Young doctors make far fewer mistakes when their hours are restricted to let them get enough sleep, according to the first study to directly examine the issue.

The study of 24 student doctors caring for seriously ill patients in a hospital found that those who were restricted to working no more than 16 hours without a break made about one-third fewer serious errors that could harm patients.

Since doctors-in-training provide much of the medical care in American hospitals, the findings suggest that current guidelines that allow interns and residents to work long hours without a break are endangering patient safety, the researchers concluded in research published in today's New England Journal of Medicine.

"The tradition of working 30 hours in a row may be the Achilles' heel of the medical education system," said Charles A. Czeisler of Brigham and Women's Hospital in Boston, who helped conduct the research. "Our study challenges the notion that it's safe to deliver patient care under those circumstances."

Czeisler and his colleagues first studied 20 student doctors during their first year after medical school while they worked two three-week rotations in an intensive care unit and a cardiac care unit at Brigham and Women's Hospital in Boston. The researchers compared the impact of a traditional schedule of more than 80 hours per week, including some shifts of 24 hours, with a restricted schedule of less than 80 hours per week with no shifts exceeding 16 hours.

During the restricted schedule, the interns worked 19.5 fewer hours and slept 5.8 more hours each week, and they slept more in the 24 hours preceding each working hour. They were also more alert, nodding off about half as often while on duty.

The researchers then compared the number of mistakes 24 interns made over a year when they rotated between the two schedules for three weeks at a time.

During the traditional schedule, the interns made 35.9 percent more serious errors, including misdiagnosing patients, ordering the wrong medication or dose, interpreting test results incorrectly, or making a mistake during a procedure, the researchers found.

There was no significant difference in the number of patients who died, but that may have been largely because a special panel of doctors was monitoring the interns and caught many of the errors.

Patient-safety advocates hailed the research as a milestone in the long debate over student doctors' hours.

"What we have here is the smoking gun," said Lucian L. Leape, a health care quality expert at the Harvard School of Public Health. Leape was a co-author of an influential 1999 Institute of Medicine report that estimated between 44,000 and 98,000 hospitalized Americans die each year from medical errors. "People have said over and over that a lack of sleep does not inhibit doctor performance. This is the first rigorous study that shows that it does. This is a landmark study."

The Accreditation Council for Graduate Medical Education, which oversees 7,300 medical residency programs nationwide, will examine the new data to see if the current guidelines should be adjusted, said David C. Leach, the group's executive director.

"This may call for a further refinement of our standards," Leach said. He noted that the group last year limited hours to no more than 80 per week, with no more than 24 hours in a row, though another six hours can be added for paperwork and classes. Many factors need to be taken into consideration before further changes are mandated, he said.

"I cannot emphasize enough that this situation is more complicated than just one variable. I don't know if it's as simple as reducing hours," Leach said. "We could end up doing more harm than good."

The issue of excessively long hours for young doctors came into the spotlight in 1984, when the death of Libby Zion, 18, was blamed on overworked, unsupervised medical residents in a New York hospital. It has continued to be the subject of intense debate, with some arguing that long hours enable doctors to bond with their patients and provide a crucial continuity of care.

"The tradition has been for doctors to work these extended shifts because it was felt that it was beneficial for their education and for patient care," said Christopher P. Landrigan of Brigham and Women's Hospital, who led the second study. "The presumption has been a doctor who knows the patient better will provide better care. What our study found is that once a certain biological threshold has been crossed, that is no longer true."

The findings indicate that in addition to limiting the total number of hours interns should work, more attention needs to be paid to the length of shifts, the researchers said.

"Prior to this, there was never been a study demonstrating there's any risk to the patient," Czeisler said. "Now the burden of proof has been shifted to those defending the tradition of working these long hours."