University of Michigan Learning Health System

Joshua Rubin, JD, MBA, MPP, MPH is the Program Officer for Learning Health System Initiatives at the University of Michigan Medical School Department of Learning Health Sciences (among other roles, discussed further below); he joins us to discuss the LHS, emphasizing the need for all stakeholders to work collaboratively and embrace rapidly changing technologies in order to achieve the goal of interoperable health information exchange for all.

Learning Health System

The Learning Health System (LHS) was a concept first expressed in 2007. The vision of the LHS is for a transformative health system in which the power of shared data and analytics are leveraged to learn from every experience of every (consenting) patient, and lessons learned are shared with patients, clinicians, and other stakeholders so that every decision affecting health is an informed one.

In my efforts to help advance the Learning Health System (LHS) vision I work in three distinct but synergistic roles, collaborating with remarkable teams in each capacity:

  • Program Officer for Learning Health System Initiatives at the first-of-its-kind Department of Learning Health Sciences at the University of Michigan Medical School. For the most part, I am responding to these questions wearing my University of Michigan “hat.”
  • Vice President of the Board of Directors of the Kanter Health Foundation, the only philanthropic foundation founded by a patient whose overarching mission is realizing a person-centered LHS. 
  • A founding member of the Interim Steering Committee of the Learning Health Community, a multi-stakeholder international movement, bonded together by a set of shared consensus LHS Core Values, working together to deliver on the potential of the transformative LHS vision.

I became involved with HIMSS pursuant to my role in the Learning Health Community, before joining Michigan. HIMSS participated in an invitational Learning Health System Summit in 2012 sponsored by the Kanter Health Foundation and contributed to developing the multi-stakeholder consensus LHS Core Values that gave rise to the Learning Health Community. HIMSS was one of the early endorsers.

Since then, HIMSS has been beyond supportive of the LHS movement. I was invited along with Dr. Charles Friedman (now Chair of the Department of Learning Health Sciences at Michigan) to give a presentation on the LHS at HIMSS13 – and even got to meet former President Bill Clinton (the HIMSS13 keynote speaker)!  We collaborated closely with key informatics leadership at HIMSS: the entire April 2013 issue of HIMSS Clinical Informatics Insights was devoted to the LHS; Dr. Friedman and I were offered the opportunity to publish and to help guide the issue. HIMSS has been beyond generous and beyond helpful and we are grateful.

That collaboration to advance the LHS movement, to share the vision, and to engage diverse HIMSS members in working to realize the LHS together, has been among the most rewarding aspects of my involvement with HIMSS for the past half-decade.

Progress to date toward realizing the LHS vision stems from contributions of so many people and organizations, separately and together. Over the past decade:

  • The thought leaders at the Institute of Medicine (IOM) / National Academy of Medicine (NAM) and elsewhere have collectively published volumes.
  • Health systems nationally and internationally have transformed themselves into rapid learning entities (and private and public funders have supported and connected such efforts).
  • Researchers and technologists have developed innovative methodologies.
  • US policymakers and influencers have transformed the LHS from a promising idea into the seminal goal of federal health IT policy (with comparable developments around the globe).
  • Public health professionals have illuminated the unsustainability of the status quo.
  • Patient activists have advocated steadfastly for a health system centered around patients that empowers patients with information to make better informed decisions.
  • LHS champions from Joe Kanter to the late Hunt Blair have simply refused to take “it’s impossible” for an answer when it’s imperative and it could change lives and save lives.

Additionally, there are some major milestones in Michigan’s LHS efforts worth sharing; I believe these LHS milestones transcend Michigan:

  • The 2014 renaming and reimagining of the University of Michigan Medical School Department of Medical Education into the Department of Learning Health Sciences. It transitioned from a department that revolved around learning at the individual and team levels to a basic science department dedicated to research, service, and education.
  • The launch of our first-of-their-kind graduate education programs: both Master’s and PhD programs in Health Infrastructures and Learning Systems (HILS).  These novel interdisciplinary programs at Michigan promise to educate a future generation of health “infrastructuralists.” The caliber of the students we have already attracted nationally and globally is simply extraordinary.
  • Michigan and our partners have received funding from the National Science Foundation (NSF) and the Computing Community Consortium (CCC) to convene a series of interdisciplinary scientific workshops. What began as a workshop grew into a realization that something approaching a new interdisciplinary science may be needed to address them. This work has attracted the expertise of diverse prominent scientists nationally and globally to envision a science of cyber-social learning systems.
  • The launch of the open access peer-reviewed scientific journal “Learning Health Systems” published by John Wiley & Sons, Inc. We just published our first issue. I am the guest editor for an envisioned future issue about patient empowerment and the LHS.

While I have certainly heard no shortage of frustration expressed by clinicians and patients about the usability of systems and accessibility of data, I nonetheless believe the past decade has been transformational.

Technology that can digitally capture human health experiences, “conserve” and share this data has grown rapidly in its adoption, plus in our awareness of and demand for it. That includes adoption of EHR systems and patients demanding their data. It also includes the growth of smartwatches and other wearables, PHRs, implantable devices, and more.

As we move toward an interconnected world of Internet of Things (IoT) and “big data,” we are starting to create the raw material to empower a LHS and more. By 2020, there may be 50 times as much data as there was in 2015, and medical knowledge generated from this data may be doubling every 73 days.

In a general sense, the LHS is about realizing the promise of this future to touch people’s lives and health and to empower all stakeholders. It should truly empower patients and families, and foster collaboration with clinical care teams, researchers, public health professionals, and other key stakeholders – all while advancing the Triple Aim of better care and health experiences, better health of individuals and populations, and lower costs per capita.

I would say the overall trend seems to be a shift from interoperability (a capability) to truly driving interoperation. That requires considering the incentives and needs of all stakeholders, and aligning them so that, beyond just having the capability to share, a trust fabric and corresponding incentives are engendered to truly support such sharing.

What is shared will grow beyond what is traditionally thought of as health data; what happens in the doctor’s office is just the tip of the iceberg.  Consider, for instance, all the factors that play into social determinants of health. Data about human health experiences that can drive more robust research in these areas and put information in the hands of individuals and organizations that can make a difference, may involve a number of social services that are not traditionally connected with health, but have enormous potential to affect health of individuals, families, communities, and populations.

Colleagues at major standards development organizations are engaged in the LHS movement, and the Learning Health Community has launched a multi-stakeholder standards initiative called the Essential Standards/Structures to Enable Learning (ESTEL) initiative.

Many organizations in the private and public sectors are working on leveraging analytics and processing power to transform “big data” into knowledge. Even at current rates of knowledge generation, there is an oft-cited 17-year “bench to bedside” lethal lag between knowledge generation and its broad application, plagues medicine. Some experts estimate that, even as quickly as we generate knowledge, health quality still takes 35 years to double.  As we get better at generating knowledge faster (i.e., biomedical knowledge doubling every 73 days by 2020), that problem could actually be exacerbated.

At Michigan, we are taking our strengths in relevant psychological, information, engineering, integration, and implementation sciences, and bringing them together to work to address this key piece of the puzzle. Think of learning as a loop that involves taking experience or performance and transforming it into captured data (P2D), using research methods and analytics and processing power to transform data into knowledge (D2K) and ultimately curating, tailoring and mobilizing knowledge to individuals to inform decisions that lead to actions or performance (K2P). Michigan is launching innovative efforts on the too-often-neglected K2P portion of the loop.

Building upon and integrating Michigan’s expertise in other domains, we are working on ways to rapidly and efficiently turn knowledge into digital knowledge objects that can be curated, updated and tailored for the appropriate recipient of the knowledge and ultimately put into action. We are working to create an open source smart grid for health knowledge generated by diverse sources at ever-increasing rates.

I will take off my Michigan “hat” for a moment and mention the Learning Health Community. The LHS Core Values are grounded in collaboration and the facilitation of sharing. The over 100 organizational endorsers of the LHS Core Values include stakeholders at every point on the compass. They include stakeholder types that at times naturally mistrust one another, and organizations that compete vigorously with one another, all coming together in a grassroots movement to work together on key pieces of the multi-faceted puzzle. These include grassroots initiatives related to standards, governance, policy, and a trust fabric to facilitate such exchange in pursuit of learning and better informed decisions engendering better health for all.

Michigan, an endorser of these LHS Core Values, embraces this spirit of collaboration internally and externally; indeed, these models of collaboration make Michigan a powerhouse. Perhaps such collaboration with other organizations is most evident in the service component of the Department of Learning Health Sciences’ tripartite mission of research, service, and education. By service, we work to realize learning health systems at various levels of scale, including within Michigan Medicine, at the state level (Learning Health for Michigan), at the national level, and internationally.

We have formed collaborations with other academic medical centers, as well as private companies and public agencies nationally and internationally. We are part of the PCORI-funded Patient-Centered Network of Learning Health Systems (LHSNet), with collaborators spanning ten states including the Mayo Clinic, Intermountain Healthcare, and even Ohio State University. We have collaborated with data-holders across the state of Michigan on a number of efforts aimed at clinical quality improvement, and at transforming into a learning health state to learn from and improve the health of Michigan’s almost 10 million residents, while serving as a model other states can ultimately emulate (Learning Health for Michigan). Globally, our collaborations have taken us to several continents; and we have hosted colleagues from these several continents at Michigan.

A key barrier involved coalescing a critical mass of believers to illuminate that the LHS is more than a promising idea and to reach sufficient buy-in to make it a working reality that touches people’s lives and health. Not only are there now many healthcare organizations and networks working to become learning systems, but the LHS vision has become the overarching end goal of federal health IT policy, and seen as where healthcare needs to go by many people in positions to help move it there.

Another key challenge is that in developing the basic science stems from its interdisciplinary nature. The challenges to be addressed require knowledge from diverse disciplines and perspectives no one individual is likely to possess. For instance, solving certain challenges may require individuals with experience in biomedical research, quantitative methods, data science, law, policy, ethics, economics, social psychology, and patient/caregiver/family advocacy to all work together. Empowering them to work together can be challenging.

At Michigan, across our own campus, we have coalesced, sustained, and grown a collaborator spanning a dozen or so diverse academic units, focused on the LHS. We have applied lessons learned to bringing together scientific and research communities of interest spanning multiple organizations nationally and globally.

I think a lot of it comes down to incentives. What gets rewarded and what gets measured is what gets done, especially in an environment where many stakeholders are overloaded and feel pulled in too many directions simultaneously.

At Michigan, as a starting point for our LHS endeavors, we work to pull together communities of interest bound together by common problems of interest.  As each team is pulled together, we go to substantial lengths to understand the priorities of the team members (who may come from different organizations and seemingly divergent perspectives).  Consequently, these teams bond together around solving problems about which they are passionate that mean something to them.  Every person starts to see themselves and their interests in the collaborative solution. So doing the right thing related to sharing data, information, knowledge, and even wisdom also gets interwoven into how success is measured by the individuals and by the teams they are parts of.

We also strive to make doing the right thing the easy thing to do. We seek to work with all stakeholders to collaboratively realize and implement such infrastructures in ways that not only solve common problems, but also genuinely facilitate easier and smoother sharing and exchange among those individuals and organizations using and affected by them. Ultimately, a key goal is that what is envisioned simply becomes part of the culture; the way things are done.

The best advice I have received in this space comes not from technologists or even people who work in health information exchange, but from patient and caregiver advocates. These activists underscore why patient-centered and person-centered need to be more than just buzzwords. Their advice includes:

  • Calls for a consumer-driven health information economy.
  • Illuminating the importance and potential of putting patients and caregivers in the driver’s seat.
  • Making compelling cases about how everyone on a care team performs better when they are all better informed.
  • Calling for privacy protection.
  • Reminding us that “patient” is not a third-person word since every one of us has been or will be a patient.
  • Sharing both horror stories of not being able to get their data and beacon examples of when data liquidity (coupled with health literacy and tools to make use of the data) just works.

Some have analogized that, much like prescribing medication, mobilizing health information to the right person, at the right time, in the right form, in the appropriate amount or dose, for the right reason, is foundational to the envisioned transformation.

Embracing this thinking about data (and information, knowledge, and wisdom) not as an end in itself but as a force for driving patient-centricity or person-centricity, starts to align all other stakeholders on the importance of furthering health information exchange.

At Michigan, thinking that puts the patient into the center has been ever increasingly integrated into our efforts. Granted, this shift can take time at a large academic medical center, but our efforts to involve patients in the very heart of more and more of what we do continue to expand. Doing so has the bonus impact of reminding clinicians, administrators, technology professionals, and researchers of the human element why data from myriad sources need to be shared and meaningfully integrated around the patient.

Michigan was one of the first institutions to endorse a set of consensus LHS Core Values, the first of which is “Person-Focused.” The Preamble to these LHS Core Values recognizes that, “The proximal goal of the LHS is to efficiently and equitably serve the learning needs of all participants, as well as the overall public good. The LHS offers an important opportunity to facilitate sharing of data in order to serve this goal, aiming to surmount obstacles to such sharing.” Doing precisely that, in order to accelerate valid and continuous learning that can ultimately (and rapidly) be shared in ways that inform decisions touching health, empowering all stakeholders to improve the health of individuals, families, communities, populations, and the general public, is interwoven into the fabric that binds together the first-of-its-kind University of Michigan Medical School Department of Learning Health Sciences.