The sums involved are modest by NHS standards. Seven grants, the largest £132,000, the smallest £36,000, awarded through the North West Secure Data Environment's accelerator fund to trusts and university teams working on mental health records, cancer data and neuroscience datasets. Cheshire and Wirral Partnership will use its £83,000 to link mental health records with GP, hospital and community data, giving researchers de-identified access to material that has never before sat in one place. Set against the £330 million Federated Data Platform contract and the controversy that has followed it through much of this year, the North West's awards barely register.

In July, NHS England was forced to correct a data protection document that had wrongly stated only NHS staff could see identifiable patient information on the Federated Data Platform. Three Palantir engineers, it transpired, held administrative-level access to the national tenant where data sits before pseudonymisation, with dozens more holding narrower project access. The correction came after months in which MPs, the British Medical Association and the National Data Guardian had already voiced serious doubts about public and clinical trust in the platform, doubts that predate this episode but were sharpened considerably by it. Fewer than a third of English trusts use the platform in any meaningful way, and some, including Leeds and Greater Manchester, have told NHS England directly that adopting it would represent a functional step backward.

The Secure Data Environment network operates on a different premise. Rather than a single national platform run through one supplier relationship, it is a federation of regional and national environments, each holding pseudonymised data under local governance, each answerable to the communities whose data it holds. The North West's accelerator fund is a small mechanism within that architecture, designed to seed novel research uses rather than to consolidate control. It sits alongside a parallel set of developments this year: the Department of Health and Social Care's engagement programme asking GPs and practice managers how patient data should be used in their own record systems, the EU's UNITE fund backing cross-border data projects on heart failure and neonatal care, and the arrival of the Zenith supercomputer at Cambridge, built to process health data at a scale previous infrastructure could not manage.

None of this amounts to a rejection of the Federated Data Platform, and nobody in government has said it should. But the coexistence of two distinct data strategies, one centralised and increasingly contested, the other federated and comparatively unexamined, is not something ministers have been pressed to account for. The Data for Research and Development programme that underwrites the Secure Data Environment network was designed to complement the FDP's operational focus with a research-access layer of its own. In practice, the two have developed at different speeds and drawn very different levels of scrutiny, and the government has offered no public account of how it expects them to relate to one another as both scale.

That silence matters more now than it might have a year ago. Public trust in NHS data sharing, already fragile, has been tested repeatedly by the FDP's difficulties, and the department has limited room to spend further political capital rebuilding it. The regional model, precisely because it is smaller, slower and more locally accountable, may be better placed to do that rebuilding than a single national contract ever could. If ministers have concluded that federated regional infrastructure is where genuine public confidence in health data will be won back, they should say so, and explain what that implies for the FDP's future scope. If they have not concluded this, the drift toward two unreconciled data strategies, one attracting headlines and hostility, the other quietly accumulating funding and functionality, will eventually have to be resolved, and probably not on terms of the department's own choosing.