Perspective

Low-Cost Lessons from Grand Junction, Colorado

Thomas Bodenheimer, M.D., M.P.H., and David West, M.D.

N Engl J Med 2010; 363:1391-1393October 7, 2010

Article

In August 2009, President Barack Obama traveled to Grand Junction, Colorado, touting that community's health care system as a model for the provision of low-cost, high-quality care. According to the Dartmouth Atlas of Health Care, average per capita Medicare spending in Grand Junction was $6,599 in 2007 — 24% lower than the national average and 60% below high-cost Miami. In 2005, Grand Junction had only 60% as many coronary-artery bypass surgeries in its Medicare population as the national average, 55% as many inpatient coronary angiography procedures, and 61% as many inpatient days during the last 2 years of life. Moreover, Grand Junction scored above the national average on a number of measurements of preventive care, diabetes, asthma, and other quality metrics.

Although the Dartmouth Atlas has been criticized for failing to adjust for regional price variation and differences in health status in various populations,1 three independent observations confirm that Grand Junction provides low-cost health care. First, a 2009 analysis by the Medicare Payment Advisory Commission of regional variation in use of Medicare services, which was adjusted for local variations in prices and population health status, showed that Grand Junction had 81% of the average use nationwide and was the ninth-lowest service user among 404 U.S. geographic areas.2 Second, data on mortality and morbidity from the Robert Wood Johnson Foundation and the University of Wisconsin's County Health Rankings database show that the population of Grand Junction's Mesa County is far less healthy than those of some U.S. counties with high Medicare costs (www.countyhealthrankings.org), although it is number one in Colorado in the quality of clinical care provided. One would expect less-healthy communities to have higher costs. Third, Grand Junction has low health care utilization not only for Medicare beneficiaries, but also for Medicaid enrollees and privately insured individuals. In 2008 and 2009, Mesa County's per-enrollee expenditures for acute care for Medicaid beneficiaries were 37% of the Colorado average; other Colorado counties with healthier populations, which would be expected to have lower costs, had far higher Medicaid expenditures for acute care (http://datacenter.coloradohealthinstitute.org).

Since Gawande showcased Grand Junction in a 2009 New Yorker article,3 numerous reporters have made the pilgrimage to this low-cost mecca, attempting to explain why health care there is cheaper than elsewhere in the United States. The usual explanation is that this town of 50,000 benefits from a cooperative spirit among health plans, hospitals, and physicians, who work together to serve the population. But even if this explanation were accurate, cooperation could not be transferred to other geographic areas — nor could the small-town nature of the Grand Junction community with its relatively homogeneous racial makeup. So what aspects of the Grand Junction success story might be replicable in other communities?

We believe that seven interrelated features of the health care system that may explain the relatively low health care costs could be adopted elsewhere. These are leadership by the primary care community; a payment system involving risk sharing by physicians; equalization of physician payment for the care of Medicare, Medicaid, and privately insured patients; regionalization of services into an orderly system of primary, secondary, and tertiary care; limits on the supply of expensive resources, including specialists, beds, and equipment; payment of primary care physicians for hospital visits; and robust end-of-life care. These features could be replicated in other markets — though generally not without political battles.

The most important event in Grand Junction's health care history was the assumption of leadership by family physicians.4 In the early 1970s, a group of primary care physicians and specialists founded the physician-run Rocky Mountain Health Plans (“Rocky”) and the Mesa County Physicians Independent Practice Association (MCPIPA). Family physicians gained substantial control of these organizations and fostered a culture of incentives for cost control and cost transparency. In 2006, Grand Junction had 85% more family doctors per capita than the national average.

Rocky — which enjoyed a 60% market share, including patients covered by commercial insurance, Medicare, and Medicaid — and MCPIPA began withholding 15% of fees from physicians. Instead of receiving, for example, $20 for a visit, physicians would be paid $17, with $3 going into a risk pool held by MCPIPA. If health care costs were high, the risk pool would be depleted; if costs were kept low, physicians would receive withheld payments at the end of the year. This system created an incentive to keep costs under control.

Withholding payment alone is insufficient, however, because physicians could beat the system by churning hastily through patients' visits and performing excessive procedures. To prevent such behavior, Rocky and MCPIPA created cost profiles of each physician and made them available to all physicians. If a cardiologist, for example, performed twice as many cardiac catheterizations as his or her peers, that physician would be publicly embarrassed — and educated about community norms. If there was no self-correction and the practices did not change, primary care physicians, knowing that high specialist costs meant that they wouldn't receive their withheld money, stopped referring patients to high-cost specialists. Excessive utilization largely came to a halt.

Rocky and Grand Junction's family physicians made another unusual decision: to pay physicians Medicaid fees equal to those for other patients. As a result, Medicaid patients gained access to private primary and specialty care — and became less likely to utilize expensive care in emergency departments. This policy is probably responsible for the low per-enrollee cost for Medicaid acute care.

Many medical communities have two or more hospitals with cardiac-surgery units and other expensive services. To compete for cardiologists and cardiac surgeons, such hospitals create more cardiac-catheterization facilities and perform more coronary angiography and revascularization procedures. In Grand Junction and its surrounding hospital referral region, there is only one tertiary care hospital, St. Mary's, providing interventional cardiac care, neurosurgery, and other subspecialized services. St. Mary's is fed by smaller secondary care hospitals that do not offer expensive interventional services. Moreover, thanks in part to the control of the primary care community over Rocky, St. Mary's has kept its number of beds and amount of expensive equipment at reasonable levels. Because only one hospital can offer interventional cardiac procedures, there isn't room for many cardiologists in Grand Junction; with such limits on facilities and workforce, the rates of such procedures remain low. The same applies to other types of expensive procedures. Yet St. Mary's is rated in the top 10% of hospitals nationwide in providing patients who have had heart attacks with aspirin, beta-blockers, and smoking-cessation assistance on discharge.5

When hospitalists began to assume care for inpatients, Rocky (unlike many insurers) continued to pay primary care physicians for visiting their patients in the hospital. This policy, which improves the continuity of care and facilitates care coordination at the time of discharge, helps to reduce the likelihood of unnecessary readmissions.

Finally, with the encouragement of family physicians, low-cost end-of-life care became a prominent part of the region's health care system. Thanks to the area's single nonprofit hospice, which also offers palliative care, physicians are educated about initiating discussions with elderly patients about advance directives, and the public is informed about end-of-life choices. As a result, Grand Junction's population spends 40% fewer days in the hospital during the last 6 months of life and 74% more days in hospice than the national averages, and 50% fewer deaths than average occur in the hospital.

Leaders in cities and towns throughout the United States might consider the feasibility of introducing these seven features into their own health care systems. The first step is for primary care physicians to come together and explore ways in which Grand Junction's story could be adapted to their local realities.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

This article (10.1056/NEJMp1008450) was published on September 29, 2010, at NEJM.org.

Source Information

From the Center for Excellence in Primary Care in the Department of Family and Community Medicine, University of California, San Francisco, School of Medicine, San Francisco (T.B.); and St. Mary's Hospital, Grand Junction, CO (D.W.).

References

References

1
Berenson RA. Implementing health care reform -- why Medicare matters. N Engl J Med 2010;363:101-103
Full Text | Web of Science | Medline

2
Report to the Congress: measuring regional variation in service use. Washington, DC: Medicare Payment Advisory Commission, December 2009.

3
Gawande A. The cost conundrum. The New Yorker. June 1, 2009.

4
West D. Mesa County, Colorado, health care: the best health care in the United States. Aurora: Colorado Academy of Family Physicians, August 2009. (http://www.coloradoafp.org.)