In This Section      

Archives - The Big Squeeze

Winter 2014

The Big Squeeze

Date: February 1, 2014

“It’s my feeling, and the feeling of many of my colleagues,” says Sanjay Desai, “that this 2011 policy is just not informed with enough data.”

While in the thick of it, Sanjay Desai never got too worked up about the ungodly hours he was racking up at the hospital. He finished his residency in internal medicine at Johns Hopkins in 2003, just before the Accreditation Council for Graduate Medical Education (ACGME) first imposed restrictions on duty hours.

"Most people back then weren't asking for limits, including myself," he says. "It was the way things had always been done, and so we did it."

Oftentimes, that meant working 36-hour shifts, broken up here and there by naps that invariably turned out to be brief. At the time, Desai's fiancée was living in Philadelphia. There were days when he'd get off one of those marathon shifts and hop in his car to head north on Interstate 95.

"It was risky for me, especially in terms of driving," he says. "As I reflect back on it now, I can see that it was just not a safe situation. And I can see, too, that there were definitely times when my fatigue was not good for my patients and was compromising my education."

Today, Desai is the director of that same residency program. Now he's facing the challenge from the other end, striving to make sure residents get all the education they need and all the patient-care experience they can while working in accord with strict limits on hours.

It's been a bumpy ride at times, especially after 2011, when the ACGME imposed a second and more detailed round of restrictions on first-year interns. The experience of adjusting to that set of rules has Desai working toward the launch of what may well be the biggest randomized trial ever undertaken in the field of medical education. He'll be aiming to deliver definitive answers about which set of duty-hour restrictions delivers better results for both patients and residents.

"It's my feeling, and the feeling of many of my colleagues, that this 2011 policy is just not informed with enough data," he says. "It needs to be studied in a more rigorous way."

Back in 2003, the ACGME's stated goal in imposing duty-hour restrictions for the first time was to improve patient safety. If residents were less fatigued, the thinking went, they would make fewer mistakes, and patient outcomes would improve.

That's not the way things worked out. Data gathered across a number of specialties after those restrictions took effect showed patient outcomes staying basically the same, getting neither significantly better nor significantly worse. That came as reassuring news to residency directors, Desai says, as it indicated that systems previously in place to catch and prevent fatigue-related errors were successful.

"I always thought it was a bit of a farce to think that the quality of patient care would get better because of the restrictions," agrees Pamela Lipsett, the program director for general surgery at Hopkins and the winner of several outstanding teaching awards over the years. "The supervisory structure in surgery has always been quite strong."

The 2003 rules initially generated a great deal of resistance and pushback from medical educators. The rules capped workweeks at 80 hours and shifts at 24 hours (with six more hours allowed when needed for continuity of care). Traditionally, residents had logged a lot of 100-hour weeks, and it wasn't unusual to hit 120 hours occasionally.

Opponents of the new restrictions did the math and argued that cutting training hours by 20 percent would translate into young physicians heading out into practice with only 80 percent of the training they need.

Another worry centered on an underlying message the new rules might send to trainees. Historically, residency programs have been the medical version of military boot camps. They were the place where trainees got pushed to the outer limits of endurance. They were the ordeal that trainees endured on the way to becoming properly indoctrinated with the ethos of putting patient care first.

"Now we're telling the interns that they're just shift workers," says Lipsett. "That can be a bad thing, I think. It delays the development in them of that all-out attention to the patient."

But as residency directors restructured their programs, they also found that the 2003 rules sparked some positive changes. Forced to hone their focus on the core goals, program directors needed to take a hard look at all the tasks traditionally assigned to residents. Some of them were extraneous.

"It used to be that residents would spend hours on the phone doing things like scheduling radiology tests," says Julia McMillan, a pediatrician who serves as associate dean for graduate medical education in the School of Medicine. "They're not doing those things so much anymore."

Of course, someone still had to complete those tasks. That generally meant hiring more nurse practitioners, physician assistants, and administrative staff. "In that sense, this has cost the institution and the departments an incredible amount of money," Lipsett says.

Making broad generalizations about the impacts of the rules is exceedingly difficult, McMillan warns. Residencies are structured in different ways in different specialties and at different institutions. Some programs hit all kinds of bumps in the road as the rules took effect, while others had much smoother rides.

But with that caveat issued, McMillan does say that it has become clearer over time that the restrictions are improving resident quality of life.

"I don't think there is any question whether it has been better for their personal lives, and studies are starting to verify that," McMillan says. "I can see it when we have our noon conference in pediatrics every day. No one is ever sleeping during that meeting anymore."

Lipsett agrees. "In hindsight, the hours residents used to work were ridiculous, and too many of those hours were spent on things that weren't really about doctoring," she says. "These rules have made the lives of residents better, and I think it's about time we paid attention to that."

Desai noticed an interesting phenomenon in internal medicine, where opposition to the 2003 rules was initially quite strong. Even after data came in showing no improvements in the stated goal of patient safety, there weren't that many folks in his field demanding a reversal of the rules.

"Once the change was absorbed, I think people began to see more clearly that some things maybe weren't perfect with the way things were done before," he says. "The idea that fatigue management was important became more accepted. Over time, the 2003 rules started to feel like a reasonable compromise between fatigue management and our educational mission."

Again, though, the experience can vary widely by specialty. In surgery, Lipsett reports that opposition to the restrictions remains quite strong today. She finds herself in a small minority among her program-director colleagues.

"The way I look at it is a little different," she says, "If we can't train somebody in 80 hours a week, there's something wrong with us."

The second and stricter round of duty-hour rules imposed by ACGME focused on first-year interns. Starting in 2011, they have been limited to 16-hour shifts, down from a maximum of 30. The rules surrounding the way interns are supervised were tightened as well, with supervisors required to be at least on site and available if not right at the residents' sides. Intern supervision from offsite over the telephone is no longer allowed.

The tightening came in reaction to a 2008 report by the Institute of Medicine (IOM) that cited "considerable scientific evidence" that 30 hours of continuous wakefulness leads to a dangerous level of fatigue. Desai, however, is not at all sure that the evidence IOM relied on relates these data to the effects the new rules may have on patient safety.

"The problem is that when you shorten duty hours to reduce risks from fatigue, you also increase provider handoffs, and those transitions of care can create new risks," Desai says. "The IOM was primarily talking about one study that was done in internal medicine, just one study, and none of the errors observed would have ever reached the level of the patient."

"The study is necessarily un-blinded," he continues, "and the residents themselves described in an editorial a change in behavior that may have influenced the study. It's good evidence, but it doesn't seem to me to be enough evidence on which to base national policy for everyone in postgraduate training all across the board."

Lauren Block finished her residency in internal medicine at Johns Hopkins just before the second round of rules went into effect. She is now at the Hofstra North Shore LIJ School of Medicine in New York.

"We residents were seeing this big change happening, and it just sort of naturally became a topic that was near and dear to my heart," Block says. So she and several colleagues launched a series of studies designed to measure the impact of the new rules.

One of those studies was a time motion analysis. Trained observers followed 29 internal medicine interns at Johns Hopkins and the University of Maryland over the course of three weeks in 2012, tracking minute by minute how the interns spent their time.

The results were published last summer in the Journal of General Internal Medicine: Interns were spending just 12 percent of their time with patients, compared with a whopping 40 percent of their time on computers.

"That's the number from all this research that is really resonating with people," Block says. "Residents are spending about eight minutes per day with each of eight patients. It's something everyone finds frustrating, the way time with patients seems to keep getting squeezed."

Comparable studies conducted back in 1989 and 1993 found that residents back then were also spending more time on documentation than with patients, but the patient time was at 20 percent, not 12. Both Block and McMillan caution that the recent findings may not be attributable solely to work hour restrictions. The two mention a number of other potential factors, including the introduction of electronic medical records and shorter hospital stays.

"An awful lot of other things have changed in medicine and medical education over the last 20 years," McMillan says, "and many of those changes are creating more computer work."

One key going forward, Block says, is that residency programs must continue to fine-tune the ways they manage resident time under the new rules.

"There are opportunities out there for these programs to develop creative solutions," she says. One surprising result of the time-motion study is that interns spent 7 percent of their time walking from place to place in the hospital. Can beds be rearranged or schedules adjusted to free up some of that time and devote it to patient visits?

Even if patient encounters remain brief, Block points out, residency programs could focus more intensely on making the most of those encounters. Another of her team's studies found that first-year interns at Hopkins were unlikely to introduce themselves to patients and that they seldom sat down during examinations. Both practices are correlated highly with a better patient experience.

"It's no wonder if patients don't feel connected to what we are telling them because many times we are not doing as much as we could to make that connection," Block says.

For Desai, the adjustment in internal medicine to the 2011 rules proved difficult, and he tried to approach it in an evidence-based way. In the months leading up to the change, he randomly assigned interns to one of three different schedules-a 2003-compliant model of being on call every fourth night with a 30-hour limit, or one of two different 2011-compliant schedules.

The latter included being on call every fifth night but working only 16 hours straight or a so-called "night float schedule," with interns working a regular week of night shifts not exceeding 16 hours.

His results were published earlier this year in JAMA Internal Medicine. The key finding here was that the number of patient handoffs from one doctor to another skyrocketed under the 2011 rules, from an average of three on the 2003-
compliant schedule to as many as nine on the shorter shifts.

Handoffs are dangerous, with each new one carrying risks of communication breakdowns. What if a rushed physician forgets to include a key bit of information? Or what if she neglects to mention that an important lab result is outstanding? In 2005, a Joint Commission analysis found that such communication problems were the primary cause of 70 percent of "sentinel events"-the unexpected occurrence in patients of death and serious physical or psychological injuries. Fully half of those communication failures happened during handoffs, the commission found.

Keep in mind that here, too, the experiences differ from specialty to specialty. The pediatrics residency program, for example, restructured the way it set up shifts. Residents now work in 12-hour shifts for a series of consecutive days, with the same two residents assigned throughout that stretch to the same ward.

"So while it's true that they sign out twice a day, the resident they're signing out to has only been gone for 12 hours," McMillan says. "In the past, residents did rotations every fourth night. Yes, there were fewer sign-outs then. But there were bigger risks because there wasn't this continuity on the care team."

The 2011 restrictions created other problems as well. With no phase-in period for the rules, Desai says he and colleagues around the country found themselves asking, "'How are we even going to do this?' When you limit duty hours, you have effectively reduced the number of full-time equivalents that we have available," he explains. "But no one is reducing the number of patients we care for-or the time it takes to care for them."

The ACGME regarded that as an institutional issue, Desai says, leaving it up to hospitals to find more money and hire more help.

"But in institutions like this, that just doesn't happen overnight," he explains. "The vast majority of program directors out there will tell you they were underresourced for this transition."

One resulting problem has been dubbed "work compression." In the absence of additional staff and dollars, interns essentially had 16 hours to do what they used to do in 24 or 30. It's unclear at this point how much or whether the need to compress work into shorter shifts might put patients at risk-that's one of the questions Desai hopes to answer going forward.

"You can just imagine how that changes the job," Desai says. "There's no time. You're constantly looking at the clock. You're constantly making decisions about what needs to be prioritized. Patient care is going to win that priority every time, and it should. But the problem is, educational activities will often go out the window."

Indeed, physicians at Brown University conducted a national survey of residents that was published in the New England Journal of Medicine in June 2012. Nearly 41 percent of respondents reported that the education they were receiving was getting worse under the 2011 rules. Only 16 percent thought it was getting better.

Desai thinks that with the second set of rules, the pendulum swung too far and the balance that he had come to appreciate in the 2003 rules was lost. The new regulations were too proscriptive, and they basically eliminated the flexibility that residency directors need to pursue true excellence and innovation.

"Every program has certain things that they think make them excellent," he says. "These rules have now become so rigid that every program has to look the same. We're losing the chance to creatively construct a program that plays to local strengths and reduces its weaknesses. You can't do that anymore. There are only one or two ways to get this done."

And so Desai is now working to win approvals and gain financing for an ambitious new research project. His hypothesis for internal medicine is that longer hours with innovative fatigue management mechanisms will be as safe as the 2011 rules-and will be better for education and certain aspects of sleep habits.

"It seems to many people, myself included, that the 2011 rules have gone too far and the balance between shorter hours and increased transitions of care is now lost," he says. "The risk of the handoffs and of what happens to education because of work compression is now outweighing the benefits of the restrictions."

In February 2013, Desai began reaching out to the ACGME. Both sides agreed that they'd like to see more and better data about the 2011 rules. And recently, the organization signed off on Desai's proposal for how to gather that data.

He wants the study to encompass 75 hospitals around the country. Programs would be randomly assigned either a 28-hour duty limit with an embedded period of protected sleep or the 2011 rules.

Desai seems confident the project will win funding from the National Institutes of Health. "It's being submitted as we speak," he says.

The two-year trial would deliver any number of measures on the patient care side-mortality, length of hospitalization, rehospitalizations, and more. It would also closely track educational results, which would range from standardized test scores to results of ACGME competence testing and evaluations.

"This should deliver a very robust set of outcomes," says Desai, "and I'm hopeful that it will really deliver the data that we need going forward to inform this important national policy."