Voice of EHDEN

We welcome Eric Sutherland, Senior Health Economist at the Organisation for Economic Co-operation and Development (OECD), which we covered in episode 2 of season 2 with Eric's predecessor, Jillian Oderkirk. Eric spent 20 years with a global bank in the financial sector, but wanted to fulfil his wish to be engaged and find meaning in his work, transitioning to roles locally and federally in health on data strategy. Having worked with the Canadian government on health data strategy development, Eric recently moved to the OECD, in particular focusing on the impact of digital tools and data on heath care delivery and outcomes within the overall health ecosystem.

In this conversation we start discussing Eric's definition of digital health, in his new role as Digital Health Lead at the OECD, and recognising there is no universal definition, but with a particular focus on technologies, interoperability and governance, linkage, quality of data, analytics and various intended use of health data. Against a context of incoherent fragmentation, Eric is focused on being able to produce useful guidance for countries to implement integrated digital health ecosystems, based on the three layers of digital tools, integrated data and responsible analytics.

Beyond this Eric outlines what he thinks is working today in digital health, and what's not, using an analogy of water, pipes and outlets to describe this, where we find ourselves too often drowning in data, but thirsting for evidence. Continuing to use this analogy we discuss how this all operates, who is responsible for the different elements, and some best practice use cases, such as OHDSI, or the FAIR principles, with ideas to promote conscious collaboration in making this all work. A lot of what we are trying to do is grow the digital health ecosystem by addressing technical fragmentation; the social, cultural, policy and process challenges; and trustworthiness (as opposed to trust alone).

In the end phase of this conversation, Eric outlines what he thinks are the priorities for Digital Health, with the three layers in mind, and a need for integrated care via multidisciplinary providers supported by integrated data, with a pivot to person-centric versus a facilities-centric philosophy. Inclusivity of new technologies to reduce non-value added tasks and activities while rationalising better data governance and data capture requirements could certainly beneficially impact on the workforce, lowering workloads and enhancing care delivery as one example. And as Eric points out, 'if we want to go fast, go alone, if you want to go far, go together'.

Some additional resources from the OECD:

The views expressed by the participants are personal and not necessarily reflective of their organisations.

The Voice of EHDEN podcast is pleased to continue our occasional ELSI series, and this is the first episode of season 4, focusing on the concept of 'trust' within the context of health and health research, and in particular real world data. In this episode we are joined by Dr Mackenzie Graham, Senior Research Fellow in Data Ethics, Wellcome Centre for Ethics and Humanities, University of Oxford. Mackenzie has a facinating background in neuroscience, data and data ethics, and also collaborates with Dr Mark Sheehan and Dr Richard Milne, our participants in the first episode of our ELSI series.

In the discussion, we explore what is meant by, 'trust', and also trustworthiness, and reliance on appropriate governance systems to support individuals and patients being confident in who and how they can trust. Within this we investigate underlying concepts that result in trust, what it is and what it takes for trust and trustworthiness, such as motives and prior experience, very much within a motivation-based model.

Beyond this, we focus in on aspects of data use, the role of Tech Titans and trade offs we all make in society around the use of our personal data, through to sociotechnical architectures, e.g., Trusted Research Environments (TREs), federated data networks, like EHDEN, DARWIN EU(R) and the European Health Data Space (EHDS). From this the attributes, roles and actor's responsibilities are articulated, and a delineation of both passive and active activity around trust. Finally, we land on the counter factual concept of 'distrust' and the need to explore this further, especially in the context of health and health data use for research.

Specifically the role of TREs is explored in terms of trust, and Mackenzie and colleagues recently published in the Journal of Medical Ethics on, 'Trust and the Goldacre Review: why trusted research environments are not about trust', available here, and speaking to a number of aspects discussed in this episode.

The views expressed by the participants are personal and not necessarily reflective of their organisations.

In our last episode of season 3 (delayed due to scheduling issues), we are delighted to invite back Prof Peter Rijnbeek, Chair of Medical Informatics at Erasmus Medical Centre, Rotterdam, Coordinator of EHDEN, Director of the DARWIN EU Coordination Centre, and Lead for OHDSI Europe, who usually ends each season of this podcast.

For this episode Peter reflects on the success of the OHDSI Europe symposium (24-26th June), held aboard the SS Rotterdam in the harbour and at Erasmus MC, with 350 registrants, 24 plenaries, 80 posters, a workshop and multiple OHDSI working group meetings. It was also an important timing post-COVID 19 lockdowns and was one of the first in-person OHDSI meetings, facilitating a return to networking and conversation not via video screens. The next symposium will be held in Rotterdam in June 2023, details will follow (keep an eye on the OHDSI Europe site).

Peter outlined the work of his group at Erasmus, and video interviews with his Post-Docs were conducted late in 2022 and are very informative on the their work in federated learning, AI/ML and utilising OMOP-mapped datasets, within Work Package 3 of EHDEN (here). In EHDEN we have reached 187 Data Partners in 29 countries, and 64 SMEs in 22 countries, a new community that did not exist pre-EHDEN, and now a new phase in completing industrial mapping to the OMOP CDM, but importantly the evidence generation phase local and regional, with emerging collaboration, sub-networks and national nodes already emerging.

In the conversation Peter outlines his own career development and the impact on his work focus and collaborations at Erasmus, in EHDEN and DARWIN EU, OHDSI Europe and Global. Using the experience also of the COVID-19 pandemic and the progress in EHDEN, we explore the advances and current challenges in the expansion of the use of the OMOP CDM, which is the glue in Peter's mind for the work we are all engaged in, and the post-ETL phase of using CDM-mapped datasets, iterating on tools, skills and methods and upskilling (via the EHDEN Academy). Needing high quality, rapid evidence, is a ubiquitous need, well characterised by the COVID-19 global public health emergency, where the opportunity for classical studies was limited.

In the last part of the discussion Peter speaks to his vision for the future of EHDEN, the focus on scientific advice and services, as well as training, and his confidence in sustainability. Furthermore, continued work within OHDSI, and in DARWIN EU, all point to a remarkable opportunity for European open science research in health, also with transition from the IMI EHDEN to the not-for-profit EHDEN.

The views expressed by the participants are personal and not necessarily reflective of their organisations.

The Voice of EHDEN will return in March for season 4.

In our penultimate episode of season 3, and this year, we are delighted to be speaking with David Seymour, Director of Infrastructure and Services, and Paola Quattroni, Head of Alliance Strategy & Engagement, at the Health Data Research UK (HDR UK) network, Data Science Institute, and UK Health Data Research Alliance. Both initally outline their careers to date, and both David and Paola have very applicable experience in working to ensure data utilisation by researchers in elderly care and oncology respectively.

We start with the history of HDR UK, the national institute of health data science, with a mission to unite health relevant data to enable discoveries that improve peoples' lives, with a 20-year vision to ensure advanced analytics to benefit every clinical interaction, clinical trials and public health (as seen with COVID-19). Founded in 2018, HDR UK is undergoing its 5-year review of what it has achieved and what it proposes to achive over the next 5 years, focused on three pillars, (1) delivering ground breaking research via its driver programmes, (2), bringing together infrastructure and services across the UK and internationally enabling large-scale data use in a trustworthy way, and (3), working through partnerships at a local level in the UK, as well as internationally, to meet its goals.

In the remainder of the episode, we discuss these there domains further, with respect to HDR UK, but also the development of the UK Health Data Research Alliance, in particular to ensure trustworthy use of health data, including patient and public involvement. We explore some of the challenges HDRUK and the Alliance have faced, and how the institution has responded to them, for instance with the use of the '5 Safes' framework.

EHDEN is delighted to be collaborating with HDR UK on the final Data Partner Call at time of this episode, with HDR UK supporting some of the UK applicants if selected and ongoing partnership with our combined network Data Partners in the UK and EU.

The Voice of EHDEN podcast is pleased to publish a new limited series on ethical, legal and social implications (ELSI) of working with health data for research, and this first episode with Dr Richard Milne, social scientist, Wellcome Connecting Science and the Kavli Centre for Ethics, Science and the Public, and Dr Mark Sheehan, philosopher, Oxford Biomedical Research Centre Ethics Fellow, both in the UK.

In our discussion, both Richard and Mark outline the work of the EHDEN Ethics Advisory Board/ELSI Group, and in particular publication of the development of a conceptual framework to assist the research community in understanding how to balance protections with value of the data for research. In a person-centric model, the Concentric Circles View (CCV), provides a structure where the researcher's relative position to the patient or subject, in a concentric circle, illustrates the extent of duty of care required. 

The closer a researcher is to the subject in the model, the greater granularity and need for protection of the subject (compliance with law, privacy and confidentiality), whereas the further out, the data is often less granular, with less protections. It may be in the future that metadata will contain such relevant information to aid a researcher in understanding both the constraints on the use of health data, and their responsibilities in using it.

The EHDEN EAB/ELSI Group are looking further to explore the application of the CCV, especially in federated data networks like EHDEN, which are designed to address some of the challenges outlined in the CCV. The publication, in Frontiers of Big Data, can be found here, to accompany listening to this episode. A lay, non-expert version will also be produced in due course.

In this fourth episode we have a conversation with Magda Chlebus, Executive Director, Science, Policy & Regulatory Affairs at EFPIA, exploring both her career and work in relation to health data, public private partnerships (IMI, IHI) and the importance of both to the Pharmaceutical Industry.

EFPIA is the representative federation of R&D-based pharmaceutical companies, creating a collaborative environment for its members to innovate, discover, develop and deliver new therapies and vaccines for the people of Europe. More recently, via IMI2, Partners in Research was created to also include non-pharmaceutical partners of EFPIA, recognising the diverse nature of collaboration today.

We start with exploring Magda's career, her current position leading the science and regulatory team, focusing on innovation policies and the regulatory process from development through to post-marketing authorisation of pharmaceutical products. Initially Magda started in public affairs in EFPIA, and points out that the reality we live in today for the pharmaceutical industry, started around 1995 with the creation of the European Medicines Agency, multiple legislation in clinical trials, paediatrics, patents, etc., with review of some of these, such as the pharmaceutial legislation review currently.

The Innovative Medicines Initiative (IMI) was also set up with the European Commission, and EPFIA, for a unique public private partnership programme, which Magda has been a long term champion of, especially in moving the 'data agenda' forward.

In the middle of the episode, we discuss the role of real world data from an EFPIA and pharmaceutical industry perspective, with evolving technologies and opportunities. Magda sees two key points for use of health data, in modernising the development of therapeutics, and in evaluating patient outcomes, with better delivery of care. We evaluate current developments, particularly moving from expert-led medicine, to evidence-based medicine, through to learning-based medicine.

Later we discuss aspects of trust, trustworthiness and transparency of purpose for the Industry, using benchmarks such as EHDEN, or IMI2 H2O, or indeed EMA's DARWIN EU as concrete use cases across the R&D continuum, as well as initiatives such as DataSavesLives. Towards the end of the episode Magda outlines her vision for the next five years or so, with a dream that the various strategies and documents are realised and do not stay on paper, being a big win for all stakeholders.

Episode 3 of season 3, and a conversation with Prof Paul Nagy, Deputy Director, Johns Hopkins Medicine Technology Innovation Center, Associate Professor of Radiology and Radiological Science. Paul outlines his career trajectory that led to his current roles within Johns Hopkins, but also collaborating with OHDSI across multiple workgroups, inclusive more recently the Education WG. Paul started out in Radiology, originally as a Diagnostic Physicist, and was instrumental in developing international education programmes within biomedical sciences and informatics, and he has a real passion not only as an educator, but also in in empowering clinicians with data science and technologies.

In our conversation, Paul describes the multidisciplinary approach, team science, to support clinicians in reengineering care provision and improving patient outcomes. In OHDSI, he sees the strength of the various disciplines coming together to advance open science, again team science, and in utilising health data. Ultimately, being able to enhance the, 'wow', or, 'eureka' moment in the diagnostic experience will lead to more attuned diagnostic skill.

A particular focus for Paul is education metrics, and specifically psychometrics with reference to implementing, providing and importantly, evaluating educational interventions. A reference paper on psychometric evaluation, Developing and Verifying the Psychometric Integrity of the Certification Examination for Imaging Informatics Professionals | SpringerLink, is a must read in terms of understanding the perspective of Paul's and his colleagues' work in designing and reiterating design modifications to improve the outcome of education and its outcomes, for application by clinicians. Within the discussion, Paul utilises the imaging and radiology context to describe a real world use case of this work.

We go on to explore the developments in OHDSI, and more recent work in learning from this international system, such as the public psychometric dashboards on YouTube, PubMed, GitHub, MS Teams, etc.,  via https://dash.ohdsi.org, that can inform our understanding of applied education. Further work will assist in the design of OHDSI educational programmes, but also in facilitating how OHDSI can focus on reproducible evidence generation, via large-scale network studies, but also the tools, methods and skills used by OHDSI researchers. In the future, Paul would very much like to see how we can generate interoperability and connectivity of the OHDSI standardised analytical tools and educational resources for learning as you apply your skills and methods.

In the second episode of season 3 we meet with Assistant Professor Liesbet Peeters, leading a research group at the Biomedical Research Institute & Data Science Institute, Hasselt University, Belgium, and the lead for the Multiple Sclerosis Data Alliance (MSDA). Liesbet has been described as a 'rising star' in the domain of real world data/evidence, and she had a very interesting path to her current position with a background in bioengineering, gene and cell technology, animal production, and veterinary science and focus on data.

We explore Liesbet's trajectory and work in the field of multiple sclerosis, the challenge with working with relevant and representative data, and in particular an 'accidental' interaction with the European Medicines Agency through a workshop, with key recommendations on enhancing the FAIR-ness of MS data. What has frustrated Liesbet is the inability to answer questions we readily ought to be able to, whether in MS or more widely. In working with colleagues, for instance the European MS Platform (EMSP), European Charcot Foundation. and others, Liesbet took the lead in creating the MSDA to respond to these frustrations over the last five years. Alongside this, Liesbet has been collaborating with e.g., EBRAINs, ELIXIR Belgium, DataSavesLives, et al, internationally and locally to address wider aspects of working with RWD/RWE. Liesbet outlines the details of collaborating and building and the complexity involved in the episode.

While Liesbet provides a fascinating insight into building a startup in working with data, she goes on to explore the recent work on COVID-19 and MS, answering critical questions quickly to address patient and clinician concerns as the pandemic emerged internationally. Beyond this Liesbet outlines her vision for the future, the collaboration with multiple stakeholders, initiatives and networks, but most importantly working with, 'amazing people'.

A future meeting, led by Liesbet and colleagues on 16th November in Baveno, Italy, on the, 'Use of Artificial Intelligence and Great Data to reveal progression in MS' is still open for registrations.

Welcome to this first episode of season 3, and a great way to start with a conversation with Prof Dipak Kalra, formerly a Primary Care Practitioner in London, UK, academic at University College London, and Ghent University, and President of the Institute for Innovation through Health Data (i~HD). Dipak is internationally known and renowned with a reputation for cogent and clear thinking around health data, data interoperability and quality, and policy development, with i~HD being a leading European-based organisation dedicated to advancing the field. 

In this episode Dipak reverts back to his experiences in general practice in the 1980s, and the advent of electronic health records systems being introduced also to support use of data for research, albeit with underlying challenges around ethics and privacy governance, which of course still pervade to today. We discuss some of the challenges around ethics of data use, and privacy preserving aspects, as well as some of the more recent innovations, while focusing on the balance between risks of harms, versus benefits for patients themselves.

Dipak outlines the vision and direction of i~HD, from its origin in the IMI EHR4CR project, today and in the future, particularly regarding collaboration and safer use of health data, data quality, best practice, and policy development in this domain at a European level. Ultimately, i~HD acts as a neutral catalyst in addressing development and innovation. Lastly, Dipak describes the vision he has for health data use, RWD/RWE and a 21st century frameworks to meet the explosive need for evidence, , avoiding 'bureaucratic processes' and ensuring enhanced efficiency in generating RWE.

In our last episode of season 2, we are delighted to invite back Prof Peter Rijnbeek, Chair of Medical Informatics at Erasmus Medical Centre, Rotterdam, Coordinator of EHDEN, Director of the DARWIN EU Coordination Centre, and Lead for OHDSI Europe, who ended season one of this podcast discussing predictive analytics, and ends now season 2. 

For this episode Peter reflects on the success of the OHDSI Europe symposium (24-26th June), held aboard the SS Rotterdam in the harbour and at Erasmus MC, with 350 registrants, 24 plenaries, 80 posters, a workshop and multiple OHDSI working group meetings, all for the back in person after the COVID-19 lockdowns. A number of highlights, in the science, in networking and collaboration and with regards to future developments are discussed, in particular the session on 10 countries in a 'cruise around Europe' and the growth of national networks or nodes, and evidence generation at scale. Peter goes on to explain OHDSI Europe, its work as a regional chapter of OHDSI, support of symposia, but also development of national nodes across multiple European countries, and reflecting on methodological and technical developments required for Europe itself in the CDM and standardised analytical pipeline.

In the middle of the episode, Peter explores the many proud achievements made by EHDEN, from the network development of Data Partners, training and certification of SMEs, implementation of the CDM at an industrial scale, and the move now to large-scale, rapid network analysis. The launch of the EHDEN Portal at the OHDSI Europe meeting is also a recent highlight. Ultimately, the remaining time of the IMI phase of EHDEN will be focused on illustrating evidence generation to improve patient outcomes, but also sustainability via an EHDEN not-for-profit that will parallel and then take over in 2024.

Lastly, we discuss the difference from the launch of EHDEN in 2018, to now and to 2024/2025 at the end of the IMI phase and beyond and what that may look like in a new Europe with mass adoption of the CDM, also exemplified with the recent initiation of the European Medicines Agency's DARWIN EU initiative, also based on the OMOP CDM and a federated model. Perhaps content for a future episode.

The views expressed by the participants are personal and not necessarily reflective of their organisations.

The Voice of EHDEN will return in September.