Medicare, 25 years old, needs a mid-course correction to assure good care. Such is the recent conclusion of a group commissioned by Congress to seek ways to guarantee high-quality care to Medicare's 33 million recipients.

The Health Care Financing Administration -- which administers Medicare for the elderly and disabled, as well as Medicaid for some of the poor -- lacks "the capacity" to assure first-rate care, said a 17-member committee of the Institute of Medicine of the National Academy of Sciences.

"Care of Medicare patients is generally good," committee member Paul Griner, University of Rochester medical professor, summed up. "But there are major weaknesses. There are patients not receiving care. Even more are receiving unnecessary or inappropriate or flawed care."

Chairman Steven Schroeder, president of the Robert Wood Johnson Foundation, until July chief of medicine at the University of California at San Francisco, said: "There's a lot of poor performance out there . . . poor performance by physicians, over-use of surgery, prescription drugs, diagnostic technology . . . each with a finite chance of patient harm."

These blunt conclusions have implications for the care all Americans get. This is because this committee's work is but one part of two growing movements aimed at improving that care -- and at helping avoid its widespread rationing.

One of these movements seeks to improve doctors' performances, partly by spotting and eliminating "bad apples" but mainly by giving doctors the information to give first-class care.

The other seeks to get that information, to learn what treatments really help patients. Doctors mainly practice good medicine, the committee said, but too much is hit-and-miss, the result not of bad intentions but of a lack of facts.

Study after study shows that much care is unnecessary or inappropriate. Studies also show wide variations in what doctors in adjacent states, even in adjacent counties, think necessary. Care in one town thus may cost far more than care in another, with no apparent differences in results.

Medicare already has an elaborate care-monitoring program, conducted largely by a national network of Peer Review Organizations (PROs).

"In fact," said the Institute of Medicine's study director, Kathleen Lohr, that effort has "ended up as a program for saving money . . . We've tried to shift the emphasis to quality assurance."

Many students of medical care -- though not all -- believe true quality assurance might save even more money. All believe that patients should benefit.

The committee recommended:

A 10-year step-by-step "major redirection" of the way Medicare services are monitored -- to emphasize good care as well as preventing costly and medically dangerous overuse.

Conversion of the present network of PROs -- which spend most of their time trying to prevent unneeded hospital- ization -- into reoriented Medical Quality Review Organizations (MQROs).

Collection and analysis by the new MQROs of data on patient outcomes and the performance both of physicians and hospitals.

Regular feedback of the data to hospitals and doctors, in the belief that doctors often don't know how their care stacks up and that self-review and improvement will work better than over-policing.

Use of corrective action -- including financial penalties or expulsion from the Medicare program -- when self-improvement fails.

Such steps would help "redress the balance between those who give care and those who receive it" -- the persons with the greatest stake in the outcomes, the patients -- said committee member Jerome Grossman, chief executive of the New England Medical Center in Boston.

Gail Wilensky, Medicare administrator, disagreed with the recommendations for change within HCFA. She said they would duplicate and possibly slow the agency's own efforts to monitor care.

By one means or another, however, Medicare's review may evolve into the kind of data-collecting and disseminating that the Institute of Medicine committee suggested. "Quality improvement" and "outcome management" programs are multiplying throughout the medical world.

Nineteen months ago, HCFA proposed using "cost-effectiveness" as a test for funding expensive new technologies or procedures. HCFA has approved the necessary regulations. Final approval is now up to the Bush administration.

Another Institute of Medicine group has been helping HCFA concentrate on three areas -- treatment of breast cancer, heart attacks and hip fractures -- to learn what kind of care works well and what doesn't. The project was launched two years ago by former HCFA Administrator William Roper.

An enlarged federal Agency for Health Care Policy and Research has set Jan. 1 as its deadline to recommend treatment guidelines for three conditions among seven important ones it is studying: chronic pain, bedsores, depression, urinary incontinence, cataracts, sickle-cell disease and prostate enlargement.

The American Medical Association, Rand Corp. and several medical centers are working on similar practice guidelines for a number of procedures. Some medical specialty groups are doing the same.

The Joint Commission on Accreditation of Healthcare Organizations is developing "performance indicators" to measure the results of hospital care.

Some health insurance companies, health plans and employers are already using various systems to measure quality of care and patients' outcomes -- both to decide what to pay for and to try to buy good care rather than pay whether care is good, bad or indifferent.

Critics say some of these systems and programs are trying to do the impossible: decide what doctors should be doing when a true knowledge base has not yet been built.

Others warn that building anything like a complete knowledge base -- and getting habit-bound doctors to apply it -- will take years. Doctors learn to be conservative about change. Too many highly touted new ways turn out after a while to be something less. This makes even the least hidebound physicians tend to stick to old ways until they are truly convinced that new ways are better. And some doctors, it is also true, stick too long.

One of the country's most widely respected economists, Columbia University's Eli Ginzberg, warns that outcome assessment and practice improvement may indeed improve care but are not likely to cut costs very much. He and others say they actually may show that much care is now skimpy and needs to be more extensive, therefore more expensive.

HCFA's Wilensky guesses that the new movements will cut costs but only "by 5 or 10 percent." But any such estimates, she warns, are pure guesses at this point, based on "absolutely nothing" in the way of solid data.

True, various studies find that 5 to 40 percent of various kinds of medical care are "unnecessary" or "inappropriate." But no human endeavor ever achieves perfection, and Harvard physician and analyst Arnold Epstein warns: "Progress will be made, but slowly. Our expectations must be modest if we are to avoid disappointment."

Modest or not, today's visionaries see a day not much farther ahead than the dawn of the 21st century when virtually every doctor will be on a computer network that gives both physician and patient a list of informed options for care, together with the advantages and disadvantages of each and statistics on what has happened to patients who have made each choice.

"We've reached a magic moment" for change, said New England Medical Center physician Grossman. "Everyone is miserable."

If the vision comes true -- every doctor truly informed by facts at a fingertip -- the ultimate effect for patients can hardly be modest.