33
Examination of witnesses
Preet Kaur Gill, Fiona Bride and Rob Thompson.
Q236 The Chair: Good morning. Welcome to the Science and Technology Select Committee. We are proceeding with our inquiry into innovation in the NHS, specifically concerning personalised medicines and AI.
We are very pleased to have as our witness this morning Preet Kaur Gill, who is the Parliamentary Under-Secretary of State, Department of Health and Social Care. With her are two of her colleagues: Fiona Bride, the interim director general for commercial and growth, Department of Health and Social Care and NHS England, and Rob Thompson, chief digital, data and technology officer, Department of Health and Social Care and NHS England. Thank you all for coming. We have quite a number of questions for you.
As you know, our inquiry has chosen to look at innovation in the NHS through the lens of two major healthcare technologies, personalised medicines and AI for healthcare. The 10-year health plan describes an ambition to establish a “genomics population health service” and to “make the NHS the most AI-enabled care system in the world”. The first question that we have is: what are the Government’s ambitions for these two technologies and what progress are you making to deliver on them? Minister, would you like to start?
Preet Kaur Gill: Yes. Thank you so much, Chair, and thank you for the opportunity to give evidence today. As you know, we are at a very important moment for the NHS, because innovation is not just about new technology, it is about how we improve outcomes for patients, how we support the workforce and how we build a health service that is able to prevent illness earlier and clearly deliver care more effectively. That is the premise on which I approach this.
When we talk about innovation, it is important to separate the two areas. While they are often discussed together—and I get why they are—they are not the same. You have personalised medicine and then you have artificial intelligence. Of course they are both built on the advances of science, of data and technology, but they answer different questions. Let me take those separately.
Personalised medicine is about getting the right treatment to the right person at the right time. We know that healthcare has often relied far too much on a one size fits all approach, but every patient is different. Our genetics, our biology and how we respond to treatment can vary significantly. The opportunity is there to use some of the advances in genomics and life sciences to better understand these differences so we can predict risk earlier, prevent avoidable harm and provide treatments that are far more effective for individual patients. That is why personalised medicine, as you said, Chair, is such an important part of the Government’s 10-year plan.
A good example of this—because it is always good to have a real-life example of what personalised care looks like—is pharmacogenomics, because we know that some babies, for example, carry a genetic variant that puts them at a much higher risk of permanent hearing loss if they receive certain antibiotics. From the work that the NIHR Manchester Biomedical Research Centre did, we now have a rapid test that can identify that risk before treatment is given. That is a powerful example of what personalised medicines means in practice.
We are also seeing huge opportunities in cancer pathways, and I am sure your committee has heard some of that evidence. The NHS Cancer Vaccine Launch Pad, which was established three years ago, is helping accelerate access to personalised cancer treatments with clinical trials. We are working with industry partners such as BioNTech, for example, and we want to provide up to 10,000 patients here in the United Kingdom with personalised cancer immunotherapies by 2030. Patients who are undergoing treatment at the moment can access clinical trials. It is very important that they have the opportunity to be part of trials, and of course it widens access across NHS trusts.
Artificial intelligence, as you said, is quite a different opportunity. I want to make it clear that AI is not about replacing clinicians, because far too often this is a concern, and a genuine concern that I take on board. It is about giving clinicians better tools, whether it is making decisions or reducing some of the unnecessary burden and freeing up some of their time so that they can focus on patients.
A good example of this at the moment—which I am very excited about—is ambient voice technology. This is supporting clinicians by helping with note-taking during appointments. That means they can have eye-to-eye contact; they can be actively listening with patients. At Manchester University NHS Foundation Trust, trials have suggested that we are saving around five minutes per patient. That might seem quite small, but if you think about the millions of appointments, it means that clinicians can spend more time on care, for example.
NHS England is supporting some of the safe scaling of these technologies through the front-line productivity programme, which is starting in outpatient settings. We are learning what works before we roll it out, which I think is very important. That is the approach we want. We do not just want pilots that never spread, but innovation that is properly evaluated and then it is adopted at scale. That is the approach we will take.
The Government and I want the NHS to be the most enabled AI health system in the world. That ambition has to come with responsibility. Finally, before I bring in my officials, let me say that the question is not just simply can we use AI, but how do we use it safely and effectively, in a way that patients and clinicians trust? That is very important. The key challenge is that AI can change over time. We see technology advancing far quicker than we can keep up to date. We recognise that a single assessment at one moment is not enough. That is why we need a lifecycle approach to how we adopt AI into the NHS. That means assessing it before deployment and it means monitoring how it performs in real NHS settings. We have examples: understanding what the benefits and potential risks are and ensuring that we are continuously evaluating and have clear accountability.
I know that this is the thing on many constituents’ minds when I speak to them about their data and about what happens to it, so the evidence threshold has to be proportionate. I recognise the need to enable innovation. That is why the work on the national AI road map, which is the regulatory framework—and it is Professor Alastair Denniston’s review—will be very important in helping us build a system where innovation and safety move much closer together.
Chair, the opportunity is very significant. Personalised medicine just allows us to move from treating the average patient to treating the individual and both are essential to the future of healthcare.
Q237 The Chair: Thank you very much. Before we move on, your 10-year health plan promises universal newborn genomic testing and population-wide polygenic risk scoring. In both these cases there is a scientific and medical debate about how useful those measures will be. How are you gathering evidence on this and how are you assessing the threshold at which the Government will begin to implement these measures?
Preet Kaur Gill: As I have already said in my remarks, it is about continuous evaluation, assessing what the opportunities and potential risks are before there is wider rollout and adoption. Clinical trials are an important part of the work that we are doing in making sure that we have access to innovative medicines. Equally, the genomics around prediction is very clear. There are quite a lot of safeguards. I do not know whether either of you want to say something on that.
Fiona Bride: Certainly the Genomic Medicine Service and how we bring that data together is a fundamental part, and an investment that this Government have made to ensure that we iterate carefully and with the right MHRA safeguards. I know you heard from Lawrence Tallon previously, but I think it is about stepwise iteration and that is how the research will allow us to progress.
What you talked about in terms of some of the pharmacogenomic aspects and how we start to risk score and assess is also a separate entity. That is the role of AI, which will help us to learn as we go and inform the decisions over the next 10 years.
The Chair: Do you anticipate polygenic risk scoring happening quite soon?
Fiona Bride: We need to be careful in terms of how we move, but as we see with most AI iterations, there is a learning process. As we embed some of those learnings, we will look at how we progress.
Q238 Lord Winston: Briefly, can you tell us where you think AI is already working well in the health service?
Preet Kaur Gill: There are loads of examples. I have been to the Royal Surrey County Hospital and Chelsea and Westminster Hospital and I have seen the FDP being utilised, which is the federated data platform. This is about helping clinicians see their data in real time, bringing it all together, and how GPs interact with consultants by referring someone on to a waiting list, for example, through the EPR system. The consultant can see that referral in real time, give advice back to the GP and say whether they should be on a waiting list or not, and then the system enables the clinician to input them on a theatre utilisation list.
Lord Winston: We get the message. Thank you very much.
Preet Kaur Gill: Also chest X-rays. I went to see how we support lung cancer screening, for example, and what the human eye may miss on an X-ray machine. It is an AI tool, an X-ray tool, which is a second pair of eyes for a clinician. There are lots of examples. I will bring Rob in to share some of those.
Lord Winston: Would it interest you to know that most radiologists are absolutely crying out for more access to AI? There are a whole range of radiological investigations, as there also are in pathology, for example, where AI is already being used remarkably significantly and reducing the number of mistakes and increasing diagnosis more effectively. What it does not do is screen. It is not good for screening, which might be worth bearing in mind. That is what they are saying to me, anyway. I think there is a real need to understand why this is so poorly supported in the health service, because currently it is.
Preet Kaur Gill: It is in half of the trusts at the moment. We want to have full rollout by the end of this Parliament so that everybody benefits, no matter where they are. I spoke to clinicians who are using it in terms of things that were missed by the human eye. You are right, it is an absolute game-changer.
Lord Winston: It is a bit more than that.
Preet Kaur Gill: It is more than that; I am simplifying it. Yes, of course the technology can do so much. That is why we need to see adoption. Of course, adoption varies across the country because different ecosystems and different IT abilities exist. The rollout on the chest X-ray AI should be by the end of this Parliament.
Q239 Lord Duncan of Springbank: I want to pick up on Ms Bride’s point about the risk register and the concept there. How is the risk register generated? What is the genesis of that? I can see a medicinal part of that. Is there also an economic part and a political part that you are alert to?
Fiona Bride: At the moment we are very much looking at this from a patient outcome perspective, from the health merit that it brings. In fact, in the obesity arena we are already starting to look at the productivity impact and collecting some data. There is a strong initiative up in Manchester already looking at collecting different data. As we progress, it is about learning as we go and not having it all written down. I am absolutely sure there will be wider benefits that we will bring in.
Lord Duncan of Springbank: I have no problem with the benefits and I have no problem with it being patient-outward, but we have already heard evidence that much of this technology and innovation will be quite costly. You will hit up against buffers very quickly, because the notion that you could do a national rollout of almost any individual part would be quite unlikely. Therefore, within the risk register you must also be looking at the economic constraints of what will be extraordinary innovations and extraordinarily powerful outcomes, but may hit up against there simply not being enough cash.
Fiona Bride: Like most innovations, what we need—and maybe it comes to the point of adoption and where we really focus—is to be clear about the priority demand signal, if you like, that the NHS must serve to solve. When we look at the modern service framework as a recent initiative, transforming the outcome in cardiovascular disease as a priority, what are the outcomes we want for patients in the next 10 years? What are the innovations that will transform those pathways? That is where I think we will bring all of this innovation to transform that patient pathway as a priority.
The Chair: We will be coming on to funding innovation in a few questions. Mr Thompson, do you want to come in?
Rob Thompson: Yes, I want to pick up on some of the AI points. It is a rapidly developing technology. I think you all know that; I think the public know that as well. Its application is very specific around the specific use that you want to apply it to. That is where some of the complexity is. You cannot just take AI and uniformly throw it at a system and then expect it to stick. You have to look specifically at what you are trying to do.
I will give you three different examples of work that we are doing right now, just to highlight the difference. I am going to come back to the ambient voice technology element. That is a fantastic piece of labour-saving technology there. My visits to acute and general district hospitals have been unequivocally supportive of its deployment. When I speak to clinical people in emergency departments and elective care, they are very adamant and can demonstrate that, as a labour-saving device, it is highly effective. That is very different from a speech-to-text old-style piece of technology. It is able to identify context, communicate and summarise in a very elegant way. It is demonstrably saving two to three minutes of time on a 10-minute consultation for elective and 30% to 50% saving of time for a clinical person in an ED department.
The most important point though is non-efficiency. It is the removal of the cognitive load on the clinical person. That is the thing that jumped out at me. Instead of them being sat there trying to listen to somebody while simultaneously trying to diagnose and understand, while simultaneously trying to figure out what it is they might type, while also simultaneously typing, they are able to give the person in front of them their whole attention. The patient has a better experience. They can enact their job in a far more effective and elegant way as well.
I think there is a qualitative beneficial aspect to AVT. I would say it is not a particularly sophisticated application of AI, but it is very productive and useful. That probably underlines where, in some regards, the health system is at the moment. We are in the foothills of using AI. We are exploring where we can use it most effectively. Understandably, one of the areas that we turn to most quickly is in and around productivity and efficiency.
The Chair: Thank you. The next question is coming from Lord Booth, who is coming in online.
Q240 Lord Booth: Good morning, Minister Gill, and your colleagues, and thank you for coming to give evidence to us this morning. My question is about the failure to scale pilots across the system. We continually hear in this inquiry that the NHS struggles to adopt proven innovations and that it can take between something like nine and 16 years of clinical research findings to enter routine care. In particular, we have heard about the so-called “valley of death” between piloting a new innovation or medicine and procurement of that innovation becoming routine. Often trusts can afford to do the pilots, but their budget for adopting innovation as routine is much more limited. This is not a new problem. What are the Government doing about it?
Preet Kaur Gill: Thank you for your question, Lord Booth. This is an important question and it is the one thing that I have been picking up, because we are brilliant at innovating. There is no shortage of our brilliant science, our researchers, our SMEs or start-ups having that innovation. However, it is the scaling of pilots and it is the adoption within the NHS where we are clearly struggling. What are we doing nationally? We are putting in place the infrastructure so that the NHS can identify, assess and adopt proven innovation faster.
We already understand the barriers. I am sure you understand them and I am not going to go through them. The life sciences sector plan is a year on and this is the year for delivery. It is an absolutely ambitious part of the industrial strategy. It is an area of growth, health and tech. It is certainly an area that the Government take very seriously.
Nobody wants to see a product just be in a pilot, because if it improves outcomes, if it reduces pressure on staff, if it saves money across the pathway, systems do have to recognise that value. I think there is an ad hoc approach in different parts of the country wherever you go. That is why the department has published the value-based procurement guidance. The whole purpose of that guidance is about moving away from what something costs, because we do get caught up in cost, as per recognising what value it creates for patients and the health system. That is the question that I want us to answer.
We are also creating much clearer innovation pathways. I have lots of medtech companies in my constituency, for example. We get this question from SMEs about, “We have this brilliant product, but just cannot get it into the NHS”. It is not quite streamlined in the way that medicines are, where the pathway is quite clear. We do not have that for healthtech. That is definitely something that we are going to address. We are creating the new rules-based innovation pathway for medtech, including the national healthtech access programme, which is a much clearer route. What is the aim? It identifies what works, it assesses it properly, making sure that it has gone through NICE and MHRA processes, and then we want to support adoption at scale.
It also means moving away from every trust having to repeat the same assessment of the same technology because, honestly, that has been a big issue. We do not want everybody doing that. I think trusts are going to be central to seeing this come to fruition. We have 15 health innovation networks across the country. They work locally, regionally and nationally. What do they do? They identify, test and spread information. That is their whole purpose. They help organisations understand what the problem is they are trying to solve and whether there is evidence that the technology works, because clinicians can be very sceptical, as you know. Also, how do we implement it safely and how do we share learning?
I have been to some amazing hospitals such as Royal Surrey and Chelsea and Westminster. Robotics is exciting, seeing prostate cancer surgery, but I want clinicians to share that best practice so other parts of the country benefit from clinicians saying that this technology is good. This improves pathways. We have to bring the investment into areas. There is a lot we are doing and that is why I am taking a bit longer to answer the question.
We then have the NHS Innovation Accelerator, which has already helped scale more than 100 evidence-based innovations across England. We already have that. Those innovations have reached over 10 million patients, with around 3,500 sites using NIA innovations, and about £243 million external funding has been raised to support that. That shows that we can scale and we now need to make that the norm. That is what I want to do. This is not the exception. This has to be standard NHS business that we are looking at.
There is another programme called NHS Insights. What does that do? It brings 18 member organisations, representing about 155,000 staff with over 35 million patients, to build skills and capability around innovation adoption. There is no point just having the adoption. We have to have the skillsets in those places so that they are able to adopt that innovation, but clearly be able to say whether it is working, or whether it needs changing or reviewing. That includes hospitals in places such as Birmingham, Bradford, Leeds, Essex, Newcastle and Milton Keynes.
A good case study I want to give you about AI reducing missed appointments is in Mid and South Essex NHS Foundation Trust. It has an AI solution to reduce missed appointments and improve waiting times. Deep Medical’s AI was integrated into patient communication systems and it analysed patient data to offer alternative bookings. That contributed to a nearly one-third reduction in missed appointments. NHS England is now planning further expansion across many more trusts.
That journey from evidence to adoption and scale is not something we just want in one hospital. We want that adoption and that scaling of pilots to be much more uniform across the NHS trusts. I will probably stop there in case you have any further follow-ups.
The Chair: I have a question, Minister. I think what you have correctly done is identified the problem. What we are interested in is how to solve it. Professor Andrew Beggs, who is an academic at Birmingham—you perhaps know him through your constituency—when he gave evidence to us said there was a local problem. He said that whenever you take an innovation into the NHS, the first answer is always “no”. What is your comment on that?
Preet Kaur Gill: I would say in week 7 of this role, absolutely right, and I am hearing this. I did the industry engagement yesterday. I have an industry round table with healthtech later on today. That is exactly what I am being told. The pathway is the solution that I have talked about and apologies if it was not clear. It is about helping companies get access into the NHS through adoption. Some of that is national and some of that will be local and regional, because the ecosystem that exists will be different in different parts of the country. Certainly on accelerating, once they have gone through NICE and MHRA processes, it will be easier to procure and adopt into the NHS.
I recognise the concerns that are being raised and we have to move at a much faster pace because nobody wants to see innovation not reaching patients. That is the whole point of all the innovation that we are doing. We have some remarkable work in this country. We just need to get that into trusts and adopted much more widely. I think that is why the clinician voice of sharing good practice really does help. Unfortunately there is a lot to do in this space. I do take that on board and rest assured we are doing that. I am happy to come back and update you on the work we are doing.
The Chair: Thank you. Lord Willis is coming in online.
Q241 Lord Willis of Knaresborough: Thank you very much and good morning to you. I was particularly pleased that you said England is brilliant at innovation, then you went on to say that it needs to be spread to all patients. My concern is: how do we ensure that all patients, wherever they are in England, can benefit from innovation? It comes back to the issue of funding this nationally. The 10-year health plan sets out a vision where a shift to prevention—particularly achieved through technologies such as genomics or AI—secures the sustainability of the healthcare system for the future.
At the trust level, innovation has to compete for funding and other types of transformation, and transformation budgets themselves are often redeployed to cover spending pressures. The vicious cycle is that trusts are often unable to invest in innovation that could make them even more efficient. How are we intending to get rid of that?
Preet Kaur Gill: That is a very important point. Let me be clear: there is the AI diagnostic fund. This is just one example of a funding pot that is available. I appreciate there are a number, because the life sciences sector plan sits across DSIT and DHSC. The fund has been deployed across imaging networks at the moment, covering about 66 acute and specialist trusts. Some of this is about supporting what I talked about with the innovation around faster lung cancer diagnosis. There are already around 3.2 million chest X-rays and about 80,000 CT scans that have been processed through using some of these tools. The evidence, as I have said earlier, shows that AI can help radiologists analyse scans in around four days, compared with six days previously, and the most complex cases, helping get patients get the answers much sooner around their diagnosis.
Because of that progress, we recently announced a further £20 million to expand the AI chest X-ray tool, so that by 2029 every NHS trust in England will absolutely have it adopted. That is taking what works, what is proven and then showing that we can roll it out and scale it up so that every patient, no matter where you are in the country, should absolutely benefit from that. We are also investing £8.1 million through NIHR, which is part of my department, the National Institute for Research. What we are doing there is testing six more AI and digital technologies across the NHS site.
This shows you what we are doing nationally. We are taking innovation. We are now doing this pilot on six further AI technologies and these technologies will support the NHS staff. What are we looking at at the moment? We are looking at analysed CT scans, ECGs and X-rays, and we are also looking at digital therapies. What we are doing is identifying the most urgent patients for prioritisation as part of this pilot. If this works, as we have done with the chest X-ray, it is about how we scale that and make sure that that adoption and innovation benefits patients across the country.
There is so much innovation and there are so many different areas. At the moment the focus is on heart failure, stroke, lung cancer, lung infections and TEC disorders, where we know that faster treatment in these areas saves lives. That is why we have to do this prioritisation. That is why we have to do this national rollout. I want to assure you there are funding commitments there, which takes that innovation, looks at the safety of it and the adoption of that much more broadly.
Q242 Lord Willis of Knaresborough: We have spoken to a number of people during this inquiry [Inaudible] our hospitals and within our practices are not there for significant numbers of people. Is there a decision by the Government to say everybody must have the basic technology, full stop?
Preet Kaur Gill: Yes, absolutely. Look, there has not been a premise of mandating trusts. Rob and I on the FDP, for example, will be writing to trusts telling them what three tools they must use in order to improve productivity and patient outcomes, supporting clinicians in terms of that administrative role—you are absolutely right—where we have proven technologies. Look, if we have to get into the mandated space, so be it. It is about showing trusts about the benefits. Sometimes we are not sharing best practice: the fact that the FTPs can pull into a national dashboard, so in real time we can see how many ambulances are sitting outside a hospital, what bed capacity is available, what the theatre utilisation is, how we are reducing waiting lists so that we prioritise people, not on how long they have been on it, but around their medical needs and prioritisation. I absolutely take your point. I am very much of the view that we absolutely have to do this at pace, but be very clear what works and what we want adopted across the country.
Q243 Lord Winston: Briefly, you mentioned robotic surgery. That is why I am taking it up with you. There has been huge emphasis on it for 30 years now. That is quite a long time. Can you say whether, in your experience, robotic surgery in randomised controlled trials is significantly better than just non-invasive surgery, which uses laparoscopy or other endoscopic techniques? Robotic surgery is very expensive, but maybe it is not being taken up for other reasons.
Preet Kaur Gill: This is about trust leadership as well. The Royal Surrey has five robots. It has consciously decided to make an investment into that. What I found quite remarkable is that you are right, it is expensive technology, but if we think about the patient outcomes, it means that the patient can be discharged on the same day. They will not occupy a bed for four or five days, whatever it is. The best thing about it was that the prostate, when it was removed, could there and then be analysed to check whether all the cancer cells around it had been removed, so the patient does not have to be brought back and then have a further operation because something has been left and it has not all been cleared.
Lord Winston: That could be done with non-invasive surgery.
Preet Kaur Gill: If we look at it from a patient perspective, the patient will have to have two operations, as opposed to having one. The patient will be able to go home on that day. That means we are saving a bed for about five days. If we start thinking about it from the patient perspective, the patient outcomes are absolutely clear. The more we start doing something, investing into something and bringing the cost down and it becomes quite normal practice, that is where we need to get to. It is showing trusts that investment in robotics, how many more surgeries can you do as a result of that, versus having to wait for surgeons. We know that cancellations happen all the time for various reasons. The idea that you can know in real time whether that surgery has been successful, so you are not having to bring the patient back again to have a further invasive procedure, is so important.
Lord Winston: We talk about non-invasive, which means using the same instruments but having the advantage of the surgeon not being separated from the patient, but actually connected to the patient. That makes a very big difference. That shortens the time rather than lengthens it, on the whole.
I have watched robotic surgery in this country, in America, in Calgary, New York and many other places as well, because I am interested in that area. On the whole, it is extremely complex sometimes. I am still convinced that there are good trials that show that it is really justified, though of course when I look at the literature that comes from the Department of Health, I see that there is a huge focus on robotic surgery, and I wondered what the scientific basis for that was.
Preet Kaur Gill: I accept your point, but I think that this is where we need clinicians talking to other surgeons across the country as to what they see as the benefits of that. There is not much of that happening. It cannot just be that the Royal Surrey is where everybody wants to go for certain procedures because it has access to this robotic surgery. It is proven, by the way.
Lord Winston: In private practice.
Preet Kaur Gill: In private practice, exactly. We have to look at what the investment is, equally with the CT scans that are AI enabled. That is why we want to roll out the chest X-ray. I think it is where innovation is showing it works, where it can improve productivity, it can have good patient outcomes and it can support clinicians in terms of more capacity for theatres, for example. It is the evidence base that we need to see as to why we should put the investment into purchase, which is very expensive. I hear what you are saying, for non-invasive it can be complex.
When I went to the Royal Surrey, all the clinicians were in the room and absolutely checking with the patient. They are on site, of course they are managing the robotics, but very cognisant of what is happening for that patient, so that the clinician is not separated. They are pretty much directing, which I know you know. I do not know if you want to add anything more.
The Chair: Minister, I am just worried because we have a great deal more questions. Can we move on?
Preet Kaur Gill: Sure.
The Chair: Lord Drayson wants to come in.
Q244 Lord Drayson: Ms Bride, as chief commercial officer for the NHS with responsibility for £30 billion spent, we have heard from witnesses—all previously working in the pharmaceutical industry—that it is well established that the NHS has a reputation for implementing schemes that are subscale, which give good announcements to be made by politicians. The Minister has mentioned allocating £43 million for an accelerator and £8 million for a pilot. These are tiny amounts of money. How do you square that prioritisation balance? You are making the decisions about where to spend this money. How do you defend that?
Fiona Bride: This is a recognised challenge and we embrace it fully. I think it is widely recognised that we have not leveraged the value that the NHS can command as a national procurement approach. What that does not mean is that we buy it all centrally. Clearly the local decision-maker holds much of the actual accountability for the finance. We do this in medicines already and I think we can learn as to how we approach this towards healthtech and medtech in particular. Interestingly, robotics is one of the areas we will leverage this in. You can imagine if you are at trust level procurement buying one robot, the ability to negotiate the best value is very difficult because you are buying one. If we can start to leverage an at-scale value procurement—as we do in medicines already, we set one national price—to enable the choice to be local still, the decision for the clinician to make, we recognise that we can then do much more of a value-based procurement focused more on outcomes and leveraging the at-scale procurement that the NHS should and can realise. What that means is we have frameworks where we set a more national approach to that. That is work in progress.
We are iterating. We have currently started on some of the less high-tech innovations to look at how we help choice as well, so that we focus on what best value looks like, not just about price, but the outcome it generates and how we can embrace that at scale and pace.
Q245 Baroness Willis of Summertown: I want to move on, but I think some of these questions have already been slightly touched on. We have heard a lot about the pilotitis, where you have these trusts doing these excellent pilot projects and the innovations being adopted by one trust. You then move to another trust and you have to start all over again. It is interesting that you are picking up on procurement because that seems to be one of the big blockages. That is not quite the right word, but you know what I mean, you begin all over again. Of course, if you are that life sciences innovator, you give up in the end, and you have to keep going round in the same circle and with the variations that each trust wants something slightly different.
What are the Government doing to ensure that innovations that have been adopted by one trust can then be rolled out across other trusts without having to go back through this loop again?
Preet Kaur Gill: Through the life sciences sector plan we have set this out in terms of what our plan is, including our 10-year health plan, so we have a number of things that will address this. The national healthtech access programme is supposed to identify the well-evidenced technologies that meet some of the NHS priorities, as set out in the 10-year plan. What we are supposed to be doing is support them through the assessment, through the NICE appraisal, and then get to adoption.
It is taking on board your point that if something works, all patients have to benefit everywhere. You are right, it cannot be that just one trust is procuring one thing, we know that it works and it saves so much time, and then another trust has not. That is why the national procurement on some things is quite critical, as is the local stuff. I will bring in Rob to answer that further.
Rob Thompson: I will bring it back to an AI context, because it is easier to see the change. The technology is moving so fast that it is forcing us to move fast as well, or certainly a lot faster than we used to do, by all accounts. We have mentioned the national AI road map already. It is about a structure. It is about putting structures in place that allow innovation to thrive and then propagate, as appropriate. Under the national AI road map, there are three types of work. The first type of work is work that my team and I do, which are AI enablers that we propagate into the system. As the Minister said earlier, we can still be more assertive about the adoption of it, so it is not optional, but getting close to mandation for some things where we believe there is a lot of value. There is a second—
Baroness Willis of Summertown: Can I just stop you there? Who mandates it? Who is the person responsible at the top to say, “We are going to adopt this”? What we are hearing from evidence is you will have a clinician or a team of clinicians pushing back saying, “We do not want that here” or even, “We do not have the bandwidth to take this on board”. There is a—
Rob Thompson: The structure for that would be that I would probably make a recommendation to the Minister and the Secretary of State and they would say, “Yes, go ahead” or, “No”.
Fiona Bride: Just to come in there, Rob, you have already heard about the national healthtech access programme. As we see in medicines, when NICE determines the product to be cost effective, there is a funding mandate that attributes to that. The same will be true—and I think this is a major change—for any of the healthtech solutions reviewed through NHAP will come with a funding mandate.
Baroness Willis of Summertown: They will all behave like NICE and—
Fiona Bride: Not everything. Some of the AI may not be appropriate to go through that. We need to look at a route B. We have already started that work, as we see in medicines, where we look at national clinical guidance to inform. I think there are two other parameters here. One is national clinical leadership. Dr Frankie Swords, the national medical director, joined us recently. She has joined the innovation board to ensure that that clinical leadership as a national voice is also heard.
Also, the innovation passporting speaks exactly to your point. What do we do once well to assure the system, but also how do we simplify what the system has to do to assure themselves that it can be legally rolled out? There are two parameters that need to be done, but simplifying that: the poor supplier runs around 200 different people with different requirements at the moment. We have to eliminate that and I think there is a real commitment to that.
Preet Kaur Gill: Can I just assure you, Ministers and the Secretary of State are clear, because this has to come from the top of the leadership. While we want that flexibility around what is adopted locally, recognising that, absolutely, the national leadership has to identify the innovation that works. That is in some of the examples that I have given you. I am equally—as I said, in week 7—very much of the view that there are some things that need to be adopted nationally. The work that we are doing is looking at that and engaging with trusts.
Rob Thompson: Can I go through the other two categories? The next one is where something is locally innovated and then we look at it and we go, “That has national worth”. Then we move into a support role where we probably take the technology, work with the local area and then look to propagate it.
The announcement at the weekend around AI triage in the NHS app is a perfect example of that. That was a piece of technology procured by a GP; I think it was a group of GPs in the Sussex area, covering 23,000 patients. We supported them in running a trial to see whether that AI triage was effective for them. It turns out that as a result of introducing it, the amount of people in the queue on a Monday, Tuesday and Wednesday morning trying to get a GP appointment dropped by 30% because the triage activity was doing the job necessary to direct them to the right care pathway, whether that was self-care, go to a pharmacy, have a GP appointment or go to A&E. We are now working with them about taking that technology, embedding it in the NHS app—which is used by 41 million people—and then working that through for national propagation.
The next step is to do that in London with about 200,000 people. We will get a larger base of evidence suggesting the extent of the benefits and then move into a national rollout, which we anticipate in April 2028. That is a two-year cycle to go from a local innovation, where it was done discretely at grass-roots level, to a national propagation. That is significantly quicker than the nine to 16 years that I mentioned earlier in the discussion. That is the second category.
The third category is where we are asking local areas to innovate and helping them with the structures and technology to be able to do that, getting the baseline technology in place to support them. The AVT announcement, which was also part of the weekend, is an excellent example of that. I am giving them guidelines saying, “Use AVTs integrated into your electronic patient record systems” because that is the best way of getting the data in, while giving them latitude to select who they want to trade with, who they want to buy from and implement it locally.
We have three different structures that we are bringing to life now. I think that solves some of the issues that Lord Willis mentioned earlier about how we avoid a future where people just do pilots and then nothing happens. It is because we are being assertive about innovating and because we are putting structures in place to help it.
Baroness Willis of Summertown: I want to move on to the next question, but perhaps I can just come back. Your third example fills me with dread, because to me it sounds as though you are going to have everybody trying their own innovations. We have heard time and again that the areas where you get local innovation coming through are the areas where you have a university or people who are much more research active, and the areas that are often in the lower socioeconomic income bracket are the areas where you do not get that local innovation coming through. It does fill me with some concern.
Rob Thompson: In the detail of the AVT deployment, what we are putting out as guidance is that if you are capable of doing this yourself—Manchester is a good example, where they are very capable of doing it, as are many of the advanced trusts—that is fine, but we will provide financial support, expertise and change and transformation support for those areas that are not able to do that right now. There is additional financial support to help that.
The Chair: Do you have more questions?
Baroness Willis of Summertown: I have the next two.
The Chair: Please be as brief as you can.
Q246 Baroness Willis of Summertown: I will be very quick. I suppose it is easy to keep saying what the problems are, but we do have examples where things are working well, one of them being the Genomic Medicine Service, where you have that national commissioning model delivered regionally through hubs. That is top-down. Will the Government consider extending that kind of approach? Is that in your plan?
Preet Kaur Gill: Last month the NHS Genomic Medicine Service reached the milestone of 200,000 whole-genome sequencings since the service was launched. I think something like 52,000 whole-genome equivalents were sent for sequencing last year and this year. This is personalised medicine in practice. It is building on the evidence base of the research of what we are able to achieve and being able to roll that out more broadly.
Fiona Bride: Perhaps just to add that where we look at personalised medicines, often rolled out through specialised commissioning routes with central funding, we see the uptake of that innovation transformed in terms of consistency and equitable patient uptake. It is not appropriate that we have all medicines funding in one central place and clearly local delegation of that fund is important. However, I think that is the opportunity as we move towards more prioritised demand signalling—how do we give greater clarity?
The single national formulary is a significant step forward. The last analysis we did found 51 different ways of categorising whether a medicine that NICE had endorsed could or should be used in a local situation. The single national formulary will stop that variation and enable one clear sense of formulary approach. There is a step towards, while not having everything centrally.
Q247 Baroness Willis of Summertown: I have one more question, which is about innovation metrics. We have heard a lot about what is going to happen, but there is nothing like a target to make people focus on what they are trying to do. Is there an idea of embedding KPIs for innovation within the National Health Service?
Rob Thompson: Yes, but they are not specific at the moment, if that makes sense. What we are typically doing is trying to draw a line back from the innovation activity to performance against statutory measures—the 18-week RTT figure or the four-week ED figure—and also trying to codify performance benefits across the system. There is nothing in the plan currently to say we are going to specifically introduce KPIs A, B, C and D for innovation. I think that is largely down to where we are in our own development of the national AI road map, the strategy and plans for it, but I do concur that it is a good thing to do.
The Chair: Lord Drayson, I think we have covered the questions that you were thinking about asking in terms of who in the NHS takes responsibility. I think those questions have already been asked. Is there anything you want to add?
Q248 Lord Drayson: Mr Thompson, you basically made it clear that in the digital arena, it is you who are responsible for signalling to, let us call it, the innovation community unmet needs within the NHS. Is that correct?
Rob Thompson: Yes, I am accountable for digital, data and technology for the system.
Lord Drayson: Could you quickly give us three current unmet needs within the NHS that technology in the digital realm could potentially make a difference to for which you do not see a solution at present?
Rob Thompson: The three that are active at the moment: first, there is an unmet need specifically to be able to see more patients more effectively, which is fundamentally the AVT element of it. Second, there is the example the Minister gave earlier about the introduction of AI to support radiologists in making assessments about X-rays and CT scans, which I think was rolled out to 66 of the acute trusts and we are now looking to roll out further. That is a decision support tool, using AI to identify issues within the scans. The third is a different kind of need, mentioned earlier in the FDP discussion, which is using AI and machine learning to help in the optimisation of the running of the system.
Those are primarily focused at the moment on what I would call operational performance or decision support. I think this is a very important point. My observation, having come into the health system fairly recently from other departments, is that we seem to be worried that we are taking decision-making away from clinical people, but I have seen no evidence of that in the use of AI and digital. If anything, we are trying to use that technology as a support tool and a decision assistance tool rather than a decision-making tool in its own right. I think that is the right side of the line to be on.
Q249 Lord Drayson: Final question, Minister: the closure of NHS England and its integration into the department, how is that going to affect this process?
Preet Kaur Gill: It should not affect this process, because this is a clear commitment in the 10-year health plan; the life sciences sector plan rollout also encompasses this. As part of the Government’s industrial strategy, life sciences is an important part of that. All the work we are doing is about not stopping and starting or changing course, because you cannot do that with big transformation projects. The work of NHS England and what is going to happen is ongoing, and that is for the Permanent Secretary’s piece of work that she is undertaking.
I can assure you, given that this is in my brief, that I want us to accelerate and go at pace, because I hear your concerns around innovation and adoption not getting into the NHS. We are in a unique position here in the United Kingdom in that we have the NHS, where innovation should be adopted. I have learned in the brief about the phenomenal work in science and research that we are doing and why the UK is the best place to come and not just discover, but to scale, adopt and manufacture the whole of it, whether that is medicines or medtech or healthtech.
I do accept that there are areas where there will be unmet needs, and in the end we have to look at the areas where mortality is highest and prioritise, whether that is heart disease or other such areas. It is about that whole journey: how do you support clinicians to do their jobs much better; how do we increase productivity? The reason the FDP is important is that people have sat on a waiting list that is stuck in a drawer, and what AI can help you do now is prioritise the patient whose symptoms or clear concerns are serious, so that they know that their referral will get bumped up because they need treatment right now. I can assure you that I am excited about this brief and I want us to be rolling out much more innovation across the country.
Q250 Lord Ranger of Northwood: Good morning, Minister, and to your colleagues. Thank you for your time today. A few questions about the Health Data Research Service, into which the Government have announced £600 million of investment and which is due to be operational this year. Many people are asking the service to do many things already. What will be your key priorities for the HDRS? What do you see it achieving over the coming years, considering the importance of data and research?
Preet Kaur Gill: If we want to lead on personalised medicine, if we want to lead on AI, then access to high-quality data is essential. You simply cannot do that without it. HDRS is going to help researchers generate the evidence they need to understand whether new treatments work, whether technologies are safe and how we can get proven innovation to patients much more quickly. HDRS will build on our existing strengths: the genomic data we have, the diagnostic data, the lab data and the many other health research assets. That will primarily create the foundation. The HDRS is now incorporated, so this is going to be an integral year in terms of getting things moving.
I also have to consider public trust and the safeguards around HDRS in my role. Yes, this is an AI national body of data, and rightfully many of you and many of our constituents will want to know how that data will be used and whether it will be safe. Public trust is something that is on my mind quite a lot. Inevitably, patients do want their data to be used to improve outcomes and they want it for everybody. When we explain to people how the data are used and what the safeguards are, the British public are pretty much on our side because they can see what it means for future outcomes. Baroness Blackwood is the chair of the HDRS—
The Chair: Which has come to our committee already.
Preet Kaur Gill: This is about bringing our world-leading science. I want to see companies come here and invest, not just to take our innovation and discovery and manufacture elsewhere, but to do everything here. HDRS is a game-changer for the UK being the place to come and invest. I was at a biotech conference in San Diego and so many companies are looking at what this will mean. Let me be clear: it is not just about having more data. It is about how we use the data safely to deliver better care. I will stop there.
Q251 Lord Ranger of Northwood: Thank you for your point on public trust. I think one of my colleagues will come back to you on that. On the self-funding decision that has been made after the initial budget is invested, how did the Government come to that decision to enable HDRS to become self-funded? If that does happen, how will that be balanced with the wider economic benefits, but also the potential barriers that that economic model could present to people becoming involved in terms of access fees and charges to that data?
Preet Kaur Gill: This is an important point. The conversation I have had is that there has to be a commercial model. I know that so many companies around the world will pay for this, and so they should, because this is money that goes back into our health service. I have also raised the issue of making sure that SMEs can access it. It is not just about universities that know how to understand the data or the information it provides. We have to democratise the data. That is important in order for it to be available not just to the big companies. It is about the SMEs and making sure that they equally have access to it.
I am sure you will agree that with the investment that the Government are putting in, seldom do we think about the commercial incentives of this. I know people want their data to be used for good, but this is taxpayers’ money and we are helping innovation and positioning the United Kingdom. There has to be a commercial model. We do not yet know what that looks like—this is the conversation I am certainly having—but I am very much of the view that we need to have something that is fair and that equally recognises who is going to be accessing and using it.
Q252 Lord Ranger of Northwood: Would there be any consideration about sovereign businesses—UK-based businesses, especially SMEs—where we know there is a challenge in the UK? Could that be an approach to that model? Has there been some consideration of that?
Preet Kaur Gill: Yes, that is an important point. Let me be clear: we want British businesses to benefit from this data and help them, whether that is in research, scaling, adoption or clinical trials. It does not matter where they are in the service. It is the whole industry, those that need access to it. British businesses will definitely be integral to this.
The Chair: Thank you. You have mentioned public trust. Lord Berkeley has a question.
Q253 Lord Berkeley: Minister, you mentioned quite a lot about public trust in your last answer. Occasionally there is a breakdown of public trust. What are you doing to maintain the public’s trust in how health data will be used with HDRS?
Preet Kaur Gill: We are still working on the detail. Cyber security will be on many people’s minds: what if there were a breach; who will have access; how will it be used; will it be anonymised data; do people have to opt in? All these questions will form part of how we engage with everybody, GPs, trade unions, the whole clinician workforce. We have to do better communication on this, because we want the British public to be proud of the wonderful work that we are doing. Quite often many people simply do not know what we are doing and we have a wonderful story to tell. We must tell it.
Equally, we must have the safeguards and we must listen to people’s concerns when they raise legitimate questions around the safety and trust of their data. To start those communications now before we start delivering things and bring the British public onside, that is integral. I am minded to get on with doing that and not waiting, because there is a conversation to be had with the country on something that is so innovative and unique that positions us as the place to come and invest and bring your innovation to.
Q254 Baroness Jones of Whitchurch: A number of the people we have heard evidence from have said that there is a problem with accessing data because of the legal framework that currently exists. During Covid we managed to overcome that. The Secretary of State intervened and people had access to the data very quickly, but we are not in that position now. Do you see HDRS as having a responsibility as the data controller for that evidence going forward? One of the problems we have had is access to GP data, the 6,000 GPs. Will the HDRS be able to provide some cover for them in terms of bringing all that data together to make sure that there is a proper legal framework and that GPs do not have responsibility for making those day-to-day decisions about how to share their data? What kind of framework are you imagining for that?
Preet Kaur Gill: There is a lot of work that needs to be done. All of the different data systems will integrate with everything else we are doing in the data and technology brief. The electronic patient record, the single patient record, the HDRS and NHS Online will all have to have an integrated system. On the electronic patient record, the data controller is still the GP. They will have to give permission to share that data as part of the single patient record so it can be pulled into the NHS app and be part of NHS Online. We have to make sure that the legal elements do not become barriers and make it is very clear who the data controller is and how that data, when it is shared, is going to be used.
That is the work we are doing now, because we have not yet built the single patient record. We are working on the architecture of that alongside the NHS app. All that work is just starting, and that is why the communications piece of getting the country engaged with this is so important. There are a lot of elements that all interact but are likely to be coming out at different points. That is why I think we have a big piece to do. Let me bring in Rob, because that is his team’s work.
The Chair: Before Rob comes in, can I be clear? Following up from Baroness Jones’s question, is it your intention that HDRS will be the data controller for GP data?
Rob Thompson: May I answer that question?
Preet Kaur Gill: Yes.
Rob Thompson: We have to work out who is the data controller for what and you have to get very specific. If HDRS’s role is predominantly to provide or work on large datasets that are anonymised for longitudinal, multimodal studies, they would be the data controller for that dataset, which is probably anonymised. When it comes to the specifics around the GP and the patient record, that will be handled as part of the single patient record programme, which is under my portfolio. GPs will still remain data controllers for GP records.
When we pull in that data as part of the single patient record, we will also become a data controller at that point, and I suspect there will be a technical mechanism that transitions that data for secondary purposes to HDRS. The nature of that data and how it is stored will determine whether it is anonymised or not. My sense is that, architecturally and technically, it will be anonymised data that they are working on, not the actual patient record itself.
The Chair: Is that a problem? We have heard several witnesses say that GPs controlling the data is a major barrier. Do you agree with that?
Rob Thompson: GPs are currently the controllers for their patient record. Some of them, as I understand it, are probably uncertain about what that means in terms of the single patient record and HDRS. I have to go out and resolve that problem, frankly, which means making clear what the information governance, the data controllership and the data processor relationships are, and then the actual architecture for moving the data around the whole system. That is one of my key things to fix.
Preet Kaur Gill: Chair, you raise important points, because not all GPs are in the same space. Some will want to share and give permission, retaining data controller status, but sharing as part of the single patient record. That is why the communications piece is so important, so we can address any concerns that GPs may have about how their data are amalgamated into the single patient record.
The Chair: You would not go as far as saying that you would mandate that GPs should not have control of the data?
Rob Thompson: I think it is in the Bill. The secondary uses are covered in the Health Bill under the single patient record, so if the Bill goes through with the secondary uses provisions accepted, then by definition that would be in legislation at that point.
Q255 Baroness Jones of Whitchurch: You have already said, Minister, that there should be some sovereign AI capability. Can you explain the scale of that? There is a huge dilemma about how much we harness the great advantages of the data we have in the UK and use it to our own advantage versus the extent to which we share it with the world. There is a significant ethical question here. Can you tell us how the Government are going about solving that? If we did agree that there should be a significant sovereign AI capability, what would be the governance of that? How is that going to be controlled?
Preet Kaur Gill: I will bring in my official on that. These are important points because, for example, many universities will know very well how to use the data versus others, and that is where the sovereign part becomes a challenge. That is why democratising the data—making it usable and understanding what researchers or companies or SMEs need from the system—is essential. We are making sure that the data are usable, because at the moment you would have to be an absolute expert in terms of being able to download that data and knowing how to use that.
This is where the communication with industry, with SMEs and with universities is integral. How do we make sure the data is focused on the things that we need it to be focused on; how do we make sure that SMEs have access? That work with the industry is important, understanding some of the barriers they have faced and the concerns that they have. This is the work that is starting. This is the conversation I am having with Dr Ivarsson. It has taken a while for HDRS to get incorporated, but this is a very important point.
The Chair: Minister, we have more questions. Can we move on to the next question? Rob, do you have something very brief to add?
Rob Thompson: It is a very technically complex area. Data ownership and data control sit with the NHS. Hosting of technical infrastructure—which is what I think people think of when they say “sovereign”—is generally with the cloud providers, which are US-based at the moment because they are providing the best cloud services in the world. The AI models can come from anywhere. We could have sovereign AI models used in the system or AI models from different states or different companies. They are all different parts of the structure and we have to think carefully about how we build it.
The Chair: Let us move on to regulators and regulation.
Q256 Lord Stern of Brentford: Thank you again for coming. The MHRA and NICE are obviously under a lot of pressure around resources, particularly in the context of the areas we have been discussing, which are moving so quickly with personalised medicine and AI. Lawrence Tallon, the CEO of the MHRA, came to us last week and had a range of interesting recommendations around funding, including, for example, charging a bit more for some timely, quick responses. He also raised issues—which you mentioned, Ms Bride—around other regulators being involved in AI because it is device-like rather than drug-like. Will you be engaging with those ideas? I am sure you must have heard of them.
Preet Kaur Gill: I joined Lawrence Tallon in San Diego at the biotech conference. The regulator is moving at pace and recognising some of the challenges. The fact that the FDA will now have an office in London and the MHRA will have an office in Washington, recognising FDA-approved products, is a real game-changer. It is recognising where we can streamline and work together much more effectively. Let me bring in Fiona.
Fiona Bride: I think this is fundamental. We have to speed up and clearly Lawrence is championing that. It is also about bringing in NICE as the health economic evaluator and NHS England. What we are seeing much more closely working is that integrated approach, particularly in medicines, but increasingly in the healthtech space, looking at horizon scanning to determine what innovations are going to be transformative and where we want to lean in, go faster and look at a different, more risk-share, gain-share type innovative procurement, as we do in medicines, where we may go into a managed access forum or look at collecting more evidence to further inform what then needs to happen.
The same can be true in some of the healthtech space as well. It links back to how we need to evolve how AI starts to inform what is possible. What Lawrence is clearly doing well is championing the need to evolve. We cannot have a process that was fit for purpose 25 years ago and is no longer fit for purpose now. That is the evolution we are committed to seeing.
Rob Thompson: Lawrence and I are highly aligned on the AI side of things. It is about being pragmatic. For example, we have made some decisions around the ambient voice technology, which we both recognise is an administrative tool and therefore would not warrant severe scrutiny from the MHRA. Conversely, we both recognise that when it comes to anything that is doing diagnostic or triage-related work, there does have to be some oversight, but the point-in-time approval is not at all appropriate for the pace of innovation. We are both aligned on a process of what we call continuous assurance and evaluation.
Q257 Lord Stern of Brentford: He also emphasised reflecting that need for speed in the charging process itself. My next question is about NICE.
You will have to excuse me, I am a professor of economics and I am being slightly nerdy here, but it matters. It is a big point. NICE works mostly with QALYs, quality-adjusted life years, which are directly focused on the individual gains from treatment. Of course, we all know that there are economy-wide gains from treatment. Mental health is a big example, but there are many more. It seems very odd—and I have to say I used to be in charge of the Green Book when I was head of the Government Economic Service 20 years ago—in an assessment of the value of treatments to leave out those wider benefits.
We were told by the CEO of NICE that that was a matter of government policy. You must have a chief economist or somebody in the department who would be quite capable of working out those higher benefits. As I say, I was in charge of all the chief economists and my encouragement then would have been, “Get on with it. Do not wait for anybody to give you permission to do that” because it is so substantial. What do you think?
Fiona Bride: This is a recognised challenge. We are only looking at the value in isolation of QALYs, as you quite rightly say. You will have noted last week—certainly in the medicine space with the conclusion of the sprints that came out, further to the collaborative approach with industry—that NICE has committed to looking at some of those productivity gains, piloting some of that and also relooking at all of those aspects. We recognise that we need to think about the wider societal benefits, whether it is workforce productivity or maybe carer benefit. We need to make sure the mechanism is correct. Clearly the value and the pricing associated have a direct correlation there.
There is a question of whether we have the methodology right for how we apply it to healthtech. Is it the same methodology? I know that Jonathan Benger has already started to think about what needs to be true for the NHAP process and whether it should follow medicines as closely as it has done to date.
In terms of the Green Book, I will not specifically comment on whether that is right or wrong. I know people are looking at it and clearly that will be something that is ongoing.
Lord Stern of Brentford: If this is a major change in the way that NICE evaluates, it could make a very big difference to not only overall resources, but allocation of resources. If I understand it, you are saying that you are getting on with that at pace and you are encouraging that change to take place.
Fiona Bride: Yes, and the commitment was made last week with the conclusion of the sprint process further to the US‑UK medicines agreement.
Q258 Lord Berkeley: We have heard from some witnesses that the regulatory approval of an innovation does not lead reliably to NHS deployment. Is there a gap between MHRA or NICE approval and the actual commissioning and procurement and what are the Government doing to close that gap? It is of concern to many people.
Preet Kaur Gill: It is an important point and it comes back to the point of the adoption per se across a number of issues in the NHS. The MHRA has done some very good work. I have to say it is a regulator that lots of people around the world respect and Lawrence Tallon has done a lot of good work. The MHRA has the AI airlock, which is helping especially to support confidence in terms of innovation. What does that do? It allows innovative AI products to be tested in controlled environments and brings together the regulator, clinicians, industry and technical experts so that healthcare settings can absolutely have confidence that the AI performs safely in real-world settings. That is the concern of a lot of the adoption barriers.
The MHRA has also established the National Commission into the Regulation of AI in Healthcare. Again, this is about bringing together NHS organisations and technical experts to feed in on what the future regulatory framework could look like. Those are the conversations we need to have because we know, as Rob said, that this is a fast-moving area and regulation has to keep pace while we protect patients. In the end it is about patient safety too. I take on board the point that you make, because adoption is the bit that we absolutely have to get right.
Lord Berkeley: Is there a timescale on any of this?
Preet Kaur Gill: I definitely want to put some timelines on it. There are lots of pieces of the jigsaw that all fit together, especially in the datatech brief on AI adoption and innovation. It comes back to the point about how we make sure that what we are scaling up works effectively, shows the gains that we want to make and then we do the rollout. Certainly that is the kind of shift we are taking.
The Chair: Minister, we are conscious that you have limited time. We have a few more questions we would like to ask. Are you able to stay for five or 10 minutes longer than the schedule? Thank you very much.
Q259 Lord Verjee: Thank you, Minister. The Government have set up a National Commission into the Regulation of AI in Healthcare. When do you expect to have a proper framework for regulating the use of AI in healthcare? Will this involve moving regulation of AI tools into MHRA and away from the current system? How do you deal with the challenge of such a fast-moving pace of innovation and regulating that?
Rob Thompson: I will pick that up. Lawrence has commissioned this work himself, which I think is a good thing, and it is needed. I understand that the review and the report are likely to come out in September—that is the last time I had heard—when he will formalise what they have found. They have been good in outlining what they are finding on an interim basis and there are three things that come from there.
The first is that they are not only considering the risk of introducing AI, but are also looking at the risk of not introducing it and the negative impact that has on patients. That is a profound shift, because ultimately it is looking at the counterfactual and asking, “If we do not do this, how much harm potentially continues?” That is a very bold and ambitious thing for Lawrence to do and it is badly needed, given the pace of change in this arena.
I am in agreement with Lawrence and the interim findings that we need to move away from the point-in-time approval and accreditation process and find a way of continuously improving AIs that are deployed, recognising that they will in themselves evolve and get better. I think Lawrence’s current view of this is to frame the nature of how much an AI can evolve without it having a full review. We will have to learn by doing on that because that will be an experiential thing.
The Chair: Are there any further points or can we move on?
Rob Thompson: The third point is that regulation should be context-specific for the AI. In summary: review in September, the recommendations will come out and we will probably build a plan from there.
Q260 Lord Winston: We have heard repeatedly in this committee about the time that clinicians are having to devote to things, which prevents them from deploying innovations in the National Health Service. Would you care to comment?
We have also heard that one of the most important ways of ensuring better medicine generally is to have more academic clinicians. Let me define what I mean by an academic clinician. That is somebody who is on a career path to a consultant appointment, generally in the NHS, or to an equivalent scientific appointment, who has training in both science and medicine—a PhD, for example—and is also a member of one of the Royal Colleges. There is a massive shortage of clinical academics and certainly I think that is a major problem. They spend at least five years in post and they are not being properly funded. I must say NIHR has done a very good job in trying to fund them, but on the whole there is completely inadequate support, as is shown by the Academy of Medical Sciences.
What do you think we could do to ensure that that is increased in the NHS? They have a massive effect on departments. They also do important clinical work at the highest level.
Preet Kaur Gill: This is an important point because clinical academics, as you say, are vital to UK health research. I am very privileged to represent a constituency that has the largest hospital trust in the country. It sits next to the University of Birmingham. You then have the Health Innovation Campus, which is a training and research hospital trust. However, that is not replicated everywhere and we do not have the benefits of that. The National Institute for Health and Care Research, as you have rightly recommended, is the largest funder of clinical academic training. That is important. It invests over £230 million every year in research training programmes.
To your point, it is about how we retain and support those clinicians who have both of those different skill sets. There will be a new research workforce update because there is a recognition that across the country, the landscape varies. There is no fixed timeline for publishing that, but I think in the coming months—in a couple of months—there is a view to bring that in.
The NIHR also has an academy. There are lots of different things that the NIHR is doing. It is responsible for the development and the co-ordination of some of the academic training that will be available, but it also does research capacity and career development. There is a lot of work that is being done in this area. There is also money. In the personalised medicine space there is about £17.9 million to do this piece of research.
Lord Winston: I am grateful to you for giving us so much time, but I want to concentrate on the specific issue of clinical academics because I think they are the leaders in this area and in so many major advances in the health service. For example, care of neonates was completely changed by clinical academics. This applies to many areas of surgery. Certainly one of the most interesting aspects of innovation has been in vitro fertilisation, which is now a $35 billion industry. We get no money at all from that, even though we in fact developed it.
What has happened—I think partly because of regulation that has been misapplied—is that it has become more and more difficult to get licences and more and more difficult to get the work in clinical medicine inside the NHS. That is now ignored completely, and that is completely inadequate if we want innovation. We should be encouraging it.
There is a huge amount of innovation to be done, for example in genetic medicine. One of the things we have heard again and again is that the two greatest, most important areas in the workforce are, first, imaging—which of course includes what we have already talked about, not only radiology, but mainly radiology—and secondly genomics and the use of genetics generally. It seems to me that those two areas are being pushed mostly by these sorts of employees, but there just is not enough money in the current pot. There used to be other moneys from research councils, which is no longer available, nor indeed from the Wellcome Trust. It is urgent that we have some recognition of that quite separately in the health service, in my view.
Preet Kaur Gill: Let me be clear. The points you raise are very important and I recognise that we have to strengthen this workforce, as you say, if we want to sustain the capability of the United Kingdom. That is why the workforce update piece is there in terms of how we recruit, retain and fund the areas that we are pushing through. I can share with you when that is likely to come, but it is due in a couple of months.
The Chair: Baroness Willis, you have a point about training.
Q261 Baroness Willis of Summertown: I have a very small point. These are the current medics and current clinicians, but of course you have those going through their six-year training programme, where none of this is covered, at least in my experience and in the evidence we have taken from others. How will you weave this into the training programme now so that the next generation who come through are able to embrace these technologies without seeing them as something that is not for them?
Preet Kaur Gill: This is a live conversation. You are absolutely right, the whole workforce piece—who we need where, with what skills—is integral to being able to do the transformation work that we absolutely want to do. There is a focus on the genomics capability and how we build that workforce. There is a structured NHS genomics education programme offer that is available, and that is very much about working closely with neighbourhood health teams and recognising how genomics improves the preventive and pre-emptive healthcare piece. The programme is directed very much in that space.
Baroness Willis of Summertown: It is an offer. Does every health trust need to take that up? In every medical school, will they be teaching this?
Preet Kaur Gill: That is something I will take back. It is a valid point and I am happy to.
Q262 Lord Winston: We have heard that the two great advances have been in imaging and genetics and that in both of those areas the staffing is inadequate in the health service. There is a desperate shortage of clinical geneticists. On the imaging side, according to the Royal College of Radiologists, we may be as many as 3,000 jobs short at consultant level. That is a huge proportion that is needed in that area. That also applies to pathologists, which is the other area that is closely related. We still have large numbers of chairs—maybe 20% or more—not filled at that level because in fact we are not doing enough. That is a responsibility partly to ensure that it is integrated in the health service.
Preet Kaur Gill: The department’s overall workforce strategy is due imminently. It is ready to be announced, I think. That is recognising where we need everybody and where those gaps exist. Particularly in terms of this area that you focus on, there is a recognition that we need to do more.
Lord Winston: If they have a need for better innovation, surely this is an essential requirement.
Preet Kaur Gill: You are absolutely right, it is, and there is a recognition that underneath the overall workforce strategy we absolutely need to have the research workforce up front. That is the bit we are working on, as you allude to.
Q263 Lord Ranger of Northwood: Minister, I want to congratulate you. It has been seven weeks and you seem to have approached this brief with gusto. Your excitement is palpable. You have a lot to deal with. We all know the NHS: the technology, the innovation, the bureaucracy and the organisations. My question is about the voice of the patient. We have had quite a broad-ranging discussion today, but even in the 10-year strategy, where is the voice of the patient in terms of their experience? I say this as someone who has spent the best part of 30 years delivering major public services using technology. I am not a techie and I am not a commercial person—you have those people with you—but the voice of the patient, the user experience, the customer experience, or in this case the patient experience, will be fundamental to any success that we see happening in this area.
You mentioned the 41 million people using the NHS app. I have used it recently. My experience is not great from the app. You have mentioned comms a few times as well, and if I was to say that my average fitness app, with much less data, gives me a much better experience and understanding of what is going on with me than the NHS app does, I think you would know where I am coming from.
Where is the focus that you want to have in all the things you are doing on what we call the UX, the patient experience, how that can be fed back in and how they can feel that they have some control and understanding of what is coming? We talk about precision medicine. Is this not about personal well-being and the language that we use to describe what will happen?
Preet Kaur Gill: Yes. Lord Ranger, thank you very much for your kind words. I have to say I have probably one of the best briefs across Whitehall. Health touches everybody’s lives in some way and it is a real privilege to do this brief. It is exciting because I think the work that we are doing, especially in the datatech space, is revolutionary. It will change people’s experience of the NHS and how they navigate various treatment pathways.
The question you raise is fundamental to everything that I do. My brief is Minister for Innovation and Patient Safety, and it is patient safety from the constituency angle as a Member of Parliament that I bring to every single question. What does this mean for the patient? How is it improving patient outcomes? Where is the voice of the patient in terms of the things we are doing? That is why I think the communications piece is vital now, having the national conversation about what patients accept.
You are absolutely right about the NHS app. There are some good things it does. You can get a GP appointment, brilliant; you can see your blood tests; you can see your appointment letters. However, there are lots of people who are excluded digitally as well. Recognising rural areas where wi-fi is not always brilliant and the infrastructure is not there, how do we make sure that those people are equally very much part of being able to have access to healthcare?
Health and well-being is interesting. I want to be ambitious and the NHS app has to do a lot of things—it will be part of the single patient record, NHS Online and the HDRS—but we have to take the country with us. We have to be talking to groups. We have to talk to the Lords, very important, with lots of expertise and experience in how you use the app versus what clinicians think. How easy is the interface? You have shared about your Fitbits and things like that. You are absolutely right, what that can produce versus what should be on the NHS app is important.
Let me be clear, I am absolutely ambitious. I want the patient voice to be central to this. It is a consultation starting now, having that national conversation, engaging young people as well. We can learn from the social media ban piece of consultation and work that was done in DSIT just how integral that was and how integral it is to bring people with us. All of us want to see outcomes improve for patients. Rob, can I bring you in?
Rob Thompson: The Minister makes a good point about what people think generally en masse about the NHS app. There is high variability in there. I own the delivery of the app; I own that team. There are 170 user-centred design people who work in that area and they have access to about 50,000 people through focus groups, experiments and labs to be able to figure out just how people will or will not use the service. When it comes to UX-type activities, I feel that I have good depth in being able to make the app as good as possible.
One of the challenges that we have is connectivity to data and other systems within the health system, particularly in and around GP IT systems, which is where a lot of the records sit. That is why you get variability in the app experience across the country, because some GP systems are not very well connected in and others are.
Lord Ranger of Northwood: Mr Thompson, may I interrupt you for a second. When did you last use the app?
Rob Thompson: Last week.
The Chair: Lord Ranger, we have so many more questions.
Lord Ranger of Northwood: Just one point. On blood tests, you can get results back but they are not in a language that a general person could read and then they send you off to NICE to work out what it means. With all those 170 people, there is something about how you get that experience from a user perspective that I think is slightly missing there.
The Chair: We must move on. Minister, you and your officials have been extremely patient. We are almost reaching the end of our questions.
Q264 Lord Drayson: This committee has looked—not just in this inquiry, but in previous inquiries—at how the Government supports domestic UK companies, including within life sciences companies. The evidence we have received is that the experience of UK companies is very negative, so much so that if a company sets as its business strategy to focus on being a supplier to the NHS, the venture capital community does not want to invest in it. We have seen evidence from leading medtech, AI and life sciences companies that have had much more success in achieving procurement wins outside of the UK.
Minister, you have joined a department that has a track record of not buying British. In the IT space, Mr Thompson, you say that cloud hyperscalers are world-leading. They are world-leading in the UK now because they so undercut the domestic industry 15 years ago that the UK companies could not compete. We are now going through that again with AI.
Minister, to the point of trust, an attitude survey on AI in the UK just two weeks ago showed that the British public want politicians who are in control of the way in which AI is deployed in their country and who respect the issues that are important to the general public and therefore as patients. They are very concerned, from this survey, as to the way in which the alignment between the US Administration and US big tech has been used negatively. We see how the Government have recently adjusted the NICE value for money threshold as part of their pharma deal with the United States.
I would like to hear from you, Minister, as a relatively new Minister in the department, your enthusiasm or otherwise for looking again at the way in which the NHS procures technology/medicines across the piece to help favour domestic industry. What are you doing now to change the environment such that it aligns more with what the British public want?
Preet Kaur Gill: Thank you so much for your question. It is an important one. The first thing I did when I was newly appointed was ask, “Is there a British alternative? Can a British company do this?” Absolutely, we must buy British. There are lots of very good companies here. In the end, the NHS is procuring capability and it is not focused on who the company is. There are companies that are not British companies that partner and have a consortium with UK companies, which UK employees are working on. The FDP is a good example of that. You are absolutely right. I want us to give confidence to British companies and be tapping them in terms of letting them know that there are these procurement opportunities so that they can bid for them successfully.
Lord Drayson: Minister, a previous Labour Government made a decision that they needed to maintain a nuclear capability by providing orders to Barrow-in-Furness shipyards so they could continue to know how to build submarines. If we want to have industry in this country able to meet the strategic needs of the NHS, your department has to spend money with them. I am looking for what action is being taken now, because this is an urgent, here-and-now problem. What is being done now to change the environment, such that these businesses can have confidence that the NHS is not going to be buying American tech?
Preet Kaur Gill: This is a very important point and what the life sciences sector plan is seeking to do. We have to make sure that the whole pathway is available. I do not want to see, in my university or in my area, medtech companies or start-ups having to sell their innovation, which is then taken out of the UK and manufactured out of here. I want us to be the best in clinical trials.
Lord Drayson: I am sure you do Minister, but—
Preet Kaur Gill: This is the work we are doing.
Lord Drayson: Genomics England is storing all of its patient data on an Amazon Web Services datacentre, which comes under the US CLOUD Act and therefore is under the control of the US Administration. Unless your department takes decisions to build this infrastructure, you are only left with the choice to use these solutions, which go against what the public wants.
Preet Kaur Gill: Some of it is about the value for the taxpayer. There are two questions that you asked me. Let me address the first one. In terms of the life sciences sector plan and supporting start-ups to get access to cash and have the innovation to build the ecosystem and service industry around life sciences, we have a lot of brilliant UK companies doing some great innovation. There are the national wealth fund moneys; there is money across DSIT to support this and the manufacturing fund. There is a lot available. What is not always clear is how all of this integrates and how SMEs or British companies can scale up in the way that they want to and get the adoption that they need into the NHS. These are the things that we need to absolutely get right.
On sovereign capability, we have to make sure that companies can deliver. There are a lot of companies that will be pitching for the procurement on single patient record. Absolutely, I want to be clear that I want British companies to be bidding for that when it comes out. It is not all about the tilt to looking at other companies doing the piece of work in the NHS, in the way that is shared. We have to look at where there is capability and capacity as well.
I take on board your point that there are amazing companies that want to get into the NHS and are struggling. That is why the MHRA and NICE are doing work collectively. Having a pathway and being very clear about what adoption looks like is integral. All of this work seems to be happening simultaneously, but I am cognisant of the issues that you raise.
Q265 Lord Drayson: Thank you. That is very good to hear. Could we turn to clinical trials? The UK is currently undertaking about half the number of phase 1 clinical trials that it undertook 10 years ago. This is a pretty catastrophic decline in the picture, but there are early signs under the current Labour Government of that starting to stabilise. Could you tell us what you are doing to reform this, particularly in the area of personalised medicine and AI, where there is clearly an opportunity for major breakthroughs that will accrue to the UK only if the UK is the place where these new innovations are trialled?
Preet Kaur Gill: Absolutely. Coming back from San Diego biotech, I want the UK to be the best place in the country for clinical trials. The Prime Minister set a timescale of 150 days. We have managed to bring that down to 122. That is not good enough. The recent Quell clinical trial shows you what we can do. In 57 days we were able to set up the first trial and we did it simultaneously in 10 sites across the UK. The NIHR has an industry hub and the idea of the industry hub is to support those clinical trials taking place, but 122 days is not good enough.
We have to be competitive. I do not want to hear people telling me that Australia or Spain can do it better than us. We can do it. This is certainly the conversation I am having with the department. How do we bring those clinical trials here so that the UK is the best place to come and do those trials, not just phase 1 but the whole of them? That is important and I know we have the capability of doing it.
Q266 Lord Drayson: The last question is comparators and what other countries are doing in the context of the massive change that is taking place, not just in technology but geopolitics and the relationship between nations. It is interesting that other nations, particularly European nations, are taking a much more muscular and robust approach in the way in which they choose their own industries when making procurement decisions. We have heard that there is a cultural difference between the way civil servants within the UK—not just in the NHS, but including within the NHS—and civil servants within other countries, such as France and Germany, behave towards their home-grown businesses. Is that something you accept is a difference and is that something that in the current geopolitics you agree needs to change?
Preet Kaur Gill: I agree that it needs to change. Absolutely, we should be procuring from British industries and helping that wider supply chain. This is about creating jobs. It is about making sure that the innovation stays in the UK and that we can have the wider supply chain that supports that ecosystem, whether it is manufacturing, legal services or in the labs. This is a sector of growth. It is part of our industrial strategy. It is absolutely right that we should be procuring British and I am glad the Chancellor set that across Whitehall as a requirement. We hear you loudly and clearly.
The Chair: Minister, you have been excellent, and thank you for extending the time that has been allocated for this session. You and your officials have been wonderful in answering so many of our questions. You said earlier that you have one of the best briefs across Whitehall. We wish you very well in your endeavours on that. Thank you so much for coming.
Preet Kaur Gill: Thank you so much.
