IN A DEC. 26, 2007 LETTER to The Journal of the American Medical Association, Bertrand Bell, the physician who chaired a 1987 New York State commission on residency training for doctors, described the startlingly offhand calculation used to decide how long residents could work without endangering patients or themselves. The Bell Commission was created in the wake of the death of a young woman in a New York City hospital, under the care of two unsupervised and apparently overworked residents. The commission’s recommendation of an 80-hour workweek not only became state law in 1989 but, in 2003, also formed the basis of national rules mandated by the Accreditation Council for Graduate Medical Education.

“The specific ‘80-hour week’ was actually determined by a colleague on my porch,” Bell wrote in his letter to JAMA, “and was based on the following informal reasoning: There are 168 hours in a week. It is reasonable for residents to work a 10-hour day for five days a week. It is humane for people to work every fourth night. If you subtract the 50-hour workweek from 168 hours, you end up with 118 hours. If you then divide 118 by four (every fourth night), it equals 30. If you then add 50 to 30, then eureka, that equals an 80-hour week.”

Informal reasoning, indeed. Yet now, in a further attempt to safeguard patients and residents, ACGME is instituting additional limits. This July, new standards for the nation’s 114,115 physicians-in-training go into effect, restricting the number of consecutive hours that interns (first-year residents) may work without sleep, from the customary 30 hours to 16 hours. Meanwhile, more experienced residents will be limited to 24-hour shifts, and they’ll be urged to pursue “strategic napping.”

How these tighter restrictions will affect the many teaching hospitals that depend on residents as a vital resource remains to be seen. “I don’t think anyone would argue against the notion that well-rested trainees can and will provide better care,” says John Co, director of graduate medical education at Partners HealthCare, which has more than 1,700 residents and clinical fellows at Massachusetts General Hospital and Brigham and Women’s Hospital in Boston. “But the question is, how do you implement that in practical terms?”

Perhaps more crucially, medical educators worry that the reduced shifts will provide an insufficient educational experience. “I’m concerned whether residents will see enough patients,” says Craig Brater, a physician and dean of the Indiana University School of Medicine, which has more than 1,000 residents. “Will residencies need to be longer?”

Others are worried that shorter hours might even increase medical errors if the change results in patients being handed off more frequently from one resident to another. And so far, there has been little conclusive evidence to validate Bell’s on-the-porch estimate or to gauge the impact of the 2003 reforms. “We’re making a very big, expensive change in residency programs, and the problem is we don’t have enough high-quality data from real residency programs to know how to do this and improve outcomes,” says Vineet Arora, a physician and associate professor at the Pritzker School of Medicine of the University of Chicago.

For soon-to-be residents, that adds up to an uncomfortable level of uncertainty. “If I’m in the hospital less, I’m afraid I’m going to learn less,” says Celine Goetz, who this summer will start her residency at NewYork-Presbyterian Hospital/Weill Cornell Medical Center. “I interviewed with a dozen residency programs, and the subject of the new rules came up in every one. Every institution seems to have a different philosophy about the changes, and many emphasized that no one really knows what the impact will be for medical education. My class is going to be the guinea pigs.”


SINCE THE EARLY TWENTIETH CENTURY, when U.S. medical school graduates began competing for a scarce number of positions that offered a year or two of living and working in hospitals as “house pupils,” there has been little question about what residency would hold: a rite of passage notorious for hard work and extreme hours. After the First World War, this system of medical apprenticeship evolved into a hospital-based educational program with conferences, clinical rounds, lectures and other types of formal and informal instruction, according to physician and medical historian Kenneth Ludmerer’s Time to Heal: American Medical Education From the Turn of the Century to the Era of Managed Care.

As medical specialties established themselves, they spawned residencies that could last as long as seven years. Medical schools looked upon this additional training as an essential complement to what they provided, Ludmerer explains, while hospitals saw residents as inexpensive providers of up-to-date care for a growing patient population. “That has been the tension from the very beginning, the fact that hospitals have benefited from residents versus the needs of residents to have a genuine educational experience,” says Ludmerer, a professor of medical history at Washington University in St. Louis.

Before long, residents were regularly working 36-hour shifts and 120-hour weeks, in what most physicians considered apt, if grueling, preparation for the realities of professional practice. “I remember being an intern and not having a single day off during 50 weeks and being so tired I couldn’t examine a patient,” says Ludmerer, who trained during the 1970s. Although concerns about resident burnout arose as early as the 1950s, it was several decades before anything was done to address the issue.

According to Ludmerer, it took a “confluence of forces”—including the patient safety movement, advances in data collection and the rise of medical consumerism—to raise public awareness of residency issues. But it was the death of Libby Zion in 1984 that really made people think something had to be done. When Zion, 18, was admitted to New York Hospital in New York City with a 103-degree fever, there was no attending physician on duty. An intern and a second-year resident, working 36-hour shifts, were unable to come up with a diagnosis, and they prescribed a painkiller and a sedative. Zion’s temperature ultimately spiked to 108 degrees, and she died within hours.

Zion’s father, a prominent New York journalist, launched a campaign for greater resident supervision and brought charges against the hospital and the residents. (A grand jury concluded that Zion had died as a result of resident error, but no one was criminally charged.) Then came the Bell Commission and the Bell Regulations, the rules requiring that residents in New York State work no more than an average of 80 hours a week, that shifts be a maximum of 24 hours and that residents be supervised by senior physicians present at the hospital. New York became the first (and is still the only) state to regulate residents’ hours.

On the national front, pressure for reform began mounting as well. In 2001, Public Citizen, a consumer advocacy group, led a coalition that petitioned the U.S. Occupational Safety and Health Administration to regulate resident work hours, and congressional legislation mandating work-hour limits was introduced the same year. OSHA, aware that ACGME—a private nonprofit organization with a board that includes representatives from trade associations for hospitals, physicians and medical schools—was working on new rules, turned down the petition in 2002. A year later, ACGME announced regulations that restricted interns to 80-hour workweeks with one full day off each week and no single shift exceeding 30 hours.

Hospitals struggled to adapt, redrawing work schedules and hiring additional staff to make up for the missing resident hours. But although residents worked shorter shifts, their workloads typically didn’t decline in terms of the number of patients they admitted or managed. “We had shorter hours for residents with more work to do, and it just added to the chaos,” Ludmerer says. What’s more, compliance with the rules was hardly universal. In 2006 and 2007, 16% of sponsoring institutions had racked up at least one duty-hour violation, and by academic year 2010–2011, 56 residency programs were on probation with ACGME for work-hour violations and similar issues.

YET EVEN AS HOSPITALS STRUGGLED TO COMPLY WITH THE 2003 REGULATIONS, evidence mounted that long residency shifts continued to bring the risk of serious medical error. In a 2006 study by the Harvard Work Hours Health and Safety Group at Brigham and Women’s Hospital, for example, one of five resident physicians acknowledged making a fatigue-related error that harmed a patient, and one in 20 said such a mistake had led to a patient’s death. A separate 2006 study by the same group found that residents who worked more than 20 hours at a stretch were 73% more likely to injure themselves with a needle or scalpel than those whose shifts lasted 12 hours.

In 2007, prompted in part by such findings, a subcommittee of the U.S. House of Representatives Committee on Energy and Commerce investigating medical errors requested that the Institute of Medicine determine whether long resident hours were indeed a threat to patient safety. In a 400-page report issued in 2008, the IOM concluded that 30-hour shifts “promote conditions for fatigue-related errors that pose risks to both patients and residents.” The report recommended that residents be restricted to working just 16 hours—or, if doing 30-hour shifts, that they be given five hours of “protected” time for sleep in the hospital after 16 hours of work, and that they not be allowed to admit new patients during the second portion of such shifts. The report also said residents should get a 24-hour period away from the hospital once every seven days and a 48-hour break once a month.

In September 2010, after asking more than 140 medical organizations to weigh in on the conclusions of the IOM report, ACGME published a final version of new rules that included some modifications of the IOM recommendations. In particular, the council chose to focus on interns, limiting first-year residents to 16-hour shifts. (According to ACGME chief executive officer Thomas Nasca, writing in 2010 in The New England Journal of Medicine, the emphasis on interns was guided by surveys suggesting that they routinely log the most hours.) More experienced residents, by contrast, can stay four hours beyond their 24-hour shifts to facilitate the transfer of a patient to another physician’s care. And in “unusual circumstances,” residents may delay their exit to care for a single patient when there are clinical, academic or humanitarian reasons to do so. Other changes to the requirements act on IOM recommendations regarding on-site supervision for interns and providing residents with more time for rest, but the council rejected such measures as granting residents a weekend off each month.

All of these changes come even though research evaluating the effects of the 2003 duty-hour limitations on patient safety and mortality hasn’t been conclusive. One study, published in JAMA, found that shorter shifts for physicians training at the most teaching-intensive Veterans Health Administration hospitals were associated with lower mortality rates in patients with acute myocardial infarction, gastrointestinal bleeding or stroke, among other conditions. But a second study, also published in JAMA, showed that during the first two years of implementation of the 2003 rules, the reduced hours neither worsened nor improved mortality for Medicare patients.

Meanwhile, the science of determining how long residents should work seems incomplete. The IOM report equates the impairment of being awake more than 16 hours to a blood alcohol level of 0.05% to 0.10% (0.08% is the usual legal limit for driving). The IOM also considered such studies as one showing that among interns in intensive care, those working traditional 24-hour shifts made more than five times as many serious diagnostic mistakes as those on the job no more than 16 hours. But Arora of the University of Chicago says there’s scant evidence suggesting that 16 hours is the optimal shift length. “The magic number is one in which residents are well rested and also have an optimal clinical experience,” she notes. “From the data we’ve looked at, based mostly on three small studies, 16 doesn’t seem to be that number. We really need to study this further.”


WHAT’S ALREADY CLEAR IS THAT THE LATEST CHANGES ARE PUTTING SEVERE PRESSURE, financial and otherwise, on hospitals. In its report, the IOM estimated that U.S. teaching hospitals may need to pay as much as $1.7 billion more in labor costs to cover new hiring because of the rule shift, while in a separate report, ACGME’s estimates range from about $400 million to more than $1 billion, depending on how hospitals reallocate staff. Although advocates say these costs pale in comparison to the financial and ethical costs of preventable medical errors, hospitals are getting no help from the federal government to pay for implementing the changes.

Among a handful of specialties, including emergency medicine and anesthesiology, standards have already evolved to cap residents’ hours at levels that fall within the new rules. But programs in such disciplines as internal medicine, pediatrics, surgery and psychiatry are implementing strategies to achieve compliance with the new rules while maintaining educational quality. Among the steps are discontinuing nonessential training; hiring nurse practitioners, physician assistants and physicians to pick up clinical duties; shifting patient care to more senior residents and faculty (without increasing their work hours); and switching patients to nonteaching units (eliminating the need for residents to staff the units and faculty members to oversee them).

New York-Presbyterian Hospital, which sponsors 119 specialty residency programs shared by the Weill Cornell Medical College and Columbia University College of Physicians and Surgeons, has more than 1,600 residents, and because of the state’s long-established duty-hour and resident oversight regulations, “we started ahead of the game,” says Richard Liebowitz, associate chief medical officer of the hospital system, regarding the new ACGME rules. But the system is scrambling to hire at least 75 physician assistants, physicians and other personnel by July to cover the manpower shortage that the tighter restrictions will create. “We’re easily going to spend $7 million to $8 million on new hires alone,” Liebowitz says. “And we’re competing against other hospitals to get the people we need.”

Some changes are happening before the new rules go into effect. “It used to be that everyone admitted to the hospital had a resident taking care of him or her,” says Liebowitz. “Now some will be looked after by hospitalists [physicians who work mainly in the hospital] working with physician assistants. In one of our intensive care units, we already have nurse practitioners taking the place of interns working as part of a care team.” He says that residents are also being taken off some rotations. And he has been working with the heads of other residency programs in the hospital to determine the impact of proposed changes on resident education. “Are residents graduating from our programs now as prepared to act independently as residents were 10 years ago?” he asks. “I believe the answer is yes, but there’s a new concern about what this latest change in regulations will do.”

“Safety net” hospitals, which serve poor and uninsured patients, are particularly dependent on interns and residents as a low-cost labor source. And with public funding already severely compromised, the new ACGME rules pose special challenges for such organizations as Los Angeles County+University of Southern California Medical Center. The system employs 836 residents, most of them county employees. “The county of Los Angeles and the state of California are facing a severe budget crisis, so we’re not expecting much new funding to fill in for lost service time by residents,” says Lawrence Opas, a physician and associate dean at the Keck School of Medicine at USC.

Lacking additional money, the medical center is reconfiguring schedules in internal medicine and pediatrics, for example, to include interns and residents in daytime shifts, a bridge shift between daytime and evening, and another shift bridging late evening and overnight, with one resident and one intern present at each handoff. “We’re trying to provide continuity of care and maintain a team concept,” Opas says. But he expects some residents and faculty members—those with the most experience—to have to shoulder increased workloads.

At Yale–New Haven Medical Center, where residency programs span 75 specialties and include more than 900 residents, implementing the new rules will mean hiring additional physicians, nurse practitioners and physician assistants, as well as other changes. Attending psychiatrists or senior-level residents who are in the hospital’s emergency room can no longer supervise night-shift interns in the psychiatry wards—because the ER is a block away from the main hospital and the new rules require on-site supervision. So Yale–New Haven will replace interns on night shifts in those wards with senior residents.

In the internal medicine department at the University of Chicago Medical Center, the biggest impact of the new rules is on how teams are structured, according to Arora. Current teams include a resident and two interns who stay overnight every fourth night and work from 7 a.m. to noon the following day. Other days, team members have shorter shifts and receive training in various clinical specialties. Though that system will have to change because interns and residents will no longer work the same shifts, “we wanted to preserve the team concept in which a resident teaches interns,” says Arora. So beginning in July, the senior resident on each team will continue to work a 24-hour shift, with one intern beginning a 14-hour shift in the morning and a second starting a 16-hour shift in the evening.

Another consequence, says Arora, is that workloads for interns in internal medicine may need to be reduced. Current rules let them admit as many as five new patients and two transfer patients during a single shift. “To keep the same cap on admissions could create a lot of work compression and undermine any benefit in terms of having better-rested interns,” she says.

To comply with the new rules, MGH’s surgical residency program has been hiring more physician assistants and nurse practitioners as well as revising elements of how residents are trained. Because of necessary hour cuts, general surgical residents can no longer rotate through neurosurgery, urology, orthopedics and anesthesia. Interns have been pulled off transplantation surgery, which are optional rounds, and added to night shifts in mandatory pediatric surgery.

Andrew Warshaw, former surgeon-in-chief at MGH, worries that restricting resident hours could mean that surgical residencies have to be extended beyond the current five to seven years. “There’s an exploding universe of knowledge in medicine, and with more to learn in less time, a lot of residents are finding they can’t get everything they need,” he says. A 2008 poll of chief surgery residents (the first whose entire residencies had been run with the 2003 duty-hour restrictions in place) showed that nearly two-thirds were entering fellowships to further their training—adding more cost to their education and slowing the physician pipeline.

PHYSICIANS ALREADY IN PRACTICE WORRY THAT THE U.S. SYSTEM could be moving away from a tradition of taking individual responsibility for patients. “None of us can picture practicing as they do in other countries, where there’s a 48-hour-per-week limit and physicians basically say, ‘My shift is over,’ ” says Joanne Conroy, a physician and chief health officer for the Association of American Medical Colleges. She notes that Europe’s duty-hour limits have contributed to physician shortages and longer training periods and have spurred concerns about continuity and mastery of skills. Yet there’s little likelihood that hospitals will find the funding to create the additional 8,247 residency positions that the IOM report said would be needed to pick up the slack.

At the same time, health advocates and other groups continue to push for further reforms. Public Citizen, the Committee of Interns and Residents, and the WakeUpDoctor coalition all support a 16-hour shift limit for all residents, but Helen Haskell, president of the patient advocacy group Mothers Against Medical Error, says 16 hours is still too long. “It isn’t safe to have people working more than 12 hours,” says Haskell, who maintains that resident fatigue and lack of supervision played a role in the death of her 15-year-old son Lewis after surgery at the Medical University of South Carolina in 2000 (the school settled with Haskell for $950,000).

There’s also a renewed push from Public Citizen and other groups for OSHA to impose federal oversight on resident duty hours. “The agency is conducting a full and complete review of the petition,” says Kimberly Tucker, an OSHA spokesperson.

Yet while advocates for stricter rules and a federal role cite lax enforcement by ACGME, many residency programs have been disciplined by that body. The MGH surgery program, for example, is on probation because of duty-hour violations, though it has quickly moved into compliance, and Yale’s general surgery program, put on probation in 2004 for similar reasons, had to enact changes to preserve accreditation.

In coming years, research may help establish just how long residents can safely work. A study under way at the University of Pennsylvania, for example, will evaluate how much sleep residents obtain at work and when they’re off duty, both before and after implementation of the 2011 rules, according toDavid Dinges, chief of sleep and chronobiology at the University of Pennsylvania, who served on the committee for the IOM report. “We need to further our understanding of how sleep makes a difference,” says Dinges, who notes that there’s no evidence that trainees take advantage of shorter work hours to sleep more. “Residents must prioritize sleep,” he says. “What residents do in their discretionary time has to be part of the solution.”