Ignore Palantir’s political opponents and look at the data: this technology helps patients
It’s all very well for a country that can’t build a railway, a reservoir or a functioning hospital discharge process to congratulate itself on its scepticism, but the Health Foundation is wrong – Palantir’s NHS tool works, says Louis Mosley
There is a ritual in British public life. A government programme shows signs of working. Its opponents, who objected on political grounds before a line of code was written, go looking for a study. A think-tank obliges. A newspaper amplifies. And a country that struggles to build a railway, a reservoir or a functioning hospital discharge process congratulates itself on its scepticism.
This week it is the turn of Optica to face this merry-go-round. Optica is a discharge tool built by an NHS body, North of England Care System Support, on the NHS Federated Data Platform, supplied by Palantir. The Health Foundation, a respected think tank that does serious work, published an analysis finding “no noticeable improvement” in discharge delays from using Optica.
The Health Foundation found this because it seemingly doesn’t understand what the tool was designed for, and therefore which patient cohort to base its study on.
Hospitals discharge two kinds of patient. Most discharges are simple: the ward sends someone home and there is no care to arrange. The guidance says these cases should move fast, and they generally do. But complex discharges are different. These are typically patients who have spent weeks in hospital. Someone must book them a care home place, order equipment, arrange transport and brief a district nurse. This required coordination across the NHS, the local authority, independent providers and charities. These patients are fewer in number, but they wait the longest.
Judge us on results
Jonathan is the kind of patient Optica was built for. “When I suffered life-threatening complications following a rare abdominal cancer, I wanted to get straight to hospital,” he says. “But once I started feeling better I just wanted to recover at home. The hospital and social care teams could all see and update my data so they could work together to see if they could discharge me more quickly.”
In a whole-hospital average, Jonathan vanishes. For long-stay patients like him, NHS England‘s analysis shows a 14.1–14.9 per cent fall in delay-days using Optica. Even when NHS England reran the Health Foundation’s own, different, method — but restricted to that same long-stay population — it still showed a positive effect.
Two different ways of counting land on the same conclusion: a real impact, even if the two methods disagree on its size.
The wider platform should be judged the same way. Trusts using its theatre tool have recorded 110,000 additional operations since they began using that tool. Similarly, Trusts using its cancer tool have seen a 6.8 per cent improvement against the 28-day diagnosis standard. The programme holds a green delivery rating, one of 30 among more than 200 major government projects, and is forecast to return £5 for every £1 spent.
Can we prove no other factor moved those numbers? No. And we say so in writing: since the statistics regulator’s review, every figure the programme publishes carries a causal caveat. Elsewhere, the benefits of virtual wards or electronic patient record systems, for example, carry none. And where a rigorous test of Optica has been run, the benefits held: an independent Imperial evaluation of the theatre tool found it improved theatre use.
Software cannot conjure a care home bed. Where social care capacity is the bottleneck, no technology can fix that. What ours does is make the problem visible: who is waiting, why, and where the blocker sits, so discharge teams stop losing days to spreadsheets and chasing by phone.
This is about patients: elderly and vulnerable people getting home sooner and in better health. We will keep backing the clinicians and care coordinators who arrange that care every day.