In 2008, Lawrence Casalino and I wrote about the U.S. health care delivery system reform: “At the heart of the challenge is transforming a 19th century craft-oriented delivery system to provide 21st century biomedical science and technology.” Evidence from the past 12 years suggests some progress is occurring, with, for instance, accountable care organizations (ACOs) beginning to reduce spending and achieve savings in Medicare. Further, research shows that in California — which has a high penetration of medical groups using integrated care delivery models — physician organizations operating under capitated budgets have lower costs and higher quality of care than those primarily using fee-for-service. Yet, nationally we continue to waste resources, with health care costs exceeding GDP and affordability a concern for many Americans. In short, the required changes appear to be too slow in coming and the incentives and capabilities for change too weak.

With an eye toward addressing this challenge, we linked data from a national survey of more than 1,000 physician practices to Medicare claims data to look at the relationship between practices and outcomes of care. We thought that patients, particularly high-need or complex patients who received care from practices with higher scores on their use of recommended care management processes, greater electronic record functionality, and greater use of quality improvement processes would have lower ambulatory care–sensitive admission rates such as for diabetes or asthma, lower 30-day unplanned readmission rates, and lower spending per beneficiary. But the data suggest we were largely wrong, with the exception that greater use of care management processes over time was associated with fewer ambulatory care–sensitive admission rates. We also found that larger practices (50 or more physicians) had higher spending per beneficiary, contrary to expectation. This was particularly true for high-need and complex patients compared with smaller practices, with no evidence of higher quality-of-care. This may be partially explained by the fact that even practices scoring in the top quartile were doing 50 percent or less of the recommended capabilities. In brief, we are not yet at the tipping point of what is needed to improve health care quality or contain costs.

Some candidates in the upcoming presidential election are focusing on various proposals for universal coverage, but the ongoing need for delivery system reform is being neglected. Policymakers, payers, and providers should consider the following five actions:

  1. Move more quickly to capitated global budgets for a defined population of patients. These budgets can be negotiated between health insurers and provider organizations, and the process facilitated by all insurers adopting standard encounter forms, billing mechanisms, and quality reporting metrics to reduce administrative costs and burden. Kaiser Permanente has been operating under a global budget since its inception in the 1940s. In addition, the Centers for Medicare and Medicaid Services has launched payment reform initiatives moving toward full risk-bearing contracts for provider organizations. Lessons from these organizations should be captured quickly as we move toward eliminating fee-for-service–based payment.
  2. Provide more care in lower-cost settings. Advances in treatment technologies have made it possible to provide more care in less expensive outpatient settings, rather than in hospital beds. Further, advances in digital technologies — including use of telehealth — allow more care to be provided in patients’ homes or work settings. For higher-need patients, “Hospital at Home” initiatives that allow for daily monitoring and nurse visits have shown to provide high-quality care at a lower cost.
  3. Use lower-cost personnel. In addition to taking advantage of the full competencies of nurses, physician assistants, and pharmacists, greater use can be made of medical assistants and community health workers. Kaiser Permanente has developed a “leveraged primary care model” that makes extensive use of medical assistants and pharmacists combined with real-time electronic health record information in managing complex, high-cost patients.
  4. Put quality improvement processes “on steroids” by adopting a radically new and different way of leading and managing. Often referred to as the “lean management system,” this approach is based on empowering frontline workers (e.g., nurses, physicians, and other direct caregivers) to solve problems; standardizing work processes to eliminate waste and unnecessary variability in performance; implementing a daily management system; using a measurement system that allows shortcomings in performance to be addressed; and having organization leaders serve as coaches to ensure that all activities serving patients are aligned with strategic priorities. Adopting this approach will involve difficult cultural changes but may be the most important missing ingredient in getting to the tipping point.
  5. Address social determinants of health. Social determinants of health — for example, housing, nutrition, education, and community development — have the biggest impact on overall population health. Health care provider organizations and payers need to form effective partnerships with organizations in these other sectors to address the root causes of health care costs and population health status.

The above actions are interdependent; each will be needed for the U.S. health care system to reach the tipping point of sustainable affordability and continuous quality improvement.