AI in publishing, the Darzi report, and population levels
01:08:19 Share

Hello and welcome to The Lancet Voice. I'm Gavin Cleaver and I'm here with my co-host Jessamyn Baganel and editor-in-chief of The Lancet, Richard Horton. There's another bit of a bonus episode for you today as we enjoyed making the last one so much.

We sat down for an informal chat in the Lanset offices in central London and ended up covering xenotransplantation, AI in publishing, the Darcy Report on the UK's national health system, and falling population levels over the next few decades. We hope you enjoy listening in, and please do send me any thoughts or feedback you have, or anything you'd like to hear us talk about, to podcasts at lancet.com.

Right, well, everybody's brought a lot more notes with them this time, so you would hope that the podcast is going to be a little bit more organised than the last one, I guess. Richard, you wanted to kick us off by talking about a meeting you attended recently in Paris on xenotransplantation. The meeting was held in the absolutely magnificent, typically...

National Academy of Medicine. The fall where Louis Pasteur and Marie Curie gave lectures. So immediately you feel that you're in incredible company. Sitting at desks that remind me of when I was a pupil at school. Wooden desks with your name on. In the 1800s. When I was at school, exactly. Thank you for that, Gavin. And it was...

And I wasn't actually expecting the meeting to be quite as transformative as it was, to be honest. Whenever you're invited to a meeting in Paris, you just accept because it is the most beautiful city in the world. But this was incredible because I had not appreciated that the science of xenotransplantation over the last three years, only the last three years, has improved.

undergone a step change such that we are now very, very, very close to being able to transplant. I won't say unlimited, but transplant successfully modified, genetically modified pig organs, genetically modified pig hearts, pig livers.

whatever you care to think of. So I would consider myself relatively up on the news, but I've literally not heard anything about that. Has this been churning along in the background, just constant incremental improvements? No, it's, well, yes and no. So about 30 years ago, there was a burst of interest in xenotransplantation and people like Tom Stiles, who was a pioneer in liver transplants,

He started transplanting baboon livers into people. But we didn't understand anything about the immunology, the reasons for rejection, anything at all. And so they all failed. So people thought xenotransplantation was a dead end. So nothing happened for 30 years.

And then one or two stunning discoveries were made in relation to pig organs to understand why rejection took place. And once people understood that, and then you could genetically modify pigs, so you could produce knockout pigs, in other words, where you've knocked out a particular gene, three genes, in fact.

and then you start putting other genes into the pig, and there are up to about 10 or 11, 12 genes into these pigs, then the organs now don't get rejected, they're tolerated, and they work.

They're not perfect yet, but this is all in the last three or four years, and it's completely changed the field. That is some full-on Blade Runner stuff, isn't that? It really is. It really reminded me of the Keziah Ishiguro novel, the idea where we're going to have animal farms. We're going to have farms where we breed animals, we breed pigs, with the single intention of harvesting their organs.

So as much as it's a medical breakthrough, which will completely change the field of transplantation and all the people who are on waiting lists and unfortunately who do not survive being on waiting lists, we're going to meet a massive global need. But it's going to change our relationship with animals. It's going to raise new moral questions about how we use animals. At the moment, if I give an organ to you,

It's an act of altruism. It's me giving you a gift and I don't expect you to give me anything in return. But as soon as we have farms of animals with genetically modified hearts and livers, well, then we're going to get into the fact that, well, I can buy that heart, I can buy that liver. So we're going to have a market in organs and it's not going to be just a domestic market. It's going to be an international market.

So you can see that suddenly this rather lovely gift relationship that we have at the moment is going to disappear. And we're going to have a very, very capitalistic market in organ transplantation. So there's this immense opportunity, but this rather fearful opportunity.

about what might happen as a result. Unless, of course, it's well controlled. Unless it's well controlled. But my worry is that the science has moved so fast, as you say, you didn't know about it. I certainly didn't understand that it was moving this fast, that the science and the practice could outpace the regulation.

Because we have not had any discussion about this in society at all. I mean, I've not read any newspaper articles. I've not seen any discussion in the media about the fact that we're going to very soon have pig farms with hearts available on demand. I wonder if scientific research more commonly outpaces regulation these days.

Well, it's interesting, isn't it? I mean, sometimes people say there's an ethics industry and they bemoan ethics because we over-regulate research. And I've certainly heard clinical trialists very upset by the regulatory environment, which actually stops them doing research rather than encourages them doing research. There used to be, when we were in the European Union, various clinical trials directives, which were very burdensome.

And I can remember people like Richard Peto and Rory Collins going over to the EU to lobby to try and reduce those regulatory burdens.

So, but here, we don't have anything at all. So it's, yes, I think there are some parts of medicine where we over-regulate and then other parts where there's just nothing. You know, there's another area is biosafety level laboratories. There's no international regulatory system for biosafety level laboratories, BSL-4, BSL-3 labs.

People are doing all kinds of stuff and we don't know. And that's one reason why everybody is so anxious about what happened in Wuhan. Is this a Brexit bonus? No, I don't think it's a Brexit bonus. Because we've been looking for them for a while. We've been searching for them, but I don't think so because we've still got massive regulatory overload in research. But I think in this area, it sure is in need of some

Let's put it like that. I think essentially if it's a medicine and we conventionally think of it as a therapeutic, then there's a fair amount of regulation. But essentially now with exponential growth in biotechnology, in digital health, in organ transplantation, all of these other areas that don't really fit into a kind of here's a pill, you can take it devices, then there's almost there's an exponential gap.

between the growth of what we know and what we can do and policy and regulation to try and think about how we should be controlling and implementing.

Yeah, I think on devices, that's exactly right. I mean, with devices, you just have to be able to show that your device is kind of approximately similar to the device that exists already. And you don't have to prove its efficacy in the same way as you do with a medicine in a placebo-controlled randomized trial. And that's why we got into all the trouble with breast implants in France several years ago.

because breast implants were being produced from all kinds of manufacturers, and there was absolutely no quality control. There was absolutely no proof of efficacy. And so who suffered? The women who received the breast implants.

So I think also when you've got technology now which can be used within devices, then actually it's very difficult to regulate anyway because it's a changing scale because it might be that you have software updates or the actual very fundamental basis of what that technology is doing can change from month to month because of the way that it's being developed and improved. And that's incredibly difficult to regulate.

Especially in a kind of situation where security updates can often malfunction. You know what I mean? It's kind of incremental progress. As we saw a couple of months ago when about a sixth of the world's laptops shut down, it's quite often not only difficult to regulate, as you say, but it's quite often difficult to manage for the actual companies that are putting out the monthly updates as well. It's a bit of a minefield. Yeah, true.

So what have you been up to the last month, Jessie? I have not been anywhere as glamorous in pretending to be anywhere. But I've been thinking and talking and writing a bit about generative AI and our policies here and what other journals are doing, which has been very interesting.

You're our local expert on generative AI. I don't know whether I say that, but I've been very much thinking about it recently. Where are you right now on generative AI? I think it's a very good efficiency tool. And there's huge opportunities to make researchers' lives easier, to make publishing better,

But there's, I was on a panel recently and this isn't my own idea, but I'm going to steal it from someone else. Essentially, you know, our mission is the best science for the best alliance, right? Or, you know, so, but every single thing that we do in that value chain has to do with text. So now we have this innovation that can generate text. There's no part of what we do that will not be disrupted or transformed by this technology. And that's incredibly exciting.

But it's also challenging. And so it will transform the way that we work and the way scientific publishing behaves. How much do you think our authors are using AI tools at the moment? Well, we've implemented this little checkbox. And at the moment, it looks like 6% are declaring it.

When I use it... Is that in research articles? It's more in green content. So that's our clinical education reviews. But we do also, there is a feeling across our journals that more and more people are using it but not declaring it.

And that often results particularly in editorials or correspondence in lots of sort of generalized flowery language. But you can see it's sort of probability linked words put together, but it actually has very little meaning or insight or novelty. And so it's not really adding to the scientific literature.

And so I think I'm thinking more and more of it as efficiency tools. And I'm sure that there's much more that you can do with it. But for things where there's human interpretation, where there's context, where there's a sort of necessity to really understand the landscape and to look forward.

I have not found it to be helpful or useful. Those 6%, do you think it's mostly, and I can see that it would be helpful here, because if I was trying to, let's say, write for a Japanese journal, I wouldn't quite know where to start. So do you think it's for non-English language people?

speakers. But also for English language speakers who might be neurodivergent, who might be dyslexic. There are lots of clinicians and healthcare professionals who are dyslexic or have other neurodivergent issues. And I think that for them, being able to put something through a large language model that will formulate it in a way that's sort of socially acceptable, then that's a

an incredible tool, you know, and will help with inclusion and diversity. I think it's a really good blank page destroyer as well. That's what I often use it for. If I just can't think of how to start a particular thing, I'll go to a large language model and say something like,

How would you start this? Or, you know, like brainstorming, brainstorm, like write the first paragraph of this. If you were me, you know what I mean? And then rather than copy and paste that, which is very lazy, you can actually finally look at the thing. And it's almost like you've started writing the article. And often I find the starting is the issue. And then once I've got the start, you know, the rest comes on. So I often use it as kind of like a prompt machine more than anything.

Yeah, no, I mean, it's interesting. The way into a subject when you're writing is so important. And you don't want your first paragraph to be like a cough. That is just a sort of boring, you know, rendition of a summary. On 8th August 2016. So you want to try and find...

Um, Kingsley Amis always used to say that he should begin a novel by something like a shot rang out. In other words, you should have something that immediately catches the reader's attention. And that's just as true in scientific writing as it is in novelistic writing, I think. So, and I don't know whether a large language model can do that. Um,

I think it's very good at doing that, particularly if you can give it examples and prime it before and say, this is the style that you should train it. This is the style that I want you to make a big impact. But I think Gavin's point is, you know, that's the ideal where human creativity is in harm, but the oversight and the intelligence and the...

the kind of, and the authority, I suppose, remains human. So I go off to say to my multimedia colleagues these days, we have to be on top of the machines by this point. Otherwise, the machines are going to do everything for us. The other thing that I think is really interesting is that it's having an effect the whole way down, not just scientific research, but how people earn qualifications as well. It's having a major effect. Anecdotally, my partner is a course director at a local university and

And it's got so bad there that they have switched to entirely in-person assessment. So they replaced all of their essay writing for one course this year.

with an in-person science day where students, undergraduates had to come and present a particular scientific paper as if it was a poster at a scientific conference and then talk lecturers through that, which ended up being fantastic because also they got local school children in to see these presentations and then they could see people that looked like them presenting scientific research and things like that. So very wonderful.

But it's striking how quickly it's kind of infected assessment at a university level. Because everybody was using it to write their essay. Yeah. Yeah, that's pretty scary.

I think there's so much to talk about on this topic, because there is a huge element of de-skilling, whether that's researchers, whether that's at higher education, whether that's in medicine. What happens when we rely on things to such an extent that actually we don't? You could get people to actually write their essays in an exam situation. We did this at A-level going back a few centuries.

where you'd actually have the essay topic and you wrote it by hand. You know, you have to write for an hour and a half. This is a sign of two people that didn't do a humanities degree. Yeah, that's how we did all the S6. All of my exams were turn up and pick two essay topics and you've got four hours. Exactly. So that would be... And if you have to write them by hand, then you can't use gerative AI for that. No. So...

Well, then it's so funny, isn't it? Because there's this, there's a diversion there. Because on the one hand, we've got, you know, this, where there's generative AI basically being able to write university level essays. And on the other hand, I've got my kids who are at a state primary school and learning exactly the same thing that I did in exactly the same way as I did, you know, 30 years ago, which also doesn't seem quite right. Good progress. Well,

If there's something that works, then why change it? Well, it's interesting though, isn't it? Because when they grow up and reach the workforce, they're never not going to have access to these tools. And so sometimes you can look at like, for example, a school smartphone ban and say, well, actually...

you know, they're always going to have this to hand. They're never not going to be able to just pull something out of their pocket that's got a calculator on it, for example. Can I just say, I haven't done this. I ought to try it. But if I wrote in to chat GPT, please generate me a data set for a randomized trial in 5,000. It would do it. Yeah, it would do it. It would do it. There's a great study, I think Nature, that did this. They had a false hypothesis

something to do with ophthalmology and said please generate a manuscript with a data set that shows that the hypothesis is correct and it did it perfectly. The models have even taken in so much of for example the Lancet that you could tell it to output this piece of scientific writing in a Lancet house style. Or a Richard Horton style. Yeah it can do that.

Incomprehensible. So how do we, what checks? So we have, at the moment, we have authenticate for checking for plagiarism. There are various image manipulation tools. Are there any tools out there to check to see if a paper, so you said the self-declaration is 6%.

Are there any tools? We don't have any reliable technology that will pick up when generative AI has been used. I think there's different claims on the market that tools are more successful. But my understanding is that that tends to be that they get high percentages of pickup because they're using a very small data set to train it on and to validate it on. And then when they actually put it out into the world,

it doesn't you know picks up 50 or something like that so there's nothing that we can reliably use but i think that's a you know that's obviously a huge area of activity for someone like us

That's also become one of the problems, again, at university level is they have lots of checkers that they run students' essays through. It comes back and says this essay was 90% AI generated and the student goes, well, no, it wasn't. And, you know, you can't go into their history and look at it. You can't prove anything. So what we're going to be asking for is the prompt that you use, the large language model version that you use and how you asked it to...

change your manuscript or to help you develop your manuscript. And I think we as editors then, we're going to have to ask the question in the manuscript meeting when the paper comes up, you know, could, whether AI has been self-declared or not, could AI have been used in a malign way in this paper? And I guess we have to think about what malign means, because actually using it to enhance a

formatted, structured article that hasn't changed in centuries, is that so bad to make it more readable? As long as there's been human oversight, as long as it hasn't done the data interpretation, as long as everything's accurate. So data interpretation, that would be more, wouldn't it?

I think it would be, and our policy is that it shouldn't do data interpretation. But, you know, again, to what degree? So if you say, if you've got a huge data set and you say to a generator, can you help sort this out? Can you clean it? Can you divide it into samples, which will make it more easy for human methodologists to interpret?

You know, there's a... Or if you put it... If you feed it a base and say, draw me five conclusions from this. That would... We would feel that would be... It's sort of the case where you ask it to do things beyond cleaning and sorting data is that's when it can start hallucinating. And that's when, especially in large data sets, you could get just random numbers introduced because it's...

all it is really is a predictive tax model so it puts the numbers there where it thinks those sort of numbers should go but at that point if you've got a live data set and you're the person you know doing the research you don't know which numbers are hallucinated or not so that's when the danger comes in I'm feeling more anxious as this conversation proceeds I think we should be you do okay then we have to think what do we do about this well we can do for now spend more time on papers

I mean, think about them more than we probably do. I mean, we think about them a lot, but we need to think about them more because...

We have highly experienced editors who will read them. And for the most part, even if AI hasn't been declared, they'll be able to make a judgment as to whether this is novel or whether this is actually adding anything to that debate.

But I just wonder, you talked about the self-declaration. Maybe we need to ask the direct question. Have you used AI? That's what we ask. Oh, so they ask that. So when they submit. So when they say no. When they submit, they have to say, has this been used in any part of your work? And then if they click yes, they have to describe how. So if they said no, we hope they're not lying. That's essentially as far as we can go. But we think that some are.

Across the journals, editors feel that they particularly are in correspondence

and commentary because they have seen an increase in content particularly from some authors who might not have any expertise in that area who are sending in a correspondence or a commentary on a topic actually i've noticed that in the letters because monique looks after all of our letters on the weekly but all the letters that she shortlists she shares with me

And I see letters from some places where they're commenting on something and you see the authors and where they're from and they've got absolutely no particular reason to be commenting or expertise on that particular subject. Exactly. And you do think that's weird. Yeah.

I don't know particularly what the phrases are as well, but I get the impression that there are some phrases in AI-generated work that are hallmarks of AI-generated work that aren't often used by people, but for some reason are in the AI data sets. And so apparently that's another way that you can spot it is these particular phrases. We've seen that in some peer reviews, haven't we? We've seen some peer reviews which have clearly been generated through some sort of chat GPT model.

where it in the way it's written it's it's almost like written this person's this ethereal character the the should consider it sounds very nice quite hyperbolic very flowery language but you'd never write that but you'd never write you know we would never write the author should consider getting a statistician to look at these data i mean you'd never write that but

But that's, I've seen nothing. You just write stats in caps. You just write that this is clearly mistaken and a frightening new world. It is. Talking of frightening new worlds, Richard, the Darcy report came out in between podcasts. What did you make of it? Well, I know Aaron Darcy quite well because I've followed his career over many decades and

from a stunningly successful surgeon to becoming a Labour minister in the Tony Blair government or Gordon Brown government. And he was put into the House of Lords when he became a minister. And he really championed the idea of two things in particular, one quality and second, which got general practitioners up in arms, polyclinics. And then...

when Labour went out of government, he kind of went quiet for a while and then came back, as you say, with this report on the NHS. He said that the NHS is in a critical condition, but the vital signs are looking OK. And if you read his report, he

I would say the emphasis is more on the critical condition than the vital signs being okay. What's particularly interesting about his report is not the report itself, but the annex, which is 330 pages of tables and graphs, which just shows how decimated in almost every department of the NHS care has become.

And he lays it out completely dispassionately. There's no text written in the annex. It's literally just tables and graphs. But as you read through them, you just see how really it's been over the last decade in particular, how the NHS has become so critically eroded. And he places particular blame

for that on Andrew Lansley's Health and Social Care Act 2012, because that was the reorganization, which he has some very choice words to say about it. He calls it calamity. That really was the starting pistol, because once that happened, it was a

disorganization, not a reorganization of the NHS. And it never recovered from that. And then after the decade of five prime ministers under different conservative governments, the NHS has indeed reached this critical position. And it's not clear at the moment what is going to happen because Keir Starmer said,

at the launch of Arradaza's report, Reform or Die, which is a great phrase, but what reform?

I mean, it used to be the case in UK politics that we would talk about, there's no magic money tree. People would, you know, say, well, we've got this. And in fact, that happened a lot during the election campaign. Everyone was just talking about how they would make efficiency cuts to fund all of their particular pledges. But I think the new kind of watchword or the new kind of catchphrase in British politics these days is the reform fairy, which is just the idea that we can wave a magic wand and do this magic thing called reform.

and then everything will be all right. Whereas actually, it's the capital investment that's required. So, Arafat Darzi emphasised the importance of capital investment into infrastructure. At the end of his report, there's a very brief section on what should be done. To be fair to him, he wasn't asked to write a report on what should be done. He was asked to do a diagnostic audit of the NHS. But Arafat

What worries me is talking about the idea that technology is going to solve the problems of the NHS, greater patient and staff engagement is going to solve the problems in the NHS, having a moving money closer to where the patient needs to be cared for, creating a neighbourhood NHS. These are very fine phrases, but what does it really mean? Because at the moment you have hospitals...

delivering care, are you really going to start closing hospitals and moving money into community care and primary care, maybe reinvent the idea of the polyclinic? It's not that easy. And it's certainly, he's talking about a 10 year period of NHS reform.

I can't see clearly what that the shape of that reform he makes. He does say very, very definitely in his report, there should be no Lansley like top down reorganization again. I think we can all agree with that. And the integrated care system model that we have at the moment is fantastic.

It's not perfect, but it's a lot better than what we had before. And the idea of integrating care is absolutely the right idea. But in terms of the reforms that need to be implemented now, we really have no clue. And so we've got the diagnosis and then there's a whopping vacuum when it comes to what do we do next. Go on then, you're West Streeting, what do you do next?

Well, what West Streeting has done is he's created another group that's meant to come up with the answers. Is that what you would have done? No. And I said this to Ara directly face to face. I said, well, you've produced the diagnosis and now you're going to walk away and leave it to somebody else to come up with the answer. That seems completely crazy to me. You know now better than anybody where the problems lie.

lie. So you should be involved somehow in working out what the solution is. But that's not what Wes Streetsing has done. He's handed that over to a different group. And I think that's a mistake. Because the people he's handed it over to are not working full time in the NHS. Aradazi is working full time in the NHS. So he has a stake in the solution. So I am concerned. I think what I...

You can't do everything all at once. So there's got to be a few objectives. And probably the biggest concern at the moment is, and this comes out in the 330 page annex, is the really catastrophic length of waiting times that are facing, well,

every everybody from cradle to grave i mean if you look i looked at the data for children you know tens of thousands of children waiting um over a year for an appointment is just unacceptable in a health system like the uk's so uh i think that um i think my focus would be on how do you get the waiting times down so it's access to care and at the moment

I think the evidence is pretty clear that there are so many barriers to access the cloud that that's the problem. Because my reading of his report, the vital signs being good, what he's saying there is we do have fantastic NHS staff. And we do have more staff than we did in the very recent past as well. And that's one of the fascinating things.

facts that comes out of his report. We've had a 17% increase in NHS staff between 2019 and 2023, but productivity hasn't gone up.

Why hasn't productivity gone up? Because there hasn't been the infrastructure investment and so on. Addressing infrastructure, but that takes a longer time. I think it's getting the care pathways worked out so you can get people in to see a primary care doctor, into community care or into a hospital. And

If you can address that as the number one priority, then let's then think about what the other phases are. But the most important thing has got to be surely in the National Health Service, that when a patient needs care, they have access to care. And for literally tens of thousands of people, if not hundreds of thousands of people, they're not getting access to care. And it feeds into the other problem that the UK seems to be having, which is a lack of productivity caused by so many people unable to work.

These people are waiting on waiting lists. They're not participating in the workforce anymore. Growth across the country is sluggish. I think we have a record number of people currently out of work and not seeking work.

And Aradazi makes the point that NHS should be fixed because it is an engine of prosperity for addressing exactly that point, Gavin. And what's interesting about what West Streeting has done in terms of prioritising the parts of the country where you can't do everything all at once, so where do you begin? And he's arguing you begin where you have most people who are

not in work because of sickness. So that's a very interesting way of prioritizing your NHS resources. So they're making a very explicit link between health and the economy. And for a government that's put the priority on growth, you can understand why they might have done that.

At the end of our last podcast, I was slightly sceptical about Darcy's report because I felt we have heard he's been involved in UK health for a long time. Surely we needed some fresh voices. But actually, the report is absolutely critical. And it's a fascinating and brilliant report that we haven't had this level of transparency on where the NHS is for 15 years. And we need it.

And I think that for me, what came across is this capital aspect.

The productivity, I hate the word productivity. For the Conservative government, it always felt like asking NHS staff to do much more for less. But actually, when you look at the hard data, you know, there are 7% fewer daily outpatient appointments for each consultant. There are 12% less surgical activities for each surgeon and 18% less activity for each clinician working in an emergency medicine department. That is an extraordinary decrease in productivity. I mean...

it really fathoms the brain how that can happen when we've had such technological advances, when we have incredible staff dedicated to working, committed to working. And when I look at that annex,

The capital issue does seem to be such an important one because that 37 billion of being able to have more hospitals, when you look at the investment that happened under the Labour government in those first 10 years, it doubled. It was extraordinary. That's how you're able to do so much more. And I don't think that we can get away from that. So although last week, Wes Streeting at the Labour conference, it was a great speech. He said all of the right things. But

But actually, and I'm delighted that he's negotiated a salary increase for resident doctors. Brilliant that, you know, we're going to say that care workers need to be paid more. This is all fantastic. But there was nothing about increased capital. There was nothing about an actual social care system that we were going to pay for. And without that, it's very hard to see how these three, you know, tilting to technology from sickness to prevention,

How are we going to do that? Yeah, I agree. I agree. I agree. Take one specific example then, because you come from the world of surgery. So the point about we've got more surgeons and we've got less units of surgical output, as you just quoted. So what...

What's the reason for that? Is it that we don't have enough operating theatres, or is it that we're not using our operating theatres efficiently? Are some of them not being used during the day, or do we not have enough nurses?

Should we be using them at weekends? What's the sort of reason? I think of the whole mixture of things. I think when you look at this report, you do see the huge impact that social care is having of stalling up beds. So you actually don't have beds to move people into. We don't have the right infrastructure in terms of surgical pathways of people coming in and being efficiently, you know, operated on. And we've seen that now with these new vanguard systems.

areas where surgical teams are doing high productivity and they are decreasing the surgical list, you know, very impressively. And then they're, you know, doing very low sort of not difficult operations, but lots of them and having a high output. So there are a lot, I think there's a lot of room to be able to turn that around. But we don't,

haven't focused on that because I know you know when I was um as we now call them a resident doctor

you would literally, I mean, the idea of day case surgery hadn't even been introduced. And so patients would come in and they come in on a Sunday night and then you Clark them and they'd have their operation on the Monday. Um, and then they'd be in all week and then you'd hope that you could kick them out at the end of the week. But, um, and mostly you did now, of course, everything's day case and that's laparoscopic. Um,

So the possibility of increasing the productivity, this word, I agree with you, it's not a pleasant word to use in health, but in terms of the units of output per person should be enormous. But we're not, for some reason, I'd see the point about social care, but is it right?

I mean, that's one part of the population, but it's not everybody. It's not, but I also think it's crisis. And for each, you know, we have a health system, but the boundaries of each health system changes. You know, there's a hospital, there's a hospital trust, there's the out there. So when a hospital trust and every single one in the country has been in crisis for about 10 years,

You mean financially? Financially, but also from a position... From a capacity point of view. From a capacity point of view. How do you ever have the headspace? How do you have the resources and the capacity to change things around to make things more productive? You don't. You're just on black alert the whole time. You're just trying to...

survive and that's where I feel I was at an event a couple of weeks ago on waiting lists and I was having this conversation with other NHS leaders and it's really hard when you I mean these are such complex organisations that are in chronologically how do you actually then

take a step back, divert resources to say, okay, we're going to do the whole of our surgical department differently. Yeah. It's incredibly difficult. And then maybe there does need to be a complete change in the way we think that actually we have operating theatres open 24-7. Yeah, maybe. Weekends. Maybe. Yeah.

I mean, then you would expand your capacity considerably. You'd sweat the assets, for sure. You would, you would, you would. But... For a long time, like Jess was saying, there's still no beds to discharge people into. That capacity is not there.

I remember when we did a series with one of our colleagues to be a client on the German health system. This was obviously pre-COVID. One of the great strengths of the gym, although they saw it as a negative at the time, although as it turned out with COVID, it turned out to be a great positive. They had a huge amount of spare bed capacity in their hospital system, which that's

they didn't use at all. And so they didn't think they were very efficient. Right. So I've been reading some articles about this this week, and that was the major complaint about the NHS in the late 70s, early 80s, wasn't it? That there were too many spare beds. And so from 1987 to 2019, beds per capita in the UK decreased by 53% to try and meet the kind of... Say that figure again. From 1987 to 2019, beds per person in the UK decreased by 53%.

My God, that's quite a figure. France currently has three times the capacity per person and Germany currently has four times the capacity per person that the UK does, which are astonishing figures.

And, you know, maybe from the hospital point of view, maybe that's the route of one of the routes at all. I mean, also, when you walk around our hospitals, you know, it would be so much easier to just build a new hospital. When you're talking about Northwick Park, we talk about High Wycombe Hospital. These are dilapidated buildings which have been built on over time where there's bits of flooding coming through and.

how do we make these into places that are meant to deliver care in the 20s? So you would recommend then, I mean from a hospital point of view,

You would recommend a massive hospital building program, massive investment. Well, it's not even building at this point, is it? It's just rebuilding what we currently have because it's all falling down. Well, okay. Actually, I've got a good figure for this as well. At the end of 2023, guess what the urgent backlog of maintenance costs was in the NHS? The backlog of maintenance? Things that needed to be just maintained right now. This is a money figure. Yeah.

A billion. Ten billion pounds. It's enormous. It's in such... So I think you're right when you say the vital signs are so strong because we have such an incredible... It's the people. ...star. It's the people. You know, people... That's the energy. That's the energy. But our actual reason... But the problem with that argument is that while I...

Having worked in a hospital, I love hospitals. They're palaces to modern medicine. They can't be the place. I agree. Every report that's written, whether it's the U.S. or the U.S. It's all about prevention. It's all about primary care. But that's where the reimagination and a reform of how we're viewing health starts, doesn't it? They both have to come together.

Because the reimagination of bringing health closer to communities, of having neighbourhood centres, you know, all the rest of it in the IPPR report, fantastic, but that's going to take time. And meanwhile, we also have to have places where we're delivering care, just as you say. Because primary care is falling apart even more. Also falling apart. Well, we were talking about staff rising, weren't we? But actually the number of GPs has fallen, which has a big knock-on effect.

And what's fascinating is the encroaching of the private sector into that. So Bupa is one private sector organization has introduced this

system where you can pay it's like a Netflix subscription every month and for that you get access to a GP by phone and if you pay a little bit more you can get place to base access to a GP I mean if we're all buying on Netflix and Amazon Prime subscriptions 25 quid a month isn't you know for a certain proportion of the population it's going to be affordable and then you end up with

a two-tier system that you're going to have people who can afford to pay the 25 quid a month and get to see their GP whenever they want and everybody else has to use the NHS which has got has been

But it's also not like we're magic. It's not like these private companies are magicing these stuff out of thin air. No, it's the same pool. Yeah, exactly. Can I take us on to another intractable problem that I was looking at when we... You mean we haven't solved the last one? We're not going to solve the other one today, but maybe tomorrow. But in that appendix, and people often say, you know, we're a sicker, thicker, poorer nation. Yeah.

These poverty and deep poverty rates in children, we've flatlined since 2001. There are so many figures in that appendix where we're just flatlining. We haven't done anything for 20 years on some of these massive issues. And yes, things have got worse, but we've also had a political system which has totally...

If this was a video podcast, I would see Bridget and Jasmine wading through pages and pages of very serious-looking graphs. Yeah, this is a country that's making no progress. This is not me. We've made no progress in 2001. Yeah, no, absolutely. This is a country that's making no progress at all. And it's interesting because West Streeting was...

quite heavily criticized for saying that the NHS is broken and that that's just too much doom and gloom. But when you actually look at the data, let's not say the NHS then, but our health system is broken. It's not delivering for people because either waiting times are getting worse, we're getting sicker,

Access to care is reduced. Quality of care is mixed. Productivity has gone down. I mean, there are very few signals apart from the science. And this is where I think we do need to be optimistic because science is delivering for patients. You know, the transformation...

in care across almost all domains of medicine in the last decade or so. It's incredible the contribution that scientific advance has made, whether it's in medicines, whether it's in diagnostics, surgical treatments. This is the cutting edge. And

I think, you know, one issue that's often not talked about is how can we make sure that we rapidly translate scientific discovery into clinical care? And I don't feel that we talk about that enough. And in America, they do that not everywhere, but they do it very well. If you go to the Cleveland Clinic, literally everything.

I mean, perhaps not every single patient, but it feels like every single patient is entered into a randomized trial. They mix together and it's in the signs all over the Cleveland Clinic. Care, education, research,

There's a sort of holy trinity. So this is an institution that's all about serving the patient. It's an institution that's a learning institution, and it's an institution that's generating discovery science. And those values at the heart of an institution, you really feel that it's like

working full speed. It's really sharp as an institution. And the hospitals I go into in the NHS, you do not feel that. What you feel is this is an institution full of way too many people just struggling to

as you say, to keep its head above water. But you don't feel that it's about serving the patient. It's a learning institution and it's all about cutting edge science. And if we could institute those, you know, one of the areas where the UK could be proudest of any country in the world is the quality of its science and its quality of its life sciences and clinical sciences.

in particular, but they seem to be utterly divorced from the NHS. And I do not understand that. I do not understand how somehow we've got a management structure for the NHS, which is utterly separate from this amazing science research ecosystem that we've created. And somehow we have to bring the two together. And I do think that the values that drive discovery science and make us

literally the leader in the world per headcount of scientists. I do not understand why we cannot transfer those values into the NHS. And if we could, I think that would be transformation. And Ara doesn't talk about that in his report, actually. But maybe that would be my prescription.

Very good. I've got a good link here. Richard, you mentioned headcount just then. One thing you wanted to talk about before we wrap up today was Hungary. Why don't you tell us a little bit about that? Well, one of the great mysteries of this coming century is going to be the demographic time bomb that is going to explode as the century proceeds. And...

We've known about this demographic time bomb for some years now. I remember studying it at school. Exactly. So, you know, it's not a surprise. It's not a surprise. But we published a paper a few years ago from our friends at the University of Washington, IHME, which looked at the projections of population at 2100.

And it's a paper that is one of the most important papers I think we've published, not in terms of medical science and the impact on patients, but in terms of just the way the planet is going to evolve. Because so many countries are going to see their populations collapse. By the end of the century, countries like China, Japan, South Korea, China,

Countries in Europe like Poland, many of the former Eastern Bloc countries, Spain, Portugal, Italy, you're going to see their populations half, literally half. And what that means is you're going to have an elderly population with a population

dramatically reduce working age population and the social contract that we have at the moment whereby the working age population generates the tax revenue that creates the welfare state and keeps everybody in particular older age populations safe and secure that social contract is going to break down it's not going to work because you're not going to have a working age population that's going to be able to create the tax revenue that supports the welfare state that supports the rest of the population who's not working

So the countries that we currently think of as really critical to our future, countries in Europe, countries like Japan and China, that isn't going to be the case by the end of the century. So the point is that we have this problem.

we're not producing enough children. Now, this is a very, very difficult area to talk about because where you want to start off with is a position where you're talking about sexual reproductive health and rights. It's not about coercing families or women in particular to have more children. It's about protecting the rights of individuals and

but recognizing that we have this challenge. So just in the last few days, Vladimir Putin has talked about the catastrophe, as he calls it, in Russia, because Russia is going to be one of those countries that's going to see a dramatic fall in its population over the course of this century to essentially unsustainable levels. Now, the response in Russia...

has been exactly the wrong response. It's basically a response which is highly coercive. It's attacking feminism. It's attacking sexual reproductive rights. It's talking about how you can restrict abortion.

very, very aggressive withdrawal of rights from the population. Interestingly, in Hungary, which is not what you would expect, Viktor Orban isn't exactly the...

Progressive.

There are tax breaks. There are benefits on mortgages. There are low-interest loans and grants, all given to families who have more children. So he's tried to create a set of pro-natalist policies that

Because Hungary is one of those countries that's going to see a catastrophic collapse in its population. And if they don't do anything about it, they will be an unsustainable country. So it's very interesting what you're saying at the moment. This, you know, China's plan was to go from a one child country that it went to two children. And I think it's gone to three countries.

But it hasn't worked. You're still seeing a very rapid decline in fertility. But also in Hungary, it's done subtle. In Hungary, it worked. Well, it worked for a short term. It worked in the short term. And now it's gone back to roughly where it was. Which is a lot. Which is a lot of money. It's a lot of money. It's true that it's fallen back a little, but it did advance significantly.

little as well. So they went from 1.2 births per woman to 1.59 births per woman. Now, what do we do about this? There are two solutions to a catastrophic population collapse, as we're going to see in many countries. One is what Victor Auburn was doing to try and have pro-natalist policies.

Basically strengthening the welfare state to encourage families to have more children. And the second is another hot topic, immigration. And the solution surely has got to be a mix of the two. You strengthen the welfare state so that families are safe and secure to have children. And you attack this vicious system

racist, anti-immigration movement, which has swept across the world. Because unless China is willing to accept more migrant migrants, unless Japan is, unless much of Europe is, then it's not going to be sustainable. So it's a very, I find this whole, you know, demography is not destiny, but it sure is a big part of destiny. And

It's going to affect our future in ways that, again, we don't talk about. It's not in the public debate. Well, the other issue affecting countries like Hungary, of course, is that often skilled workers emigrate using freedom of movement across the EU to countries where they can expect to earn a better wage and have better working protections. The brain drain. Yes. We need an ethical framework to be able to...

sort of shape immigration to around the world to balance out these populations. But I went back to our GBD, obviously our Vival, about these pronatal policies. And I, because I just think it's such a fascinating area and some of these policies are so interesting. So in general, okay, there's very little data on whether they work or not, but there's

IHME say that such policies have led to strong, like very few have ever led to strong sustained rebounds. And most empirical evidence suggests that there's an effect size of no more than 0.2 additional live births per female for any pronatal policy. So when they change their projections to 2100, where the replacement rate is 2.1 total fertility rate,

And if we have no pronatal policies, then we'll be at 1.59 average globally in 2100. If we do pronatal policies, that will only take us up to 1.68. So we have to have immigration. We have to have immigration. But I would say we also have to make sure that we're creating a welfare state which protects women and families so that they can safely have...

children. And look, this is a hot political issue. You know, in the last two days, we've seen a contender for the Conservative Party, Kemi Badenoch, talk about the fact that maternity... Come, let's let it go. Yeah, basically. It's too general. Maternity pay is too burdensome for companies. One of the most incredible

correlations I've ever heard a politician make, which was Kemi Badenoch, when she said that we didn't used to have maternity pay and women had lots of children. Yeah, I know, I know, I know. So if you imagine that you're a woman in her 20s in a cost of living crisis and you're thinking about having a family and

And you've got Kerry Badenot saying that she's going to withdraw, and you're working, and your company provides certain maternity benefits, but she's saying we want to withdraw those, see the reduce. Is that going to encourage you to have a child? No, it's not. Of course it's not. So the stronger the social protections are,

the more that's supporting the rights of women to choose. That's what we're trying to do. We're trying to strengthen the rights for women to make the choices they want to make. And at the moment, we're not doing that. I agree. I agree, except...

I do also think, because this came, so there was the Budapest Demographic Summit, where all these right-wing leaders go across and talk about how they needed a normal baby. Yes, I have some choice quotes. Oh, yeah, go on. One of the featured speakers was Jordan Peterson, which is always a sign of a good event. And he said, the proper encapsulating structure around the infant are united and combined parents, man and woman. All alternatives to that are worse. Single people, divorced people, gay people all deviate from that.

Okay, so this is my point, is that this is our approach to it, which is policy-based, which is evidence-based. But there is actually a whole group of people who come at this from a completely different perspective. And many of them are women as well. They're not just men that are saying this. There are many women who think that,

that wokeism is bad, that we should all go back to being married, that we should stay in that. I don't understand how that's the conclusion that you come to, but it is a conclusion and it is an approach for a really large proportion of most high-income countries. But I want to argue a rights-based approach to strengthening social protections

as a as a means i want to argue with you i agree with you but there are too many jordan peets yeah no it's it's uh but if we if we don't do something we're already 2024 um the world's population is going to peak i think that the ihme said around you know 2060 something yeah um

So we really have to start bending these curves the other way for many of these countries. Otherwise, we're really going to be in trouble. But to go back to the beginning of what is now an extraordinarily long podcast, this is where we get the robots to help out, right? Yeah.

How are the robots going to help? They'll automate productivity in some way. The robots will earn the money and we can all sit around at the seaside. 2160 is when it's going to peak. But before that, you're going to have these dramatic changes in high income countries who are also playing with the idea of populism. And so how do you both tackle? Because it's a sort of downward spiral.

Because you've got demographic changes, puts pressure on the economy. There's less money that's going around. We've already talked about how basically in the UK, we've made no progress on child poverty. We've made no progress on people living in poverty. So for this group of people over the last 20 years, nothing has worked for them in the centrist approach to politics. Now we've got this very changing demographic climate change. All of that's going to make it worse. Right.

And the only place for them to go to, for the most part, is a populist approach. Because a centrist or even a progressive approach, which is policy-based, probably doesn't sound as enticing as we think it should do. It's also, I think, the case is that a lot of politicians just spend their whole time kicking that particular generation, let's call them Gen Z, for want of a better phrase, and then wonder why they aren't doing the things that they say.

Having a child is far less affordable than it used to be. Raising a family on only one income is almost impossible in a lot of high-income countries.

And then politicians, a lot of whom contributed towards the conditions of unaffordability, of the constant need for growth. I like, well, why is no one having kids anymore? While they look at, you know, two people together who are both working full-time careers in order to afford even the basics. Well, it's very interesting you raise that because I had a very interesting conversation with somebody here. I hadn't thought of it this way. We were discussing...

Why has there been, because we're about to publish a commission on self-harm in a few weeks, and we were discussing why have the rates of self-harm among women, young women, adolescents, increased so dramatically. And of course, there has been this discussion about the role of social media. And interestingly, what she said to me was she didn't buy the story that it's social media.

And she's younger. She's in her, I think, early mid-mum, baby mid-thirties. And her argument was, she said hers was the first generation, which every generation before hers expected to be.

earn more, succeed more than their parents. And hers was the first generation where she had to look at her future and realize that it may not be as secure or as prosperous as her parents. And that, she said to me, that people like me had to understand the psychological shock

of that realization that that understanding that your future may not be as secure as your parents future is a hell of a realization to have to confront um and of course in that case then the idea of families children and all this sort of thing is completely off the table because it is you know you just don't have a

You don't have that. Again, it's about choices. Your choices are contracting. You've got fewer choices today than you did 20 years ago. And that's... So that's a policy question then. How does a government expand choices for people? And part of that is about

social protection of which the health system is a part. Yeah, I agree. I 100% agree. Any functioning capitalist society has to have an extremely strong role for the welfare state. That's just how it works. There's winners and losers in a capitalist society and the losers need to be supported. But that's where this demographic issue then becomes so vital because the social contract that underpins that welfare state, those social protections, is under threat. And I don't understand why we're not talking about that more.

Luckily, well, we are, but the wrong people. Exactly, exactly, exactly, exactly. Fortunately, the UK's population is pretty stable at 2100. Germany's population is predicted to go down by about 20 million. France's is pretty stable. America, which is still, despite all the political argument, a country of immigration, is stable. So there are points there.

or a stability for the future, but there's a hell of a lot of change coming. Well, it seems an upbeat place to leave it. Jessamy, Richard, thank you as ever. And thanks to you for listening if you're still here after an hour. Thanks so much for listening to this episode of The Lancet Voice. If you'd like to find out more about any of the podcasts that The Lancet puts out, you can find them all now on thelancet.com slash multimedia.

Remember you can subscribe to The Lancet Voice wherever you usually get your podcasts and we'll see you again next time.