In a hard-hitting article in the new issue of Health Affairs, RAND's Elizabeth A. McGlynn and Robert H. Brook, two of the nation's leading experts on health care quality, describe the quality deficit, explain why attention to the problem has never been sustained, and suggest initial steps to improve quality at home and abroad. The goal, they argue, should be a complete "reengineering" of health care practice. To get there, "we need a war on poor quality that has the same level of public commitment as the war on cancer or the campaign to put a man on the moon."

How bad is our quality of care? We don't really know, the authors observe, but the scientific literature is broadly suggestive. Medical errors are rife. Only half the U.S. population receives needed preventive care. Just 60 percent of the chronically ill get the care they need. About one third of the care delivered for some problems (antibiotics prescribed for the common cold, for instance) is not needed and may actually be harmful. Invasive diagnostic tests (angiography, for example) are often performed so unskillfully that the results cannot be accurately interpreted. Moreover, the story in other developed countries with supposedly advanced health care systems is sadly similar.

Why do such basic, dangerous defects persist? McGlynn and Brook respond by citing a series of factors that differentiate the health care system from other economic sectors.

• Diffuse responsibility. Problems with the processes or outcomes of care are rarely the fault of a single company, like Firestone or Boeing. Instead, responsibility is diffused - and solutions must be developed - across thousands of hospitals and practices. "There is rarely a threat that poor-quality providers will be driven out of business or even suffer a significant loss of revenue."

• Cognitive dissonance. Most people assume that their own doctor is excellent and most doctors assume that they deliver good care. Reports on medical errors have come closest to breaking through this barrier, but only briefly.

• Outmoded system design. Despite amazing advances in medical equipment, drugs and techniques, most physicians and hospitals rely on illegible handwritten notes to track a patient's care. Much of the medical establishment dismisses attempts to introduce systems principles. Physicians dismiss implementation of practice guidelines as "cookbook" medicine, as if consistent delivery of best practices is a bad thing. "We cannot recall defective medical care the way we can recall a car," the authors point out. "Systems must be in place to guide doctors' actions while the patient is being seen."

• Information void. There is no national tracking system for identifying defects and correcting them before patients die, few early warning systems in place to ensure implementation of best practices, little systematic information about what reengineering strategies are likely to work on a large scale. Scattered, individual improvement projects can be found but generalized knowledge is lacking.

• A shoot-the-messenger attitude. Doctors and health system administrators often devote more energy to undermining findings about poor quality than to addressing solutions.

Changing the system isn't hopeless, the authors assert, but it will take strategies to generate sustained public interest as well as strong public and private sector leadership. For starters, they suggest these:

• Create quality "champions." Advocacy organizations dedicated to curing specific diseases should also assume the mission of pressuring health systems and public and private purchasers to reward high-quality care consistent with best practices. Medicare could become a potent quality champion in its own right.

• Develop a functional computerized information system. Adequate clinical management hardware and software could become a condition of licensure, contracting, malpractice insurance policies and reimbursement. With help of government and private sector financial incentives - including tax breaks and rewards for quality - a system could be in place within five years.

• Routinely monitor and report on performance. Trends in quality should be the subject of national, federally-funded reports compiled by an independent group. Public reports should be made available on the quality of health care practiced by hospitals, health systems and providers and available in communities.

• Ensure adequate funding for quality measurement. The cost would be several billion dollars per year, a fraction of the budget of the National Institutes of Health ($19 billion in 2001) and a drop in the bucket of the nation's total annual health care bill (more than $1 trillion).

"We must find a way to keep quality of care at the top of the health policy agenda," the authors emphasize. "After providing insurance to all Americans, there is no issue of equal importance."

Elizabeth McGlynn is the director of RAND's Center for Research on the Quality of Health Care. Robert Brook is a RAND vice president and the director of its health program. He is also professor of medicine and health services at UCLA's Center for Health Sciences.