Disputed Trust Figures Disappear From NHS Data Platform Website – Distilled Post

For a time, a page on the NHS England website carried a list of 137 hospital trusts described as live with the Federated Data Platform and realising benefits from it. Then 29 of those trusts told researchers that the system was not live in their organisation or that they used none of its applications. The list came down. A shorter page remains, saying that early users are already seeing real results. It names no trusts and gives no figures.
A deleted webpage would ordinarily be a matter of housekeeping. This one formed part of the case for a £330m contract that ministers must now decide whether to extend. The initial term ends in March 2027, with a break clause exercisable the previous month, and reports suggest ministers favour a further period running to 2029 while a UK-based alternative is prepared. NHS England has said its review will weigh uptake and value for money. Uptake is the obvious starting point, and it is the measure the withdrawn page can no longer support.
The page is not the first difficulty. The Office for Statistics Regulation had already warned that there was no evidence behind the claim that the platform enabled more than 100,000 extra operations, and it is now examining the adoption claims. Two overstatements in succession carry a cost that outlasts either correction. Ben Coleman, a Labour member of the health select committee, argues that overclaiming could put hospitals off using the system and threaten its viability. His point holds whatever one thinks of the supplier. A platform that depends on trusts choosing to use it cannot afford a promoter whom trusts cannot rely on.
Consider the position of a trust chief operating officer. Waiting lists remain the government's most visible test, budgets are tight and analytical teams are small. Time spent configuring a national tool is taken from local work, and the decision to spend it rests on whether peers report gains in theatre scheduling or discharge. A national list that overstated live use hands that officer a false benchmark. Trusts that adopted early and saw little return have no incentive to say so publicly, and those that held back may wrongly conclude they are laggards. The operational learning the platform was meant to speed up depends on honest reporting of what is actually being used.
The politics make the evidential gap more consequential. Palantir says that ending the contract would put politics above patients. It has hired a Labour peer as a full-time lobbyist, and he has written to the health secretary offering help to build sovereign technological capability. Unison, Doctors' Association UK, Medact, Amnesty UK and a patients' group oppose the firm on grounds of data sovereignty, security and its work for the Israeli military and US immigration enforcement. Both camps appeal to values, which is natural when the benefit to patients has not been established to a standard anyone can check. In that vacuum, lobbying and campaigning will carry the decision, which is a poor basis for committing public money for two more years.
The exit question is also partly contractual. Campaigners argue that a termination assistance period would allow trusts to move away in 2027, making an extension unnecessary. A plan to extend while building a UK replacement may be defensible, yet ministers would invite the charge of saying one thing and doing another if the extra years deepen dependence in the meantime. Whether they do depends on usage data that NHS England has just shown it cannot yet supply.
Several practical consequences follow. Ministers should require an audited, trust-by-trust account of use before deciding, published with its method. NHS England should treat statistical scrutiny as a design requirement for every claim it makes. British health-tech and life sciences firms, whom the government wants anchored in a sovereign data strategy, need open standards and interoperable architecture to compete for work, and Coleman's proposal of an open platform on which trusts and domestic companies build their own tools speaks directly to that. Patients, whose consent underpins every data ambition in the life sciences agenda, will notice if official claims keep failing checks. Reports that Palantir lobbied for its tools to reach into GP records sharpen that sensitivity.
The scrutiny began with freedom of information requests sent to every hospital and integrated care board in England by Medact. It took a campaign group, not the health service's own monitoring, to establish how far the platform had spread. That sequence is itself a finding about how the system checks its own claims.
A platform built to give the NHS one reliable view of its operations has been promoted with an unreliable view of itself. A revised webpage will not repair that. It requires figures that trusts, regulators and the public can test. Until they exist, any extension will be a decision made on assurance alone, in a health service that has already spent much of the credit that assurance once carried.
‍

source
This is a newsfeed from leading technology publications. No additional editorial review has been performed before posting.