Resident study: Hours worked do make a difference
■ Medicine mulls a trade-off between a reduction in medical errors and less time for medical training.
By Myrle Croasdale — Posted Dec. 6, 2004
During his first year at Brigham and Women's Hospital in Boston, every medical mistake Christian Ruff, MD, and his fellow interns made in the intensive care unit was tracked as part of the Harvard Work Hours, Health and Safety Study.
Although most mistakes were caught before patients were harmed, the mistakes themselves were preventable, researchers say, because it was primarily fatigue that caused the interns to make them.
Two studies, one measuring the impact of sleep deprivation on medical errors, the other examining how much interns slept and how often they nodded off while on duty, were recently published in the New England Journal of Medicine.
Some physician leaders say the studies vindicate the time limits placed on all residents a year and a half ago. Others say the amount of improvement in patient safety might not justify the cost to resident education. But all agree that the data provide fresh reason to re-evaluate patient care and the hours all physicians, not just residents, work.
For one of the studies, interns worked an average of 80 hours a week, with call periods of 24 hours and longer, then switched to a schedule of 63 hours a week, with call limited to 16 hours.
The interns made 36% more serious medical errors while working the longer schedule than during the shorter one. They also made twice as many serious diagnostic errors when working longer hours.
Ironically, the 80-hour schedule was put into place just 18 months ago in an effort to improve conditions for both patient and resident safety, the researchers noted.
"One of the principles we learned is that it's not the total hours but the number of consecutive hours worked that is the most important," said Charles Czeisler, MD, PhD, division chief of sleep medicine at Brigham and Women's Hospital.
Just taking a nap mid-day before overnight call made a big difference for the interns, Dr. Czeisler said.
Since July 1, 2003, the Accreditation Council for Graduate Medical Education has restricted work hours to an average of 80 per week, with in-hospital call, including patient transfers, capped at 30 hours.
Christopher Landrigan, MD, MPH, lead author of the study, said serious medical mistakes happened more often during the longer call shifts.
"We believe that 30 hours in a row is the genesis for most medical errors," said Dr. Landrigan, who added that there's been almost no data on why those errors occurred. "Our study suggests sleep deprivation is a substantial cause."
Feeling less informed
But reducing residents' hours comes at a price. Dr. Ruff said he felt less informed about his patients when working 63 hours a week.
"Often you felt you didn't know the patients as well because you didn't admit them. You were unfamiliar with the details you would normally know," he said.
Shorter hours also meant fewer educational opportunities. "Attendings teach about patients in the morning. If you aren't there, you miss the game plan for the patient and the didactics," he said. "The attendings need to do two different didactics so both sets of teams are learning from the patient."
Dr. Landrigan agreed, saying shorter workweeks require a change in the culture of medicine. Historically, doctors have considered the patients they admit as their personal responsibility, he said. If they're going home while the patient is still being treated, this attitude must change.
"For optimal patient care, you need to consider yourself one of many taking care of the patient," he said.
Peter W. Carmel, MD, an American Medical Association trustee and the chair of neurosurgery at the University of Medicine and Dentistry of New Jersey, said the studies in the NEJM prove that shorter hours make for safer patients.
The AMA supports the ACGME's work-hour reforms, but there are trade-offs. To accommodate the 80-hour schedule, programs already have increased staff. A further cut in hours could raise personnel costs by a third, if the staffing models in the Harvard study were followed.
A cost to education
There's also the cost to residents' educational experience, particularly those in surgery, who have fewer hours to learn.
"We either have to say we'll lengthen the course of their training or we're satisfied with residents who have operated on fewer cases," Dr. Carmel said.
With the advent of 80-hour weeks, more neurosurgical programs are extending training from five years to six, Dr. Carmel said.
"It may be we're happy with the trade-off and the price we're paying for marginal increases in safety, but the medical budget of the country is already busting," he said. "You can argue if we saved just one life, then it's worth it, but funding for this change will have to be reflected in either increased government funding or decreased available money somewhere else in the system."