Graham Silk at Google DeepMind’s Patient and Public Engagement Event
On the 20th September Data4Health's Graham Silk spoke at Google DeepMind's Patient and Public Engagement Event. The event was live streamed and you can watch it here (Graham's part starts at around 1:33:00). Please find his speech below (check against delivery).
Thank you very much indeed. It's a great pleasure to be invited to speak on something as important as patient engagement, specifically on the subject of patient data and the effect it will have on patients and health outcomes as we really see a move towards 21st-century medicine.
My qualification for being involved in the Health and data arena stems purely from the fact that I am a patient in the NHS who 16 years ago was diagnosed with leukaemia, given three years to live but got extremely lucky.
I got lucky because of many factors but the main one being that I was in the right place at the right time to get access to the very first example of 21st-century precision medicine with the drug Gleevec. So in 2001, with only 1:10 people surviving for five years, I was lucky enough to be one of only 40 people out of 130,000 with leukaemia in the UK, to get onto the clinical trial. The drug saved my life and, together with the professor who helped me get access to the drug, we started a charity in 2001.
This charity started in a world where medical treatment was beginning to move forward rapidly and is now becoming very different to the one we left behind in the last century. Precision, stratified or personalised medicine (take your pick) is about developing drugs, devices and diagnostics like the Genomics 100k Genomes project. This enables scientists and clinicians to define a patients disease much more specifically and consequently allows them to deliver an increasing supply of much more precise treatments, leading in turn to more effective outcomes. Sadly, though, in many disease areas it's very slow in coming through to patient benefit.
So, I was very keen that our charity should have a focus with a specific target in mind which we wanted to achieve. Our focus was to develop an organised structure which had a purpose and an end goal - and the end goal that we targeted was that, having been so lucky myself, we now wanted to create an environment that would actually take luck out of the equation for other patients in getting access to new treatments more easily.
Our charity raised funds and began developing an infrastructure which we called a TAP – or a trial acceleration program. This has enabled a great number of trials to be established by paying for trial research nurses, teams of trial coordinators and data managers but has crucially also linked together numerous hospitals across the country, which has given clinicians and researchers access to what is now over 20 million detailed patient records.
This has allowed the trials teams to recruit patients with specific disease diagnoses, very quickly. They have found that trial times - this includes writing trial protocols, recruiting patients and then publishing the results - have reduced from an average of 6-7 years down to 2-3 years. Consequently, patients get access to new drugs much more quickly and have a greater chance of surviving a disease like leukaemia. This demonstrates how crucially important the use and ability to share data is in driving new and better patient outcomes. The TAP is now an knowledged and proven model.
And this is now finally beginning to extend to a number of other disease disciplines such as Rheumatolgy and certain children's diseases.
Indeed, many diseases within Leukaemia are now seeing survival rates move from the one in 10 in 2001 for diseases like mine to nearer 9 in 10 people surviving for five years. One of the most important factors in all of this is the ability to link together hospitals within the NHS and to utilise their patient data which is in this case is quite literally, the lifeblood to enable better health delivery.
However, it is interesting that while all these patients and their families are presumably delighted to be benefiting from these advancements, very few of them are probably aware how their patient data has been made available to researchers and clinicians. I don't know for sure, but they're probably big supporters of this process. The people who maybe less supportive, may well be those who don't yet see the benefits of shared data - people who maybe haven't yet had a need for it. The task therefore is to make the case as to how data will help create a better and more effective 21st-century health system, one that will benefit everyone but without exposing people to the dangers of data misuse, which the anti-data lobby says will happen.
Last year I wrote the life sciences commission report for the Midlands, evaluating how we could deliver accelerated access in a better manner and I made a number of recommendations to government.
One of the main areas of recommendation was around being smarter with the available assets that we have in order to make our health system more sustainable so that more people will be able to access this new age of medicine in an affordable NHS. At the heart of this has been data.
A major element in driving investment, economic growth and a more efficient and effective NHS is the development and enhancement of the power of data and digital technology. The ability to better utilise data will stimulate efficiency not only in terms of reducing costs but also by allowing earlier and better diagnoses, improved care pathways and, most importantly, eventually the treatment of patients in a more targeted and effective manner.
One way of achieving this is to take anonymized data of people like me, combine it with thousands of other patients and then share it with scientists, researchers and industry to accelerate the development of new and better drugs and treatments.
This will achieve two things. Firstly, it will enable scientists to understand more about disease progression and so help create, better, more effective solutions to health problems and to do so more quickly. Secondly, because Industry will subsequently make a lot of money by utilising this information to develop better drugs, the NHS should, as a quid pro quo, effectively commercialise this data asset by charging Industry for being allowed to access the data that they use. This could potentially leverage billions of pounds. We could then utilise this money by ploughing it back into driving a more effective and, very importantly, more sustainable 21st-century health system.
We've actually seen an example of greater commercialisation from within the now proven model of the TAP. Over the last five years the small disease arena of leukaemia alone has seen trials leverage over £150 million worth of drugs, free to Patients and the NHS, put into the trials by industry. If we can do this in other disease disciplines and (as I said, we are starting to) we can really drive improved patient benefit and at the same time, significantly improve economic efficiency. It will be very probable that we can create a far bigger cancer drugs fund, which will be a multiple of the £350 million fund that we currently have and which the Government says is unaffordable, but at no cost to the Treasury.
There have also been hundreds of high value jobs created as well as significant improved cost efficiencies for the NHS as patients move to being out patients, often back working and with their families, rather than a greater cost as an in-patient. Beds are also freed up for other patients which might otherwise not have been available earlier. This is a really good example of a process which is BETTER - CHEAPER - FASTER.
Earlier this year I co-founded an organisation called Empower: Data4Health. The purpose of this organisation is to press the government to drive forward health data reforms so that clinicians, academics and the pharmaceutical sector can use data far more effectively and so that ultimately treatment can be improved for patients across the country.
Frustratingly, much of the interest from the media around the data topic has been around how to assure people that their data will be safe.
Over the last couple of months I've been on programs like BBC Breakfast, the World at One and 5live and they've all asked this same question. I've begun answering them by saying they're actually asking the wrong question.
We are putting the cart before the horse when we suggest not using data because of potential misuse. If we really want to help improve people’s health outcomes then we have to utilise this precious asset.
There are already laws in place which prevent our data being sold or misused - what we have to do as patients is to tell government to ensure that they cast these protections and rules in stone so that we can remove the subject of its misuse from the agenda and concentrate on creating better treatment options to keep people alive and healthier for longer. Data is probably the most important prize that we have.
I found it very interesting that after The DeepMind announcement of their data sharing agreement with the Royal Free around AKI, the Times ran an article. I was the lone voice in support of this great initiative that they quoted. Indeed what was worrying was the tone of some of the negative responses. You've heard that around 10,000 lives a year could be saved by utilising this data properly and yet a doctor in Oxford actually stated that it was a disgrace that deep mind had been given this information without proper regulatory approval (which actually of course is wrong) – and that they too should have to wait 12 months to dot I's and cross t's before getting permission. What we've already established is that 10,000 people will be dead if we wait that long and I'm pleased to say that the Dr had to retract what was a preposterous statement.
What was perhaps just as worrying, was that a patient group interviewed said that they didn't want to see patients being treated by software. This was presumably a reference to new diagnostic technology such as the potential utilisation of Genomics.
But I find it interesting that if you have a problem with your car you take it into a dealership. They sit you down, offer you a coffee and plug your car into the diagnostic machine. Within minutes you know what the problem is, what the cost will be and how long it will take to fix, which is usually not very long. It's almost certainly better - cheaper - faster. What you don't do is say "noI want a mechanic to get a spanner and go under the bonnet."
We expect to take advantage of modern advances in diagnostics in other areas so why aren't we prepared to do that in health?
As I said before, it seems that people are worried the data will be misused and that the system can't be trusted.
Interestingly though, every single day millions of us go online onto sites like Amazon or ASOS in order to ensure that someone we love or a good friend, get the exact present they want for their birthday. In order to complete the order, we happily give all of our Financial data to a faceless computer program, located God knows where, without worrying. We give them our full name, address, phone numbers, we give them entire bank account details - including that supposedly secret three digit code on the back of the card, despite everyone telling us that we should never tell anybody - and we do it because we TRUST that system. And yet if your loved one was diagnosed with a life threatening or debilitating condition you would be screaming for that data to be used to give them the best chance of a successful and better health outcome.
It's bizarre that we don't seem to trust health data when it actually has the potential to do the most good.
We have to get the message over that unleashing the power of data in the NHS to improve individual care, system monitoring performance, research and the ability to create a more sustainable and affordable system is absolutely vital and it's the basis of what will keep the nation healthier for longer. They tell us that one in three people born post 2000 will live healthier lives to 100 – that's a brilliant target but if we don't start utilising an asset like data properly and fully, it just isn't going to happen. The Empower: Data4Health campaign intends to continue working with partners such as Google Deepmind to ensure that it does. Thank you