Data infrastructure for Internet age healthcare: an introduction

This is the first of a series of posts exploring patterns that could help solve some of the biggest healthcare data challenges. I'll be posting a new post every day or so over the next couple of weeks. I'll update this page with links to the other posts when they are published.

0. Introduction
1. Data minimisation

Blockchain will save us all (!)

Over the past few days there has been much talk in UK government tech circles about using blockchain technology beyond Bitcoin.

This was triggered by the release of a report from the Government Chief Scientific Advisor that explores potential uses of blockchain in government. What piqued my interest especially was the mention of the application of blockchain technology to manage healthcare data:

In the NHS, the technology offers the potential to improve health care by improving and authenticating the delivery of services and by sharing records securely according to exact rules.

This report wasn't the first time blockchain has been suggested as a solution to some of our healthcare data challenges.

The idea has been floated in various articles, and by startups and big companies alike. Even I am guilty of giving it passing mention in my last blog post.

On the government blockchain report, there have been some sensible words written, so I won't reherse the arguments here (risks to privacy, techno-utopianist fetishes of the new, reinvention of wheels etc).

However, one comment by Jeni Tennison struck me especially:

The challenge comes first then the tool comes second.

I couldn't agree more. We should be in the business of solving problems, not searching for nails to hammer just because we can.

Problem or hypothesis driven design is a much better way to develop new ideas, products or services than solution driven design. It's the double diamond again.

Too often I've seen project doomed from the start because its sponsor wants a thing (read: usually an app or a vanity website) rather than an outcome.

Data challenges in the NHS

With that in mind, I've been thinking quite a lot recently about the healthcare data problems we need to solve.

At UKHealthCamp, the dominant conversation was not about digital transformation, apps or wearables, but - perhaps inevitably - data.

Following on from UKHealthCamp, Jen Persson published a fantastic exploration of some of the 'hard' data issues facing the NHS. It's a long read but well worth it.

I tried to sum up some of these challenges when I spoke at a UK Data Service event last month.

In the talk, I spoke of the need for a new data infrastructure that could start to tackle some of our data challenges such as interoperability, patient privacy and data sharing.

Data infrastructure is an awful piece of verbiage with no satisfactory definition. I think of it as all the tools, rules and patterns that we need to meet the needs of the users of data. It's not just technology - it is about how we design our services and interface with data.

This may seem like a dry topic, but it's important. As I've written about before, while the NHS has been busy going 'paperless', the rest of the world has moved to the Internet.

It's worth making the point that the current data infrastructure the NHS relies on is barely meeting the needs of today's users.

We live in the world of bulk data sharing, 'extractions' (sounds painful), opt-outs and complex, proprietary data standards. And it's not good enough.

How can we expect this kind of infrastructure to cope with the huge challenges the Internet is going to pose to the delivery of care over the coming years?

The patterns

In this series of posts, I'll showcase some of the patterns that could help address our healthcare data challenges; the kinds of things we need to do with our data to bring the NHS into the Internet age.

By patterns, I mean general, reusable solutions to a common problem. Things like how to ask permission to use personal data, and what to do with it once you've captured it.

I'll attempt to do this in as much Plain English and as little technical jargon as possible.

None of this is new. Some of these ideas have been around for a while. Others have been the subject of research and prototyping by the UK government over the past few months.

All these posts aim to do is explain the patterns, put them all in one place and speculate on how they could be used in health and care.

I will also focus on England's NHS. Sorry - I know there is a world of healthcare data beyond this but this is my main frame of reference.

Of course these patterns have a broader application, but I figured it was useful to think about their use in health as I've heard little public conversation about their use in this domain and healthcare seems to be the arena where challenges around data are felt most acutely.

As Richard Pope put it in his submission to the UK government digital strategy consultation:

there is a component missing from our national data infrastructure: tools and standards for managing and sharing that personal data that are understood and trusted by citizens.

My hope is that when the next UKHealthCamp comes around we can talk about how these patterns have been put into practice and how they're working. I'd like us to start exploring some of these in our work on NHS.UK.

Thanks to Paul, Richard and Tom for shining light on many of these.

Please get in touch if you have any suggestions, ideas or corrections.

Read the first pattern: data minimisation