Intended for healthcare professionals

Opinion Primary Colour

Helen Salisbury: Continuing disquiet on the data front

BMJ 2023; 382 doi: (Published 01 August 2023) Cite this as: BMJ 2023;382:p1761
  1. Helen Salisbury, GP
  1. Oxford
  1. helen.salisbury{at}
    Follow Helen on Twitter: @HelenRSalisbury

People talk to their doctors in confidence, expecting the details they share about their weight, alcohol use, terminations of pregnancy, or thoughts of suicide to remain in the consulting room. They also assume that if they need emergency care in another part of the health system the details of their past operations, current medicines, and allergies will be available to the doctors treating them. Although many patients are happy for their medical information to be used for a wide variety of purposes, other people wish it to be used only for their personal direct care, even if it’s anonymised.

Researchers working on new treatments for old diseases—or perhaps repurposing old treatments for new diseases, as happened with covid-19—need access to datasets that are as rich and comprehensive as possible. Planners tasked with predicting future needs and providing for them in the NHS also require data. A contract is currently being advertised to provide data services to the NHS, and the frontrunner—widely expected to win the tender—is the tech and analytics company Palantir.12

The NHS has a long and inglorious history of bungled IT procurements, with a repeated pattern of misplaced confidence in suppliers that were ultimately unable to deliver. The ambitious National Programme for IT in the NHS was dismantled in 2011 after it had spent £9.8bn and failed to win the support of doctors or produce meaningful results.34

More limited attempts to bring data together for research and planning have foundered through a lack of patient trust. The fiasco of 2014 saw patients and GPs alike opting out of data collection because they weren’t reassured about who would use the information and for what purposes.5 Lessons were seemingly not learnt, and those events are closely mirrored in the currently stalled programme on General Practice Data for Planning and Research.6

Many unanswered questions surround this latest IT procurement. Will it work? The news from health trusts already using Palantir’s Foundry software isn’t encouraging, with 11 pilot projects either paused or suspended.7 Complaints include the fact that the tech wasn’t user friendly and didn’t meet the trusts’ needs.7

Will it offer value for money? Tech insiders are raising their eyebrows at the £480m price tag, with suggestions that other suppliers could do the work for less.8 There are also worries about a lack of transparency and interoperability with other systems raising the prospect of “vendor lock-in”—the risk that, once we start down this route, it will be very difficult for trusts or integrated care systems to pull out if they don’t find it useful.

As my patients’ advocate I’d like to be able to tell them that the opt-outs they signed in the past (indicating their express dissent for the use of their data for anything other than their direct care) will still have force. So far, I haven’t been able to find this reassurance that patients will continue to have the right to choose what happens to their medical information.