Health care has been called one of the most complex sectors of the U.S. economy. Driven largely by robust innovation in treatments and interventions, this complexity has created an increased need for evidence about what works best for whom in order to inform decisions that lead to safe, efficient, effective, and affordable care. At the same time, traditional approaches to clinical research are straining to keep pace with these demands. Calls for approaches that draw from and better inform real-world practice, that leverage the increasingly available vast amounts of digital health data, and that are more cost-effective have been on the rise. These approaches are at the foundation of a learning health system, one that continuously and seamlessly generates knowledge from the practice of health care to answer important questions that matter to patients, their health care providers, and stakeholders system-wide.
As health care becomes more digital, clinical datasets are becoming larger and more numerous. These data, many of them gathered through the normal course of health care, offer great potential for extracting useful knowledge to achieve the “triple aim”—improved care, better health for populations, and reduced health care costs. In a continuously learning health system, data from sources such as electronic health record (EHR)
1 The planning committee’s role was limited to planning the workshop, and the workshop summary has been prepared by the workshop rapporteurs as a factual summary of what occurred at the workshop. Statements, recommendations, and opinions expressed are those of individual presenters and participants and are not necessarily endorsed or verified by the Institute of Medicine, and they should not be construed as reflecting any group consensus.
systems used to manage patient care, claims data necessary for billing purposes, and increasingly patient-generated sources of data such as patient portals, surveys, and online communities are used to inform questions of operations, to guide care, to further scientific understanding, and to power innovation. This approach differs from traditional approaches to clinical research, which are often removed from the clinical experience both in terms of the questions asked and the environment in which they are carried out, require large amounts of additional data collection, can take several years to complete, can be very expensive, and are often criticized for producing evidence that is not easily generalizable to broader populations or easily implementable in real-world settings.
By realizing the potential of knowledge generation that is more closely integrated with the practice of care, it should be possible not only to produce more usable evidence to inform decisions but also to increase the efficiency and decrease the costs of doing clinical research. Delivering on this promise will depend on certain technical capabilities, but, more important, ensuring the sustainability of this approach will require the delivery of value to stakeholders who are engaged in these processes. Among these stakeholders are the patients whose experiences are captured; clinicians who deliver care, collect routine information, and implement results; researchers who are experts at using routinely collected data to answer questions; and system leaders who allocate resources and set institutional priorities and culture.
All of these groups of critical stakeholders have potentially different value propositions for participating in or supporting practice-integrated knowledge generation. For health system leaders, the current trends of increasing complexity and paucity of the proper evidence to inform care coincide with heightened pressures to reduce unsustainable health care costs. Health delivery system leaders are facing shrinking operating margins and pressure to do more with less, while at the same time contending with the resource- and knowledge-intensive demands of moving to value-based reimbursement approaches that emphasize population management. Additionally, as is the case with clinicians and patients, system leaders are often faced with a lack of evidence on which to base their operational decisions. This can be the result of a misalignment of priorities among leaders, researchers, and funders or of differences in the timeliness requirements for results.
Digital health data are increasingly being used by organizations to manage patient populations, to fulfill reporting requirements, to carry out local quality improvement activities, to support externally funded research projects, and to share with other organizations in inter-organizational collaborations for quality improvement and research. What underlie all of these activities are robust informatics capabilities that are often designated for clinical, operational, or research purposes. However, in some cases the
infrastructure is common across activities and provides opportunities for greater efficiency and value in its use.
Although the effort to use clinical and patient-generated data to answer important research questions is still in its infancy, investigators have already demonstrated the potential for using routinely collected clinical information to detect and respond to disease outbreaks (Al-Samarrai et al., 2013), to target medical services to those who need them most (Stephens et al., 2014), to help patients and clinicians make better decisions (Ray et al., 2012), to avoid errors that can harm patients (Zhou et al., 2014), and to speed medical research (Borch et al., 2011). Many of these efforts are done locally within a single health system, but there have been a number of efforts to share information across systems in order to compare data for quality improvement purposes and benchmarking, such as the Minnesota Community Measurement initiative; for public health and drug safety surveillance, such as Mini-Sentinel; for the conduct of pragmatic clinical trials, such as the Health Systems Research Collaboratory of the National Institutes of Health (NIH); and, most recently, for the conduct of comparative effectiveness research (CER) through the National Patient-Centered Clinical Research Network, or PCORnet.
PCORnet (see Box 1-1) is a nationwide patient-centered clinical research network funded by the Patient-Centered Outcomes Research Institute (PCORI), which consists of clinical data research networks (CDRNs), each spanning at least two health care delivery systems, and patient-powered research networks (PPRNs), whose focus is the collection and use of patient-generated information. Together these networks are intended to form a resource of clinical, administrative, and patient data that can be used to carry out observational and interventional research studies and enhance the use of clinical data to advance the learning health care system. The primary goal of the first phase of PCORnet will be to establish the data infrastructure necessary to do such research. This includes getting to harmonized, comparable, and consistent data; working out privacy and security issues; and establishing the trust needed to work out data-sharing agreements across participants. In order to achieve this, the network has made robust stakeholder engagement a priority, including requiring the involvement of health system leadership in CDRN governance.
PCORnet symbolizes a new approach to clinical research, one that is integrated into the delivery of care and that leverages its experiences, rather than creating a set of parallel infrastructures and processes. Critical to the long-term sustainability of such a network of networks will be demonstrating its value to its many stakeholders. For health system leaders, such demonstrations may include building or improving their systems’ data infrastructure to help with data-intensive functions such as reporting and identifying variation and areas for improvement; enabling health systems
To facilitate more efficient comparative-effectiveness research that could significantly increase the amount of information available to health care decision makers and also increase the speed at which this information is generated, the Patient-Centered Outcomes Research Institute (PCORI) has invested more than $100 million in the development of the National Patient-Centered Clinical Research Network (PCORnet).
PCORnet will be a large, highly representative national network of health care information networks that will be used to conduct large-scale clinical outcomes research by establishing a resource of clinical data gathered in real time and in real-world setting such as hospitals and clinics. Data will be collected and stored in standardized, interoperable formats under rigorous security protocols, and data sharing across the network will be accomplished using a variety of methods that ensure confidentiality by preventing patient identification. A hallmark of PCORnet is its requirement that patients, clinicians, and health care systems that provide the research data housed in each constituent network be involved in the governance and use of the data. PCORnet aims to advance the shift in clinical research from investigator-driven to patient-centered studies.
During an 18-month development phase, 29 health data networks—11 clinical data research networks (CDRNs) and 18 patient-powered research networks (PPRNs)—will work closely with a national coordinating center and other stakeholders to refine the capabilities and capacity of the individual constituent networks. By the end of this first phase PCORI expects that a fully functional research network will be in place and ready to support comparative-effectiveness research. PCORnet, even in its formative phase, is creating a unique opportunity to make a real difference in the lives of patients and their families. By building clinical research into the health care process and by working directly with patients and their advocates, PCORnet will be able to provide the answers that patients need quickly, efficiently, and at lower costs than previously possible.
to leverage their data to be more active partners in generating the evidence used by their providers and patients to inform care; and helping health systems better manage their populations toward higher-value care.
In April and June 2014 the Institute of Medicine’s (IOM’s) Roundtable on Value & Science-Driven Health Care convened two workshops aimed at accelerating progress toward real-time knowledge generation through the seamless integration of clinical practice and research, one of the funda-
Statement of Task
An ad hoc committee will plan two workshops to explore the engagement of health system leaders as active proponents of greater integration of research activities with clinical care processes and discuss specific insights for advancing a learning health care system through the work of the Patient-Centered Outcomes Research Institute’s (PCORI’s) National Patient-Centered Clinical Research Network (PCORnet). The first workshop will integrate issues, opportunities, and strategic options by bringing together health care system leaders, both administrative and clinical, and researchers, including those from PCORnet grantee institutions, the NIH Health Systems Research Collaboratory, and the Innovation Collaboratives of the IOM Roundtable on Value & Science-Driven Health Care. The workshop will consider issues and strategic priorities for building a successful and durable PCORnet and facilitating progress toward a continuously learning health care system more broadly, including issues related to science, technology, ethics, business, regulatory oversight, sustainability, and governance. A follow-on workshop focusing on implementation approaches will convene health system CEOs to consider strategic priorities and explore approaches to implementation. Discussions will inform the decisions of field leaders moving forward, including PCORI, the PCORnet steering committee, and PCORnet grantees (Phase 1 and 2). An individually authored workshop summary of the workshop series presentations and discussions will be prepared by a designated rapporteur in accordance with institutional guidelines.
mental concepts of a continuously learning health system, centered on the development of the PCORnet (see Box 1-2).
The first workshop brought together health care system leaders, both administrative and clinical, and researchers, including those from CDRN and PPRN grantees, the NIH Health Systems Research Collaboratory, and the IOM Innovation Collaboratives, to identify and consider the issues and strategic priorities for facilitating progress toward sustainably integrating PCORnet into a continuously learning health care system, including issues related to science, technology, ethics, business, regulatory oversight, sustainability, and governance. The goals of this workshop, developed by the planning committee, were
- Broaden and deepen health systems’ leadership awareness of the prospects for and from a continuously learning health system.
- Foster the development of a shared commitment, vision, and strategy among health system leaders building a national clinical research network.
- Identify common applications in meeting health systems’ responsibilities for science, technology, ethics, regulatory oversight, business, and governance.
- Consider and learn from models and examples of productive integration of research with care delivery programs.
- Explore strategic opportunities for executive, clinical, and research leaders to forge working partnerships for progress.
- Consider the approach and desirable outcomes of a meeting of chief executive officer (CEO) leadership in building, growing, and making full use of the infrastructure necessary.
The second workshop convened health system CEOs to consider strategic priorities and explore approaches to implementation that will inform the decisions of field leaders moving forward. The goals for the second workshop, developed by the planning committee, were
- Continuous learning infrastructure and business case: What are the key infrastructure, value proposition, and business case implications in integrating research and practice as the foundation of a continuously learning health system?
- Aligning continuous improvement and knowledge generation: What infrastructure commonalities exist in aligning executive agendas and knowledge generation priorities and in driving continuous improvement through learning?
- Institutional opportunities: Consider common principles and strategies for participants to move priorities forward in their own institutions.
- PCORI contributions: Reflect on strategic infrastructure and research opportunities for PCORI that can support delivery systems in evolving toward learning health systems.
A major premise that served as the foundation for the two workshops is that the continuous and seamless assessment of the effectiveness and efficiency of care is basic to a continuously learning and constantly improving health care system. Advancements in the digital infrastructure and the development of innovative methods for research and learning now make this aim achievable in health care, as it has already been achieved in many other sectors of the economy. PCORI and its Methodology Committee are committed to accelerating this progress, including prior work with the IOM in conducting a workshop exploring the role of observational studies in continuous learning (IOM, 2013). A foundational focus for PCORI’s PCORnet (see Box 1-1) is developing a large, highly representative, nationally linked, and coordinated clinical research program with multiple
collaborating large-system networks to improve the nation’s capacity to conduct comparative clinical-outcomes research. The issues are complex, including matters of ethics, governance, sustainability, and stakeholder engagement. PCORI is committed to active stakeholder engagement, and these two workshops were aimed, in particular, at informing and engaging health system leaders as essential partners in building the necessary and sustainable infrastructure to power a continuously learning health system. As Joe Selby, PCORI’s executive director, said at the opening of the first workshop, “The entire premise of creating PCORnet is that it does make sense to engage health care systems, their clinicians, and their patients in a process of putting data together, standardizing it, and using it for quality improvement and clinical research.”
In his introductory remarks to the first workshop, IOM president Harvey Fineberg said that his hope is that the domains of medical practice and medical research will be seen not as parallel, independent enterprises, but rather as tightly integrated efforts. “This is a way in which we can reduce the traditional distinction between research that shows what could work and research that demonstrates what does work in practice,” Fineberg said. “If we can bring those more tightly together over time, I believe we have the opportunity not only to make faster progress, but to make the kind of progress in health care that really pays dividends immediately and over the long term for the health consequences of our population.”
At the start of the second workshop, IOM president-elect Victor Dzau, who previously served as the CEO of the Duke University Health System, said that he believes that the health care system is now positioned to finally join other industries as an enterprise of continuous learning, but that this will not happen without the seamless integration of research and practice. He noted, too, that the main challenge that he faces as a health system CEO is to bring together learning and innovation that can disrupt the way things are done today and at the same time create standards for more effective and efficient health care.
A learning health system can create an ideal environment in which to take advantage of the advent of “big data” and to conduct the type of outcomes research that the health care system needs in order to become both more efficient and more effective. The challenge, Selby said, will be to develop a learning health care system that is sustainable so that the data infrastructure that PCORI is helping establish to enable large-scale outcomes research will not require continuous infusions of funds to maintain operations. Doing so will require establishing a business case, not only for the delivery system and for funders but also for patients and the clinicians who will populate the data infrastructure with their patients’ data. Speaking about the business case for a learning health system, Selby said that there are many advantages for a health care system in supporting this kind of
research. “The experience of collaborating across institutions can bring a lot of value and insight, and the entire process can position delivery systems well for driving towards value-focused payments,” he said.
The Roundtable on Value & Science-Driven Health Care provides a trusted venue for national leaders in health and health care to work cooperatively toward their common commitment to effective, innovative care that consistently adds value to patients and society. The Roundtable explores concerns that, despite the world’s best care being available in the United States, in certain circumstances health in America falls far short on important measures of outcomes, value, and equity. Care that is important is often not delivered, and care that is delivered is often not important. Roundtable members are leaders from core stakeholder communities, including clinicians, patients, health care institutions, employers, manufacturers, insurers, health information technology companies, researchers, and policy makers, brought together by their common commitment to steward the advances in science, value, and culture necessary for a health system that continuously learns and improves in fostering healthier people.
The Roundtable’s vision is that the nation will develop a continuously learning health system in which science, informatics, incentives, and culture are aligned for continuous improvement and innovation. In this continuously learning health system, best practices will be seamlessly embedded in the care process, patients and families will be active participants in all elements of the health system, and new knowledge will be captured as an integral byproduct of the care experience. This vision includes an “afferent” arm of data collection, analysis and learning, as well as an “efferent” arm of dissemination and implementation of learning and best practices. The Roundtable’s goal is to promote collective action and progress so that by the year 2020, 90 percent of clinical decisions will reflect the best available evidence. The Roundtable aims to meet this goal through stakeholder workshops and meetings designed to accelerate understanding and progress toward the vision of a continuously learning health system and through joint projects conducted by affinity-group Innovation Collaboratives focused on best clinical practices, clinical effectiveness research, the communication of medical evidence, digital technology for health, incentives for value in health care, and systems engineering for health improvement.
This publication summarizes the presentations and discussions that occurred during the two workshops (see Appendix A for the agendas), highlighting the key lessons presented, practical strategies, and the needs and opportunities for future leadership. The perspectives included in this summary reflect the experience of the workshop attendees, who included a preponderance of system leaders committed to developing learning systems in their organizations. Their perspectives may not represent other stakeholder groups or of all system leaders. Chapter 2 offers examples of organizations that are on the leading edge of integrating care delivery and research in a way that leads to greater efficiency, better value, and improved health care. Chapter 3 provides a brief introduction to the vision for a continuously learning health system, including a description of the value proposition for various constituencies, and Chapter 4 explores the business and financial issues and opportunities that arise as organizations move toward continuous learning and improvement. Chapter 5 looks at the challenges and opportunities surrounding the legal and ethical oversight of integrating care and research opportunities, and Chapter 6 focuses on issues of institutional governance of continuous learning activities. Chapter 7 discusses the challenges and opportunities that come with efforts to engage clinicians, patients, families, and the public in integrating care and research efforts. Chapter 8 identifies and prioritizes the key issues for health systems leadership in moving toward greater integration of care and knowledge-generating activities. Chapter 9 discusses infrastructure needs for a continuous learning health system and how that infrastructure can have several uses within an organization’s operations, and Chapter 10 explores how continuous learning can become an executive agenda priority. Chapter 11 provides a summary of key points made by speakers from the two workshops.
This page intentionally left blank.