edical errors are killing tens of thousands of Americans each year and harming countless more, so it has been a salutary trend that many medical, academic and business groups have been developing ways to reduce the dangers. But now a survey of practicing physicians has revealed disheartening evidence that the doctors themselves may be the biggest obstacle to effective reform.

Three years ago the Institute of Medicine estimated that 44,000 to 98,000 patients die each year because of medical mistakes — more than are killed annually by automobile accidents. The numbers seemed so staggering that many medical practitioners thought them inflated. But the survey of physicians, published in The New England Journal of Medicine last week, has offered corroborating evidence that, whatever the number of deaths might be, there are an awful lot of medical mistakes causing an awful lot of damage.

The survey, conducted by the Harvard School of Public Health and the Henry J. Kaiser Family Foundation, examined the views of more than 800 American doctors and 1,200 other adults. Fully 35 percent of the doctors said that either they or members of their families had experienced medical errors in the course of being treated, and most said the errors had "serious health consequences," such as death, long-term disability or severe pain. Three in ten had seen an error that caused serious harm to patients outside their families in the past year.

The critical issue, of course, is how to prevent harm, and here the survey found troubling attitudes. Although studies have demonstrated that various technological and procedural changes can cut the error rates in hospitals, the practicing physicians were lukewarm toward many reforms. Only a third of the physicians, for example, consider that reducing the work hours of young doctors in training to avoid fatigue is a "very effective" strategy to cut errors.

Less than a quarter of the doctors think it would be very effective to use computers instead of paper forms to order drugs or to include pharmacists on hospital rounds, two approaches that have been shown to reduce medication errors in hospitals. Nor were they enthusiastic about using only specially trained physicians on intensive care wards, or about limiting high-risk medical procedures to hospitals that do a lot of them, despite evidence that expertise and frequent practice are key ingredients in successful medicine.

With the evidence growing ever stronger that medical errors are a danger to many patients, it is disturbing to find such retrograde attitudes among physicians. Reform can succeed only if the medical profession gets behind changes that expert groups and plain common sense suggest could significantly reduce the harm caused by medical errors.