5

INFRASTRUCTURE

CONTEXT

Infrastructure represents the foundational, organizational, and structural elements that enable a society to function. In the context of health and health care, the cornerstone elements of infrastructure include the diverse systems and settings in which care occurs; data generated by individuals and populations; policies that shape how health care is organized, financed, and delivered; the health care workforce; and the ability to utilize information to inform and improve outcomes. Infrastructure is also the backbone of connectivity, enabling individual patients, caregivers, and their communities, as well as other stakeholders, to engage with one another and to use information in their daily lives. Infrastructure as a whole is complex—including myriad elements ranging from the available diagnostic and therapeutic modalities available to the communication networks and systems, and on to the human capital involved—and these must work together seamlessly in pursuit of optimal health for all. The twin forces of the COVID-19 pandemic and rising awareness of ubiquitous, long-standing health disparities brought on by entrenched systemic racism have sharpened focus on the U.S. health care infrastructure that is brittle, fragmented, and unevenly distributed. Among other challenges, the pandemic illuminated key shortcomings in public health infrastructure, not only in the ability to exchange data and information rapidly to track cases and optimize care but also to quickly implement effective treatments and COVID-19 vaccines. Moreover, while the contributions of biomedical research grow exponentially, with thousands of rigorous peer-reviewed articles produced annually, leveraging the evidence at the point of care is an uneven proposition due to the heterogeneous nature and variable capacities of health care settings. The need to invest in health care infrastructure as a step toward realization of a patient-centered learning health system has only intensified in recent years.

COMPELLING RESEARCH QUESTIONS

Readying for another pandemic is widely viewed as an imperative in light of the impact of COVID-19 and includes attention and investments at the state and federal level focused on data, workforce, policies, and material. Concurrently, further research on how to actualize the learning health system will yield widespread benefits and ensure that biomedical research successes reap their full potential. This could have the corollary benefit of invigorating health care providers, in that clinicians (e.g., doctors, nurse, pharmacists, behavioral health practitioners) could see more immediate and tangible benefits of applying knowledge at the point of care. Finally, thoughtful attention to the emergent field of data justice (exploration of how the generation, collection, and use of data intersect with societal and structural inequalities), coupled with scrutiny of embedded biases in artificial intelligence, are two of many needed steps in effectively using data to mitigate health inequity. Compelling questions may include the following:

• What should a learning health care system look like to meet patients’ needs in the next 5 years?
• What would be necessary to mount a coordinated and accelerated response to the next major public health challenge?
• What are the most effective approaches to dissemination and implementation of evidence, given the heterogeneity of U.S. health care?
• How can the wide range of data sources be leveraged most effectively to accelerate evidence generation and translation?

DISCUSSION HIGHLIGHTS

Rainu Kaushal, M.D., M.P.H., from Weill Cornell Medicine, provided the overview presentation on infrastructure at the March meeting, emphasizing five essential components of research infrastructure that have bearing on health and health care: (1) people, (2) real-world data, (3) evidence generation, and (4) evidence implementation. Dr. Kaushal linked these infrastructure components to the learning health system, showing their relationships and interdependence. Dr. Kaushal observed that some elements of the U.S. infrastructure can be robust, but other parts need considerable attention, especially when the full continuum of prevention, screening, diagnosis, and treatment, as well as health equity, is considered. The pandemic was particularly revealing in this regard. Development of COVID-19 diagnostic and treatment strategies worked comparatively well with the extant infrastructure; however, disparities were magnified. Dr. Kaushal

described the INSIGHT Clinical Research Network (which she leads as part of PCORnet^®) as an infrastructure exemplar. At the outset, she garnered cooperation from New York City health systems, which typically compete for market share, by establishing the value proposition that sharing clinical data via this network would improve patient care, population health, and research, thus bringing new ideas and expertise into New York’s health care ecosystem. INSIGHT’s infrastructure consists of a large research data mart and attendant logistical support; engagement of health system leaders, researchers, participants, and other stakeholders; and commitment to sharing expertise and supporting other people’s research to further an active implementation and improvement capability.

Dr. Kaushal concluded by describing the union of precision medicine and precision prevention as the ultimate use case for infrastructure. By bringing real-world data to bear on diagnoses and treatments of diseases and applying those real-world data at scale for entire populations equitably, precision health becomes the blended product of informed and guided public health and health care delivery systems. Infrastructure for precision health spurs synergies and holistic thinking about how to make real-world data readily available for research; ensures that patients are engaged in evidence generation with an equity perspective; and then implements evidence at the point of need.

The infrastructure breakout group built on many themes from Dr. Kaushal’s presentation, noting the importance of retaining and sustaining the research efficiencies spurred by the pandemic, including administrative, contractual, and oversight-related improvements; embrace of remote monitoring for clinical trials; and redoubled efforts to collaborate. The group noted that a key aspect of an effective infrastructure is diffusion of evidence to broader audiences in an understandable way, with attention to health literacy and data literacy. Reflecting on misinformation challenges of the pandemic, the group underscored the importance of communication and engagement as a part of an optimal infrastructure. The inherent complexity of the extant health care infrastructure creates challenges for patient engagement, such as fatigue from having to navigate the fragmented system. The group’s concluding insight was that all stakeholders need a roadmap and clarity about the destination: What does a robust infrastructure for a patient-centered learning health system look like, how do we ensure that this infrastructure is further enhanced and sustained, and how can PCORI bring attention here and invest in weak spots?

Accordingly, development of a patient-centered learning health system was the second topical focus for the April meeting. Participants in that meeting used the frame “what would it take?” as they discussed the learning health system, its infrastructure, and PCORI’s role. A notable theme was PCORI’s ability to func-

tion as a connector that can drive research insights into practice, because this capability does not really exist anywhere in the health ecosystem. In this respect, both Sachin Jain, M.D., M.B.A., SCAN Group and Health Plan, and Atul Butte, M.D., Ph.D., University of California, San Francisco, noted the PCORI’s potential role as an “effector arm” for the learning health system. PCORI’s unique mission and attributes (e.g., PCORnet^® data collaboration model, dissemination and implementation funding mechanism, engaged stakeholders) could combine to help it fulfill this function across the research enterprise and clarify the ideal infrastructure for learning what works best for whom.