NHS England has not yet proven its case on FDP

With the award of the Federated Data Platform contract said to be delayed until mid-November, NHS England appears to be seeking to drum up support for the FDP and secure public backing from health system leaders for its £480m flagship data platform project.

The belated efforts to win public endorsements comes just weeks before the expected award of the FDP contract, the selection process for which has been completed.

So far though, public support for the project outside of NHSE and some of the NHS pilot sites has been limited and lukewarm.

Earlier in October, 14 medical directors from trusts involved in FDP pilots agreed to put their names to an NHSE open letter supportive of FDP.

Digital Health News understands Integrated Care Board chief officers are now being asked to sign a similar letter, with suggestions that some are proving reluctant to do so, and with only about half of the 42 ICB chiefs thought to be willing so far to add their names.

If true, this would suggest significant reservations exist among ICB leaders, particularly concerning as ICBs are meant to be some of the main future users of the platform. At the national level it’s also striking that none of the Royal Colleges, the BMA or big medical charities have publicly backed the programme.

In a June BMJ article BMA deputy chair Dr David Wrigley wrote, “The British Medical Association is calling for NHS England and the Department of Health and Social Care to have an open discussion with us clearly outlining the scope of data they plan to collect, how exactly they plan to use it, what role Palantir will play and how it meets the NHS’s core values.”

This call for greater transparency echoed some of the questions raised by the Faculty of Clinical Informatics (FCI) in a position paper published earlier this year, that put FDP into the context of past and current data sharing arrangements and initiatives within the NHS.

“Overall, there is a general lack of transparency about many aspects of the FDP which is reminiscent of the recent history of Care.data, GPDPR [General Practice Data for Planning and Research]… There is a need to build trust by taking concrete action on privacy and transparency; trust cannot be earned through communications and public engagement alone.”

The FCI was not thanked by NHS England for its well-informed position paper, being told the paper was “unhelpful”.  Some FCI members even believe that publication of the paper contributed to the FCI subsequently losing a previously promised NHSE grant.

In a new NHSE blog post attributed to Vin Diwakar, director of transformation NHSE, Professor Stephen Powis, medical director NHSE and Rachel Power, chief executive of the Patient Association (PA), promises to build on research by the PA to develop a new pact with patients on how their data is used is promised.  Though the PA stops short of endorsing FDP.

FDP scope not clearly defined

Perhaps the key recurring problem with a mission to win hearts and minds on FDP is that there has been lack of consistency or clarity on what FDP is for and what its scope covers, and the confusion appears to grow the more NHSE seeks to explain.  This has very direct parallels with Care.data where the scope and purpose of the project and who would have access to patient data repeatedly changed, fatally undermining public confidence.

NHSE’s original published material on FDP said that it was to cover five areas:

1. Elective recovery – to address the backlog of people waiting for appointments or treatments.
2. Care coordination – to enable the effective coordination of care between local health and care organisations and services, reducing the number of long stays in hospital.
3. Vaccination and immunisation – to continue to support the vaccination and immunisation of vulnerable people while ensuring fair and equal access and uptake across different communities.
4. Population health management – to help integrated care systems proactively plan services that meet the needs of their population.
5. Supply chain management – to help the NHS put resources where they are needed most and buy smarter so that we get the best value for money.

But it now appears that these were rather loose initial use case scenarios and the longer-term ambition is much wider, having previously been described as a ‘future operating system for the NHS’. More recently, it is being described as software to join up the NHS and enable direct clinical care.

Addressing the October NHSE board meeting Dr Diwakar described FDP as “software that enables you to view data from different sources in one place on your computer screen.  That makes it easier for you to manage your patients, whether that is bringing down waiting lists or making it easier to discharge patients from hospital, so all caregivers have access to the same information,” clearly indicating that it was envisaged as a tool clinicians would use to manage their patients.

The new blog post further suggests that it will mean patients’ “telling their story once”, suggesting something approaching an electronic patient record.  A laudable aim but a very significant scope creep, taking it far beyond being about analytics and resource coordination and into direct clinical care.

It is precisely this lack of clarity of purpose, and apparent mixing up of very different data types with specific and different legal and information governance requirements, that is fuelling anxieties.  Responding to the new blog clinical informatics veteran Dr Ian McNicholl posted on X, formerly Twitter: “It is mixing up operational service data e.g. bed state, with direct care data, such as people telling their story once, which I understand was not in the scope of FDP.”

Ele H meanwhile posted on X: “NHSE is conflating direct patient care and secondary uses of healthcare data which means that either the scope of FDP has drastically changed compared to the tender documentation or NHSE doesn’t know the difference.”

The same NHSE blog says the purpose of the FDP is to “enable NHS organisations to bring together operational data – currently stored in separate systems – to support staff to access the information they need in one safe and secure environment. This could be the number of beds in a hospital, the size of waiting lists for elective care services, or the availability of medical supplies”.

Limited evidence on benefits

A further problem undercutting the case for FDP is the lack of openly published peer-reviewed evidence on benefits.  In presentations and blogs NHSE has cited a range of anecdotal benefits from pilot sites, and extrapolated hundreds of millions in potential savings that could be achieved, but not published any hard evidence to back this up.

Examples from the NHSE website of benefits claimed include:

“Chelsea and Westminster NHS Foundation Trust has used new software in its gynaecology department to track patients with suspected cancer through to diagnosis and first treatment, which has reduced the waiting time for a first appointment by an average of two days and cut the overall time to diagnosis.”

This sounds a tremendous result for patients with suspected cancer, but there is as yet no indication of how software helped reduce time to first appointments.  Did it do so in isolation? What part did accompanying process improvement and staffing play?  How can others achieve the same benefits?

A further example of benefits achieved is given as North Tees and Hartlepool NHS Foundation Trust, where a patient flow application called Opica developed on the Palantir Foundry platform, is said to have “freed up almost 10 hours of clinician time each week and resulted in patients going home faster, with long stays (over 21 days) falling by more than a third (36%).”

Ukauthority.com on quoted Ayub Bhayat, NHSE’s director of data services and deputy chief data and analytics officer, forecasting that extrapolated on a national scale it [the application] “could save 1.8 million delay days over 21 days per patient, equivalent to £700m”.

Again, the results certainly sound impressive, but they appear so far at least to be largely anecdotal.  What was the cohort of patients studied? In what specialities? Is it realistic to extrapolate such big savings from a small number of pilot sites?

Another pilot cited is University Hospitals Dorset NHS Foundation Trust, where “teams are using the software to align their waiting list with information such as staff rosters and annual leave to improve scheduling for patients and staff”.

This is said to have “enabled them to see extra operating theatre capacity, which is being used to provide care to those patients who are most in urgent need as well as the longest waiting patients”.

Once again the evidence of these benefits and how they have been achieved is anecdotal so far.

A source close to Guys and St Thomas’, another pilot site, say there is no doubt there are benefits being achieved, but acknowledge they have yet to be systematically captured or written up.

The problems of building evidence and developing and publishing quality studies and taking significant time and effort is by no means unique to FDP.  It is a common feature of evaluating all digital health initiatives.  But just because it’s difficult and time-consuming doesn’t mean it should be bypassed.

The reality is that objectively FDP should still be considered very much firmly at the pilot stage and developing a solid body of published evidence and evaluation will take time.  While early results sound promising, there is no indication that there is sufficient evidence of benefits to yet proceed with awarding a half-billion pound contract, whatever the political pressure to do so.

FDP may or may not prove an important development, it is too early to tell.  But in a system as complex and varied as the NHS anyone offering a near magic data elixir should be treated with scepticism and have to convincingly prove their claims with robust evidence that can be independently verified.  At this point FDP rather appears far more a data panacea for NHSE, and possibly even dangerous snake oil.

If public trust is to be gained and the NHS is to have a realistic chance of gaining the potential benefits of new data platforms and realising the value of its rich data assets, then NHS England and the Secretary of State should hold off on any contract aware and take more time.  They should have the courage to invest in proper evaluation of the pilots, dissemination of lessons and findings and then use this as the basis for engaging with patients.

And rather than seek to get rubber stamp names on letters of support from NHS leaders as a fig leave of communications and engagement, they should actively set out the case and win public confidence.  Failing to do so in advance of a contract award risks repeating the mistakes of Care.data and poisoning the well of public trust in use of NHS data at scale for another decade.

However, if the Royal Colleges, BMA, Patient Association and medical charities can be convinced to support FDP, the public and professions will surely not be far behind.