What Can the Rest of the Health Care System Learn from the VA's Quality and Safety Transformation?
Five years after the landmark Crossing the Quality Chasm report by the Institute of Medicine (IOM), the quality and safety of health care in the United States remains far from ideal.(1) It is easy to feel pessimistic. Can health care organizations really change? Is it possible to provide effective, safe care, and to do so without increasing costs even more? While providers and policy makers grapple with these questions, there are lessons to be learned from one success story: the Veterans Health Administration.
For decades, the Veterans Health Administration within the Department of Veterans Affairs (VA) had a dreadful reputation as a provider of care. Although this reputation was likely exaggerated, as recently as the mid-1990s, VA was certainly no model of high-quality care.
What a difference a decade makes.
By the mid 1990s, VA faced major challenges. A Republican congress was looking askance at large federal spending programs like VA and questioning their value. Politicians from both sides of the political spectrum agitated to let more veterans access the private health care system by using vouchers. VA's very survival was threatened, and the system was in dire need of bold leadership. Luckily, a change agent arrived in 1994 in the form of Ken Kizer. Kizer, a former Navy diver, experienced administrator, and visionary leader, arrived and set out to transform the VA system to try to improve its quality and safety. The largely successful experience holds lessons for other systems, most of which are probably less challenging to change than VA, aiming for the same goals.
Although VA's transformation was multifaceted, four changes are particularly noteworthy: decentralization of decision-making, ongoing data collection and feedback, performance contracts with incentives, and the implementation of a sophisticated clinical information system. Let us go through each of these in greater detail.
Decentralization.—VA replaced its centralized, often overlapping bureaucracy with 22 regional networks, known as Veterans Integrated Service Networks (VISNs). The VISNs became the locus of decision-making and accountability. Each VISN was in charge of all care provided to veterans in its region, and each network was funded on a capitated basis rather than based on historical costs (ie, VISNs received money based on the number of veterans they cared for). This rationalization of funding realigned incentives for the networks to focus on the long-term welfare of the patient: The hope was that keeping patients healthy by providing preventive services and better disease management would lead not only to better outcomes but also to lower costs through reduced complications and fewer hospitalizations.
Data collection and feedback.—Starting in the mid-1990s, VA leadership began the External Peer Review Program (EPRP). The purpose of EPRP was to collect data on clinical care as well as on structural features, such as waiting times to be seen in primary care. The data were then fed back to each VISN and facility. Clinical leaders could compare their performance with those of other VA facilities. This had a profound impact on the way providers and clinical leaders behaved. Until the EPRP program, VA clinical leaders knew little about how their performance compared to others, and had little motivation to improve.
Performance contracts and incentives.—These contracts were initially signed by each VISN director and VA Undersecretary for Health. They created specific goals for networks on both clinical and structural features. For example, they set benchmarks for the percentage of diabetic patients who needed to have their hemoglobin A1c controlled to less than 9.0 mg/dL (81% was the national benchmark in 2003), or of the percentage of patients who needed to be seen in primary care within 30 days of a requested appointment. The contracts came with incentives for both the VISNs and the directors themselves to meet pre-specified goals. All the performance data were available and shared not just internally but also with members of Congress and interested outside stakeholders, such as veteran service organizations (VSOs).
Information systems.—Finally, VA rolled out a comprehensive electronic health record (EHR) across all VA sites. In 1996, VA created the Veterans Health Information System and Technology Architecture (VISTA). Using this architecture, VA launched the Computerized Patient Record System (CPRS), an EHR for clinicians throughout the VA health care system, in 1997. Although CPRS was created at the national level, each facility was given some flexibility to customize the system to meet its own needs. This mix of top-down implementation and bottom-up customization has been instrumental in ensuring that the system is still integrated, allows for data sharing across VA sites, and is widely accepted by end users.
Were these programs successful? By all accounts, wildly so. Quality metrics improved rapidly (2-4), hospitalization rates fell, and clinical outcomes improved.(5) The number of veterans who were using VA services grew dramatically, patient satisfaction increased, and yet the cost of caring for each veteran declined dramatically.(6) What lessons does this hold for the rest of the health care system?
Drivers and Facilitators of Quality ImprovementSome of VA's programs drove quality improvement, while others facilitated quality improvement. Data collection and feedback, along with performance contracts and incentives, moved organizations to improve their care. Providing current performance data motivated clinicians and leaders to improve by appealing to their competitive nature: clinicians at one facility could see if their performance was lower than their colleagues in a nearby facility, a powerful motivator for change. Moreover, performance contracts and incentives focused local leadership to put resources into improving care. However, these drivers were not enough.
Decentralization efforts within VA are often underappreciated. By allowing each VISN (and usually each facility) to decide how it should meet its performance targets, VA leadership harnessed local culture and ideas to improve care. The EHR system, CPRS, gave providers a powerful tool for change. Individual facilities could customize their improvement approach. For example, while some facilities might have focused on electronic reminders, others used patient registries to identify high-risk patients and improve their care.
Although much has been written about VA's strides in improving quality, we know less about its efforts to improve patient safety. Over the past decade, VA has implemented an impressive array of programs to improve safety, focusing on a non-punitive approach to build a culture that prioritizes it. These efforts include increasing disclosure of adverse events; root cause analyses; and programs for preventing falls, wrong-site surgery, and nosocomial infections. VA has installed bar-coded medication administration (BCMA) in nearly all hospitals, and most clinicians order medications electronically, which has been shown to reduce medication errors.
However, despite this impressive set of programs, there have been few critical evaluations of the effectiveness of VA's patient safety efforts, at least in the published literature. Although it is possible, even likely, that patient safety has improved in VA since 1999 (when VA's National Center for Patient Safety was launched), at this point, the evidence is more impressionistic and anecdotal than data-based, making it challenging to draw firm conclusions.
Lessons for the Rest of UsIn many ways, VA has served as a laboratory for the IOM's recommendations from Crossing the Quality Chasm. Well before the IOM report, VA had already done many of the things the IOM would later suggest that most organizations adopt: identify priority conditions, take a long-term view of patient care, align incentives with quality, and use clinical information systems. One need not be a public entity with salaried staff clinicians to implement these highly effective programs. Most of VA's lessons are transferable to other organizations.
Many of these same programs are now being implemented in the private sector. For example, the Hospital Quality Alliance (7) and the upcoming Ambulatory Quality Alliance will make data collection and public reporting of quality information routine. Many programs, including Medicare, are experimenting with pay-for-performance plans. Additionally, the President and leaders in Congress have called for increased adoption of clinical information systems. Will the combination of these efforts improve the care for all Americans? Over the long run, they probably will. However, it will surely take time and patience, as these changes will bring as many failures as successes along the way. Furthermore, it must be noted that VA was able to achieve success much more quickly for several reasons that will be hard to replicate: VA has been blessed with inspired leadership dedicated to quality (from Ken Kizer through today's leader, Jon Perlin), and it has cultivated a culture of learning and change. Many VA hospitals are teaching facilities, which may have shortened the provider learning curve for CPOE/EMR use. VA employs its physicians, which made physician buy-in to safety and quality programs far easier than in the more common community-physician model. In the face of these attributes, some private sector organizations will be able to follow, whereas others will surely stumble. But, we now have a proven path to improving health care. And, as we struggle to improve health care, it seems quite fair, even noble, to have begun with our veterans.
Ashish K. Jha, MD, MPHAssistant Professor of Health PolicyHarvard School of Public HealthStaff Physician, VA Boston Healthcare System
1. Corrigan JM, Donaldson MS, Kohn LT, eds. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001.
3. Asch SM, McGlynn EA, Hogan MM, et al. Comparison of quality of care for patients in the Veterans Health Administration and patients in a national sample. Ann Intern Med. 2004;141:938-945. [go to PubMed]
4. Kerr EA, Gerzoff RB, Krein SL, et al. Diabetes care quality in the Veterans Affairs Health Care System and commercial managed care: the TRIAD study. Ann Intern Med. 2004;141:272-281. [go to PubMed]
6. Perlin JB, Kolodner RM, Roswell RH. The Veterans Health Administration: quality, value, accountability, and information as transforming strategies for patient-centered care. Am J Manag Care. 2004;10:828-836. [go to PubMed]