A wake-up call

Dr. David Andorsky, a resident at Brigham and Women's Hospital, woke up from a two-hour nap in the on-call room at 6:30 a.m. last Tuesday. Andorsky had been on duty since 7 a.m. the previous day and still had a few hours to go caring for patients in the intensive care unit.

But because of the hospital's ground-breaking research on medical errors, Andorsky, 29, was not allowed to order medications for patients during morning rounds.

Instead, that task went to Dr. Anne Liu, who was fresh from a full night's sleep at home.

The hospital is placing new restrictions on how many hours doctors can work, the result of findings published by Brigham researchers in today's New England Journal of Medicine. Researchers found that first-year residents, or interns, who worked 24 to 34 consecutive hours every three days committed 35 percent more serious medical errors than when they worked a maximum of 17 hours.

While working the longer shifts, interns made 136 serious errors for every 1,000 patient days, compared to 100 errors on the shorter shifts. Patient days are the combined number of days all patients spent in the ICU during the study period. On both shifts, the vast majority of errors involved medications. Usually interns prescribed the wrong dose.

Andorsky, who participated in the study, is now a third-year resident. "I feel awake right now," he said Tuesday morning after his short rest. "I don't see why I shouldn't be able to enter a simple order. But that's why we do these studies. Because everyone thinks they can do it. It's easy to put in the wrong medication for a patient when you've been up all night."

To many people, the Brigham findings may sound supremely obvious. Anyone who's stayed up all night knows how hard it is to stay alert the next day. But the medical profession has been slow to acknowledge the dangers of sleep deprivation, and scientific studies on the impact of long hours are rare.

This is partly because of how medical training evolved. Residents once lived at their hospitals and followed cases from beginning to end as part of their training. Getting to know patients and seeing them through their hospital stays still is considered an essential part of residents' education.

Some older doctors also believe that the punishing hours weed out young medical school graduates who aren't hardy enough for the profession's demands. And even physicians who are open to change fear that shorter hours will hurt residents' education, as well as increase "hand-off errors," which occur during the confusion when a doctor arriving at work takes over patients for doctors going home.

But the Brigham is trying to take a hard look at medical errors, even if it means opening the institution to public scrutiny. "The hospital is very courageous in allowing us to put the institution under a microscope," said Dr. Charles Czeisler, one of the authors and a nationally known sleep researcher.

The results have national implications, as the 100,000 residents now training in US teaching hospitals routinely work more than 24 hours straight. Errors could be even higher at other hospitals, as the Brigham has a sophisticated computer software system that in many cases stops doctors from ordering the wrong doses or drugs for patients.

In the study, senior residents, doctors, and nurses intercepted more than half of the interns' errors before they reached patients, suggesting that a strong safety net exists in at least some hospitals. In one instance, an intern prepared to slide a needle into the left side of patient's chest and draw out fluid. But a senior resident walked into the room, saw what was happening, and told the intern the fluid was on the right side.

But senior doctors did not catch all of the errors, including one in which an intern ordered an antibiotic for a patient -- even though he had a known allergy to the drug. The patient received one dose before the mistake was discovered, but he did not have an allergic reaction.

Dr. Christopher Landrigan, one of the researchers and director of the hospital's sleep and patient safety program, said all of the errors combined caused harm to a patient in the two ICUs every three days. But researchers defined harm very broadly, and in most cases, harm meant a patient had an allergic reaction, an irregular heart beat, or that treatment was delayed. No intern errors led to a patient's death, he said.

In an editorial accompanying the Brigham studies, Dr. Jeffrey Drazen, editor of the New England Journal of Medicine, said the current system is flawed, but he raised concerns about shorter shifts. Drazen, who worked as an attending physician in the Brigham ICU during the study, said that the interns working shorter shifts were "more awake." But he said they "often knew very little about patients admitted the night before they came on duty," indicating that hospitals must find ways to provide residents with more complete information about patients' histories as they move to shorter shifts. The study did not specifically record so-called hand-off errors.

While he praised the Brigham for opening the ICUs to public scrutiny, he pointed out the number of errors even when interns worked just 16 consecutive hours, saying that "the overall ICU performance in this trial was not exactly stellar." Given this, he said, additional solutions are needed, such as more rigorous teamwork.

During the study, 24 interns in the hospital's medical and cardiac intensive care units were studied during three weeks of a traditional schedule: from 7 a.m. the first day to 1 p.m. the following day, every three days. For another three weeks, interns worked a modified schedule: researchers added an extra intern to the rotation and divided the overnight shift between two interns. In the end interns on the modified schedule slept an extra six hours per week and worked about 60 hours total. They worked about 80 hours a week on the traditional schedule.

Interns who participated in the study from July 2002 to June 2003 allowed researchers to glue 10 electrodes to their heads for long stretches to measure their drowsiness. They wore the contraptions while talking to patients, watching movies, eating in restaurants, and sleeping. Portable monitors attached to the electrodes recorded more than twice as many instances of profound fatigue at night--measured by interns' eyes rolling back in their heads--in first-year residents working the longer shifts.

Dr. Aaron Kesselheim, now a third-year resident, said the strangest part of participating in the study was seeing "The Matrix" on opening night with electrodes pasted to his skull.

"The data doesn't surprise me," he said. "It's common sense if you're awake 24 to 26 hours you're going to be really tired. But people in medicine don't like to change their practices without good hard data."

Kesselheim said he's been so tired he's fallen asleep on morning rounds, while standing up. Before the study began, Andorsky said, he once ordered the wrong antibiotic for a patient after being awake all night. But, he said, he realized the error and corrected it before sending the order to nurses.

Landrigan said that the study has limitations. Being done in one hospital, it was too small to really measure patient harm, since such adverse events are rare to begin with.

But Landrigan said it was a significant step and taught doctors several lessons.

In July 2003, the Accreditation Council for Graduate Medical Education adopted new limits on resident work hours, restricting their average week to 80 hours on duty. But the new rules still allow residents to work up to 30 consecutive hours -- which Landrigan said appears to be too much.

In the past several months, the Brigham began prohibiting all residents on duty 24 hours or longer from ordering medications, said chief medical officer Dr. Andy Whittemore.

And while most surgery shifts are now 12 hours because of the new ACGME rules, the hospital plans to reduce shifts for all residents to no more than 16 hours by next summer -- a plan that will cost hundreds of thousands of dollars to hire additional staff.

"On the basis of four data right now, interns in high-intensity settings should not be working shifts longer than 16 hours, and really the number may be lower than that," Landrigan said. "My sense is that five or so years from now, people are going to have to prove that working these extended shifts is actually worth it."

Liz Kowalczyk can be reached at kowalczyk@globe.com.