Professor Nina Hallowell is a member of the Wellcome Centre for Ethics and Humanities in the Nuffield Department of Population Health at the University of Oxford, and a member of the Steering Committee for the Centre for Personalised Medicine at St Anne’s College Oxford. In this guest blog post, she writes about what she learned from holding a citizens’ jury to understand how people feel about the way different kinds of health data are used.
Given the speed at which research on data-intensive technologies is advancing, it is important to regularly gauge people’s views on the use of healthcare data. As a group of researchers who either use healthcare data in our research or are interested in publics’ views about this issue* we decided to run a citizens’ jury in early 2020.
Our aim was to gain some understanding of what members of the public think about the use of different types of healthcare data (MRI images, digital images of pathology slides and genomic data, i.e. DNA sequences) in research. Specifically, we were interested in people’s views on the benefits and challenges of sharing these different data types, and the acceptability of sharing healthcare data with commercial companies. We also wanted to trial the citizens' jury method of public deliberation.
Citizens’ juries involve members of the public who come together to hear expert evidence about a topic of interest, deliberate the issue, (hopefully) reach a consensus and deliver their verdict.
So over the last couple of months of 2019 we advertised in local media for jurors to participate in the 'Debating Data Citizens' Jury' in early 2020.
Data access on trial
On February 15th 2020, with the COVID-19 pandemic reaching the UK and storm Desmond making landfall overnight, twenty people braved the elements to attend our citizens’ jury in Oxford Town Hall. The jury members heard evidence about data collection, data analysis and data access from a series of expert witnesses in the fields of digital pathology, MRI imaging, genomics research and bioethics. The jury was given the opportunity to cross examine the witnesses, and then broke off into smaller groups to discuss a number of questions, including:
- In which ways are these different types of data similar/different and why?
- What relevance do these types of data have for you?
- What ways should these different types of data be used?
- Should these different types of data have different ‘rules’ about how they’re used and shared?
- Who should/should not be able to access your data?
Finally, each group was asked to reach a verdict on data use in research by answering the following question: How should the different data types be used, who should they be used by and why?
The jurors returned a series of statements about data use and commercial involvement in research:
- Genomic data is different from MRI and pathology image data. It is more personal (it can identify individuals), could be misused, and therefore deserves greater protection and oversight.
- Patients should be able to control uses of their data and should give consent for use of their healthcare data in research.
- Private companies should be allowed to access healthcare data as we need them for their technological expertise, but their research should be more closely monitored and subject to more stringent oversight than publicly funded research.
In addition to finding out what the jurors thought about researchers using healthcare data we asked them to reflect on participating in this event. Most were very positive, and said they had learnt a lot about the use of different types of data in research and appreciated the fact that we were interested in hearing their views. One participant said they: “learned much more about underlying science” and “felt better able to take an informed view.”
Not all were so happy though. Another participant said that they felt the information they received was biased in favour of the use of data for research (although we did instruct witnesses to be as neutral as possible). This juror also felt there was not enough focus upon wider discussions about the use of personal data by the state and the politics of the NHS.
This was the first citizens' jury the team had organised, so Debating Data was a learning exercise for us too. At the end of the day we were relieved it appeared to go so smoothly. We were grateful that so many people had attended, given the weather. We were also pleased to hear a number of participants say they felt the day had given them a good opportunity to air their views.
We noted a number of improvements for implementation in future events, including: more targeted recruitment to get a more representative sample of jurors, allowing more time for the witnesses to give evidence, and for jurors to cross-examine, deliberate and reach their verdicts.
What will we do next? As you may have spotted, one of the jury’s conclusions was that people should give consent for the use of data in research. There are many ways to give permission for the use of data, including: opt-in, opt-out, broad or narrow consent and dynamic consent, to name but a few. As it stands, the National Data Opt-Out allows people to opt-out of having their confidential patient data being used in healthcare research and for planning services. But if people do not opt-out does this really mean they consent to their data being used? This sounds like an important question for a citizens’ jury, and we have plans to explore this in the future.
* Members of the Wellcome Centre for Ethics and Humanities and the Centre for Personalized Medicine developed the event and members of two of the AI Centres of Excelllence - NPIC and NCIMI - helped us to run the event on the day.