Using health data for scientific research isn’t that simple
While health data is essential for scientific research, that data also needs to be protected. In her doctoral thesis, Irith Kist found a balance between protecting the individual and exchanging health data.
When Dr Kist answers our early-morning Teams call, we can clearly see a hospital in the background. It turns out not to be a standard filter – it’s actually her regular place of work. She’s speaking to us from Antoni van Leeuwenhoekziekenhuis, the specialist cancer hospital in Amsterdam where she works as a privacy expert. Alongside her job at one of the Netherlands’ top cancer treatment and research centres, Dr Kist has written her doctoral thesis on the use of patient data.
Coronavirus: more cancer deaths than COVID-19 deaths
Dr Kist began her PhD research back in February 2020, just before the start of the coronavirus pandemic. It would be reasonable to assume that the timing gave her extra motivation, as scientific research using health data suddenly became extremely relevant. After all, we needed a suitable coronavirus vaccine as soon as possible. But Dr Kist puts the impact of the pandemic on her research into perspective: ‘My supervisor, Professor Marjanka Schmidt, rightly pointed out that more people died from cancer during the coronavirus pandemic than from COVID-19. In other words: before the pandemic happened, there were already countless life-threatening diseases, which were a good incentive for us to research the use of health data. But, as she said, the pandemic showed us that the process of exchanging health data for the purposes of scientific research and healthcare could be improved.’
Why is the current process of exchanging health data less than ideal?
Scientific research into diseases and drug development involves collaboration between researchers various hospitals, laboratories and pharmaceutical companies. These parties are often located in different European countries, and the problem is that they independently decide how to collect and use patient data. Dr Kist continues: ‘Patients in the Netherlands usually have to consent to the use of their personal data and bodily material for scientific research. However, giving consent isn’t as simple as saying “yes” or “no” – patients have to be informed about what they’re consenting to. It sounds logical, but at the start of research, the specific research question and any future areas for research are often still unclear. Within the first year of ‘longitudinal studies’ (longer-term research projects that can take up to 30 years) in particular, it’s unclear what scientists will specifically need to research in 30 years’ time.’
Explicit consent: skewed results and ‘reverse discrimination’
Dr Kist says that asking for the patient’s consent every single time is no solution either: ‘Patients can be difficult to reach. They might have moved house, or – if they’re no longer undergoing treatment – they might not want constant reminders of their cancer journey.’
Asking patients to participate in ‘new’ scientific research also requires a lot of time and effort. ‘You run the risk that the research population is no longer complete. Say, for example, you begin with 1,000 patients. For each research question, scientists have to contact all of those patients and ask, “Is that all right with you? And this? And also that?” Often, all of the information is provided in extensive patient information leaflets anyway – it’s not uncommon for the leaflets to be 25 pages long. Remember that these people have cancer and other illnesses, or they’ve already recovered. They have other things to think about. Suppose you end up with 500 patients in the final study: the patient population would be incomplete, which might skew the results.’
In that situation, ‘reverse discrimination’ can occur. Dr Kist explains: ‘Scientists want everyone to participate – they don’t want to exclude anyone. After all, an inclusive approach leads to the best reflection of society and therefore the most accurate results. However, we now know that some population groups are participating in these studies less often and for shorter periods of time. If, at some point, they do not feature in the research data at all, then less scientific research is conducted among those population groups – inadvertently, of course.’
What's the solution, then?
Dr Kist feels it would be helpful if, for the purposes of future scientific research, the term ‘consent’ is interpreted more broadly. That way, scientists would no longer have to request consent from each patient for every single aspect of their research. Dr Kist suggests permitting the exchange of health data for reasons of general scientific interest, as is currently the case in other European countries. Ultimately, today's patients are also benefitting from their treatment, which may never have existed in its current form without the scientific research conducted 30 years ago.
Dr Kist's recommendations may well be followed in the foreseeable future. The European Parliament recently adopted the European Health Data Space Regulation. Drawn up during the coronavirus pandemic, this legislation could simplify the process of sharing health data for the purposes of healthcare and scientific research.
Irith Kist defended her doctoral thesis entitled ‘A fair balance. Health data protection and the promotion of health data use for clinical and research purposes’ on 5 June 2024. The non-scientific thesis summary is available here.