“Unachievable” care.data programme should serve as warning to Government-as-a-Platform agenda
- Summary:
- The Major Projects Authority has given the NHS's care.data programme a red rating – and its failure should serve as a reminder for challenges facing government-as-a-platform.
And whilst this is interesting in itself, given the unique nature of the care.data programme and the national attention it has received, it's also important to consider the wider implications its possible failure has for the success of the Government-as-a-Platform (GaaP) agenda.
Care.data aims to pull together all of the UK public's GP health records, store them in a central database and share them with the Health and Social Care Information Centre (HSCIC). The NHS believes that the sharing and analysis of information across the NHS will help to ensure that the quality and safety of services is consistent across the country.
It also believes that care.data will highlight different diseases and conditions that may require more NHS investment going forward.
However, the NHS ran a poor PR campaign surrounding care.data and privacy advocates successfully argued that although the data being stored would be pseudonymised, where it would only reveal details such as age, location and gender, this wouldn’t stop the data being matched with other databases and patients being ‘re-identified’.
As concerns were mounting at the beginning of the year, a leaked NHS document outlined that the data could potentially result in patients being identified. The nail in the coffin, as it were.
And although the NHS argued that it had carried out a publicity campaign by delivering leaflets on care.data to 26 million households across the UK, polls found that two-thirds of people had not had seen them.
Despite all of this, NHS director for patients and information Tim Kelsey said at the beginning of the year that the programme was far from dead and that the big data project was in fact a “moral obligation”, given the impact it could have on healthcare and people's lives. He said:
There is an endemic problem of failure to use existing datasets, which is harming people and is not contributing to efficiency.
However, the MPA doesn't hold out much hope. A red rating from the watchdog means “successful delivery of the project appears to be unachievable” and “there are major issues with project definition, schedule, budget, quality and/or benefits delivery, which at this stage do not appear to be manageable or resolvable”.
With specific reference to the care.data programme, the MPA said NHS England needed to “clarify, agree and communicate the programme scope; appoint a full-time Senior Responsible Owner; reconstitute the Programme Board with a clear role and responsibilities; approve explicit go/no go criteria; agree and clarify finances; assign owners to key risks; and recruit key personnel”.
Basically, there's a lot to do.
A warning
The MPA's report and red rating for care.data, however, should also serve as a warning to the Government Digital Service (GDS) over its government-as-a-platform (GaaP) agenda. A core element of getting GaaP right is the governance of public data and the introduction of data sharing schemes. Although this hasn't been spoken about much yet in public, it's my understanding that the data discussion will happen and is central to GDS's plans for GaaP.
GaaP aims to introduce common systems for re-use across government, which will inevitably cut down on costs, introduce efficiencies and 'standardise' public services. GDS has already done this to some degree through the introduction of GOV.UK and its upcoming Verify system – but ultimately these are low hanging fruit.
If the government really wants to take advantage of the benefits achieved through GaaP, it needs to get its data governance right, so that systems can be built quickly and easily on top of flexible data platforms. As far as I'm aware no firm plans have been agreed yet, but discussions are happening around the structure of this.
And new Cabinet Minister Matt Hancock has already started dropping hints that this data discussion needs to take place. At a conference last week in London, Hancock said:
For the first time, we are in a position to build digital foundations: made of data not paper, holding up platforms not silos. Common registers, common payments platforms, and common license systems, all based on common data standards.
This digital infrastructure is the modern equivalent of the canals and railways that made industrialisation possible. And once it’s established we can build platforms on top: integrated public services with the citizen at their heart.
However, this won't be an easy conversation to have. The problems associated with care.data are largely down to the public's sensitivities around the government's use of their data, with many untrusting of data sharing policies.
Care.data has struggled because of a lack of transparency and a naivety on NHS England's part that the public would just accept projects of this scope. GDS needs to learn from the NHS's mistakes and realise that if it wants to make GaaP work and do some truly innovative things with data, it needs to consult the public, explain the benefits and be completely open with its plans.
My take
The public doesn't trust the government to do the right thing with its data. And yet, I believe, data
governance and sharing are essential to the true vision of GaaP.So, something needs to change. That trust needs to be won back. This will likely be done by way of one of two options – either by making the benefits so great to users that they outweigh the risks, or by giving people transparency and control. Ideally, users would like both.
I would argue that what GDS needs to do is consult on its plans and explain what it needs to do to make GaaP a success. Instead of ploughing ahead, ask users and experts what works for them. Involve people from the start.
And then give people control, or at least a view, of what's happening to their data. Who is using it? Who can see it? In what form is it being shared? Distrust comes from the unknown. Involve people with the decision making process and give them some control and success will come a lot quicker.