Article Content

Almost one third of Medicare spending for the care of the chronically ill is unnecessary, according to Dartmouth Atlas of Health Care. Improving care and investing in an infrastructure that can better coordinate and integrate care outside of hospitals, for example, home health and hospice care, could also lower costs.


Staggering variations in how hospitals care for chronically ill elderly patients indicate serious problems regarding quality of care and point toward unnecessary spending by Medicare. Lower use of acute care hospitals and physician visits could actually lead to better results for patients and prolong the solvency of the Medicare program, according to a new study by the Center for the Evaluative Clinical Sciences at Dartmouth Medical School. The study calls for overhauling how the nation manages chronic illness and proposes that hospitals take leadership in redesigning how they care for the chronically ill.


Three issues drive the differences in the cost and quality of care, according to principal investigator John E. Wennberg, MD, MPH. "Variation is the result of an unmanaged supply of resources, limited evidence about what kind of care really contributes to the health and longevity of the chronically ill, and falsely optimistic assumptions about the benefits of more aggressive treatment of people who are severely ill with medical conditions that must be managed but can't be cured," said Wennberg.


The Dartmouth Atlas Project (DAP) studied the records of 4.7 million Medicare enrollees who died from 2000 to 2003 and had at least one of 12 chronic illnesses. The study demonstrates that even within this limited patient population, Medicare could have realized substantial savings-$40 billion or nearly one third of what it spent for their care over the 4 years-if all US hospitals practiced at the high-quality, low-cost standard set by the Salt Lake City, UT region. The report comes on the heels of a report by Medicare's trustees that the insurance program will exhaust its trust fund in 2018, 2 years earlier than previously forecast.


The new research is based on Medicare claims data for more than 4300 hospitals in 306 regions, released today in a new database available at For the first time, those who use, provide, pay for, and make policy on America's healthcare system will be able to compare the efficiency of states, regions, and their individual hospitals and associated physicians in treating patients with chronic illness. The new interactive database was funded by the Robert Wood Johnson Foundation, the long-time principal underwriter of the DAP.


A fundamental problem, one that contributes to both overspending and worse outcomes, is that most acute care hospitals have become the first-line providers of services to chronically ill elderly people, whose care would be better managed, safer, and less expensive outside the hospital setting, according to the DAP.


Wennberg refers to the findings that the resources per capita allocated to managing chronic illness during the last 2 years of life are increasing steadily each year. For example, the nation's healthcare providers were using 13.6% more intensive care unit (ICU) beds in 2003 than they did in 2000. Physician labor used to manage chronic illness also increased substantially: 13.4% for medical specialists and 7.7% for primary physicians. The acceleration was greatest in regions that were already using the most care; thus, the gap between high- and low-rate regions grew greater over the 4 years.


The financial incentives used by Medicare and most other payers encourage the overuse of acute care hospital services and the proliferation of medical specialists. The care of people with chronic illness accounts for more than 75% of all US healthcare expenditures, indicating that overuse and overspending are not just a Medicare problem-the healthcare system as a whole has not developed efficient, effective ways of caring for people with severe chronic illnesses.


The study paints a picture of the healthcare system in disarray over the treatment of chronic illness. There are no recognized evidence-based guidelines for when to hospitalize, admit to intensive care, refer to medical specialists or, for most conditions, order diagnostic or imaging tests for patients at given stages of a chronic illness. Lacking this, two factors drive the following decisions: (1) both doctors and patients generally believe that more services-that is, using every available resource such as specialists, hospital and ICU beds, diagnostic tests, imaging, and others-produces better outcomes, and (2) based on this assumption, the supply of resources, not the incidence of illness, drives the use of the services. In effect, the supply of hospital beds, ICU beds, and specialty physicians creates its own demand; thus, areas with more resources per capita have higher costs per capita.


There is no better illustration of the lack of scientific consensus on how to manage chronically ill patients than the wide variations among academic medical centers. Among the U.S. News & World Report honor roll hospitals, the average number of days hospitalized during the last 6 months of life ranged from 12.9 days per decedent at St. Mary's Hospital (the principal hospital of the Mayo Clinic in Rochester, MN) to 23.9 days per decedent at Presbyterian Hospital in New York. The University of California at Los Angeles teaching hospital had the highest average number of days in ICUs during the last 6 months of life (11.4 days per decedent), a rate 3.5 times higher than the rate for patients treated at the University of California at San Francisco teaching hospital (3.3 days per decedent). Medicare enrollees who were patients of the New York University Medical Center had an average of 76.2 physician visits during their last 6 months of life, almost one third more than that of the patients at the next highest rate academic medical center, the Robert Wood Johnson University Hospital (57.7 visits per decedent). Patients of the University of Kentucky Hospital had slightly more than half as many (18.6) physician visits as the national average (33.5).


Hospitals that treat patients more intensively and spent more Medicare dollars did not get better results. Similarly, the regions with the best quality and outcomes used fewer resources relative to their high-cost counterparts. Patients in low-cost, high-quality regions such as Salt Lake City, UT, Rochester, MN, and Portland, OR, are admitted less frequently to hospitals, spend less time in ICUs, and see fewer specialists.


"This carries an important implication for healthcare policy: healthcare organizations serving these low-cost regions aren't withholding needed care," said coauthor Elliott S. Fisher, MD, MPH, senior associate at the VA Outcomes Group and professor of medicine and of community and family medicine at Dartmouth Medical School. "On the contrary, they are more efficient. They achieve equal and often better outcomes with fewer resources. These organizations offer a benchmark of performance toward which other systems should strive."


The researchers studied patients with chronic illnesses because approximately 30% to 35% of Medicare dollars are spent on people with these conditions during the last 2 years of their lives. Two thirds of those in the study were diagnosed with cancer, congestive heart failure, and/or chronic lung disease.


"The majority of acute care hospitals are applying their standard forms of 'rescue medicine' to people who are in advanced stages of diseases that can't be cured," said Wennberg. "Patients don't benefit-they can't be rescued-and the costs of such care are very high, both in dollars spent and in providing care that the majority of chronically ill patients might not want, such as admissions to intensive care and being sent to specialist after specialist."


Using this unique database, researchers compared every region in the country to three regions that provide high-quality, low-cost care: (1) Salt Lake City, UT, served primarily by Intermountain Healthcare; (2) Rochester, MN, served largely by the Mayo Clinic; and (3) Portland, OR, the largest and most metropolitan region in a state that has made improvement in end-of-life care a public policy goal. If, over the 4-year period, every region in the country had used hospitals and physician services in a manner similar to practice patterns in Salt Lake City, the $123 billion Medicare spent for these patients would have been reduced by $40 billion, nearly one third. By the Mayo Clinic benchmark, savings would have been $19 billion. By the Portland benchmark, savings would have been $38 billion.


The report clearly speaks of the need to overhaul the way chronic illness is managed-to redirect resources away from acute care and invest in an infrastructure that can better coordinate and integrate care outside of hospitals, for example, home health and hospice care. A major challenge is to develop reimbursement policies that provide a path toward transition. When payment is based solely on use, hospitals that reduce stays lose money. The report calls for reimbursement system that rewards, rather than penalizes, provider organizations that successfully reduce excessive use of services and develop broader strategies for managing their patients with chronic illness.


Extensive research, both across US regions and among leading academic medical centers, has now documented that greater use of resources is, if anything, associated with worse outcomes, poorer quality, and lower satisfaction with care. The report emphasizes the need for academic medical institutions and federal agencies, such as the National Institutes of Health, to do patient-level studies to produce detailed evidence defining the efficient clinical practices-for example, whom to hospitalize, when to schedule a revisit, or when to refer to a medical specialist, home health agency, or hospice.


The DAP began in 1993 as a study of healthcare markets in the United States, measuring variations in healthcare resources and their use by both geographic areas. More recently, the research agendum has expanded to reporting on the resources and use among patients at specific hospitals. Dartmouth Atlas Project research uses very large claims databases from the Medicare program and other sources to define where Americans seek care and what kind of care they receive and to determine whether increasing investments in healthcare resources and their use results in better health outcomes for Americans.


The study was funded by the Robert Wood Johnson Foundation, in partnership with a funding consortium that includes the Wellpoint Foundation, the Aetna Foundation, the United Health Foundation, and the California HealthCare Foundation.


The Robert Wood Johnson Foundation, based in Princeton, NJ, is the nation's largest philanthropy devoted exclusively to health and healthcare. For more information, visit