Transforming data into actionable insights

As this supplement goes to press, the Veteran Health Administration’s (VHA) electronic health record system is undergoing a critical transition; from a largely home-grown system to a widely used commercial product over a period of multiple years. This transition holds enormous promise for Veterans’ care and is deeply rooted in our mission to provide timely, safe, high quality health care to the men and women who served this nation. The articles in this supplement not only advance this essential mission but reflect a collective recognition that data science is essential for both research and health care operations.

VHA’s s pioneering work in this area began in the 1970s on the pre-internet Decentralized Hospital Computer Program (DHCP). The program was officially launched at VA medical centers in 1981 and thirteen years later, revised as the Veterans Health Information System and Technology Architecture (VistA) (see www.worldvista.org at worldvista.sourceforge.net/vista/history).

VHA centrally maintains, validates, and makes accessible secured EHR datasets for analysts, including researchers, scientists, program evaluators, clinical practitioners, and administrators. Ongoing development and support are provided by the Office of Research and Development (ORD) through direct funding for research and quality enhancement initiatives. ORD also provides technical and statistics support through a network of centers of excellence including VHA’s Information Resource Center (VIReC) and VHA’s Informatics and Computing Infrastructure (VINCI). Together, these entities ensure the security of our data, the privacy of our patients, staff, and trainees, and fully leverage our computer resources. Analysts from program offices such as the Office of Academic Affiliations’ Data Management and Support Center routinely evaluate VHA files to provide rapid responses to questions from Executive leadership, Congress, law enforcement, the General Accounting Office, and other government institutions.

Overall, these systems’ datasets afford VHA important advantages in assessing multiple dimensions of health care performance while at the same time provide the capacity to share such information enterprise-wide. VistA is directly linked with multiple operational systems that support care delivery, thereby providing information on a wide range of areas such as wait times, progress addressing prescription opioids, timely availability of supplies, patient health outcomes, health services research, and the health professions education of over 1,25,000 trainees at VA Medical Centers. At a time when U.S. and worldwide health care delivery are rapidly changing, the importance of accurate, relevant data cannot be overstated.

However, it is important to keep in mind that that easy access to reliable data is necessary, but far from sufficient. Analysts need solid methods and analytics to translate data into knowledge and actionable insights. As the nation’s largest integrated health care system, VHA faces the same challenges as the private sector (e.g. demographic shifts) as well as those more specific to Veterans’ needs. For example, VHA is challenged by the urgent need to rapidly incorporate scientific advances to benefit a growing population experiencing multiple chronic conditions. At the same time, VHA is also charged with addressing service-related injuries, including the invisible wounds that are the signature of current conflicts, as well as the capability to translate veteran-focused research into tangible care improvements. In short, VHA is a vital partner in a health care ecosystem seeking to use data science to become a learning health system [1].

As noted above, this supplement underscores the recognition that data science is essential for both research and health care operations. While each sector uses data differently, neither can succeed without an appreciation for the data source (and possible bias) as well as for solid methods that actionable insights and transparent, replicable research findings. Moreover, as articulated by the National Academy of Medicine, an adaptable, learning health system is described as ‘a system in which science, informatics, incentives and culture are aligned for continuous improvement and innovation, with best practices seamlessly embedded in the care process …  new knowledge captured as an integral by-product of the care experience [2].’ Fulfilling this vision will rely on insights derived from daily processes (encounters, referrals to specialists or the community, patient feedback) and research and evaluation.

In fullfilling our fundamental mission to veterans, it is imperative that our system be positioned for continued learning and improvement. Assuring the continuity of patient data for research as well as the incorporation of an analytics capability that can be continually enhanced will require dedication. It also will require close collaboration with numerous stakeholders as well as experts in human factors, statistics, analytics and other disciplines. The work will not be easy – but the papers in this supplement offer a strong foundation on which to build.