Kiran Gupta writes about her experiences as a resident in the Department of Medicine at Brigham and Women's Hospital. She has done research on domestic health policy issues, including health care costs, medical errors, and end-of-life care, and is a former columnist for the Financial Times Weekend Magazine. Kiran received her A.B. in Government with a certificate in Health Policy from Harvard College and earned her M.D. from Harvard Medical School.


While working a night shift recently, I was paged to see a patient who was confused. A family member sat anxiously at the bedside. Fortunately, I had a couple of minutes to glance at the patient’s chart before entering the room. She had been discharged from the hospital just days before and admitted again earlier that day for dehydration with kidney failure, which seemed the likely cause of her confusion. After reading the admitting resident’s note, I began to take a brief history.

The relative was frustrated. “How many doctors do I have to talk to? Can’t you see—it’s all in the chart. We already explained everything to the emergency room and the other doctor. And now you?”

I didn’t have the heart to say that the next morning the same questions would be asked yet again by the primary team responsible for the patient’s care.

Advertisement - Continue Reading Below

This patient’s experience with fragmented care is all too common since last July, when medical residency programs across the country revamped physician-trainees’ schedules to comply with new work-hour restrictions imposed by the Accreditation Council on Graduate Medical Education (ACMGE). Under these regulations, first-year residents were no longer permitted to work more than 16 hours at a time, marking the end of a long tradition of these “interns” taking overnight call. The changes were intended to enhance supervision of trainees by experienced physicians, improve transitions in patient care from one provider to another (“handoffs”), and alleviate concerns over residents’ waning performance due to sleep deprivation during continuous periods of duty—which previously lasted up to 30 hours.

To many outside the profession, the idea that residents ought to work such onerous hours is troubling. But I am concerned that the new restrictions, though well intended, are contributing to a work environment that compromises our clinical education and ability to provide care in a patient-centered manner. They may even have negative effects on our levels of rest and personal happiness—the very outcome these reforms were intended to address.

And I am not alone, as revealed by a recent nationwide survey of residents’ opinions on the duty-hour changes published in the New England Journal of Medicine. Only 29 percent stated that they feel more rested, and just 22 percent of residents surveyed approved of the new hour restrictions.

While some medical specialties may lend themselves better to shift-based training, not all do. Capping trainees’ consecutive hours in the hospital comes with its own drawbacks, potentially impinging on clinical education, contributing to errors due to frequent handoffs, increasing patients’ frustration at care fragmented across more providers, undermining trainees’ sense of “ownership” over patients, and generating constant pressure on residents to leave the hospital on time. Keeping residents well rested is an important goal, but the survey data make clear that these restrictions ought to be carefully studied to ensure that their benefits outweigh such costs.

My own experience leaves me unconvinced. For one, the restrictions require most residents to work more night shifts than they did in the past. Patients admitted overnight by these “nightfloat” residents are not seen by the primary senior physician and trainee caring for them until the following morning. Here’s a typical case: an afternoon-shift resident admits a patient to the hospital around 2 p.m. and then hands that patient off to me at 7 p.m., when I come in as a “nightfloat.” Since I am not part of the patient’s primary team and am covering most of the general medical service that night, as well as admitting new patients from the Emergency Department, I will likely not meet this patient during my shift unless the nurse has reason to page me. That night she doesn’t, and at 7 the next morning, I hand off the patient (or rather, his admission note) to the primary team.

As a result, the patient’s plan of care is not reassessed for nearly a full day. In the new duty-hour system, it is quite possible that patients will spend their first 12 to 18 hours in the hospital, a window in which they are often seriously ill and require frequent reassessment, without a physician re-evaluating any initial treatment that was provided upon admission unless paged by the nurse to come to the bedside.

We call this type of patient—one admitted the previous day and not seen by his or her primary care team until the next morning—a “holdover.” It’s an unfortunate word, one connoting a task that remains incomplete. But, in some ways, it is accurate. This year, medical teams are greeted each morning by a barrage of holdovers from the previous night, many more than a year ago. The early morning is spent asking these patients to repeat what they have already told multiple physicians the night before. Interns must read patients’ care plans developed by the admitting resident the previous day, rather than developing their own care plans after gathering and assessing the clinical data themselves.

In the old system, interns stayed in the hospital overnight and admitted patients while on call. There were fewer holdovers as a result. Patients were “owned” by the team directly responsible for their care upon admission. Interns, with resident supervision, had time to assess and reassess their clinical decision-making throughout the night. Did I choose the right antibiotics? What did that lab I ordered show? Did we give enough fluid? Do I need to transfer this patient to the intensive care unit? There was time to think, read, engage with more experienced residents, and most importantly, get to know patients. On morning rounds, it was rare to read an admission note that you had not written yourself. And at noon, when your 30 hours were up, you went home. The next two days were fairly short, with opportunity for instruction from senior faculty and adequate time to attend teaching conferences.

The new restrictions were supposed to increase faculty supervision of residents. In my experience, however, while working two months of night shifts admitting countless “holdover” patients this year, I was never present on teaching rounds the next morning as I was not a part of any daytime, primary patient-care team. As soon as my shift ended at 7 a.m., I would rush home to sleep so that I could get back later that evening. I never had the opportunity to discuss my choice of antibiotics with a senior attending, or participate in bedside teaching when she might review various aspects of the physical exam—an increasingly lost art—on the patients I had admitted. During those two months, I did not attend noon teaching lectures or interact with a single faculty member regarding my work caring for patients. With continuous-duty periods restricted, the move to day- and night-shifts raises a challenge: how do we ensure that residents working more nights still learn from experienced faculty? In the recent survey, about two in five residents said the quality of their education had worsened because of the hour limits, while about the same number reported no change in quality.

Every day now is a long shift. There is a constant sense of urgency to get out of the hospital on time. There are more handoffs from one resident to another and holdovers, each carrying a risk that an essential patient detail might fall by the wayside—it’s like a high-stakes game of telephone. Rounds are more harried, with less teaching time. And what about the effect on patients, those whom the work-hour limits were purported to help most? We don’t yet have adequate data to say that this new system is better for them.

Medical trainees were given the title “residents” because at one time they literally lived in the hospital. I am certainly not suggesting that we return to the days when residents, like my parents during their training, rarely saw the light of day. But we should consider whether the pendulum has swung too far in the other direction.