Published in the January 2005 issue of Today’s Hospitalist
When new rules took effect last summer requiring residency programs to limit the hours worked by housestaff, educators were concerned. While residents who work fewer hours should be able to function better, physicians worried that disruptions in continuity of care due to more frequent handoffs would hurt not only residents’ educational experience, but the quality of care at the nation’s teaching hospitals.
Even as many teaching programs continue to struggle with limits on residents’ hours, new research explores the relationship between work hours and errors. Results from the study indicate that in the end, sleep may be a more critical ingredient in patient care “and safety “than many physicians realize.
The study, which appeared in the Oct. 28 New England Journal of Medicine, compared the number of errors made by interns working in critical care units under two different schedules: a traditional schedule with shifts of 24 hours or more every other shift, and an intervention schedule that eliminated the “every third night” call schedule and reduced the total number of hours worked per week.
Researchers note that continuity of care issues surfaced in the group working fewer hours, particularly when it came to activities like morning report. They also note, however, that those problems paled in comparison when they looked at the number of errors made by residents who were working longer hours.
The results, researchers say, not only shed new light on the relationship between work hours, sleep and patient safety. They also show how medical educators need to take the issue of sleep deprivation more seriously.
The error effect
Decades of sleep deprivation research has shown that lack of sleep hurts performance in all professions, including medicine. But Christopher P. Landrigan, MD, MPH, a hospitalist at Children’s Hospital Boston and lead author of the study, says that this is the first research to conduct a systematic study of the relationship between work hours, sleep deprivation and patient safety.
Interns working the traditional schedule made 35.9 percent more serious medical errors than interns working the reduced schedule. Interns working the traditional schedule also made 20 percent more serious medication errors and more than five times as many serious diagnostic errors.
(For more detail on the types of errors caught in the study, see “Work hours and errors: a look at three types of mistakes made by interns” below.)
Dr. Landrigan, who is research director of the inpatient pediatric service at Children’s Hospital Boston and director of the sleep and patient safety program at Brigham and Women’s Hospital, where the study was conducted, says he was surprised by the results.
“If anything,” he says, “we were expecting that diagnostic errors might be increased in the intervention group because of continuity factors. We thought we might see interns fail to follow up on test results or misdiagnose patients because they didn’t know the patients well. But it turned out that sleep deprivation had an even more profound effect.”
Dr. Landrigan notes that working fewer hours did have an effect on continuity of care, but to a limited degree.
“There was a sense among both interns and staff members that there was disorganization on rounds the following morning” among interns in the intervention group, he says. “That’s because the intern who worked overnight had not done all the admissions him or herself and didn’t know all the details about the patient’s past when presenting on rounds the next day.”
Despite those disruptions, the better-rested interns still made fewer errors. “They performed better from the standpoint of safety,” Dr. Landrigan says. “They made fewer mistakes caring for patients when they were well-rested, even though they presumably knew fewer details about their patients’ past.”
He adds that the results may be a sign that sleep is more valuable than continuity in patient care. “They also suggest to me that the issue of sleep deprivation has been underestimated by medical personnel,” Dr. Landrigan explains.
While physicians need to be concerned about errors that occur because of discontinuity problems, he says, “We also need to be very cognizant about the fact that sleep deprivation is a very powerful force. I think that it can potentially overwhelm continuity errors.”
(Another study in the same issue of the New England Journal of Medicine found that interns who worked more than 80 hours a week experienced twice as many “attentional failures” while working on-call at night. Dr. Landrigan was a co-author of that study.)
Issues of continuity
While the big news from the study focused on the relationship between work hours and errors, Dr. Landrigan says that concerns about continuity are still critical, particularly as teaching hospitals try to trim their residents’ schedules.
“Although this intervention was certainly a success from the standpoint of patient safety,” he says, “as these types of schedules are implemented, there has to be a very strong focus on sign-out issues.”
As part of the study, researchers tried to address those concerns by developing a sign-out template to help streamline handoffs. The problem was that many of the interns in the study chose not to use the template.
“They were too busy, and it wasn’t part of what they expected when they walked into the rotation,” Dr. Landrigan recalls. “Because it wasn’t part of the normal way of doing business, it was difficult to implement that change when we were also changing the interns’ hours.”
And while he acknowledges that the study’s sign-out template was not a success, he says there is more that can be done. “If we do a better job of providing team training,” he explains, “we could have well-rested interns who are exchanging information regularly without much loss. It’s much like the hospitalist model, where if there’s been a good handoff between the primary care physician and the hospitalist, the care of the hospitalist can be as good or better than that of the primary care provider.”
Dr. Landrigan also points out some emerging trends, some of which involve hospitalists, may make it easier for teaching hospitals to ensure continuity of care while reducing the number of hours worked by residents. He notes that in the last year, for example, Children’s Hospital Boston has moved to 24-hour coverage. The goal is to make sure that there is direct attending supervision at all times.
“There is certainly some shift toward having more senior personnel in the new era of medicine,” he says, with hospitalists playing a major role. That trend, he says, may help ease some concerns about gaps in continuity of care.
Cost considerations
While the study found that reducing interns’ work hours helped reduce errors, it raises significant questions about costs.
In designing the intervention group, researchers opted against using traditional night fl oat coverage. Dr. Landrigan explains that night fl oat systems typically rely on residents from other services. They tend to replace tired residents with equally tired residents, except the replacement housestaff don’t even know the patients.
“There have been a couple of preliminary studies that not surprisingly suggested that under these circumstances,” he explains, “night fl oat residents are exhausted themselves and don’t perform very well.”
Night float coverage may not be ideal, but the alternative ” adding another member to the team “boosts a training program’s costs. For large teaching hospitals with many residents, that expense grows quickly.
While Dr. Landrigan acknowledges that his study’s staffing model would raise up-front staffing costs, he argues that at least some of those costs could be offset by savings due to reductions in errors.
“The typical daily cost for a 10-bed ICU is about $50,000,” he explains. “A resident’s annual salary is about $50,000. It doesn’t take many averted adverse events, which lead to increases in avoidable bed days, to pay for an additional resident’s salary.”
“If you believe that we would prevent adverse events with these kinds of schedules,” he continues, “that would presumably save costs. How those factors would balance out exactly is difficult and requires a fairly sophisticated cost-effectiveness analysis, which we have not done yet. Our suspicion, however, is that scheduling interventions could ultimately be cost-saving.”
While costs are always an issue, he says, the bottom line is that shorter shifts improve patient safety, a goal that is on the radar of hospitals everywhere. He notes that data have consistently shown that shorter shifts equal improved quality.
“Studies in other settings suggest that 12-hour shifts are safer than 16-hour shifts,” he explains, “and that eight-hour shifts are safer than 12-hour shifts. In medicine, we’re starting in a very different place, so we can only push so far at once. But if your ultimate goal is to make people as alert as possible, you would reduce the shifts even further.”
Edward Doyle is Editor of Today’s Hospitalist.
Work hours and errors: a look at the mistakes interns make
When researchers examined the difference in errors made by two groups of interns working different schedules in critical care units, the results were startling. Christopher P. Landrigan, MD, MPH, a hospitalist at Children’s Hospital Boston and the lead author of the study, says that most errors fell into three categories:
“¢ Drug errors. Problems included overdoses of medications, more than one order for the same medication, drugs ordered for administration via the wrong route, wrong drugs, and wrong-patient errors.
“¢ Procedural errors. Errors included problems in the placement of catheters and preventable injuries during placement of a central line or an arterial blood draw because of poor technique.
“¢ Diagnostic errors. Problems included errors made in the collection of information, the failure to check on critical test results, and errors in the interpretation of test results.
Dr. Landrigan notes that about half of all errors were intercepted by nurses and other clinicians before they reached patients. He also notes that most of the errors didn’t seriously harm patients.
