Claims that Palantir has been the best thing to happen to the British NHS for ages have taken a kicking after trusts admitted the numbers backing the statements are riddled with more holes than Swiss Cheese.
Hospital trusts have admitted there is a litany of mistakes in the data behind one of NHS England’s favourite claims about Palantir’s technology after the Financial Times found the numbers were about as bad as me doing my seven times table in primary school.
NHS England claimed hospitals saw a 15 per cent fall in delays to discharging patients after adopting a tool built using the US company’s software.
However, four trusts confirmed errors in the underlying hospital data after the FT spotted sudden drops in discharge delays from hundreds or thousands to zero. The numbers then shot up again.
The errors undermine NHS England’s headline claim about the overall fall in discharge delays. That figure has been widely used as evidence that Palantir’s Federated Data Platform is making the health service more efficient.
The FDP is supposed to pull together scattered data across the NHS and turn it into something useful. Palantir’s role has become increasingly controversial because of its links to the US defence sector.
NHS England’s former head of operational research and evaluation, Charles Tallack, said the evidence for the FDP’s impact on discharge delays looked “increasingly flimsy”.
“The delayed discharge dataset may be suitable for day-to-day management purposes, but not for evaluation,” Tallack said.
The benefits data comes from NHS England’s national mandatory dataset, which tracks discharge delays across all trusts. Palantir has no oversight of that dataset.
NHS England’s own website admits the dataset has limits and says there are only “minimal” quality checks. Across four years of acute discharge data, the FT found irregularities for 42 per cent of NHS hospital trusts.
The findings landed a few weeks after the UK statistics regulator began investigating key claims about the FDP. That followed the FT revealing NHS England had added a caveat to several figures, admitting the data did not prove the technology’s effectiveness.
On Thursday, MPs on the cross-party health select committee wrote to ministers raising concerns about the platform’s performance claims. They urged the UK government to break the NHS contract with Palantir next spring using a clause in the original deal.
NHS officials have talked about pressure to prove the FDP’s benefits too quickly. One official said NHS trusts “are being asked to put their name to statements about improvements before the tools are fully embedded and before the evaluations are done”.
The FT has previously reported that several NHS performance improvements credited to Palantir came from only a handful of hospitals. Some trusts have delivered fewer operations since adopting the technology.
Appearing before Parliament’s health and social care committee on Wednesday, NHS England chief executive Sir Jim Mackey said he had personally challenged whether assessments of the platform’s benefits “have been objective and can be fully stood up if challenged”.
Mackey added that an “objective” review “would be helpful and necessary”, but warned it would take months.
MPs, NHS staff and medical trade unions have urged the health service to find another option before a decision on extending the contract next spring.
In one of his final acts as UK health secretary, Wes Streeting ordered an internal review before his successor decides whether to break the contract.
A spokesperson for NHS England said: “These statistics come from nationally collected NHS operational data which is regularly reviewed and updated to maintain data quality, and an independent evaluation of [Palantir’s] Optica tool also found it is freeing up more time for staff and reducing unnecessary days patients spend in hospital following treatment.”
