RDA Meeting Jonathan Mayes Blog Post

I am a doctor working for the NHS and an Honorary Clinical Fellow with the University of Edinburgh with research interests in intensive care and anaesthesia.  Medical research is transitioning from being led by lab-based work to data science. I hope engaging with the RDA will lead me to be ahead of the curve, and become a future clinical-academic leader.

As a coffee-dependent life form, I entered the “RDA for Newcomers” meeting at 7:45 am tired and agitated with a few early signs of caffeine withdrawal. The warm welcome I received was the pick me up I needed. We were greeted, orientated and embraced by the senior RDA team. It rapidly became clear the scale and array of opportunities that would await us over the next three days. The meeting was a celebration - the RDA had reached its 10^th plenary>430 participants from 30 countries attended.

Professor Yoshua Bengio, a computer scientist from the University of Montreal, delivered the first plenary address. In healthcare, Professor Bengio discussed the use of AI to recognise colon polyp malignancy during endoscopy. The technology could recognise malignant cells more accurately than leading gastrointestinal experts. Of course, there are many things the technology cannot provide: treatment, counsel, and the procedure itself, but it allows more efficient and accurate service.  I was struck by the potential and power of health applications of AI and that it was an enabler rather than an inhibitor of doctors practice

It was emphasised that health data has its own unique challenges and hurdles. My first interest group on data policy and implementation discussed the availability of data for editors. It was highlighted in healthcare data security, reproducibility and confidentiality are often placed higher than availability and openness.

Meeting the current US and European Research Fellows was enlightening and engaging. Informal discussion took place at an organised lunch before attending the RDA Engaging Early Career Researchers and Professionals BoF group. Here the idea of a lightweight mentorship scheme was generated. This would consist of two meetings before and after a plenary, allowing introductions to the RDA key members. To me, this is the perfect service for a voluntary expert service like the RDA.

For myself, the meeting culminated in the Health Data IG. The group was created from members EU projects on biomedical informatics and has >140 members. The key goals include data access and protection, personalised medicine, and data literacy in healthcare. Initially, it was highlighted how inefficient healthcare spending is and how personalised medicine – through data - can help address this inefficient expenditure.

Reproducibility of scientific results was a key message for healthcare data. Leslie McIntosh and her group have created software “Ricochet by Ripeta”. The idea is excellent: a research paper is uploaded with a DOI this generates a reproducibility report which scores the article based on bibliographic information, database and data collection, data mining and cleaning, methods of data analyses, and data sharing/documentation. This aims to improve evidence-based science by highlighting examples of good and bad practice, meaning funding is directed towards systematic, high-quality research. The software is applicable for funders, researchers, journal editors, and academic institutions. I reflected on how this might impact my day-to-day practice as a clinician and academic improving the evidence base for clinicians to make decisions. A key task for me going forward is to communicate these tools and encourage clinicians to keep up to date with the best evidence-based clinical decision-making.

Data protection is synonymous with healthcare data. Ludovica Durst discussed how different regions are adopting diverging strategies to deal with the issue. The USA model is a sectorial self-regulatory approach. In comparison the EU model involves comprehensive data protection laws, a more homogenous approach for all EU countries. The impact of these strategies is not yet known and the IG set generating a report as a key task by the 11^th Plenary. In my role as a researcher I am frequently required to gain information governance permissions – and rightly so, I believe this is something we do well for the most part in the UK although making the process quicker and more efficient is something we should strive for.

Overall I thought the Plenary was excellent, what struck me most was the passion and enthusiasm the RDA members have for their work. That this organisation is a voluntary exercise speaks testament to the determination of those involved. I would like to thank the organising committee, – especially RDA Secretariat, the University of Montreal and all the other RDA fellows for making the experience so memorable and enjoyable. I look forward to seeing everyone at the next meeting!