Connect with us


The NHS data strategy – what should a second draft look like?



The NHS’ IT agency, NHSX, has issued a draft data strategy for health and social care. Matthew Chase, chief technology officer Healthcare UK at Fujitsu, has read it and thinks there are three key issues that a second draft needs to address


The draft NHS data strategy is a huge document, running to 65 pages by the time you reach the end of its six chapters, numerous case studies, and three annexes. Yet the first question I found myself asking at the end of all these thousands of words was: is this a strategy at all? 

Sections of ‘Data saves lives: reshaping health and social care with data’ read more like a vision, for which a plan will be published in due course. Other sections contain specific targets but are missing the plan that sits behind them to explain why they have been chosen and the detail of what will be delivered. 

When we get to a second draft, I hope we will have more clarity about what is being presented. What is the big picture? What is new and what will build on what has been done already? What is the timescale? And what are the priorities within that timescale? 

Because we can’t do everything. So, if we know that this Parliament will last until May 2024, and the Treasury is about to launch a two-year comprehensive spending review, then what should we try and do between then and now, using the money that might be available? Or, to put it another way, what is our minimum viable product? 

The strategy’s minimum viable product must start with clean data

Reading between the lines of the strategy, I think our MVP will have three features. The first is shared care records, to support integrated care and capture data for population health management. The second is a secure data repository to power national apps and research that can support care and save lives. 

And the third is patient control of data, both to maintain confidence in the system and to drive the development of patient portals, remote monitoring systems, and digital front doors, which are the hottest thing in health tech right now. 

If we are going to deliver this MVP within the strategy’s timescale, then I think there are few more things that a second draft will need to do. The first is to address data quality. This is barely mentioned in draft one, and yet it’s critical. 

Our shared care records, national data repository, and patient-held apps are all going to be drawing data from a mix of hospital, community and GP records. Plus, increasingly, they’re going to be working with data from social care, where systems are very different, and with information from medical and personal devices. 

All that data will need to be cleaned if we are going to have confidence that it is being matched to the right patient and is giving us an accurate picture of their condition and response to treatment. Otherwise, as the old saying has it: garbage in, garbage out. 

Standards, yes: but what about legacy systems and promotion? 

The draft of ‘Data saves lives’ comes close to recognising this when it talks about developing and promoting standards for data and information exchange. That’s absolutely the right way forward, but we should recognise how far we have to go. 

As things stand, we have standards for some data items, such as SNOMED CT for clinical coding, or the dm+d for medicines and devices. We have some standards for structuring documents, thanks to the work of NHS Digital and the Public Standards Records Body. 

And we have some standards for exchanging information, thanks to the work of organisations like the HL7 Foundation and DICOM. However, that’s not a complete list of all the standards we need, and we are a long way from being able to put them all together. 

Fast Healthcare Interoperability Resources, or FHIR, has taken on that job, but it is still in its infancy, and lacks a UK-specific flavour. Plus, FHIR can’t address the legacy challenge. 

Take a deceptively simple data item that we might want to ‘get’ for an SCR or a research database, like ‘allergies’. FHIR can define ‘allergies’ but our existing systems are not using FHIR, so they will all define ‘allergies’ in different ways, and some will store ‘allergies’ as a specific data field, while others will incorporate it within the patient notes. 

So, when we get to draft two of the strategy, it would be nice to know how that persistent legacy challenge will be addressed. We also need to know more about who is going to develop standards going forward and what the role of industry will be. And we need to know who is going to address the second part of the equation: promotion. 

Because, as anybody who has worked in this space for any length of time will know, we have been trying to get the NHS Number adopted as a single patient identifier since 1998, and it hasn’t happened across all organisations and all systems yet. 

Where is the architecture for security, access control and consent?  

The third thing that I think a second draft will need to do is to put a bigger emphasis on security and on patient consent. That big corpus of data to fuel central apps and research is exciting, but in the real world it is going to be a target for hackers. So we need to be ready. 

We also need to know how control and access to different parts of the record and to their presentation in open, pseudonymised and genuinely anonymised data sets will work. I don’t mind people knowing plenty of things about me from my medical history: like my name or my date of birth, or even that I’ve had a Covid-19 vaccination. 

But there are other things in my notes that I don’t want anybody except the clinician treating me to know and, at the moment, the strategy doesn’t set out how that confidentiality will be respected. One principle that might help would be to leave information where it is whenever possible, but to create a catalogue that analysts and researchers can use to locate the data they need, with guidance on how it is protected.  

Whatever solutions are adopted, when the second draft comes out, we need to know what they are – and to be able to explain them clearly and simply to the public. 

Let’s think bigger: let’s think Citizen Z 

There are some good, individual points in the draft strategy. From my perspective as a former NHS chief technology officer, I like its commitment to addressing the burden of data collections by setting up a body to review and streamline the huge amount of information that is demanded from trusts. 

From my perspective as an innovator working for Fujitsu, I like its commitment to sorting out information governance, and making sure that if we deliver a project for one organisation we can deliver it for another without being asked to meet completely different requirements. 

Overall, it is also good to see the NHS maintaining its commitment to collecting data and to using it in new and innovative ways. I love the idea of creating a corpus of data and making it available to leaders and researchers. I think the concept of building an ecosystem of providers around that to do things for the country is fantastic.

What we need now is a draft that clarifies the vision, sets out a strategy with priorities and timescales, and indicates who will do what, when. And one final thought: this is a very health focused document, but as individuals, we are not just healthcare users or potential users. We are also citizens. 

Fujitsu has a vision for Citizen Z, an individual who uses digital to get a consumer-style experience from government departments and public services of all kinds. It would be great if this strategy could be aligned with others that are being developed in Whitehall, so it’s not just healthcare specific, but part of a wider vision for digital interaction with the services that we all need to support ourselves and our families. 


Continue Reading
1 Comment

1 Comment

  1. Pingback: Why social care needs to catch up with Big Data - Health Tech World

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending stories