Webinar: Tech and NHS recovery - what you should know

so good morning everybody welcome to this health foundation webinar i'm jennifer dixon and i'm happy to be your chair this morning uh today we're focusing on technology uh the nhs as you know is the uk's biggest industry actually it's europe's biggest industry its budget is a quarter of all government spending on public services so improving quality and value for money is always front of mind isn't it but after the pandemic the pressure is really on nhs reforms as we know come and go but technology and innovation are perhaps the greatest hopes to speed recovery of the nhs and future and sustain us it for us all in the way we we want the uk as we know has world-leading science and research has shown in the last year but the in the nhs uptake on innovation has not been as fast as as all that how does an nhs organization which is so large how does it best and most intelligently get what it needs with response with respect to tech do we let innovation rip and see what happens or should there be a more obvious guiding hand and involving the nhs as staff and public and rapid testing and spreading what do we learn from the pandemic which turbo charged the use of tech and got it off the shelf and into practice in days not years do we need to take a hard look at management of red tape as well or do we need to halt and evaluate so as the nhs faces the mammoth task of recovery we're going to find some answers to these questions in today's webinar with our expert panel who i'll introduce to you all in a second so what we're going to do today the format is we're going to have a short presentation and then we're going to have some opening remarks from our panel and then we'll have a discussion on some key themes bringing out some key questions and some of your questions that you will hopefully be asking during the seminar and what we'll do is aim to wrap up by 11. so maybe just first some housekeeping firstly please ask questions uh use the q a box at the bottom of your screen to submit them you can see that i hopefully you're all now used to this q a function you can upload questions you would like to see answered by clicking on the thumbs up icon by the question in the q a function i don't think you can downvote those so just try and upload and we'll try and get the most popular and eye-catching questions uh out in the discussion and thanks for all of you who have already submitted uh the second is to say that the webinar is being recorded and the video will be available afterwards so please tell your friends who missed it and then finally if you want to treat a tweet it's nhs tech that's nhs tech so please feel free so i'm with all of that i'm delighted to welcome the four guests on our panel well actually three are here at the moment and the fourth will be arriving we hope in 10 to 15 minutes but i'll introduce them all in the order in which they'll be speaking so the first is our very own uh will warburton who is the director of improvement here at the health foundation and leads all our programs to improve frontline care will was previously a director of the world innovation summit for health and head of operations for women and children services at st mary's and queen charlotte and chelsea hospital in the nhs next to him we have sam roberts who many of you know um sam currently leads the legal in general's investment strategy across the health and care sector and many of you will know her from her previous role as chief executive of the accelerated access collaborative the national umbrella organization for health innovation hosted in nhs england and sam originally trained as a physician now indra joshi who isn't here yet but hopefully will be arriving soon is director of artificial intelligence at nhsx she leads the nhs so apologies everyone we appear to have a technology problem on our technology webinar um which was a deliberate plan um to demonstrate just how important it is to have uh backup plans around technology so um shall i rory would you like to just briefly introduce yourself then maybe i'll go straight to the presentation yeah i'm just i'm using myself i'm rory catherine jones i'm the bbc technology correspondent uh i have covered uh the the tech scene in in in the health service uh uh quite a lot during the last year not so much before that but i will also be coming to this as a consumer are on all sorts of levels of their whole service which has given me some insights into the problems of adopting technology and particularly in sharing data fantastic thank you so much for joining us today in which case um while we wait for our chair to reappear i suggest perhaps i'll go straight into a presentation um so let's get going i've a few slides so um i'm going to talk around 10 minutes just on some of the opportunities um and risks around the introduction of technology and then i'm going to say a bit about what we think is needed to realize the potential of technology to support nhs recovery and open up some questions for our for our panel so firstly the opportunity as jennifer said the pandemic has turbocharged the uptake of innovation as a whole in the nhs and technology in particular and the chances are that if you've been using healthcare over the past year or you've been delivering it you've been using tech in new ways or more than you would have before and that goes for established technologies like telephone consultations electronic prescribing services the nhs website and 111 but also some newer technologies and platforms seen increases in video conferencing both for patient clinician interaction but also for communication between professionals uh the rise of various apps the nhs app itself for appointment booking and patient record access and obviously covered contact tracing and vaccination and we've seen more remote monitoring of patients in their own homes just to select a few and there's understandable excitement about the potential when you look at the rising demand for care more than 5 million now on waiting lists large increases in in demand and primary care and mental health as well and constraints on a workforce that has been through an extraordinary period of strain constraints on finance that will inevitably come over coming years there's a lot of hope that technology can help in terms of nhs recovery and that could be ranging from giving us the ability to monitor our own health provide more personalized advice on how to stay healthy improving administration from basics of managing bookings to referral letters scheduling clinical applications like image analysis perhaps artificial intelligence offering opportunities to detect signs of breast cancer or eye disease more rapidly and effectively or the use of interactive robots um either to help alleviate stress and anxiety at home and provide care and support but at the same time we know that all of this excitement hype can sometimes outstrip the reality and while there's a lot of excitement particularly about advances in some of those fields like automation robotics and ai many of those applications require significant further development in real-world testing to prove that they're safe effective and help patients and staff a lot of testing so far has been done in the lab and expectations for what tech can do are often high based on our experiences in retail or in banking but there are some reasons to expect a different pace in healthcare sometimes the level of investment that the public sector can bring the skills available and the fact that the human dimension of healthcare means that there are some aspects of care that we might not want or might be hard to automate one size certainly doesn't fit all so uh recent survey work we did with the public for example showed that of around a thousand adults who had been using technology more during coved 42 recorded um or reported that they thought that the quality of care was worse as a result with those proportions slightly higher in older populations and people with a carer who are obviously those with uh who make up significant um proportion of health service use um and cost this shouldn't be taken as a final verdict by any means many of those same people were optimistic and wanted to see more use of technology in future but we need to make sure that we understand what works for whom and when and adapt to make sure that these technologies are accessible to all so what's required for realizing the potential i'm going to pick out four areas and first is setting the right priorities so the nhs long-term plan will set the priorities for the nhs as a whole but what are the problems that the nhs and patients really need innovators in industry and within the nhs to be focusing on so a lot of attention as we've already mentioned goes on to specific clinical applications of technology like the ability of ai to accelerate diagnosis and that's exciting important work that captures professional and public imagination but many of the major challenges that the nhs is facing over coming years include things like improving people's health managing wanting morbidity integrating care across pathways and the overall affordability of the system now can technology help those problems as well jennifer mentioned our report that came out yesterday switched on which is focused on automation and ai specifically rather than technology as a whole but within that for example uh we quoted research from oxford university showing that nearly half of administrative tasks in primary care which is critical to managing some of those challenges of integrated care multi-morbidity and so on maybe half of the administrative tasks there could be automated theoretically with technology we've got today but it would take a lot of work to do that and it would that could free time for patient care and help with system sustainability so we need some focus on these operational and perhaps less glamorous issues alongside the significant investment that's rightly going into high-end biomedical innovation secondly i think we need to develop new ways of testing and evaluating so technology moves fast um we've seen that and it can come from outside the formal health and care setting as well as being planned from within it so we need to be able to identify impact unintended consequences at a similar speed to that development to answer important questions about safety acceptability and effectiveness and we've seen that in recent months for example with the introduction of pulse oximetry at home um which is a fantastic innovation but we've seen a lot of work needing to be done rapidly to answer important questions like whether results are affected by skin color for example or in the center here are some graphs talking about the rise in online access to primary care um so many of you may have if you've been using primary care used online platforms such as econsult or ask my gp are these channels creating new demand that could be managed elsewhere are they adding to that pressure on primary care um our team here at the improvement analytics unit our data analytics teams are using real-time information to try and tackle and answer those questions more quickly so we also think that we need new ways to build that evidence quickly bringing in multiple perspectives and we're very excited about work we're supporting at this institute in cambridge university which is developing one such tool called discovery which enables multiple participants from different perspectives to help develop ways of knowing what good care looks like rapidly bringing diverse views to bear on these quality problems so the nhs is alive to these challenges and the new nice strategy for example recognizes the importance of this agenda but how we test and evaluate does need to develop rapidly if we're going to be able to move fast without breaking things thirdly we're going to need to get better at implementation and spread we've been funding programs to improve care in the nhs over the past 15 years and one thing we've learned is is that making my dear work in one place is only half the battle um often getting an intervention that's worked well in one situation or one place and then making it work in another is not simple but we're more likely to succeed if we support clinical teams managers patients to adapt those interventions to their setting um and put in place the right approach support to do so um this week we've just uh announced we're funding four systems in cambridge manchester dorset and bradford which sam kindly helped us select to help them develop their ability to systematically take up innovation um [Music] taking account of that local context so we think that the change that's required often when we're implementing technology is not just a technical one it's one that requires staff working differently it means paying attention to all the implementation factors you need to pay attention to when you do change well everything from involving users in the design to having the right culture and leadership and the right resources and infrastructure to support the change so we need to make sure that we fund the change and not just the tech and then fourthly finally in terms of this section even if we pick all the right problems we get the testing and evaluation right and we support the implementation really well then if there's if it's not acceptable to patients public and staff then we really won't get very far so we need active engagement with patients this is an example of some survey work we did recently with patients asking about four technologies listed here self checking communicating bad news by video ai triage and robotic care asking them would those approaches impact on whether they felt their care was centered around them and person-centered um and as you can see very varied responses but actually some doubt about whether or not they'd be willing to engage particularly with the communicating bad news by video interestingly when we asked people if they would be open to that for their own care rather than in principle the results did shift quite significantly which suggests that there may be more openness to people using new technologies for their own care than they might imagine when asked the same question in principle the public as a whole um feelings about automation and artificial intelligence this is from our report so again just looking at those technologies um we're evenly split in terms of how positive or negative people feel but one thing that we consistently found is that those who are more familiar with these technologies are more positive about their future potential suggesting that there's a very important role from policy makers practitioners journalists like rory to be helping educate engage um the public um uh in about these technologies and it's a similar story with professionals um with positive and negative views been very well balanced but slight difference between professional groups doctors being slightly more positive than others and likewise we found here the staff that are more familiar with the technologies are more likely to be positive about their potential and again given our experiences is that staff engagement and enthusiasm around change is one of the critical make-or-break factors for making change happen in the nhs then creating staff to space for staff to lead the introduction of these new technologies experience them for themselves will be really important so i'll draw to a close there just to mention our report that came out yesterday that's available on our website or through the link there focused on getting the best out of automation and they are in healthcare and i'll hand back to jennifer to explore some of these questions that she put to us at the beginning thank you thank you very much will and apologies to the audience the irony is that the technology cut me out as i was introducing the bbc technology correspondence so my apologies for that anyway it's great to be back and i also wanted to welcome indra who has now joined us uh from who's director of uh ai at um nhsx so i'll i'll come to indra last i did say she might be um third hello indra welcome to the webinar um thank you will and just a question for you will because it has already come up it was the first question in the q a um what we've seen in the pandemic is a phenomenal acceleration of the use of technology existing technology that happened overnight it seemed there is a meme there was a meme around which was the nhs somehow as a staff group are somehow resistant to technology and that is the problem um can you say a little bit about that given our surveys of nhs stock which you just referred to and what you can you unpack the word resistance here what do you think might be going on that was suddenly overcome during the first weeks of the pandemic thanks jennifer yeah i mean certainly it gave the lie to that to that to that meme um we did some work on trying to understand what enabled innovation as a whole to accelerate so rapidly in the pandemic and there are multiple factors obviously you know it's driven from the need to create social distancing there's a sort of common purpose that suddenly emerges that really drives all activity through some things that are quite hard to replicate in normal circumstances but we also saw some very nice collaboration at different levels of the system so something that we call top-down clarity and bottom-up agency so national bodies taking decisions and issuing guidance on things like anything from information governance and trying to reduce anxieties around that um but it's giving some leeway to people to innovate locally according to their circumstances that we think made a significant difference i think in terms of resistance as a whole you know it's interesting i mean the dynamics of resistance are very interesting terms of health care change usually resistance has a reason it's not necessarily that people are being obtuse or just saying i don't want to improve care in my context it's actually about having genuine engagement with staff genuine engagement with service users and building them into the design and development so that acceptability is part of the process that is built into the process rather than being landed on from the outside and that's where this uh this this good working between multiple levels of the system becomes so important i think yes and we'll come back to this won't we um but there's also the difference between the um there is a there isn't so much caution when it comes to trying new treatments certainly medical treatments you know we're as fast as every every other healthcare system it seems with that depending on investment but it's the pathways of care that there seems to be more you know whether where those are shifting i wonder whether there's a distinction there between our speed of innovation and um treatment anyway we'll leave that one hanging because there are other questions related to that and i'm now going to turn to sam um sam very interested in to hear what you have to say both from your position now as an investor with legal general but also in particular your previous role as chief of the accelerated access collaborative which was the collaboration of organizations trying to lean the nhs thicket to try to speed up innovation and spread so sam thank you great thanks so much um hopefully you can hear me okay uh so i suppose my comments will reflect on the differences on being two sides of the fence you know my last role was really kind of in the machine of the nhs and now completely out and working largely with innovators as an investor so i think when when i was in the machine working on the accelerated access collaborative the the central thing that we were trying to achieve is at that time there were lots of publications and kind of conventional wisdom that was that the nhs was just not good at adopting innovation and never could be and i think we did lots of work with folks like the health foundation and others to show that's just not true the nhs can be very good at adopting innovation but it is hard work that requires um a variety of levers and activity and capacity and incentives that can't only be exactly as well was saying kind of top-down directives such as an innovation payment but also needs to kind of be driven by bottom-up agency which is a combination of time as well as um you know making the connections for the social process that is the adoption of innovation to happen so i think that was our um our main work and and so that used to kind of take up all my the hours of the day thinking is there a way to make this adoption of innovation which feels so laborious and complex and easier it's been quite interesting moving away from that and now working on the investor side where we spend most of our time with [Music] new companies that are kind of at the proof of concept stage and trying to get into health systems or other investors who look across multiple health systems and so it's good to hear what they have to say and so i think a couple of reflections being on the other side are actually in the uk we are really good at a lot of stuff that we don't give us ourselves credit for so i think uh firstly quite a few people see us as leaders in the world in the world of evidence standards around digital and intra i have to shout out to you on this one because this is something that you led so we're pretty clear about you know these are the evidence standards you need for the for different types of innovation i think we were one of the first in the world to do that so that is really great the second thing is that our national programs you know when you're in the nhs most people hate national programs because it's all this top down thing but from the innovator side actually some of our national programs such as the ai award or such as the work that and has been done on remote patient monitoring have actually done exactly what will's pro described as they provided that kind of clarity of purpose and backed by uh funding that has really unleashed large parts of the industry so that's been really pretty transformative and if we think about the broader economy we're definitely speaking to companies that are saying we're coming to the uk we're creating jobs in the uk because we know you're going to help us test our ai we know you're really interested in remote patient monitoring so i wouldn't kind of underestimate the knock-on effects of that so the second thing i think we're really good at um the third thing i know we bombard the state of our data and i'm sure rory you'll have plenty to say about that but actually you know most other health systems data is also not great so when we speak to innovators they do regard the uk as as kind of world-leading in terms of data for research whether it be things like cprd or haze or whatever um and then the last thing i'd say is when i worked in the nhs we used to get a lot of grief about the nhs being so hard to navigate but actually every big organization insisted was hard to navigate farmers hard to navigate the us is hard to navigate so you know um the it's great that we have we we um try to simplify that and try to help people navigate it but it's not something that is unique to us i would i would say which was when i worked in the system but it felt like it was unique and we were sort of the problem children so those i think are the great streams i was going to spend the last couple of minutes of my comments focusing on areas where i feel like both of our you know both sides are leaving value on the table so to speak where innovators um could be um working faster could be creating more jobs for the economy and the nhs could be getting more value from these innovations and i think that they're in three areas which are kind of around a use case reimbursements the old chestnuts and then complexity of adoption so to start on the use case the thing that i hadn't really understood properly until i worked on the other side is that most innovations or most innovators we work with don't have a fixed idea of how their innovation can work within the health system they are desperately seeking guidance from folks in the health system to say actually this is where i want to use your innovation but in the absence of that they kind of make up a use case so i'll give you an example we're working with one innovation which could literally have 10 different use cases all the way from screening diagnosis monitoring clinical trials and they they are struggling to engage with clinicians and managers and payers to be get guidance on which is the use case that would benefit you the most um they're in general not concerned about what makes them the most money it's you know what is actually of utility to the patient and clinician so that's the one thing i'd um i'd love managers and clinicians on the school to be thinking about is how are you engaging with innovations and being quite purposeful and saying this is the need we have and this is how we want to use your innovation rather than being a passive recipient as it comes to you so that's kind of thing one um and when when we were on the the health foundation innovation hubs panel lots of folks were thinking about that and talking about how they would do that so i think that's really positive the second is on reimbursement and um i think we've got a kind of opportunity and a risk here so opportunity is we can look at other countries like germany with their digital health care act we can look and look at at the us with their cms breakthrough designation and see that there are some pretty clear national reimbursement strategies that actually do um well appear to because they're in the pretty early days um supports the um take-up of innovation in some settings um i think the so that's the opportunity i think the the risk is that and we as we move to integrated care systems some of the barriers to for reimbursement so things like the benefit occurred in one party but the cost was incur was accrued by the other party those um barriers should be melting away so to speak but when you're outside the system you don't quite understand what's replacing them so we i spend every time i meet somebody from an ics i say can you explain to me how i should be telling my innovators how their innovation is going to be reimbursed and nobody really seems to know so i think that's a bit of a risk at the moment is everybody's sort of waiting for some guidance so it's the second thing reimbursement you know opportunity for national we need a bit of clarity on local and then the last thing is on complexity where i think us investors and innovators we're probably just thinking about what's the use case what's the reimbursement how do we get it initially adopted but i love the work of this institute or the work of trish greenhalgh because what we end up always knowing or any of us that have implemented a um an innovation is that it's complex and the starting of the use is only the beginning of the journey and that it's um it's not just the pathway change that is difficult because that almost can be can be mapped out on a piece of paper but it's um the conflict around perception the conflict around ethics and the conflicts around what quality of care actually means who defines it and how how variable is that based on context and different people's needs all of those things are very personal um and kind of social negotiations that occur every day in our health system and i think it's so the being able to successfully navigate that is the thing that's going to make the difference for innovations that that succeed um and scale in health systems versus just there's a bit of a burst of initial adoption whether away so i'll just kind of end by saying i don't think innovators or investors focus enough on that and we should um and it's just so great to see the work of you guys and folks in oxford um on this topic because it feels like we're at the tip of the iceberg of understanding actually what part of the social process needs to be in place to make this work i'm going to stop now thank you wonderful plan lots and lots of points there and lots of questions that have been prompted by what you had to say and perhaps if i can just start one question which is a sort of very high level strategic question just thinking about the nhs for a minute having to live within its means over the next year it's given this huge sort of what's called fiscal overhang left over from the pandemic um how does the nhs itself as an organization decide and what its what its real needs are with respect to i think two things one is productivity which is going to help us get out of a mess in future and secondly to shift care out of hospital and into home settings how you know when you look at demand signaling which is the accelerated access collaborative is doing a lot of this is demand signaling around the treat better treatment for particular conditions or multi-comorbidities all very important of course but what about these two big uber shifts how could the nhs be a lot better at signaling and stimulating change on those two big things which after all might be the critical things for its survival in in future yeah so i think there's two ways in which we're doing it which i think have actually improved quite a lot over the last couple of years but i'm you know um so the first thing is there will be big national programs that no single system will ever want to take on because the cost saved occur over so many years and the complexity and risk is so high so i think we've they've been two great national examples of that um one is diabetes prevention uh you know which was commissioned nationally i think it's although there's lots of touring and throwing about i think we could actually say that as a uk success story and then the second is around the work that we're going to be doing with grail on circulating tumor dna where we say actually you know blood tests that help to identify multiple types of cancer could be completely transformative in cancer care we can't expect an individual ics to be taking a punt on these we've got to do big national programs so i think there is something we say there are transformative therapies and absolutely we should be taking a national view on that and i think they seem to kind of coalesce around the screening early detection stage because of the kind of the long payoff um but then the second thing is when we first started talking about demand signaling in the aac we would go around to different um hsns um or kind of academic groups and say what are your needs and it was surprisingly hard to get that articulated and because that was a new conversation people had not been having conversations with for example pharma companies for years saying actually my need is x and they're being quite a nimble industry that could meet that need so i think that's a new thing you know we're not used to lots of innovators meeting needs on almost like a yearly basis we're used to big kind of multi-year billion pound trials giving a new therapy so i think that's a new thing but having said that and i'm basing this on the innovation panel i set on for you know the innovation hubs panel i set on for the health foundation is i thought that um the folks that we spoke to there had quite a clear articulation on the ics level of actually we're focusing on these four things and some were very particular they were like better outpatient management of um gestational diabetes and some were pretty high level where they were you know we're looking at aging but it does feel like just the process of and i've seen this profile so please um folks on the chat tell me if i'm wrong the process of creating ics's and having the whole system have to come up with a set of priorities is actually enabling us to be locally much clearer about we want this we don't want that right okay that's great thank you so much sam so let's now move to rory and rory your viewers both a punter but also from your bird's eye view of seeing how other industries try to move forward with technology to survive and compete in future and the nhs doesn't have to compete does it but it does have to really think about productivity enhancing and quality enhancing tech in a very systematic way so the nhs must seem a complex animal to you but to your observations from from those perspectives sure thank you um yeah i come at it uh both as a technology reporter of many years uh and one that didn't focus much on the nhs until the last year and then then has been heavily involved for instance in following the development of the contact racing app which had all sorts of interesting lessons but also as a consumer with two um long-term conditions uh and i wanted to start with the negatives and then try and be a bit more positive so from my experience uh has been great on the carefront uh i always say i i go to moorfields hospital every six months and spend uh sometimes five hours in the basement in an incredibly inefficient process uh and then uh then meet the world's leading specialist in my area and get fantastic care which i think is a kind of typical nhs experience but my eye condition which is why i go to moorfields was first spotted by the private sector by my optician who has great technology interestingly and over the years until recently has been frustratingly unconnected with the health service to such an extent when uh a few years back when i i went in and she spotted something new and wanted to communicate it to my gp and onto my consultant she asked me to get my iphone out and film the screen because she was unable to share these images with my gp she said the only way to do it would be to effectively print them out and fax them over which is obviously not good um so that gave me some insight into the horrendous problems of i mean which we all know about joining up systems which seems to have got a bit better um on the positive side uh i'm seeing a vast amount of innovation particularly over the last year i've got a number of friends in the health service uh and one a palliative care doctor told me just how quickly things had changed back back last march um just as an example uh i mean there is huge frustration as far as i can see amongst health service personnel who are keen on using technology but are hamstrung by the quite understandable caution which inhibits the organization so for instance getting patients to do elderly patients to do facetime calls using the hospital network or using their own devices or devices supplied by relatives they said would have been impossible before the pandemic because it rules his rules and you were not allowed to use those devices on a hospital network and suddenly magically all those things disappeared and we as we know we had uh acres of innovation within uh within a few months um and i i'm i'm seeing all sorts of hopeful stuff uh uh experiments with with ai in terms of diagnosis and so on i'm i'm a parkinson's patient i'm on a trial involving using ai to monitor symptoms more carefully and develop wearable technology uh and all that's great but what i observed during the development of the nhs covid 19 app was it was two things that that got a terribly bad press including partly from the bbc uh although i in the end i i concluded that their the attempt had been uh more successful than than people realized um but i came away with the impression of brilliant teams within the nhs uh and some of the private contractors they hired um hamstrung by poor leadership so continually naivety amongst the political leaders about what the tech could achieve [Music] a bluetooth contact racing app was absolutely cutting edge technology nobody knew whether it worked and the technologists were saying you do realize there are huge problems with bluetooth as a way of measuring distance between people the politicians because they hadn't developed a proper manual contact racing system were too inclined to see back in march uh this as a as a silver bullet so we're responsible for hyping it beyond uh belief and therefore there was disappointment uh when it didn't immediately deliver it is sort of delivering now or at least it's delivering probably as well as any other contact racing app around the world which is which are are evidently not silver bullets and that brings me on to hype i mean there again i i get from the practitioners um at the front lines some some realism about for instance how quickly ai can deliver change um but a bit too much hype uh from the politicians uh notably i won't name the company but i've just been watching a video about one of the most well-known proponents of ai in healthcare whose outlandish claims about what their technology could do have been backed up by politicians uh and uh behind the scenes and not really trusted by doctors so i think there's that divide and the final thing i'd say is from working with uh from observing outside organizations uh commercial organizations over the years the story is big companies big organizations fail all the time the nhs shouldn't have an inferiority complex about this in our morning tech meeting this morning we were talking about microsoft and how it allowed skype basically to die it bought skype for quite a lot of money and skype effectively by decision yesterday has been consigned to the out of darkness microsoft has had loads of failures it it developed uh it it missed out on the mobile phone world it bought nokia uh for a huge amount of money and then wrote off all the money um and yet it is now valued at over two two trillion dollars so uh you're allowed to fail uh in in in in the private sector the poor old nhs uh is not allowed to fail by people like me it's it's the fault of uh the voter journalists often so uh i don't know quite what the answer is to um encouraging air experiment experimentation giving people permission to fail but there does need to be a bit more permission to fail and and what might be the positive messages that you i mean there's no other industry as large as the nhs there's probably no other industry quite as complex as the nhs given all the things it does rory but can you point to some positive examples where some big industries have really revolutionized um quite significant parts of their business that um that you that you think have particular lessons for the nhs particularly service based industries well where should we look to for some inspiration i mean we we should obviously look to the amazons of this world that they have i'm never sure that it's that easy just to translate that across healthcare as we've discussed has got all sorts of different um issues about trust um uh some a company like amazon you know is a lot less trusted than the nhs and you you can learn lessons from their ability to focus to i mean the the big lesson uh which i i think you should take is the relentless focus of a company like amazon on on customer service at every er every every stage i mean for many years amazon basically was not interested in making a profit every time it was in danger of making profit it would spend more on improving its service improving its logistics yeah um yeah of course the nhs isn't interested in making a profit so um it's got it's got something in common with amazon in its early days um but i i i think it would be glib to say yeah there are easy answers translated from the private sector yeah indeed and of course amazon the relentless um focus on service is at some cost to the staff involved as we all know from uh recent uh so thank you for that rory we'll come back and particularly with the rapid testing and i think we should we should come back to that issue so indra very pleased to have you join us you're in a rory mentioned quite a lot about ai um is it a load of hype but um i think your perspective on how ai is developing in the nhs and indeed is the investment enough i noticed that 250 million into the ai lab is that enough um how are you setting priorities at least for starters thank you indra brilliant good morning everyone um so my name's indra i'm the director of the nhs ai lab doctor by background trained for many years in the nhs in a lovely a e dotted across different country countries different cities so a little bit about the ai lab and then i'm going to just address some of the questions and share some of the findings of some of the surveys we've done as well over the past year and the ai lab as jennifer's mentioned it's a 250 million pound investment really looking to accelerate both the development and deployment of ai technologies into health and care we have five delivery programs one of which many are aware of which is called the ai and health and care award sam uh was instrumental in setting that up and then sadly left us um but that's been running really well uh over the last year we've had we funded now over 80 technologies and they are ranging from very early academic research studies all the way to technologies that have got regulatory approval and are looking to scale and with that in particular kind of going to some of the questions that have been put in the chat what we're really looking at is quite often in the nhs we we fund and then we think oh well we've done that and sort of slightly hold our hands up and one of the things we wanted to do with this technology and going back to rory's comment about there's a lot of hype and there's a lot of you know flowery noise about what ai can do is really make sure we evaluate those technologies properly look at how not only cost effective they are but if they're clinically relevant how clinically effective they are and then working with the other regulators so those we work with nice but also working with regulators such as cqc and bodies such as health education england looking at them what's the impact it might have on the workforce as they try and adopt those technologies so we have another program with our ethics um lead is really looking at how do we build up those both capabilities and capacity of the workforce to deal with these technologies so what we're trying to do with the award in particular is really make sure we've got that kind of wraparound look when we're both funding that technology and deploying it and we're looking very much to make sure that those impact of those technologies as we move forward looking at health inequalities the qualities of the data are also addressed as those technologies are deployed and scaled and looking at the questions that come through the chat on that kind of longer term one of the key things is making sure we work with those commissioners on the flip side so obviously we can't solve every single care pathway but for some of those that we have funded so for example with eye care or with some of the cancer care pathways is really working with the team centrally and saying how can we look at adopting these technologies more longer term so we don't rely on these kind of injections of money that quite often happen so that's really about about money and about technologies but then there's a whole wrap around that has to go around that what we call like an ecosystem and so we work very closely with the regulators making sure it's clear to both innovators and to the system what kind of what are the rules what are the current regulations how to navigate that slightly complex um landscape we worked we've got great partnership with yourselves at the health foundation also looking at some of those nitty-gritty issues so this is through our ethics program about making sure we address issues such as data proportionality quality bias but also on the flip side we look at working with the policy makers and the funders looking at what we call algorithmic impact assessment so again making sure as we develop those models that we can interrogate them and then make sure we um can put them into the system and into policy documents so that's just a very a very brief overview of some of the work we're doing in the nhs ai lab in particular and we do it with a huge number of partners who have mentioned yourselves the health foundation but also health education england the aac for example they deliver as do the nihr some of our award programs so what we're quite conscious of is we can't do this by ourselves by no means is this a lone feat and a small program we're just a very small piece and a very large complex jigsaw puzzle of which we just want to make sure we try and do this bit of that as right as we possibly can and then we look outside as well so we make sure that we're not just doing this kind of quite insular as colleagues will know ai technologies rely on large quantities of data but also making sure that those rules those benchmarks those processes are acknowledged more internationally and so we work quite closely with the who and they have a working group via the itu looking at processes both in clinical evaluation and benchmarking and so we really plug into those and then just putting a little bit of realism back onto this the topic of this debate which is around you know the technology in the nhs so earlier in the year we did a scan we do a yearly survey of both innovators and commissioners to say you know how has technology really helped you do you think it's it's really um do you think your technology is really ready for the market and we're quite we're quite conscious that you know the pandemic has had both a positive effect but also a negative effect and innovators have said that to us we had a third of developers for example say actually the pandemic has had a negative effect and none of this is rocket science i mean there's a lack of engagement from frontline staff naturally because they were redeployed into different areas there was also um a lack of capability to collect the data because again staff may have been redeployed or um there was just not that the manpower to do that because people quite often think building that model at the very end is i mean that's sometimes considered the very fun part but actually all the part that comes first when we look at ai technology is actually quite resource intensive so we're also quite conscious that the system has been under a huge huge amount of pressure and so what we're trying to do is really help them with some of the technologies we're funding but also some of the wrap around ecosystem development to make sure that that we're not unduly putting pressure on the system but trying to help them alleviate things sometimes not so exciting such as appointment booking or really understanding where best to um work with the care pathways they're very helpful so just just for those who may not know so much about nhs and ai could you just sort of say if if there's sort of four big areas there's sort of stimulation of innovation and working with innovators and other investors like uk ri there's the regulatory aspect there's the testing aspect and there's the spread aspect of ai how what's your balance of activities across those areas so i'd like to say that they're quite balanced um as with all things uh say for example with the regulators unfortunately we've had a global pandemic and their attention has been naturally drawn towards things such as making sure we get vaccines out there and so what we have to do is make a balance with all things with resources we're trying very hard to make sure we and i keep going on about this point about evaluation because for me i really think getting that evaluation right ensures longevity and quite often people love to give money up front and with no disrespect whatsoever to investors quite often they love to see the quick fixes or the quick wins but for those clinicians in healthcare and for anybody who's experienced we're always talking about you know having to take a video you know there's there's a longer term burn here which we need to get right and if you rush the first bit actually you lose the the integration and a very good example for example would have been the pac system i mean we all remember you know popping the x-rays up on a light box about 15 years ago but when we installed the packs looking at images on a digital screen that was a good investment it was rolled through and actually it's it's you know now obviously there's a digital age of transferring images in many different ways although rory would beg to differ with having to take his video um but that was something that added longevity and that's the kind of thing we want to see that real integration then into systems that's great thank you so much ninja right so we've had 54 questions which is fantastic so you've obviously stimulated everybody so let me try and come to some of these then for everybody so the first question and i don't you all don't have to chime in maybe just jump in if you particularly want to there's quite a few questions about the the kind of thicket idea this is this is just really complicated it's it's very seems very fragmented um you know there's ukri there's nhs there's there's there's businesses how how how do we move forwards on this thicket i suppose that is the first set of questions because if it's all that the potential for duplication is huge isn't it of effort so i think maybe that i'll just direct that again to sam first off um to see if she would just have a go at that first basic question yeah so my mental model of this is that the um the nhs is a fantastic place to validate technologies and for that the thicket is fine right you can test something in wolverhampton you can test something in cornwall you can have a new ukri guard you could have a health foundation grant i don't think the thicket gets in the way there but but just because of the amount of money we spend on healthcare relative to a system like the us we are never going to be the place that it's easiest to sell technologies into so i think it's fine to encourage the thicker the validation stage but then i think at the adoption stage it actually is quite helpful to have some clear national programs like we were talking about with diabetes let's go right this is national priority this is how we're going to select the participants etc or um some clear kind of national reimbursement process like we've done with remote patient monitoring um where we give a clear signal is this these are the few things out of the thousands that will be tested in the thicket that we really believe in and we're going to get behind as a nation and then obviously local areas like ics's can choose to pick up whatever they think is um necessary so i'm not too worried about controlling the thicket i just think what's helpful is to have much clearer um prioritization on the things that everybody has to do yes indeed and also that a few that are important not just for patients but for the sustainability of the and there's quite a few questions there about how some of the less exciting areas are sort of perhaps deprived of the type that that irritated rory and his optician for example um or that we will slow down the shift out of outpatient care into the home setting um so i think there's a sort of the setting of priority seems to be quite a critical one there and how the system works together in that um there's a whole sort of set of questions i think about and i think rory and will and sam i think you refer to this is how you fail fast and and quickly how do you assess these technologies in perhaps in a very innovative way given that they're not like a pill they have they are like complex systems you embed a new tech in a complex sort of human system which is dynamic so the idea of spending a year doing a randomized control trial is just not possible how do we use new technologies to rapidly get results very quickly so well i didn't if you could go back to your comments perhaps about this institute or and perhaps kick that off and then i'll bring others in sure so um i think um understanding the impacts in terms of the quantitative metrics is is one thing and i was talking a little bit about the improvement analytics units and there are other teams they're trying to understand impact that way but i loved what sam was saying in her earlier answer around the complexity of implementation and and and i think we need new ways to understand how these innovations and technologies are landing with different groups because uh it's both someone's asking questions around the inequalities we need to understand how they work with the diversity of the population that they land with and we need to understand how they're working with different staff groups and so on um i mentioned discovery it's only one tool being developed by this institute but the idea is there is is that you can take a quality um problem and you can get lots of people to view a video for example of um a process of care and give their perspective on it so you can bring in a safety specialist you can bring in people from different backgrounds you can bring in a clinician you can bring in the managerial kind of perspective which we haven't touched on so much actually and i think is a very relevant perspective for this conversation and then you can try and build consensus um around what the most important aspects are to pay attention to and develop that in new ways now in the past bringing everyone together to do that kind of work would take whole day workshops it would take everyone coming together at the same time asynchronous methods that allow people to participate in these kind of exercises do give us different possibilities in terms of doing that qualitative evaluation and understanding how things are really being experienced by people i think it's a very exciting area of development and evaluation yeah indra i didn't if you have any comments on that given in hsx's work yeah i mean it's quite if the if it's something because sometimes the market itself does allow for rapid failure um if i refer to apps for example um i don't know the current number right now but the last time i checked there were over sort of 370 health and wellness depending on how you want to define wellness apps out there many of which nobody knows anything about and so however people will rapidly iterate those try and put them out there and then whether they get adopted more large scale would be a question that the market itself decides but from a healthcare perspective i think and i totally agree with sam's point is that we don't quite often like to talk vocally about failure um especially when it comes to technology because it's quite often seen as well you've you've taken money and you've basically drained that money and so one of the things we've tried to do in and i know i keep focusing on the ai lab but that's that's my part of the world and there are many other parts of nhsx that are also looking at doing what we call iterative projects or iterative programs and taking some of that burden on ourselves so vast is putting it onto the healthcare system we actually take it on ourselves to try and iterate a product and see whether it works so one example we did was with kettering earlier in the pandemic hospital looking at could we automate their citr so situation um reporting process quite simple this is an ai technology by any means but just an automation of a process and saying look will it work and by actually supporting them it freed up some of their frontline staff or their um their technical staff to actually work with the team directly and make sure that that product worked for them and now what we're doing is trying to see whether we can take that code and use it more widely and so by sometimes taking the burden off as institute or a trust and taking it on yourselves is one approach of trying to say well look can we actually do this a bit more iteratively and then that cost as well is taken on internally thank you andrea and sam you work very closely with the um uk research industry as it were the particularly nihr inside the uh and i'm involved with them too and obviously uk or imrc um a lot of the medical paradigm is very much a kind of gold-plated assessment isn't it and how far do you think they have got to do some work on these more rapid iterative ways of assessing which indra is talking about and will yeah i mean i think that there's there's there's obviously a lot to do but i think technology itself enables this so if i think of some of the technologies around remote patient monitoring where you've got a digital platform that can pick up some of the questions we've had about proms and prems at exactly the same time as it's picking up your blood pressure and your respiratory rate etc so i think um the tools for data gathering are improving and we've seen adoption of that on the clinical research side just go bonkers over the last year so i think that's becoming a bit easier um i think most of the innovators we work with exist outside the nihr gold-plated ecosystem right they are never going to be doing multi-center thousand-person uh crn portfolio studies they're doing little single center 100 person test and i think that that's actually completely okay what i think um if if it were me kind of waving a a wand the thing i think that those guys are most looking for is guidance on what is the data that they need to collect that would make people believe they um the technology should be reimbursed because it's helpful for patients and clinicians and so i think that's kind of the the bit of the jigsaw that's missing on the rapid evaluation is is there going to be a nice digital guidance process you know what's the end point we're all going towards where we can go don't worry everybody somebody's looked at it and they say it's fine even though we've only done this on 100 patients yeah yeah thank you and just staying on this theme rory um if we want to fail faster some of the things we really need to do in the health service as some of the questioners are saying is to look at patient facing large volume patient facing things you know how you have your repeat prescriptions how you book your outpatient appointment how you interface with the system uh given what you said about politicians on the one hand being averse to failure and and and pro-hype um how do you think that we could as a you know the nhs as a large industry could could could could have small iterative tests in this sense and not get tripped up by so-called failure yeah i mean anyone who's worked in an organization i've worked in the bbc for nearly 40 years knows that within their that organization small teams work big teams don't so um and i i think about i was just thinking about this trial that i'm involved in uh i'm going to make another point that you know um innovators uh can come from all sorts of backgrounds and can have a huge impact so this trial i'm uh of parkinson's wearable kit using uh being taught basically uh being being taught to recognize different symptoms uh is it's a small team it's uh on the one hand a consultant a brilliant young consultant neurologist at the charing cross hospital on the other hand some some biomechanical engineers at imperial college they themselves have already started a small spin-out company um it's it's it's a kind of good model for innovation it may work it may not but there are other similar small teams doing probably identical work um and out of one of those teams something will come it's not some huge national program it's kind of random small teams that have zeroed in on this this problem yes thank you and that leads on to another question which is uh andy webster thank you so much for asking that question who says it's great talking about artificial intelligence but some some of us are still using windows 7 waiting for login running applications on creaking infrastructure is it like building a house on sand you know with martin marshall the chair of the rcgp said that he took seven passwords to get into his gp system and i remember way back two decades ago i think it was alan milburn sort of being fed up because everyone was faxing each other and uh and i think the banning of fax machines various ministers so so in a sense that there's a balance here isn't it how much do we really um sort of go off with the you know the frontiers of innovation and how much do we really sort of tackle our backyard with some of the basic stuff which after all rory did mention with respect to opticians not even mentioning being able to link to the gps so balance of investment here who should i turn to you there um sam so i mean i think indra would probably be better than me on this but but my understanding of nhsx is that they are um kind of you know radically pro fixing the basics and matthew gould the chief executive of nhs i mean that's what i hear him talk about all the time um so i i think if anything the the focus is on fixing all the basics like the single sign-on et cetera um but my only kind of counter argument to that is i remember once going to a lecture by don berwick you know the the kind of the god of ihi and he has this lovely slide on if you want to fix quality improvement you don't focus on the lowest quartile you focus on the top quartile and by making the top quartile even better you pull the whole system up so that would be my only thing is it can't ever be uh only it also has to be and also you know get the basics right but also focus on the frontier because it will bring the whole system up together yes thank you and that comment i think refers to quite another uh other points made in the chat which is about inequalities um you know there is a meme at the moment no one left behind you know w

2021-07-04 15:34

Other news