A driving goal for our Centre for Digital Transformation of Health is to partner with healthcare organizations in becoming a learning health system, and a persistent question asked from all sides is what the role of a university is in that mission. I plan to dedicate the next few blog posts to help me think through this question in the context of an insightful article on academic learning health systems and my experience in several academic health systems in the US and at the University of Melbourne in Australia.
The Ivory Tower
This question--what role does a university have in learning health systems--underpins our ability as researchers to create impact in healthcare, and it has plagued me since I started graduate school in 1994 without a healthcare or computing background but with a BA in Linguistics. It comes in many forms--from the healthcare side, I hear
You aren’t a clinician--how could you know how to solve our problems?
Research is important but it’s too far removed from real problems to matter
From the university side, I hear
That problem may be important in healthcare, but it’s not interesting methodologically
The project has no possibility of being implemented, but I got a good paper published.
Learning Health Systems (LHS)
There is truth to both sides: the incentives in healthcare delivery and academics are largely misaligned, so collaboration is fraught from the beginning. It reminded me of a viewing chamber Skyspace by American artist James Turrell in the National Gallery in Canberra called Within Without.
Because art can be interpreted by the viewer, I draw an analogy that your viewpoint on the value of research in healthcare delivery partly depends on whether you are within or without/outside of the system. But the views from both sides are valid and important. As Turrell describes his work, “It’s about your seeing, like the wordless thought that comes from looking into fire.”
The learning health system framework provides a shared incentive and common view: we want to leverage data collected from providing care to gain insight that can then inform and improve healthcare delivery and thus health. We see that shared ground in the response we have had to our Applied Learning Health System short course, where local hospitals have sent multiple cohorts of staff including clinicians, managers, data scientists, and EMR teams and in the growing cadre of clinicians being introduced to practical research methods in our Learning Health System Academy fellowship program.
But how do we go beyond teaching--which is often accepted by healthcare organizations as a university’s proper role--to partnering on this grand goal?
Academic Learning Health System
In their 2023 article, Rosenthal and team define The Academic Learning Health System as an LHS built around a robust academic community and central academic mission. They propose six ways an academic LHS enhances an LHS, providing a map for activities valued by mission of an academic health system align with clinical activities and goals of an LHS:
Capitalizes on embedded academic expertise in health system sciences;
Engages the full spectrum of translational investigation from mechanistic basic sciences to population health;
Builds pipelines of experts in LHS sciences and clinicians with fluency in practicing in
Applies core LHS principles to the development of curricula and clinical rotations for medical students, housestaff, and other learners;
Disseminates knowledge broadly to the research community to advance the evidence for clinical practice and health systems science methods; and
Addresses social determinants of health and creates community partnerships to mitigate disparities and improve health equity.
In the next few posts, I’ll share my evolving thinking about how these features of an academic LHS have played out in my experience, how they can contribute to more effective translation of AI in healthcare, and how to operationalize these principles into collaboration between academics and healthcare organizations that are of mutual value.
Something I really loved about starting my career in the U.S. Veterans Administration is that leadership was working hard to align incentives of its career research scientists with operational priorities of the VA health system. I love that we are thinking about this framework of C4DTH.
Wendy, this is a very thoughtful posting. It stimulated many responses in my mind. For the successful systems that I have worked with and preceeded the dynamic LHS and its companion AI, I have often considered/ recommended that the 'researchers' come and 'sit in the dirt with the end users' and have us all ask 'what do they see here' in the healthcare enviroments? As you are aware you linguistics background does not bar you for advanced health informatics. In fact in the 80s and early 90s some of the best health informaticians were librarians. Thye taught us clinicians a lot about information management. Terry