Impact of work-hour regulations on medical residents

Jul 12 2007

Four years ago this month, new rules went into effect that restricted the long hours worked by many physicians in training.

Now researchers at the Stanford University School of Medicine have shown that the regulations are associated with somewhat lower death rates in high-risk medical patients in teaching hospitals.

The new work hours are not, however, associated with an improved outcome for surgical patients in teaching hospitals.

"It's difficult to say, based on our findings, that the regulations are good for everyone," said the lead author of the study, Kanaka Shetty, MD. "But they do appear to have a modest impact on some."

The study, which will appear in the July 17 issue of the Annals of Internal Medicine, is the largest analysis to date of the impact of work-hour regulations on physicians in training, known as medical residents. It appears alongside a Yale University study in the same issue that found the work-hour regulations improved the medical outcomes for internal medicine patients on three of seven measures.

Medical residents, who total more than 100,000 in the United States, have historically logged long work weeks - sometimes more than 100 hours. Concern over these long hours began to grow in the mid-1980s, when a New York patient's death was attributed to medical resident error and fatigue. Numerous studies since have found that sleep deprivation among caretakers can cause clinical performance to suffer.

In 2003, the organization that evaluates and accredits medical residency programs in the United States, the Accreditation Council for Graduate Medical Education, enacted new work limits for residents of no more than 80 hours a week or more than 30 hours straight.

Shetty was doing a residency at New York-Presbyterian Hospital in New York when the regulations went into effect, and he had concerns about the new rules.

"On some level it seems intuitive that reducing hours would be a good thing, but prior studies did not show definite benefits," said Shetty, a Veterans Affairs research fellow in Stanford's Center for Health Policy/Center for Primary Care and Outcomes. But he added, "It seemed hard to believe that there would be a straight benefit and no downside. It didn't seem that the council had fully addressed concerns about continuity of care."

By cutting back hours, residents would be required to hand off patients more frequently to other clinical teams. Shetty worried that important patient details could fall through the cracks in the process, resulting in more medical errors.

Studies on the impact of work-hour regulations have been less than conclusive. A 2004 New England Journal of Medicine study, for example, found that interns working under a work-hours cap made fewer errors than did a similar group under a traditional system. But a Journal of General Internal Medicine study published the same year found no conclusive evidence that resident work-hour restrictions in New York improved outcomes among multiple patient groups.

"The fact that studies have contradicted each other indicates that it's not clear whether there is benefit or harm from the regulations," said Shetty.

Shetty teamed up with Jay Bhattacharya, MD, PhD, assistant professor of medicine with Stanford's Center for Health Policy/Center for Primary Care and Outcomes Research, to compare death rates in patients hospitalized in teaching hospitals before and after August 2003, one month after the rules went into effect. Using a nationally representative data set of hospital patients between 2001 and 2004, they looked at the clinical outcomes of 1.5 million adult patients, which included 1.3 million admitted for a medical diagnosis and about 200,000 admitted for surgery.

After running their analysis, the researchers determined that the regulations appeared to benefit medical patients: for every 400 patients admitted to a teaching hospital after the regulations went into effect, there was one fewer death than before. In other words, Shetty explained, "Before the regulations, 60 out of every 1,000 medical patients died. After the regulations, the number of deaths was 57 or 58 out of every 1,000."

The researchers also determined that the regulations did not appear to have an impact on surgical patients, a finding the authors called "striking." Shetty said there were several possible explanations: the smaller number of surgical patients may have limited the researchers' power to detect statistically significant differences. It's also possible, he said, that any errors due to fatigue may have been offset by transfer-of-care errors, or that the majority of errors occurred during surgery by attending physicians or senior residents - whose hours did not change appreciably under work-hour caps.

Shetty said more study is needed to clarify why there was no apparent impact on surgical patients and to pinpoint the source of the improvements seen in the other hospital patients. If that improvement resulted from shifting patients from inexperienced residents to more-experienced doctors, it could mean residents aren't obtaining sufficient experience and skills during training. In that case, "the regulations' long-term effect could be deleterious," the authors noted in the paper.

Shetty said it would also be important to look at the impact of the work-limit regulations on low-mortality patients, such as those in pediatrics and obstetrics. This study only focused on high-mortality - in general, older and sicker - patients.

"Extrapolating what we saw to all hospital patients is something of a leap of faith," he said. "Hopefully in the future we or someone else will study this and be able to clearly say whether the regulations have been harmful or beneficial."

http://med-www.stanford.edu/