Ben Goldacre: ‘GP data is the jewel in the crown of English health data’

Doctor, academic and best-selling author, Professor Ben Goldacre’s 2022 report on the use of health data for research and analysis helped lay the foundations of the NHS’s data strategy.  

Four years on from the Goldacre Review, director of the Bennett Institute for Applied Data Science, has delivered another landmark with OpenSafely, the secure data environment (SDE) that is now making whole population GP data available to analysts and researchers working outside Covid.

Ahead of his keynote at Digital Health Rewired 2026, Goldacre explains why he believes open working is the best way to drive accountability, and how using data is the best way to optimise care.

Are you satisfied with the progress that’s been achieved since the publication of the Goldacre Review?

When you write a report like that for a minister, you don’t expect everything you recommend to be done.

I think the recommendations that I’m most saddened to see little progress on are the ones around building capacity in the NHS analytics workforce.

There’s a huge opportunity there with very talented people doing critical, practical day-to-day work with health data to identify variation in care, to spot opportunities, to improve clinical services, to reduce inequalities, to forecast need and plan around it.

And those staff are operating outside the structures that you would expect to see for people doing such detailed technical work.

There has been substantial investment in secure data environments. Unfortunately, we now have a delivery problem

My core recommendation on data access was to move from data dissemination, which is inherently unsafe and inefficient, and instead move to working in trusted research environments (TREs), which the NHS has needlessly renamed SDEs.  That was picked up as policy.

There’s been very substantial investment in TREs, or SDEs as the NHS calls them. Unfortunately, we now have a delivery problem. I think it’s universally accepted that progress has been not as fast as people hoped.

The best way forward is to do a thoughtful audit to look at what money has been spent where and list the completed outputs from each SDE investment.

How important is open working?

Working in the open is absolutely critical.

On the OpenSafely platform we work in the open as far as humanly possible. That means all our code is shared openly and in public, but also all our technical deliberations on how we work are shared in public.

I’m a big proponent of working in the open because it’s the best way to drive delivery. Sharing your code, your deliberations and your outputs means that everybody else can see what you’re doing so they can reuse it, contribute to it and point out opportunities to improve it.

The best possible way for teams to share information with each other on technical issues like data interoperability is simply in the open. These aren’t secrets.

Working in the open also drives accountability. One of my worries with technical work in the NHS is that when you look online you often can’t find anything other than promotional 400-word blog posts or short slide decks.

If I can’t see any of the technical documentation around what somebody’s built, then it’s very difficult to think about whether there are learning opportunities.

It’s also very hard for people trying to navigate the system, and for policymakers and commissioners to understand what has and hasn’t been built.

Sometimes people suggest that the NHS needs to share technical documents internally, but there’s no such thing as ‘internal’ in the NHS.

The NHS is a shoal not a mammoth, with care and organisational management split across hundreds of organisations across the country.

The best possible way for teams to share information with each other on technical issues like data interoperability is simply in the open. These aren’t secrets.

Do we need a ‘no blame’ culture around data and technology?

It’s not unreasonable for people to raise criticisms about major public spend that doesn’t seem to have much of an output.

But I think if people have worked in the open and they’ve struggled to deliver, it’s appropriate to be thoughtful and considerate because you want to encourage openness.

Anybody who’s worked on any difficult technical question in healthcare knows that failure is an inherent part of the pathway to success. You want to fail quickly, happily and informatively. And all of that happens if you work in the open.

You want to fail quickly, happily and informatively. And all of that happens if you work in the open

It’s important to be empathic about failure, but it’s very hard to be empathic about hidden failures. Where there are large spends and there’s nothing visible, that’s what raises alarm bells for me.

Will OpenSafely be the turning point that unlocks the power of data?

I certainly hope so. Whole population GP data is now available for analysis and research through OpenSafely. This is an extraordinary landmark.

OpenSafely has been able to provision whole population GP data for the first time in history – the jewel in the crown of English health data.

It covers the entire population, so it’s got breadth, and it’s got enormous detail about every individual. We’ve also been able to link it onto secondary care data.

And we’ve been able to do with the full support of the professions and privacy campaigners. People have sometimes said that GPs are obstructive about access to data, but that has never been my experience.

We listened to concerns about the need to credibly protect patients’ privacy when making whole population data accessible, and the need for transparency. We invented new ways to work with NHS patient data to respond to those concerns.

The question is, will people use OpenSafely in the same way they used data during the Covid pandemic?

The biggest opportunities to prevent avoidable suffering and death in the British population are through optimising the delivery of care

Sometime I think data access is used as a procrastination technique in the NHS. People say ‘Oh, I would do brilliant things, if only I had access to… ’ and then they get access to that data, and nothing happens.

The power of GP data is that you can look at variation in care, you can see which types of patients are getting which treatments. You can look at inequalities between ethnic groups, between deprivation groups, between the sexes, between age groups.

You can also look at which organisations are using which treatment and you can use it to look at NICE pathway compliance or NHS England-recommended treatment pathways.

The biggest opportunities to prevent avoidable suffering and death in the British population are through optimising the delivery of care rather than inventing a new widget or drug.

Using data to optimise care is the next great horizon.

What key message will you bring to Digital Health Rewired?        

It’s important to try and give advice to government, but my personal choice has been to get on and deliver.

My message to people at Rewired is we have made national GP data available for research and service improvement through OpenSafely. Please come and use it. And if you find our working methods helpful, we’re happy to work with you to build data infrastructure yourself.

Goldacre will keynote at Digital Health Rewired, which is taking place on 24-25 March 2026 at The NEC in Birmingham. Register here.

Rewired 2026’s headline sponsors are The Access Group and Optum, who will also sponsor the Integrated Care and Digital Transformation stages respectively.