Why an open health platform is inevitable
This blog is intended to approach the concept of an open health data platform (or platforms) from a slightly different angle, using economics to argue that the arrival of an open health platform is inevitable, and that we the UK are ideally placed to get on and revolutionise the world by building the first one.
Why Information destroys market theory
Not many people realise it, but we are in a new, unprecedented and revolutionary age. Not another type of industrial revolution, but a different type of revolution altogether. An age brought about by what is probably very dear to most people reading this blog- IT. Or more specifically- information. I’ll hopefully do a decent job of explaining why information is revolutionising our world in a way that society has never experienced before. And what it means for healthcare informatics.
Economics has an enormous amount to say about how our world works and so it is a good place to start to try to understand why information is changing everything. Bear with me on this, it’s worth it as a primer to understand where we’re headed.
If you look back in history, economics has been driven by the core principle of supply and demand. The more you produce of something, the cheaper it gets and vice versa. It’s a fairly simple mechanic to understand and underpins the entire economic discipline.
Now enter computers over the last few decades of the 1900’s. Followed by networking and the internet, meaning that information can be reproduced and distributed almost without limit. And at incredibly low cost. The relatively very recent combination of digitisation and networks is changing everything.
The incredible, mind blowing (and oddly under-publicised) fact is that the ability to produce, reproduce and distribute information almost instantly and at massive scale completely blows apart the supply and demand model. Previously, the supply/demand formula has always balanced itself- scarcity increases, costs go up, scarcity decreases, costs go down.
But information is different, because it is conceptually infinite. It doesn’t wear out. So logically, if supply is infinite, then the value of the product must be zero. That’s the way the supply and demand formula works. This is obviously not good news if you’re producing anything that can be digitised (which is more or less everything). Everything you produce can be made free by digitising it. It’s also not good news for capitalism in general but that’s a separate topic you can read about here.
How people try to stop information destroying markets
If you’re making something that can be digitised nowadays, how do you stop your product becoming worthless? How do you ensure sufficient scarcity of something that can be infinitely reproduced? How do you monetise something that is essentially free? The answer has developed hand in hand with information- it’s IPR, proprietary data models and algorithms. Literally the only thing stopping the entire Beatles catalogue being distributed to everyone in the world with a networked device for free is copyright. The same for every single song, film and piece of digital information ever produced. The simple inescapable fact driven by the supply/demand model is that if you want to make money from any sort of data, you need to control supply, you need to keep it locked away.
It is worth noting at this point that IPR is a crumbling dam, it can’t possibly hold back the flood forever. Information wants to be free. All it requires is one tiny leak in the dam. Once a piece of data escapes, it will disappear and reproduce across the network. The term ‘viral’ is very apt. If you don’t believe me just look at the levels of film and music piracy, or levels of corporate hacking. Over time, the IPR dam is being continually eroded by today’s networked society and is leaking more and more over time. At some point, the dam won’t be worth maintaining and will be abandoned, but that’s a way off.
So for now, we have IPR. Which is totally fine really isn’t it, I mean if someone writes a classic song or produces an Oscar winning film, of course they deserve to be rewarded for their efforts, and no-one would do that for free right? They created the data, they decide on who makes money from it and how and when it is distributed and to who. Makes sense.
Monetising Health Data
Now let’s look at our health data. It’s data, it can be shared and copied infinitely. It should therefore have no value. So why isn’t health data everywhere and free to use (pending our consent of course)? Well, as explained above, people are monetising health data by locking it away. Inside proprietary databases. They create and manage scarcity by ensuring other people can’t get at it. I mean it’s the perfect business model, demand for health data is going through the roof. The King’s Fund reckon that by next year the amount of volume of healthcare data will double every 73 days.
So you have more and more data pouring into your lake. And increasing demand for that data. And you are the one controlling supply. The current major hoarders of health data are basically OPEC, they can control supply of health data and therefore its value. Except healthcare data isn’t going to run out like oil is. It is literally the perfect market
I understand that yep, health data can’t be sold without consent. The market doesn’t yet fully exist where people buy and sell health data in the same way as people trade information scraped from cookies. But that doesn’t mean health data doesn’t have value or isn’t increasing in value or that the market isn’t coming into existence. A recent Great North Care Record survey showed that 46% of those surveyed would share their information with medical companies. The larger the data sets become the higher the value. GlaxoSmithKline recently paid $300m to partner with 23andme, and their genome database for example.
But even if we didn’t and don’t consent to our data being sold, suppliers are still monetising data by locking it away. If our health data was freely available, would you need to buy Cerner, Epic, SystmOne etc? No, you absolutely wouldn’t. There would be a myriad of systems out there instead of the increasing monopoly of a dwindling number of supplies.
Interestingly, economic theory also predicts this monopoly- as data becomes increasingly abundant, it declines in value. So in order to be profitable, you must monopolise that data in order to secure sufficient volumes of it. If that sounds a bit unsettling, it should. Google monopolising search and mobile phone OS, Microsoft and Windows and Office, Netflix and film, Facebook and Twitter in social media are just examples.
But hang on, healthcare system suppliers are monetising our data aren’t they? They’re not the Ed Sheeran’s of health data are they? They didn’t create it, carefully craft it into something valuable. They actually put zero effort into actually creating it (that credit probably goes to your parents and NHS admin and clinical staff). Yeah, they engineered a way to create it and a receptacle to collect it, but by locking it away they are monetising and profiting from your data. Every time a clinician types your data into a proprietary system, the system supplier profits from it. Did anyone ask you if that was ok? Did they seek your permission?
You could say It’s an identical model to Facebook and Google, and every single website that stores your cookies. You let them store and sell your information in return for a more tailored online experience in general. But you don’t have a choice about your healthcare data being entered into a data prison. It’s an interesting moral question, should these companies profit from your data without your permission?
Interoperability- Prison windows
What about interoperability I hear you say? There’s a big drive towards that in the NHS, and things are looking hopeful with Hadley Beeman involved and NHSX’s stated objectives. But if you look at nearly every interoperability initiative, they are just providing a portal which can access and surface information that’s held within the proprietary databases. It’s the equivalent of a window into the data prison. You may get to see what’s inside, but you can never take it outside of the prison.
True interoperability would be where anyone could access any of the data held in any system (consented of course). This would depend on there being a standard and open clinical data model. Of course open data models exist, OpenEHR is one of the more high profile examples. Ade Byrne is doing some great stuff in Southampton. The challenge is that no one model is dominant over another, and certainly not competing with the proprietary data models of suppliers at this time. As described earlier, the trend is that open technology is becoming increasingly prevalent, information wants to be free and leaks out into the commons. And once it is there it can never go back. But this process is slow, and the monetisation of health data has some legs left in it yet.
But the prize for establishing a dominant open data model is staggeringly huge. If one existed, it would almost instantly destroy the current EPR market. Just look at Wikipedia. It has permanently destroyed the encyclopaedia market. Forever. To compete with it, you would need to somehow need to outcompete the output of 12,000 contributors who all work for free. It is the 5th most popular website in the world with more hits than both Amazon and Twitter. And made for free by the public with no motivation other than contributing to society.
There is potential to create the Wikipedia of health care data. Just imagine an open database of healthcare data (consented and secure of course), freely accessible to anyone that wanted to access it, from individual statisticians to drugs giants. The insight that could be gained from the power of the public working on it. You would instantly and permanently have the monopoly as a data platform. Imagine the explosion in innovation as people build software and algorithms on top of the platform. Not to mention saving 11% of the NHS budget. Of course it’s not quite as simple as all that, but the concept is absolutely valid and achievable. It would be truly world changing.
The challenge is, how do you push the development of an open platform so that it gains dominance over the proprietary systems and we reach that tipping point? How does what are currently a handful of companies and a few hundred people working in their spare time on Open Source health outcompete the thousands of staff working for EPR suppliers? Who has the resources to do this, and the motivation to spend money for the public good? Who is brave enough to try on a grand scale?
NHSX, D and E
The only answer is the government. They are the only body capable of doing this. This isn’t a novel or crazy idea at all. The government strategy is absolutely to open up public data. Government open data platforms exist all over the place. The range of data the UK government has opened up is both astonishing and inspiring.
So the question is, why aren’t we doing the same with health data?
Yes, it’s harder and it’s incredibly complex, but the rewards are literally unprecedented. It will happen anyway eventually, the move to open data is inexorable. It will happen and is already happening. AI is open source already. How many people knew that Android is an open source platform by the way? We have a single patient identifier, and some of the top healthcare informatics brains in the world. So why not us and why not now?