Key facts
- Internal NHS data suggests Palantir's Optica AI tool reduced hospital discharge delays by up to 28%.
- NHS England data shows a 14.1-14.9% reduction in delay days for long-stay patients using Optica, versus a 2.9-3.5% increase for non-users.
- The Health Foundation previously found 'no noticeable improvement' in hospital discharge performance linked to Optica.
- NHS England disputes the Health Foundation's methodology, arguing it included shorter stays not targeted by the tool.
- The government will review Palantir's £330m NHS contract before a February 2027 break clause.
- The broader Federated Data Platform (FDP) program's lifetime cost is projected at £1.1bn, with benefits estimated at £808m.
Internal NHS data indicates that Palantir's Optica AI tool has significantly reduced hospital discharge delays, countering recent claims of "no noticeable improvement." Trusts utilizing the tool saw a 14.1% to 14.9% decrease in delay days for long-stay patients, compared to a 2.9% to 3.5% increase at non-adopting trusts. For patients managed directly via Optica, NHS England calculated a 28% reduction in delay days.
These figures, compiled late last year, respond to an analysis by the Health Foundation, which had questioned the effectiveness of Palantir's Federated Data Platform (FDP). The Health Foundation's findings of 'no noticeable improvement' were based on a broader comparison that NHS England argues diluted the tool's impact by including shorter hospital stays for which Optica is not primarily designed.
The disagreement highlights differing methodologies for measuring the software's effectiveness. The NHS document remains unpublished, and its findings do not definitively attribute the reductions solely to Optica. The debate centers on how to measure the software's impact and which patient groups should be included in the analysis.
This situation arises as the Department of Health and Social Care prepares to review Palantir's £330m NHS contract ahead of a break clause in February 2027. A DHSC spokesperson confirmed the standard review process but defended the program, stating thousands more patients benefit monthly from the FDP, including through reduced unnecessary hospital days.
Further scrutiny is expected as official figures reveal the wider FDP program's lifetime cost has risen to an estimated £1.1bn, while its forecast financial benefits have been revised down to £808m. NISTA attributed the cost increase to scaled delivery and updated rollout plans for the reduced benefits. Despite these adjustments, the FDP maintains a green delivery rating from NISTA.
Palantir's Optica tool is designed to coordinate the discharge of patients requiring additional care. Palantir argues the Health Foundation's analysis compared 'apples and oranges' by including short-stay patients. Tom Bartlett, a former NHS data engineering deputy director, suggested issues with both methodologies and called for ward-level analysis. Dr. David Nicholl of Doctors' Association UK expressed skepticism about NHS England's FDP figures, citing staff concerns about benefits over existing software. The Office for Statistics Regulation is also reviewing separate FDP-related casework.