Evolution of the Veterans Health Administration Learning Health System: 25 years of QUERI

The Veterans Health Administration (VHA) is among the nation's largest integrated health systems and among the largest learning health systems. A learning health system combines institution-specific data generation and evaluation with context from the broader research, clinical, and policy communities to inform and implement approaches to improve access to, quality, efficiency, and cost of care.^{1, 2} Instrumental to the establishment of the VHA as a learning health system was the founding of the national Quality Enhancement Research Initiative (QUERI) program, an operational evaluation and quality improvement program. QUERI was established in 1998, and in the years since its founding, initiatives have evolved from implementation of guideline-based care for specific conditions to the conduct of system-wide quality improvement and implementation science efforts to the support of evidence-based policymaking.^3-5 This special issue highlights scientific contributions and real-world impacts emanating from 25 years of initiatives stemming from QUERI that reflect rapid translation of research into practice.

QUERI was established with a goal of becoming “…a national system to translate research discoveries and innovations into patient care and health systems improvement”.³ Initial translation efforts were focused on nine conditions or groups of conditions that are prevalent among veterans: heart disease, cerebrovascular disease, type II diabetes, colorectal cancer, prostate disease, spinal cord injury, HIV/AIDS, mental health conditions, and substance use disorders. This initial focus laid the groundwork for projects featured in an article in this issue by Damschroder, Hamilton, and colleagues, who summarize the far-reaching impacts from a decade of diabetes-focused QUERI projects.⁶ Evaluations of the implementation of the VA's Diabetes Prevention Program led to insights into how to improve the reach of other, related programs; the effectiveness of an online version of the program; and the importance of gender-specific tailoring of prevention programs.

In the past decade, QUERI has expanded its focus to address broader health system issues that affect multiple conditions and veterans' overall well-being.⁴ The Veterans Access, Choice, and Accountability Act of 2014 (Choice Act) mandated the creation of a commission to examine methods to improve veterans' access to care. In their report, the Commission on Care recommended that the VA leverage data and continuous improvement principles to ensure equitable access to high-quality care.⁷ Around the same time, the Office of Management and Budget urged federal agencies to use evidence to support budget and operational decisions.⁸ QUERI's expanded focus was in response to these recommendations, notably by using a learning health system framework for evidence generation and evaluation of quality improvement strategies to inform policy.⁴

Several articles in this issue highlight the ways in which data can be leveraged within a large health system to identify potential targets for improvement. Shannon et al analyze data from the VA's Survey of Healthcare Experiences of Patients and identify racial disparities in access to VHA care, particularly among veterans belonging to underrepresented groups.⁹ Authors of other articles in this issue leveraged data from the VHA's electronic health record (EHR). Using EHR data, Honken et al demonstrate that a program in which funding was made available to hire social workers as members of primary care teams was associated with increased social work service use among rural veterans.¹⁰ Leung et al found that integrated mental health and primary care service provision was associated with higher short-term depression follow-up care but not treatment completion.¹¹

As part of its focus on continuous quality improvement, QUERI supports studies of the effectiveness of different strategies to implement evidence-based programs and treatments. This issue features evaluations of the implementation of programs targeting obesity, homelessness, opioid use disorder, and patient safety. Damschroder and colleagues find that although a virtual quality improvement training program does not increase the reach of a weight management program, it is associated with increased use of quality improvement methods.¹² Wyse and colleagues describe the development of an implementation strategy to increase access to buprenorphine in primary care settings for veterans with opioid use disorder.¹³ Oberman and colleagues demonstrate the effectiveness of the evidence-based quality improvement strategy in increasing access to medications for opioid use disorder in primary and mental health care settings.¹⁴ The importance of collaboration and training opportunities for implementation is highlighted by Montgomery et al in their evaluation of the implementation of health care navigator services for veterans with unstable housing¹⁵ and by Sullivan et al in their evaluation of the implementation of a guidebook to standardize patient safety practices.¹⁶ Finley emphasizes the role of trust in forming effective partnerships with non-VA organizations to implement a suicide risk mitigation program.¹⁷

Continuous quality improvement also relies on the successful scale-up of evidence-based interventions. In this issue, the authors highlight the importance of flexibility when scaling up an intervention. Cordasco and colleagues outline an iterative, structured approach they used to adapt and scale up a case management implementation strategy for unstably housed veterans.¹⁸ Blok and colleagues find that early adopters of a family caregiver training intervention were more likely to make adaptations to the intervention.¹⁹ Reisinger and colleagues catalogue implementation strategies used by the VA's Office of Rural Health to see which are most useful for large enterprises implementing initiatives—using iterative strategies was one of the most frequently employed strategies.²⁰ Other work featured in this issue focuses on estimating the costs of different implementation strategies, which also affect scale-up. Daniels and colleagues estimate the start-up costs for using implementation facilitation to set up multidisciplinary pain clinics.²¹ Kaufman and colleagues compare the costs associated with two different methods of implementing a mobility improvement intervention among hospitalized older adults.²²

More recently, QUERI-funded work has focused on evidence-based policymaking. The Foundations for Evidence-based Policymaking Act of 2018 (Evidence Act) mandated that all cabinet-level agencies use evidence to support budget and policy decisions. All agencies submit annual evaluation plans and “learning agendas”, strategic documents outlining gaps in evidence needed to support policy and budget decisions, to the Office of Management and Budget to comply with the Evidence Act. The VHA's fiscal years 2022–2026 learning agendas focus on access to care, suicide risk mitigation, and pain management with opioid risk mitigation. A QUERI-funded center—the Partnered Evidence-based Policy Resource Center—coordinates the development of these documents for the VHA, and work featured in the VHA's fiscal year 2023 annual evaluation plan²³ and the access to care learning agenda is highlighted in this issue. Shafer and colleagues demonstrate that increasing mental health staffing levels is associated with the initiation of mental health care among transitioning service members.²⁴

Increasingly, QUERI investigators are being called upon to aid VHA operations leaders in responding to legislative mandates. The VA Maintaining Internal Systems and Strengthening Integrated Outside Networks Act of 2018 (MISSION Act), Commander John Scott Hannon Veterans Mental Health Care Improvement Act of 2019, and Honoring our PACT Act of 2022 are recent examples of legislation calling for additional evidence and mandated reports.^25-27 In this issue, Matthieu and colleagues describe congressionally mandated reports and the process for planning and executing evaluations so that they can effectively inform responses to the mandated reports and programmatic or policy changes.²⁸

As QUERI's focus has evolved from disease-specific to complex system-level and policy issues, its infrastructure has also evolved. Early models of continuous quality improvement were more siloed, slowing the communication of evidence and needs for evidence. In contrast, learning health system models of quality improvement rely on the formation of a community of evaluators, implementation experts, and operations leaders who are in constant communication about needs for data, development of evidence, and application of findings.²⁹ Learning health system models use an iterative process of identifying operational priorities, identifying gaps in evidence to address priorities, rapid and rigorous evaluation to address questions, and communication back to operational stakeholders.⁵ Two commentaries in this issue highlight the evolution of QUERI as a learning health system: Beck and colleagues depict QUERI's process for identifying and addressing changing VHA priorities using the example of chronic pain and opioid use disorder.³⁰ Braganza and colleagues describe QUERI's rapid response team mechanism—a novel infrastructure developed to allow investigators to address operations and policy needs in a timely manner.³¹

Over 25 years, the QUERI program has evolved to meet changing needs of the VHA health care system. Its guiding principles are now being adopted by QUERI's parent program, the VA's Health Systems Research portfolio. Specifically, Health Systems Research priorities are based on the QUERI learning health system framework for establishing evidence generation and evaluation priorities, and are aligned with the VA's Strategic Plan.⁵ Moreover, the Health Systems Research portfolio funds centers of innovation and consortia of research to support foundational learning in health systems research, namely implementation science; data science and informatics; science of patient, provider, and community engagement; systems science; and policy science and analysis.³² These foundational methods support the underlying infrastructures needed for learning health systems to implement embedded research, evaluation, and quality improvement to achieve veteran health care improvement.³³ Together, QUERI and VA Health Systems Research are committed to developing and translating evidence to improve Quintuple Aim goals, including improving care access, quality, costs and value, equity, and experience, serving as a national model for how to use learning health systems to translate evidence into practice and policy.