What does robust governance of a learning healthcare system look like?

This September, DigitalHealth.London and the Artificial Intelligence Centre (AIC) hosted the first of our three collaborative webinars. Members of the AIC team were joined by Dr Nicola Byrne, National Data Guardian for Health and Social Care England, to discuss: “What does robust governance of a learning healthcare system look like?”. This webinar was chaired by Yasmin Stinchombe, Senior Programme Manager at DigitalHealth.London, who said:

“The first webinar in the #AI4NHS series gave some fantastic insights into the challenges of implementing AI in the NHS and it was great to hear from the AI Centre about how they are building infrastructure to try and overcome some of those challenges. One of my main takeaways from the discussion was the importance of involving patients and the public in order to build trust in these new technologies.”

The September webinar used the AIC’s FLIP and AIDE platforms as case studies on the question of robust governance in AI. Robin Carpenter, Senior Research Data Governance Manager for the AIC, has been key in working transparently with the DAC (Data Access Committee) in establishing new protocols to manage these AI platforms. He highlights the importance of public involvement in these processes to ensure data is being used and managed in an ethical way:

Patient perspectives have been fundamental for keeping us on the right track. Obviously, it is lovely when we hear we are doing things well, but most of our progress has come from acting on criticism. As AI in healthcare grows so does the need for honest conversations about what patients expect and what needs to change.

– Robin Carpenter, Senior Research Data Governance Manager – AIC

Another key theme throughout this webinar is the responsibility to exceed the minimum legal requirements of “good” data governance and to be transparent with the public.

I had some experience with sitting on the Data Access Committee (DAC) – being in the room felt like a very authentic engagement with clinicians and the public in what the AIC are trying to do and working through questions beyond just the legal requirements.

–  Dr Nicola Byrne, National Data Guardian for Health and Social Care in England

As the AIC continues to integrate FLIP and AIDE into NHS Trusts we must continue to transparently work with the public and clinicians to build a trusting working relationship.

It is not enough to say there is a human in the loop – how have you developed that clinician? What skills and experiences do they have to do that role effectively? We have built in that transitionary process into the platform itself to allow clinicians to familiarise themselves with the platforms.

– Haris Shuib, Head of Clinical Scientific Computing, Guy’s and St Thomas’ NHS Foundation Trust – AI Transformation Lead AIC

The authenticity of this trust relies on communication, and this communication relies on transparency. The AIC is striving for “meaningful” transparency – meaning not just providing excessive amounts of undigestible information on our website but conveying the core concepts of the AI programs so both clinicians and the public can understand how data will be used and what the projects aim to do.

If you are interested in how the AIC and its platforms use data, you can explore our data usage page, or if you would like to opt-out of National data usage please click here.

You can watch the first webinar of our series on Youtube. The second of our series occurred on 02/11/22 and will be available to watch soon. If you are interested in signing up to our third and final webinar, “NHS in the driving seat: building a clinically-led AI ecosystem” please register here.