US medicine is exorbitantly expensive and since 1970, that cost has grown 2.4 percentage points faster than the GDP It is extremely difficult to imagine this rate of growth continuing. In 1970, health costs amounted to 7.2% of GDP. In 2005, they amounted to 16% of GDP. They are projected to amount to 19.6% of GDP in 2016 if nothing is done to control them. If health care costs continue to increase faster than does the GDP, health care will become the only thing we produce and consume in this nation. That is not possible. The ever increasing costs must be controlled.
It is not clear that the Democratic Presidential candidates recognize this or that they have the courage to confront the problem. It seems obligatory for them to call for such changes as emphasizing preventive medicine as a means of preventing more serious and costly illnesses in the future. They call for programs of chronic care management to reduce the costs of the populations that consume most health care resources. They call for the universal use of electronic medical records to prevent expensive errors and duplication. They call for new studies of "best practices" to encourage the development of guidelines to encourage providers to use the most cost-effective procedures and products.
Unfortunately these proposals address only minimally the major source of health care costs and cause of their escalation In 2005, about 52% of total health care expenditures went to hospitals (30.8%) and physicians (21.2%). Moreover, physicians are responsible for most other health care spending because they order it. Are hospital costs high; a physician probably ordered the stay. Do too many people see specialists; a physician probably made the referral. Do Americans use too many prescription drugs; a physician prescribed then. Are medical tests overused; a physician probably ordered them.
Researchers at The Dartmouth Institute for Health Policy and Clinical Practice (TDI) have complied detailed data and research reports atThe Dartmouth Atlas of Health Care describing the state of medical care in the United States. (This work is cited by Obama and Clinton.) More important, they have developed a persuasive theory to explain why medical costs are so high in this nation and why, counterintuitively, such high costs do not produce high quality medicine.
The extreme geographical differences in Medicare expenditures is striking. The Dartmouth Atlas of Health Care provides data showing wide variations among states. In 2003, the national average per capita spending was $6,968. New Jersey spent $8,565/capita, 23% above the national average. Hawaii, in contrast, spent $4778/capita, 31% below the national average.
Even more striking, costs varied as much as four times among hospitals in the same state, California. Costs even varied substantially between the University of California at Los Angeles Medical Center (UCLA) and the University of California at San Francisco Medical Center (UCSF). During the last two years of life, Medicare reimbursed UCLA $$72,794/person, while UCSF collected $56,860/person, 78% as much as UCLA collected.
So, can these differences be explained. The Dartmouth researchers think they can. They examine three kinds of care. Essential care, that for which there is substantial clinical evidence that it benefits patients, is underused to the same extent in all regions--high or low cost. The selection of "Preference-sensitive care" alternatives, for which two or more valid treatment strategies are available, perhaps surgery or medical management, does not show a correlation with spending. That is, high cost regions do not systematically choose the highest cost option.
In contrast, there is little or no evidence to guide physicians in making other decisions. How often ought a patient with a given condition see a physician; when should she be hospitalized; when should most diagnostic tests and screenings be done; which patients should be cared for in intensive care units? There is no guidance in relation to these decisions.
It is this "supply sensitive" care, depending primarily on the physician's discretion, that correlates with high spending geographic areas. Two Dartmouth studies demonstrate this correlation here and here. The first study examines the use of supply sensitive services and procedures in specific geographic regions during the final six months of Medicare beneficiaries' lives. The second examines the same variables during the last two years of life. Relying on Medicare data for people who were dead, they looked retrospectively at the medical services and procedures experienced by these beneficiaries. They concluded that
The incremental Medicare dollar spent in regions with higher-than-average spending tends to be for medical specialist visits, diagnostic tests, and use of intensive care and hospitalizations for medical conditions.
Finally, the Dartmouth researchers find no evidence that high cost areas produce better care. Others reach similar conclusions Whether measured by outcomes (i.e., mortality or morbidity) or by the expressed satisfaction of physicians and patients, quality is no better than in low spending areas and is sometimes worse.
So what accounts for this systematic variation in practice patterns. After many years of study, the Dartmouth team has concluded that the culprits are is the local supply of hospital beds and of specialists and the medical culture that creates. As Elliott Fisher, the director of TDI, testified before MedPac, which advises Congress on Medicare policy,
the likely diagnosis for --unwarranted variations in care, poor quality, and growth in spending is the problem of local capacity and culture and the fact that now no one is accountable for local capacity and political culture. Clinical evidence is an important but very limited determinant of physician practice, Physicians practice within a local organizational context in a policy environment that profoundly influences their decision-making. The payment system that we currently have ensure that we're all able to stay busy. And any new capacity, recruiting new physicians, a new orthopedic surgeon, a new interventional cardiologist, is able to stay busy as well. And that creates the culture within which these physicians are practicing. They all need to stay busy, they see their patients more frequency. The consequence is that what appear to the individual or to the system in the current payment system to be reasonable clinical or policy decisions, recruiting new physicians, lead in aggregate to the higher utilization, greater costs, and inadvertently to the lower quality care and the worse outcomes that we see.
(p. 284-85)
If it is true that health care costs vary significantly within our nation, if as healthcare researcher Uwe Reinhardt noted, the best healthcare in the world costs twice as much as the best healthcare in the world, and if the "excess" costs produce no medical advantage, health care policy needs a much more significant reform than just throwing everyone to the tender mercies of the private insurance companies and researching "evidence based" practices to be used by individual physicians.
I will examine some of those policy implications in my next diary on this topic.