SANTA MONICA, CALIF. -- In a review of thousands of case histories that raises serious questions about medical competence, Rand Corp. researchers said in a report published Friday that three common medical procedures are often performed unnecessarily -- in one instance nearly one third of the time.

The study of elderly patients, one of the most meticulous of its type ever conducted, focused on three procedures: carotid endarterectomies, the surgical removal of blockages in the major arteries to the brain; coronary angiography, an X-ray technique in which a tube is inserted into the heart arteries and dye injected through it; and upper gastrointestinal tract endoscopy, an examination of the digestive organs with a fiberoptic-lighted tube.

Unnecessary surgery was most common in the carotid endarterectomies, which are performed to reduce the risk of stroke, the study found.

Researchers concluded that 32.4 percent of the carotid endarterectomies were inappropriate and in an additional 32.3 percent of the cases, indications for the surgery were "equivocal." More than 100,000 of the carotid surgeries are performed in the United States each year.

The Santa Monica-based research team also concluded that about 17 percent of the coronary angiography procedures were unnecessary. A like percentage of inappropriate procedures was discovered for upper gastrointestinal tract endoscopy. More than 500,000 coronary angiograms and 1 million endoscopies are performed in the U.S. annually.

"This is the first study that has attempted to measure appropriateness in a medically detailed, scientific way and found this level of inappropriateness," Dr. Mark R. Chassin, the leader of the research team, said in an interview. "It raises the possibility that we can both improve quality and control costs by trying to reduce inappropriate care."

The study also undercuts the commonly held belief that high levels of inappropriate procedures are linked to the frequency with which the procedures are performed, Chassin said. The researchers found no significant difference in the percentage of unnecessary treatments in areas of the country where the procedures were most common and areas where they were most rare. The $3 million, five-year Rand project evaluated care for elderly and disabled patients insured under the federal Medicare program. It involved reviews of about 5,000 randomly selected medical records from the year 1981 in sites across the country. Panels of expert physicians developed ratings for necessary and unnecessary care based on detailed reviews of existing medical knowledge.

The 54 data collectors followed strict procedures to maintain confidentiality; the final report does not even indicate which procedures were surveyed in which areas of the country. The study was reported in a series of articles that take most of the current issue of the Journal of the American Medical Association.

In total, 819 physicians -- 90 percent of those who were approached -- agreed to the review of as many as 20 of their patients' medical charts; 227 hospitals -- 99 percent of those approached -- also agreed to participate.

A procedure was judged appropriate if its "expected health benefits," including increased life expectancy and pain relief, exceeded its "expected negative consequences," including death, infection and a worsened condition, by a "sufficiently wide margin," according to the report.

Overall, 35.3 percent of all carotid endarterectomies, a surgery with a mortality rate of between 1 and 3 percent, were judged appropriate.

About three quarters of coronary angiograms, a procedure with a death rate of two per 1,000, were judged appropriate, as were a similar number of endoscopies, a procedure with a death rate of one per 10,000.

In almost all cases, the office and hospital medical records were judged sufficiently detailed for rating the appropriateness of the procedures studied. In a related finding, there was little evidence of fraudulent billing, a problem that has plagued government health insurance programs for the poor.

The researchers specifically excluded cost considerations from their ratings scales so that the determinations of appropriateness could be made on a "purely medical" basis. The report acknowledges that the "ratings of appropriateness might have been lower" if the panelists had been asked to include cost considerations as well.

The researchers also point out that the results may be even more "worrisome" than they appear at first glance. This is because when information was ambiguous, the researchers erred on the side of assigning higher appropriateness ratings.

On the basis of the findings, Chassin advised patients not to "accept uncritically" a physicians' recommendations that they undergo risky procedures and to seek second opinions when doctors could not relieve their doubts about the value of proposed treatments.

"There is a strong interest on the part of the public to believe in the scientific basis of medicine, even ... (when) it isn't there," added Dr. John E. Wennberg of Dartmouth Medical School in Hanover, N.H., who wrote an editorial about the Rand study for the medical journal.

The Rand researchers said their data may "stimulate a productive educational dialogue" among physicians. It also may spur insurance companies, corporations and the federal Health Care Financing Administration, one of the study's sponsors, to crack down on unnecessary care through intensified reviews of elective medical and surgical procedures.

Chassin said he and his colleagues will analyze the data further, to see if, for example, specialist physicians provide more or less inappropriate care than general physicians or if rich patients receive more unnecessary care than poor patients or vice-versa. They are also preparing related studies to evaluate the quality of care in the Medicare program and the wide variations among hospitals in their death rates for elderly patients.

Dartmouth's Wennberg, an expert on geographic variations in medical care, said in an interview the Rand studies should spur the development of "a National Institute of Health devoted to health care outcomes."

He added: "The problem is disagreement and uncertainty among physicians about the best way of practicing medicine. The reason is that the studies to find out what works have not been done, and when they have been done, they often have not been done well."

Chassin, however, took issue with Wennberg's pessimistic view of the state of physician knowledge.

"It is certainly true from the most rigorous scientific viewpoint that we know little about the efficacy of most modern medical treatments, but that is not to say physicians do not know how to make people better," he said. "The enormous amount of experience that has been gained over the years is also very important."