the community about identifying those at highest priority.

48
00:08:41.340 –> 00:08:57.020
David Lowe: For undiagnosed heart failure, to bring people in. At that time, it’s actually a Nightingale hospital, but we changed how people get their heart failure diagnosis from episodic attendees to see a nurse, to get an echo, to see a consultant.

49
00:08:57.020 –> 00:08:57.409
Harry Malcolm: Gotcha.

50
00:08:57.410 –> 00:09:01.659
David Lowe: A single stop, bringing them in, point of care, biomarker tests.

51
00:09:01.920 –> 00:09:20.699
David Lowe: into seeing, getting their ECG and echo done, and then being, outcomed and initiated in therapy within that single attendance, which is really important, during COVID to do that, while in the process both validating the bearing to finding life.

52
00:09:20.700 –> 00:09:31.989
David Lowe: but also working with a company called US2I, which is from Singapore, who had an AI solution for point-of-care ECHO to support rapid reporting of echo, because again, that…

53
00:09:31.990 –> 00:09:41.889
David Lowe: block, around about access to ECHO is they can’t just simply ask the sonographers to work harder, or to work longer hours, you have to give them the tools to be able to do that. So that…

54
00:09:42.040 –> 00:09:52.049
David Lowe: pathway and those technologies have been taken forward and deployed in many places around the world now as an approach to both identification of patients.

55
00:09:52.230 –> 00:09:53.920
David Lowe: Being able to…

56
00:09:53.920 –> 00:10:18.679
David Lowe: have a rapid diagnostic pathway, which we are very much assuming will be rolled into the diagnostic, community hubs that have been rolled out in the UK, and then finally thinking about how AI echo interpretation can support, sonographers against that, ultimately that blocker, roundabout, definitive diagnosis that allows people to be initiated on therapy. So.

57
00:10:19.140 –> 00:10:27.820
David Lowe: And again, that was very much each of those parties coming together, saying, here’s a problem, here’s a set of technologies that were

58
00:10:27.940 –> 00:10:41.209
David Lowe: on the horizon for us was anything in this case, it would optimize that pathway, but how do we create evidence for clinical cost effectiveness and acceptability of that pathway, to be deployed?

59
00:10:41.690 –> 00:10:56.599
Harry Malcolm: And how does that all start? Because you’ve got, you know, you’ve got companies sitting on, great therapies, or a piece of technology, whatever else it is. How are they…

60
00:10:56.600 –> 00:11:03.759
Harry Malcolm: You know, finding out what matters to you in the healthcare system, and where you might be looking to collaborate.

61
00:11:04.190 –> 00:11:21.529
David Lowe: So I think what we’ve successfully done with a number of industry partners is created that relationship to be able to have very honest conversations about saying, here are the technologies we’ve scoped, are on our horizon, here’s our priorities around about diagnosis, case management, optimization.

62
00:11:21.530 –> 00:11:29.660
David Lowe: and equally be able to match that to capabilities, and work across different centres, so I’m,

63
00:11:29.660 –> 00:11:40.990
David Lowe: co-director within the center of Excellence for Regulation for AI, with Birmingham, and we do a lot of work with our colleagues in Oxford as well, and I think that’s really important that

64
00:11:40.990 –> 00:11:51.760
David Lowe: If we’re not going to do pilots, we need to do the meaningful evidence generation that allows us to be able to scale that technology into the UK and potentially into other markets.

65
00:11:51.760 –> 00:12:07.930
David Lowe: And that’s what we’re discussing today with our Taiwanese colleagues, is again, how do we create evidence that supports equivalence, or at least reliance that our HTA, Health Technology Agency, NICE, can look at evidence that’s been co-developed and

66
00:12:08.130 –> 00:12:18.610
David Lowe: the UK, Glasgow, and in this case, Taipei, Taiwan, to be able to build that evidence pack that means that these companies, be it big multinationals, or even

67
00:12:18.610 –> 00:12:29.310
David Lowe: smaller companies can invest in evidence creation once that allows them to get market adoption, market access, in multiple different territories, and I think

68
00:12:29.440 –> 00:12:49.040
David Lowe: it’s again understanding those priorities that, while global for any particular company might want to do a particular study in the UK or indeed anywhere else, what they are looking to do is invest to create evidence that will support their teams to be able to go into other markets as well, and I think

69
00:12:49.040 –> 00:12:51.739
David Lowe: That, again, honest discussion.

70
00:12:51.850 –> 00:13:02.649
David Lowe: building that relationship and collaboration to be able to match priorities and needs with capabilities, and that supports, again, yes.

71
00:13:02.720 –> 00:13:10.790
David Lowe: pathway change in Glasgow, because clearly, if we change a pathway in Glasgow, that can have both a significant impact to patient outcomes.

72
00:13:10.790 –> 00:13:22.949
David Lowe: But equally is a very large health board, which means that the decision-making around about therapeutic or treatment options is… can be quite significant, and therefore you can point to partner sites.

73
00:13:22.950 –> 00:13:26.790
David Lowe: But equally, you’ve got the opportunity to be able to say that this is

74
00:13:26.860 –> 00:13:29.950
David Lowe: Got the potential to scale into other healthcare systems.

75
00:13:30.260 –> 00:13:34.790
Harry Malcolm: I think that… that is… that is fascinating, you know, that… because obviously there’s a benefit

76
00:13:35.360 –> 00:13:54.160
Harry Malcolm: to the healthcare system, beyond what your, kind of, your territory, to say, right, if we do it this way, then it can get wider application. You know, that is generous and just thinking bigger about the healthcare of a nation or beyond. And of course, that then also works for the… for the suppliers, from industry, because they’re looking for scale as well.

77
00:13:54.520 –> 00:14:03.819
Harry Malcolm: how… you know, how do you share these priorities about where you’re looking to focus on? You know, if you’re a company, small or large, how do you learn about it?

78
00:14:04.480 –> 00:14:20.459
David Lowe: So, great question. So we’ve got some very established partnerships with some of the larger pharmaceutical and med tech companies. That allows us to bring their teams to us, or we go down to them to have those discussions around about

79
00:14:20.580 –> 00:14:38.029
David Lowe: where are we experiencing challenges? It’s a large health board with significant amounts of deprivation and health inequalities, and therefore, we can very easily describe opportunities to make a difference, which I think is really important that if we do something well.

80
00:14:38.030 –> 00:14:44.320
David Lowe: Then we can demonstrate a significant change around about both patients Clinical and cost outcomes.

81
00:14:44.590 –> 00:15:01.319
David Lowe: So, it starts with that, and then I think beyond that, it’s being able to, again, run the ruler over particular technologies. Often, we, especially in the AI and med tech world, will do some in silico work before, looking at performance in a retrospective dataset.

82
00:15:01.320 –> 00:15:06.570
David Lowe: To build confidence before we go forth and do the more expensive prospective study.

83
00:15:07.710 –> 00:15:25.889
David Lowe: And the third thing is around about patient engagement, and we’ve got some really established PPIE, so patient participation groups, that often bring them the concepts and saying… so we’re having a conversation today, later, around about ambient sensors to diagnose cardiovascular disease. So…

84
00:15:26.230 –> 00:15:38.329
David Lowe: Talking through what the acceptability of placement of ambient sensors in sheltered housing, and potentially in hospital waiting room environments, and other places to be able to say.

85
00:15:38.440 –> 00:15:57.770
David Lowe: this is a potentialist technology. It’s privacy protecting, there’s no details or photos of your face or anything else, but equally, this could potentially allow us to diagnose atrial fibrillation, when people just come through, into the hospital waiting areas, and that can increasingly be very important. We…

86
00:15:58.070 –> 00:16:10.219
David Lowe: very much feel that if you’re going to have, let’s say, lung cancer screening, and bring people over the age of 55 with a smoking history into a diagnostic clinic and take a CT of the thorax.

87
00:16:10.220 –> 00:16:20.700
David Lowe: we should use that optimized opportunity to identify multiple long-term conditions. So we’re doing a piece of work funded by UKRI, it’s about to start next week.

88
00:16:20.700 –> 00:16:28.350
David Lowe: That’s optimizing lung cancer screening. So rather than just looking for lung cancer, which is clearly massively important, and has reached,

89
00:16:28.350 –> 00:16:39.690
David Lowe: a nice quality threshold is to be able to look for coronary calcium, for osteoporosis, for COPD, for heart failure, for hypertension, for atrial fibrillation, in that single episode of care.

90
00:16:39.690 –> 00:16:51.970
David Lowe: recognizing that if you’re bringing, again, 55-year-old smokers in Glasgow, the ability to identify and diagnose one of these other conditions to prevent future hospitalisation

91
00:16:51.970 –> 00:17:03.519
David Lowe: and improve outcomes, so to start the phosphonates for their osteoporosis, to initiate the heart failure, COPD medication, and critically, to do that definitive diagnosis.

92
00:17:03.670 –> 00:17:14.299
David Lowe: Requires us to both transform a pathway, make our consultant colleagues comfortable that we are not going to create this fire hydrant of additional work, but we are going to be proactive

93
00:17:14.930 –> 00:17:21.130
David Lowe: but also look at technologies that support that diagnosis, so rather than the CT just saying it was emphysema.

94
00:17:21.599 –> 00:17:30.350
David Lowe: If the CT says, looks like there’s emphysema, and then AI agrees, and this is some of the validation work we’re doing at the moment, that individual goes straight to spirometry.

95
00:17:30.370 –> 00:17:46.059
David Lowe: And therefore, the referral to the prosperity colleagues is, they’ve had spirometer, here’s the CTE, here’s the structured data during this attendance, which is hugely valuable for us as a research organization who could now deeply phenotype the population.

96
00:17:46.180 –> 00:17:59.500
David Lowe: That allows us to be able to manage them more effectively and not continue to bring people in for episodic clinic review, or even worse, crash land with late-stage cancer, heart failure, COPD.

97
00:18:00.760 –> 00:18:17.749
Harry Malcolm: We’ve got a question that’s come in that I’m going to take now, because I think it’s the right timing for it, of, you know, how can the smaller industry partners, you know, without the established connections, understand your priorities and areas of focus? How can they open those conversations?

98
00:18:17.930 –> 00:18:32.619
David Lowe: Yeah, so great question. I think there’s… well, there’s three elements to that. The first is the priorities at a national level are well described, but are broad. It’s around about comorbidities, it’s about shifted digital community, and preventative.

99
00:18:33.130 –> 00:18:33.500
Harry Malcolm: Yep.

100
00:18:33.500 –> 00:18:54.769
David Lowe: In a previous role, I spent quite a lot of time trying to nail down policy around about which particular clinical problem was of greatest priority, and that’s often very difficult to do. Recognizing that sometimes our piece of work in, for instance, as Nigel Jameson is one of our leads for pancreatic cancer.

101
00:18:55.100 –> 00:19:11.460
David Lowe: obviously devastating, often late, late diagnosis, cancer. We spent a lot of time focusing on building out datasets and looking at different technologies for diagnosis, recognizing that can be applied to breast and lung cancer. So again, there’s

102
00:19:11.460 –> 00:19:35.679
David Lowe: sometimes a piece around about where’s our great data sets, where’s our great expertise, and how does that, for that SME, allow them to start the narrative about building evidence, versus going for one of these big, high-volume use cases, which is often sometimes very cluttered and everything else. And then the last point is that we’ve got business development people within the DHVL, and we’ll engage

103
00:19:35.680 –> 00:19:39.780
David Lowe: with SMEs, and that’s a key priority for us in terms of

104
00:19:39.890 –> 00:19:49.099
David Lowe: how we’ve set up Glasgow as a cluster for life sciences and med tech is to be engaged with smaller companies to help them build their product to get to market.

105
00:19:49.300 –> 00:19:56.240
David Lowe: recognizing that as NIHR has very clearly said, that they are now both a health and wealth creation.

106
00:19:56.240 –> 00:20:09.410
David Lowe: organization, and therefore, as a higher education institute funded by the NIHR in many different ways, we recognise that we have a civic responsibility as a university to improve health outcomes.

107
00:20:09.410 –> 00:20:23.760
David Lowe: But equally, as a clinician, I absolutely have, that same priority about how do we support big and large, as well as small, companies to develop products that can make a meaningful difference.

108
00:20:24.460 –> 00:20:27.579
Harry Malcolm: So if you’re sitting there, you know, as a… as a…

109
00:20:27.900 –> 00:20:32.929
Harry Malcolm: developer of a therapy or a piece of technology, you’re like, this, you know, this really should…

110
00:20:33.060 –> 00:20:40.719
Harry Malcolm: make a difference, can make a difference, it’s in line with the Scottish health priorities and so on. Where do they go? What do they do?

111
00:20:40.920 –> 00:20:55.609
David Lowe: So, for us, certainly, we’ve got the Digital Health Validation Lab that you can very easily Google, and we’ve got business development people that will engage and help scope things. And what we spent the last two days with Tyba and these colleagues, it’s very clear in my brain, is that it’s how do we support

112
00:20:55.610 –> 00:21:09.450
David Lowe: concept validation, moving into regulatory approval, and then implementation. Actually, what that looks like is evidence for decision makers, and that’s either for investment and grant, investments.

113
00:21:09.450 –> 00:21:18.479
David Lowe: Moving into regulatory approval, both in the UK or otherwise, and then market access, which is HTA, so Health Technology Assessment, or NICE,

114
00:21:18.480 –> 00:21:28.010
David Lowe: But recognizing there’s still that challenge of implementation and patient access, and that’s where we spend a lot of time, and it’s really pertinent to this discussion, is about

115
00:21:28.060 –> 00:21:32.620
David Lowe: Embedding pathway change, which often is asking the workforce

116
00:21:33.030 –> 00:21:37.809
David Lowe: To work differently, but equally to give them the tools to be able to do that.

117
00:21:37.810 –> 00:21:38.280
Harry Malcolm: Hmm.

118
00:21:38.280 –> 00:21:49.379
David Lowe: and the evidence to be able to do that, and there’s a slide I used last… yesterday that is healthcare moves at the speed of trust, but ultimately, that’s the speed of evidence creation.

119
00:21:50.060 –> 00:22:02.409
David Lowe: And ask either clinicians, or indeed patients or decision makers within our health boards to invest, engage, to adopt new technologies, new therapies, without evidence.

120
00:22:02.440 –> 00:22:20.470
David Lowe: And therefore, what’s very clear from a UK government point of view, in terms of how we’re engaging in the life science sectors, particularly around about therapeutics and technology, is how do we become pro-innovation, but equally support

121
00:22:20.510 –> 00:22:25.059
David Lowe: a streamlined approach to evidence creation to get to market, both in.

122
00:22:25.060 –> 00:22:25.420
Harry Malcolm: Nice.

123
00:22:25.420 –> 00:22:33.780
David Lowe: the UK and otherwise, and that’s becoming increasingly critical when you look at the 10-year journey from molecule to

124
00:22:33.840 –> 00:22:48.980
David Lowe: deliver that pill within a therapeutic domain, and equally similar challenges we see within med tech of great technologies, but take a long time to actually be implemented into pathways, rather than discrete technologies, and we spend

125
00:22:49.130 –> 00:23:01.519
David Lowe: a lot of time with companies at the early stage, thinking about intended use and what their value proposition is to get regulatory approval, to get investment, but then at the latter phases.

126
00:23:02.490 –> 00:23:13.130
David Lowe: trying to not retrofit how’s this going to work within a pathway, which is absolutely critical, as you’re not just throwing a technology over the fence and saying.

127
00:23:13.140 –> 00:23:29.359
David Lowe: this is going to work. And one example is Radical, which we’re about to publish around lung cancer with CURE, which is a chest X-ray eye solution that flags chest x-rays with lung cancer to optimize their diagnostic and ultimately treatment journey.

128
00:23:29.360 –> 00:23:37.000
David Lowe: And what we did with them for the trial is we did a cluster trial across Glasgow, so the three sectors, but we…

129
00:23:37.770 –> 00:23:44.310
David Lowe: not only introduce the technology, but says, if the AI flags cancer and the radiologist agrees.

130
00:23:44.530 –> 00:23:56.380
David Lowe: then we progress the same day, or within 72 hours, CT scan. So it’s not just about the, introduction of a new technology, it’s thinking about how that technology could optimize a pathway

131
00:23:56.400 –> 00:24:09.489
David Lowe: And when you’ve seen when it works, and when the rest of the system responds to that AI flag, then you’ve got quite significant changes in time to treatment, time to diagnosis, while creating

132
00:24:09.490 –> 00:24:19.109
David Lowe: great data for future studies, and that’s a small, but I think a great example around about how you have to engineer not just the technology.

133
00:24:19.110 –> 00:24:24.789
David Lowe: But the human factors, but ultimately the pathway that that technology is going to be deployed into.

134
00:24:25.460 –> 00:24:33.160
Harry Malcolm: And so, if I play this back, to check that I’m understanding, you know, if you’re sitting on a therapy or a piece of technology,

135
00:24:33.350 –> 00:24:45.739
Harry Malcolm: Like, you know, in some ways it doesn’t matter how brilliant that is, you’ve got to be thinking about how can we help the healthcare system to use it, so what is the pathway that it is a part of?

136
00:24:45.740 –> 00:24:56.079
Harry Malcolm: But you don’t need to know all of that yourself. You know, you as a… in the healthcare system are open to the discussions about, well, if this is in line with our priorities, let’s… we can figure some of that out together.

137
00:24:56.340 –> 00:25:00.520
David Lowe: Yeah, and absolutely that. So I think the first thing is understanding what the current pathway is.

138
00:25:00.520 –> 00:25:01.030
Harry Malcolm: Yeah.

139
00:25:01.030 –> 00:25:06.180
David Lowe: But the purpose is, I suppose, for decision makers, understanding the cost of that current pathway?

140
00:25:06.180 –> 00:25:06.600
Harry Malcolm: Yep.

141
00:25:06.600 –> 00:25:19.010
David Lowe: an understanding of what would you change to optimize this based upon the performance of the technology. And we, for instance, with CURE, prevalence of lung cancer is high in, virtually in the Glasgow population.

142
00:25:19.350 –> 00:25:37.759
David Lowe: equally, what our resource was for radiologists to respond was different, so… and that’s sometimes the opportunity with AI technologies. You can threshold and calibrate that technology based upon clinical resource prevalence of disease to optimize that technology being integrated into the pathway, and that’s

143
00:25:37.760 –> 00:25:39.749
David Lowe: Optimization can be both

144
00:25:39.810 –> 00:25:59.200
David Lowe: to drive acceptance, so reduce the number of false positives by the AI, because if there’s lots of flags of cancel that are not correct, then that reduces confidence in the technology. But equally, be able to think about how that matches prevalence and resource. So, we spend a lot of time thinking what the current pathway is.

145
00:25:59.200 –> 00:26:08.119
David Lowe: What the performance and intended use of that technology is, and how that actually responds to our priorities around healthcare delivery, which is.

146
00:26:08.120 –> 00:26:08.680
Harry Malcolm: Right.

147
00:26:08.680 –> 00:26:19.410
David Lowe: reduction in time to treatment, because we know that if we treat people, our mortality increases by 10% every 6 months for lung cancer, so therefore, if we can reduce

148
00:26:19.430 –> 00:26:34.359
David Lowe: time to treatment, then we can improve mortality, and that’s really critical for, from a policy point of view, that if we’re investing in technologies, it responds to a policy priority, which undoubtedly in Scotland is around about lung cancer waiting times.

149
00:26:34.800 –> 00:26:46.670
Harry Malcolm: So, whether you are at the, kind of, the head of a multinational, or you are, you know, in the field, as we say, you know, working on a kind of a daily basis, meeting clinicians.

150
00:26:46.840 –> 00:26:52.020
Harry Malcolm: Understand what the priorities are, And what the current pathway is.

151
00:26:52.350 –> 00:26:55.890
Harry Malcolm: And then think about how your solution fits into that.

152
00:26:56.420 –> 00:27:10.950
David Lowe: And who the decision makers are, and we spend a lot of time in the discussion understanding who’s your next decision maker. So for SMEs, a small to medium enterprise, is your next decision maker a grant funder or VC? And what evidence do they need, which might be about

153
00:27:11.440 –> 00:27:20.109
David Lowe: concept validation, and how this technology will scale into other markets, so not just sales to the NHS.

154
00:27:20.110 –> 00:27:35.869
David Lowe: Whereas later on, it might be a decision around about, should we acquire this technology into a larger multinational, and therefore, what’s its performance, regulatory approval, what’s its customer base, and how will it scale? So, often, despite being a clinical academic.

155
00:27:36.680 –> 00:27:53.950
David Lowe: Given that triple helix, we spend quite a lot of time thinking about value proposition for decision makers, be it investors, decision makers in the organization, to say, yes, we’re going to… if we take the example of pharma, we’re going to support this investigator-led research study.

156
00:27:53.970 –> 00:28:05.299
David Lowe: Because actually, it responds to needs, so we’re looking at workarounds about vaccination, and be able to say what other populations within the UK would benefit from

157
00:28:05.300 –> 00:28:15.960
David Lowe: acute respiratory infection vaccination to build the evidence, but that sometimes requires us to look at novel technologies that actually test for that particular virus.

158
00:28:16.060 –> 00:28:30.469
David Lowe: presentation into the meds department, so you end up with multiple different industry partners, like the heart failure opera example, that you’re actually… there’s a confluence of priorities, and it’s how do we shape them together, and I suppose

159
00:28:30.470 –> 00:28:38.180
David Lowe: My role, my role with some of my clinical academic colleagues, is to shape a protocol, the evidence generation pathway.

160
00:28:38.180 –> 00:28:50.289
David Lowe: That responds to potentially multiple partners’ needs around about validation of a technology, validation of a pathway that supports diagnosis or case finding or risk stratification.

161
00:28:50.290 –> 00:29:07.490
David Lowe: And then be able to create the evidence for who’s the decision maker? Is that JCVI for vaccinations, or is it NICE? Or often, the conversation we have with SMEs within the UK is, how do we support them getting to market into the US?

162
00:29:08.030 –> 00:29:17.619
David Lowe: Because that’s critical for them in terms of investment potential, is, yes, they can describe they might be able to sell to 240, NHS

163
00:29:17.620 –> 00:29:29.949
David Lowe: trusts or decision makers in England, or indeed the 240 GP practices, one of those numbers is wrong, within Glasgow, but you have to have a clear understanding of

164
00:29:29.960 –> 00:29:32.740
David Lowe: Who are you creating evidence for?

165
00:29:32.940 –> 00:29:43.679
David Lowe: And sometimes that’s very much internal for multinational companies, but increasingly that’s around about market access, which is whichever decision maker is going to

166
00:29:44.030 –> 00:29:48.120
David Lowe: Be part of the reimbursement journey for that particular technology or therapy.

167
00:29:48.410 –> 00:30:06.490
Harry Malcolm: I think that’s fascinating, like, there’s actually probably much more alignment in the questions that you are seeking to ask… answer, rather, on the healthcare provider side, as there is on the industry side. You know, looking at what is the viability of this, you know, this product or the service.

168
00:30:07.050 –> 00:30:07.460
David Lowe: Yep.

169
00:30:07.460 –> 00:30:13.109
Harry Malcolm: We’ve started to talk a little bit about some of the things that get in the way of, you know, the successful collaboration.

170
00:30:13.270 –> 00:30:25.290
Harry Malcolm: What, you know, where do… where do industry partners, and actually the healthcare system practitioners or decision makers, sometimes struggle in this… in this space?

171
00:30:25.690 –> 00:30:44.620
David Lowe: Yeah, so, good question. So, given current experiences, often contractually, that’s… I mean, we can spend a lot of time on contract… on… despite having templates and shared ways of working, and even well-established, prior contracts, but we can spend a lot of time in that space. Equally, we can spend…

172
00:30:44.620 –> 00:30:47.830
Harry Malcolm: I was just gonna say, what’s that… what’s the mismatch about there, often?

173
00:30:47.830 –> 00:31:01.580
David Lowe: It’s… I think often, because these are new technologies, and things like IP comes up a lot, so we have to spend a lot of time describing, is there IP being generated, or my view is increasingly

174
00:31:01.790 –> 00:31:21.439
David Lowe: both for SMEs and larger companies, we are creating evidence for a technology that’s already got established IP, and from the NHS perspective, we are not seeking to have IP, because that’s obviously damaging for future investment decisions, especially for SMEs. I think there’s… it’s multi-party, there’s lawyers involved, and things can go around.

175
00:31:21.440 –> 00:31:22.640
Harry Malcolm: Yeah, yeah, yeah.

176
00:31:22.640 –> 00:31:36.860
David Lowe: But sometimes you bring them all into a room and say, please, can we just agree this cause? And that’s a challenge. I mean, as a clinician, going into a room with a bunch of lawyers and saying, like, we just need to agree this, and help them understand

177
00:31:36.900 –> 00:31:51.789
David Lowe: the technology, the pathway, and what we are actually going to do. I think information governance, and we’ve worked very hard to create templates and approaches for data flows, and how we can send patients

178
00:31:52.750 –> 00:32:00.909
David Lowe: Or, to be able to allow access to the structured, but obviously anonymized data for training of an algorithm.

179
00:32:01.000 –> 00:32:16.830
David Lowe: And therefore, where does the data go? Does it sit in a sovereign data hub somewhere in the UK, or does it go somewhere else? And again, working with our legal colleagues, with our patients and our PPI groups to understand what’s acceptable.

180
00:32:16.870 –> 00:32:36.320
David Lowe: And then, I think the last thing is being really clear, and we spend a lot of time on this, is what does business as usual commitment look like afterwards? We’ve finished the study, we’ve finished the additional resources that have been put in, be it from grant funding or investigator-led funding.

181
00:32:36.320 –> 00:32:55.070
David Lowe: But we need to be able to describe, to give everybody confidence, what our route to business as usual looks like. Are we going to stop? Are we going to pause? Do something else? Because often we have very invested clinicians by the end of things that see significant benefits, want to continue, but we need to go out to the market to procure.

182
00:32:55.210 –> 00:33:14.289
David Lowe: Right. And again, it’s being very honest about that, to be able to say, we are going to help you build the best product we possibly can, but this is ultimately taxpayers’ money, so we’re going to build out the technical, clinical, functional requirements of a particular technology, but we will have to go back out to the market.

183
00:33:15.340 –> 00:33:18.959
David Lowe: To make sure that we’re actually getting the best we possibly can.

184
00:33:18.960 –> 00:33:19.510
Harry Malcolm: Yeah, yeah.

185
00:33:19.510 –> 00:33:25.270
David Lowe: The world changes quite quickly, especially in this space, so when you started off in something even two years ago.

186
00:33:25.650 –> 00:33:30.460
David Lowe: What you thought you were going to develop and what has been developed may not be best in class.

187
00:33:30.780 –> 00:33:31.330
Harry Malcolm: Yeah.

188
00:33:31.650 –> 00:33:35.950
David Lowe: But you’ve spent a lot of time giving clinical input, data.

189
00:33:36.170 –> 00:33:45.059
David Lowe: creating evidence that should put that company, that technology, that therapeutic in the best possible place. Yeah. Make the argument for why they are

190
00:33:45.410 –> 00:33:47.740
David Lowe: The… the one we should be procuring.

191
00:33:47.910 –> 00:33:48.450
Harry Malcolm: Yep.

192
00:33:49.340 –> 00:34:07.300
Harry Malcolm: I don’t think you’re allowed to have a webinar in 2026 without talking about AI, so we better do that. But let’s make it really practical, because yes, AI might be the future, but it’s also the now. Like, what are you seeing? What do you want to see more of when it comes to…

193
00:34:07.350 –> 00:34:12.170
Harry Malcolm: practical application of AI in improving in healthcare outcomes.

194
00:34:12.540 –> 00:34:26.729
David Lowe: Yeah, so it’s a great question. I sit within the UK Commission for the hours that we all met on Friday, and that was part of the discussion with the MHRA and other colleagues and other organisations that are decision makers, is…

195
00:34:27.030 –> 00:34:28.459
David Lowe: How do we…

196
00:34:29.590 –> 00:34:43.069
David Lowe: support adoption of AI technologies across all our health boards, our institutions, and there’s some really great examples, like ambient voice technologies. Let’s take one of the really easy examples, it’s very topical at the moment, is how do we

197
00:34:43.320 –> 00:34:56.369
David Lowe: support both our clinical colleagues across a range, from professors of, surgery all the way through to, community nurses, to be able to use that technology to, if appropriate.

198
00:34:56.370 –> 00:35:05.069
David Lowe: to reduce the burden of note-taking. So, for instance, in my emergency department, I still handwrite my notes, which are…

199
00:35:05.840 –> 00:35:25.520
David Lowe: biblically and historically awful, and probably unintelligent, and whatever nuggets of good, data that is recorded in that paper is lost, because it’s just scanned. So, for us, we’re very… it’s a very live conversation around about how do we create the evidence to be able to say this is both

200
00:35:25.520 –> 00:35:29.239
David Lowe: Cost and clinically effective, that reduces

201
00:35:29.250 –> 00:35:49.030
David Lowe: Potentially burnout, but allows our clinicians to be more productive when doing consultations in an emergency department, in an outpatient setting, in the community, and be able to, again, demonstrate clinical and cost effectiveness of adopting that technology, which is quite transformative.

202
00:35:49.170 –> 00:35:59.760
David Lowe: But then the next phase of that is, yes, looking at the performance of that technology, what’s the error rate, is it acceptable, is it non-inferior, but then moving on to the next stage of the co-pilot function.

203
00:35:59.990 –> 00:36:14.169
David Lowe: to be able to say, in your GP consultation, you’re having a discussion around about heart failure, and at the same time, it’s be able to look through your clinical record and be able to say, actually, you’re not optimized based upon your Q risk, and you should consider.

204
00:36:14.170 –> 00:36:14.720
Harry Malcolm: Hmm.

205
00:36:14.720 –> 00:36:25.509
David Lowe: Y and Z therapies, and that’s coming very quickly now. And how do we… that clinical decision support element, how do we, again, demonstrate it’s

206
00:36:25.640 –> 00:36:29.070
David Lowe: Efficacy in the broader sense, as well as

207
00:36:29.310 –> 00:36:44.159
David Lowe: because it’s in the real world, it’s effectiveness, and those are, obviously, those two aspects we often have to think about is, like I say, in a trial versus real-world data that demonstrates effectiveness. So, to your question, for AI, it’s…

208
00:36:44.400 –> 00:36:48.319
David Lowe: Be able to, again, and this was the theme of this conversation, is.

209
00:36:48.320 –> 00:36:48.650
Harry Malcolm: Hmm.

210
00:36:48.650 –> 00:37:05.409
David Lowe: evidence for decision makers, and in this case, it’s both our patient groups to be able to say, how do we… when you chop on the door into a GP surgery or come into a curtained area in an ED that you know an ambient voice technology is being used, you are accepting that it’s

211
00:37:05.410 –> 00:37:14.439
David Lowe: Been used in a safe way, both in terms of data security, but also how it’s being used to support clinical decision-making and care.

212
00:37:14.470 –> 00:37:29.289
David Lowe: And how that data that’s codified in that record, can support future research, future optimization of care, as well, be it risk stratification, case finding, or

213
00:37:29.290 –> 00:37:35.939
David Lowe: at the point of care decision-making around about therapy optimization. And that’s a really exciting space to be in.

214
00:37:36.270 –> 00:37:42.139
Harry Malcolm: I mean, that was going to be my question, like, what’s exciting you the most in the opportunity space there?

215
00:37:42.520 –> 00:37:43.660
David Lowe: So I think…

216
00:37:44.800 –> 00:38:04.520
David Lowe: I suppose we talk about AI being potentially both an exponential opportunity, but an existential challenge in many ways. These sorts of examples, like ambient voice technologies or AI in radiology, which are both being rapidly deployed in healthcare systems across the UK, Glasgow, and the world.

217
00:38:04.580 –> 00:38:12.019
David Lowe: is that… there… Systematic and systemic in their nature, that you switch it on.

218
00:38:12.160 –> 00:38:31.859
David Lowe: And then it’s quite hard to switch it off, one way or another. Clearly, the ambient voice technology, you can unplug the microphone, the chest x-ray AI, you can not look at the secondary capture, but they’ve… they are creating a data layer that can support clinical care in many different ways that are not just

219
00:38:32.120 –> 00:38:51.520
David Lowe: being a rarefied option that’s being used by a couple of clinicians. This is potentially something that is there and available to everybody, so undifferentiated chest pain or shortness of breath, it makes up 20% of ADA tendencies. They all get a chest x-ray. All these chest x-rays will be interpreted by an AI.

220
00:38:51.600 –> 00:38:56.890
David Lowe: You’ll have structured data about what’s present. How does that support public health?

221
00:38:56.890 –> 00:39:15.900
David Lowe: surveillance of potential spikes in RSV, COVID, or flu? How does that support detection of heart failure or lung cancer? How does that increase productivity of our resident doctors interpreting a chest x-ray? How does it improve safety by detecting pneumothoroses that potentially would have gone home?

222
00:39:16.320 –> 00:39:31.949
David Lowe: all questions that need to be evidence about what their potential impact is, and similar with ambient voice technologies, the same, does it support productivity, allow you to concentrate on a more meaningful conversation with patients? We’re doing a piece of work with

223
00:39:31.950 –> 00:39:41.640
David Lowe: Looking at TidalSense, which is a respiratory diagnostic tool for COPD, that if we’re able to demonstrate productivity and efficiency.

224
00:39:42.050 –> 00:39:52.679
David Lowe: can we, therefore, when we make a diagnosis of COPD and somebody’s got a smoking history, use that extra time to have a really meaningful intervention around about smoking cessation?

225
00:39:53.210 –> 00:40:01.050
David Lowe: So actually you’re seeing value into the system. It’s, yes, it’s potentially reducing burnout, as Eric Choppol saw, giving additional time.

226
00:40:01.270 –> 00:40:09.089
David Lowe: to the clinician, but equally, how do we describe pathways? If we adopt these technologies, we are able to add value?

227
00:40:09.200 –> 00:40:12.950
David Lowe: Yeah. And that might be very discrete.

228
00:40:13.220 –> 00:40:25.580
David Lowe: protection of a disease that you might have missed, or speeding up the diagnostic process, like radical in the lung cancer example, or it might be simply having a better third-party conversation, because you’ll be able to make eye contact

229
00:40:25.670 –> 00:40:38.540
David Lowe: and concentrate on the consultation, as opposed to madly trying to write on your knee while purchased on a worktop, which is what I do, unfortunately, in the emergency department.

230
00:40:39.090 –> 00:40:49.320
Harry Malcolm: But you’re asking the same questions of, you know, the AI technology, from what I’m hearing, you know, what’s the data, what’s the value that it would add, and what’s the pathway that would enable us to…

231
00:40:49.580 –> 00:40:50.139
Harry Malcolm: Use it.

232
00:40:50.140 –> 00:41:03.949
David Lowe: Correct. And NICE has got a very good framework that talks about risk, benefit, and harm that we apply to all these, and it doesn’t really matter if it’s, again, a therapy versus a diagnostic or med tech tool, it’s

233
00:41:03.980 –> 00:41:17.209
David Lowe: what’s the level of evidence that’s required based upon its potential risk-benefit ratio? And part of the conversation, I suppose, we’re having in many organizations at the moment is there is

234
00:41:17.270 –> 00:41:18.720
David Lowe: At present.

235
00:41:19.020 –> 00:41:28.359
David Lowe: Risk held by both our organizations and increasingly by our patients, having undiagnosed heart failure by being on a waiting list.

236
00:41:28.930 –> 00:41:30.160
David Lowe: it’s harmful.

237
00:41:31.180 –> 00:41:42.860
David Lowe: And therefore, how do we… having a chest x-ray that’s not been interpreted and then waiting a year in some places in the UK before it’s read by radiologists to identify lung cancer is harmful.

238
00:41:42.920 –> 00:41:53.089
David Lowe: And therefore, while the performance of a technology might not be optimal, increasing the conversation is that we don’t have optimal resource.

239
00:41:53.180 –> 00:41:58.840
David Lowe: So therefore, how do we make good decisions around the risk versus benefit of.

240
00:41:58.840 –> 00:41:59.220
Harry Malcolm: Yep.

241
00:41:59.220 –> 00:42:19.220
David Lowe: Change in pathway, adoption of new technologies, because while it’d be delightful to have expert clinicians, well-calibrated, well-caffeinated at 11 o’clock on a morning, that’s not the reality, going back to our first part of the discussion in most systems. So therefore, how do we identify those at greatest risk?

242
00:42:19.410 –> 00:42:24.019
David Lowe: Of poor outcomes, increased cost, and be able to optimize both

243
00:42:24.420 –> 00:42:31.390
David Lowe: Early goal-directed therapy, evidence-based therapy, as well as identifying and diagnose those individuals early.

244
00:42:32.000 –> 00:42:32.820
Harry Malcolm: Amazing.

245
00:42:33.250 –> 00:42:37.770
Harry Malcolm: Let’s try and bring things together before we move on to the Q&A. So we’ve got lots of…

246
00:42:38.320 –> 00:42:43.310
Harry Malcolm: Pharma company, med tech company, large and small decision makers on the webinar.

247
00:42:43.670 –> 00:42:52.520
Harry Malcolm: If you had to boil it down, what would be your ask, or your advice to them when it comes to, you know, really helping more patients to get

248
00:42:52.820 –> 00:42:54.930
Harry Malcolm: Faster access to optimal treatment.

249
00:42:55.740 –> 00:42:58.850
David Lowe: So, my general plea is around about…

250
00:43:00.510 –> 00:43:12.819
David Lowe: ultimately optimize… the best way to say this. So, I think the first piece is around about building strong collaborations, that going pilot to pilot to pilot across multiple different organizations.

251
00:43:13.160 –> 00:43:32.069
David Lowe: Is that actually achieving the goal you want to do around about pervasive adoption of a technology? So therefore, that… find a partner that you can collaborate with effectively to build evidence to support widespread adoption, which is about pathway change, not individual practice change.

252
00:43:32.280 –> 00:43:32.700
Harry Malcolm: Yep.

253
00:43:32.700 –> 00:43:39.809
David Lowe: And I think that’s got to be an absolute critical message, that you can spend a lot of time trying to persuade individual clinicians

254
00:43:39.890 –> 00:43:48.200
David Lowe: groups to do a thing, especially if you put pilot money in resource. But actually, is it more meaningful to bring that together with

255
00:43:48.270 –> 00:43:58.469
David Lowe: Potentially a number of stakeholders to do one thing that creates evidence that is pervasive and ultimately persuasive to

256
00:43:58.470 –> 00:44:09.760
David Lowe: change a pathway, rather than just change individual practice, and I think AI provides, if nothing else, an opportunity to change pathways, rather than to change individual practice.

257
00:44:10.380 –> 00:44:13.950
Harry Malcolm: There we go. Look for the… Scalable.

258
00:44:14.650 –> 00:44:15.750
Harry Malcolm: Solution.

259
00:44:16.980 –> 00:44:17.930
David Lowe: Correct, yeah.

260
00:44:18.350 –> 00:44:21.479
Harry Malcolm: Let’s go to some Q&A. So…

261
00:44:21.850 –> 00:44:40.310
Harry Malcolm: While we do that, I am going to recommend to everybody that they take a look at the momentum check, which you should be seeing on the screen. This is a pretty simple diagnosis that helps you to see where, in your account working, and in other areas, actually, you might be able to accelerate performance.

262
00:44:40.890 –> 00:44:43.960
Harry Malcolm: So take a look at that, but let’s get into some questions for David.

263
00:44:44.400 –> 00:44:45.380
Harry Malcolm: So…

264
00:44:45.910 –> 00:44:54.870
Harry Malcolm: What about looking outside of the healthcare industry for inspiration? Have you had the chance to, I suppose, copy or learn from…

265
00:44:55.060 –> 00:44:59.069
Harry Malcolm: public-private collaboration in other industries. What does that look like?

266
00:45:00.020 –> 00:45:06.080
David Lowe: So I suppose a long time ago, we looked at that from the airline industry, and that’s popping up back again around about…

267
00:45:06.530 –> 00:45:12.750
David Lowe: To resource management and what that looks like, and we had that conversation last week around about

268
00:45:12.860 –> 00:45:31.620
David Lowe: regulatory approval, so there’s definitely some synergies there, in terms of looking at how other industries regulate or create evidence around about safety, and the airline industry is one of them. I think what we’re increasingly seeing is that fusion of

269
00:45:31.670 –> 00:45:43.580
David Lowe: biotech and med tech together, and previously we were talking about in Glasgow around this therapeutic stack concept of the drug, the delivery device, and the digital companion.

270
00:45:43.590 –> 00:45:55.099
David Lowe: I’m increasingly seeing this coming in through, I suppose, GLPs, doesn’t necessarily have a delivery companion, but biologics and COP certainly does, that all those contribute data

271
00:45:55.170 –> 00:46:00.449
David Lowe: That supports decision-making, both around about therapeutic optimization.

272
00:46:01.030 –> 00:46:09.729
David Lowe: The requirement for additional diagnostics, as well as informing, kind of, future research, and allows us to be able to deeply phenotype

273
00:46:09.730 –> 00:46:29.440
David Lowe: a group of individuals, often with high-consequence disease. Let’s take CAR-T as an example that you’ve already deeply, deeply phenotyped before you initiate therapy, but then you need to continue to monitor these individuals really carefully to be able to manage their condition, and optimize their care. So that’s…

274
00:46:29.440 –> 00:46:44.569
David Lowe: increasingly that concept of how do you monitor in the community, both through physiological, biochemical biomarkers that don’t require people to come in on a Tuesday afternoon for a blood test, or to send a self-addressed envelope back with a patient-reported outcome.

275
00:46:46.860 –> 00:46:54.639
Harry Malcolm: We’ve got a, a fellow country person, or at least perhaps somebody interested in the Scottish healthcare system asking a question about

276
00:46:54.940 –> 00:47:03.399
Harry Malcolm: the Scotland-wide AI strategy being implemented into clinical pathways. What pace and what does that journey look like for change?

277
00:47:04.120 –> 00:47:09.789
David Lowe: Yeah, so great question. So, obviously, the Scottish AI strategy was announced on Friday at the same time as the year…

278
00:47:09.790 –> 00:47:11.740
Harry Malcolm: in now, then. Yeah.

279
00:47:11.740 –> 00:47:23.490
David Lowe: I actually contributed to the writing of it, as well as the one for health, but at the same time, the US strategy for AI was also announced on Friday, so I spent quite a lot of my time reading the United States one.

280
00:47:24.010 –> 00:47:41.390
David Lowe: Which are broadly similar. I think everybody has made a statement. Now, while Trump has said that he wants to win the AI race, I think Scotland’s, aspirations are somewhat, probably more realistic, but ultimately, it’s how are we going to safely adopt AI

281
00:47:41.390 –> 00:47:49.849
David Lowe: Into… into care, but actually the broader society, and that’s a kind of very live discussion around about, ultimately.

282
00:47:49.940 –> 00:47:56.580
David Lowe: The data, how data is used, how data is stored, how data is shared, across both

283
00:47:56.580 –> 00:48:15.659
David Lowe: healthcare providers and potentially industry partners that are creating that data to be able to do that. So, what we talk about, and actually we talked about a little bit with our Taiwanese colleagues, yesterday, is as we move to large data stores within Scotland, we’ll have the national digital platform, we’re going to have

284
00:48:15.660 –> 00:48:28.909
David Lowe: a companion app, similar to the NHS app in England, that allows us to transact data, both from potentially community settings, as well as secondary and primary care settings, together.

285
00:48:28.910 –> 00:48:37.970
David Lowe: And therefore, how do we create insights? Because that’s the key challenge that we see, and certainly when we talk to both our primary and secondary care colleagues, is

286
00:48:38.620 –> 00:48:42.640
David Lowe: What does the service model look like if you start to visualize

287
00:48:42.950 –> 00:48:46.639
David Lowe: Oxygen saturations for every patient with respiratory disease.

288
00:48:46.920 –> 00:49:04.680
David Lowe: Now, clearly, you didn’t… there’s no GP in the land that wants a minority report screen that shows every single person with potentially low SATs across the whole of their patch. So therefore, how do we use AI to be able to support identification of risk? How do we…

289
00:49:04.680 –> 00:49:14.320
David Lowe: help the patient, the citizen, be able to utilize that data? And how can that potentially inform clinical pathways around about risk?

290
00:49:14.330 –> 00:49:32.860
David Lowe: And that’s going to have to be relatively sophisticated, and some of the work we’re doing around about COPD is to say oxygen saturations, many patients have bought their own finger probe. That actually contributes to their understanding of their own disease, at what point, given they do have a chronic disease, at what point do they need to seek

291
00:49:32.860 –> 00:49:40.900
David Lowe: help in terms of a rescue pack for antibiotics, or indeed see the primary care clinician. But that’s…

292
00:49:40.900 –> 00:49:55.330
David Lowe: ultimately, the conversation we’re having is, how do we surface data, both to citizen healthcare providers, and how do we utilize AI to be able to manage that vast volume of data that we are now creating that can actually add value?

293
00:49:56.940 –> 00:50:01.990
Harry Malcolm: Brilliant. We’ve got another question come in. It’s a bit of a long one, so I’m going to read through.

294
00:50:09.070 –> 00:50:21.499
Harry Malcolm: Okay, I think it basically boils down to, we’ve got a good solution from the perspective of what it can do. It will improve clinical care, but also increase total cost.

295
00:50:22.410 –> 00:50:28.509
Harry Malcolm: You know, what’s… I suppose, what’s the likelihood of something like that getting into use?

296
00:50:29.630 –> 00:50:38.770
David Lowe: Yeah, so I think, again, that comes down to evidence, and it’s around about… and we’ve seen that with the opera, the heartfelt example, is ultimately, if you find people early.

297
00:50:38.810 –> 00:50:54.870
David Lowe: diagnose them early, then you increase costs because of the therapeutic, aspects. So therefore, that then leads to quality, and that’s why we talk about qualities, quality-adjusted life years, because we’re demonstrating people living healthier for longer, which is ultimately the aim

298
00:50:54.870 –> 00:51:01.229
David Lowe: Both from a policy, citizen, patient, clinician level, and…

299
00:51:01.960 –> 00:51:20.630
David Lowe: this is a challenge, and I think we have to be really clear about this, around about, let’s say we’re doing some work with a company called Eindoven, which is an ECG AI company based out of Madrid, around about identification of rare disease. And now, the rare disease has a therapeutic option that’s expensive, and therefore.

300
00:51:20.630 –> 00:51:36.299
David Lowe: But it’s gone through nice. It’s got approvals because it improves outcomes. So, we have to be very clear with our healthcare organisations when we’re doing, especially deploying AI at a population-based level for the purposes of population health and precision, both for rare disease and

301
00:51:36.320 –> 00:51:44.110
David Lowe: more common, be it acute respiratory infection or heart failure, COPD, diabetes, that we have to be very thoughtful around about

302
00:51:44.390 –> 00:52:04.130
David Lowe: both the delivery model for that diagnostic pathway, but critically also the ongoing care costs. And we’ve had some really interesting conversations with some of our partners around about if we move to a much more pervasive model of diagnosis, then can we call and cap the cost of therapy?

303
00:52:04.540 –> 00:52:15.860
David Lowe: And that’s quite a bit… it’s a very mature conversation, as well as some healthcare systems are looking at outcome-based pricing. And again, how do you model that? And Olivia Wu, who’s

304
00:52:15.860 –> 00:52:25.749
David Lowe: Prof of Health Economics that I spend a lot of time with, we’re increasingly thinking around about those health economic models, of what does pathway change look like.

305
00:52:26.070 –> 00:52:26.570
Harry Malcolm: Yep.

306
00:52:26.570 –> 00:52:42.110
David Lowe: how do we align that to qualies and decision makers in NICE and IHTA, both for therapies and for technologies, but also then think about how we have a more mature conversation based upon these outcomes around about cost of

307
00:52:42.250 –> 00:52:46.210
David Lowe: Either procurement of, again, that therapy or that technology.

308
00:52:47.480 –> 00:53:00.890
Harry Malcolm: Fantastic. I’m going to, ask the last question, and, it’s going to be from me, so a bonus of being in the conversation with you, about collaboration. You know, collaboration being real art and a science.

309
00:53:00.960 –> 00:53:16.059
Harry Malcolm: you know, what are the key roles, as you see it, to make collaborations work? So not the, you know, not which parties need to come to the table, but what are they doing? What are the things that really glue or oil the collaboration?

310
00:53:17.330 –> 00:53:31.759
David Lowe: Yeah, so great question. So I think, ultimately, it’s about relationships, and it’s building those relationships, understanding, having something really honest conversations about what we can and can’t do, rather than just say, yes, yes, we’ll do it, and that’s ultimately about building trust.

311
00:53:32.000 –> 00:53:47.440
David Lowe: I think what we’ve seen is that having some people like myself sitting somewhat uncomfortably in the middle, often, that triple helix, that understands both the economic business imperative, the industry perspective, the clinical perspective.

312
00:53:47.440 –> 00:53:48.030
Harry Malcolm: Yeah.

313
00:53:48.030 –> 00:54:01.679
David Lowe: and be able to engage with those individuals, as well as the kind of academic rigor about creating evidence for HD for decision making, and be able to create that conduit is really important, and we’re very thoughtful and

314
00:54:01.750 –> 00:54:09.289
David Lowe: Both in terms of developing fellowships, creating system leaders, and programs like the AI Fellowship is a great example

315
00:54:09.360 –> 00:54:23.589
David Lowe: of that, of how do we meaningfully understand that for the Triple Helix to work, you need to effectively communicate, and therefore you probably do need to train people to at least understand those other perspectives.

316
00:54:23.610 –> 00:54:34.029
David Lowe: But often, given the theme of translation, which has been underpinning this conversation, is you need people to be able to translate signposts along that journey.

317
00:54:34.770 –> 00:54:59.409
Harry Malcolm: Brilliant. Look, David, thank you so much, and thank you to everybody that’s joined us. I hope that you will join us for the next webinar in the series on the 6th of May with Dr. Penny Gehrigoglu, who is the Chief Clinical Officer and Consultant Oncologist at University Hospitals Coventry and Warwickshire NHS Trust. And as we move towards the autumn, we will have some guests from outside of the UK and Ireland as well.

318
00:55:00.460 –> 00:55:10.510
Harry Malcolm: Once the webinar closes, a short feedback form will pop up. It is ever so helpful to understand what attracted you all to the webinar, how valuable you found it, and some suggestions for next time.

319
00:55:11.020 –> 00:55:14.299
Harry Malcolm: So, thank you again, and bye for now.

Disclaimer: This transcript is provided for reference purposes only and may contain minor errors or omissions.