One of my worst experiences as a pediatric resident had nothing to do with the demands of a difficult service or a decision made in a challenging case. That day towards the end of my intern year, all I had to do was show up for our didactic conference to become completely demoralized.
By noon the meeting room was already packed with about 40 of my colleagues. The disposable aluminum take-out trays had already been strip-mined of protein with just a few sad broccoli heads smothered in heavy General Tso’s sauce remaining. The crowd was larger than for the typical lunch-time talks on the treatment of constipation or the latest on asthma management. Our speaker was a senior surgeon at our children’s hospital and a member of the committee on graduate medical education. The talk was billed as something of a “state of the union” address on resident education in our hospital.
In a rather pedantic tone, he proceeded to show powerpoint slides with graphs depicting the growth in the number of residents and fellows (referred to as “trainees”). He gave statistics about the amount of federal funding for medical education received by our department and hospital and claimed that this fell far short of the actual cost of educating residents. He spoke about his own experience supervising surgical trainees in performing his bread-and-butter procedures: tympanostomy tube (a.k.a. ear tubes) placements and tonsillectomies. These were procedures that he could do by himself in about 20 minutes. Floating back and forth between two operating rooms, he could easily knock out ten cases before lunch. When his ORL (otorhinolaryngology, at our hospital they would not answer to the more pedestrian name “ear, nose, and throat”) residents were working under his supervision, the typical surgery took at least twice as long. The bottom line of his presentation was that resident education was expensive and time consuming for our hospital and that residents made no demonstrable contribution to patient care.
Keep in mind that I had probably worked an overnight shift the night before and that I was fighting a potent post-lunch drive for sleep (more on this another time). This guy is getting up there to tell me that I don’t contribute substantively to patient care after I just spent the last 28 hours taking care of patients (or at least their orders, nursing issues, and paper work)! I looked around the room to try to gauge how his message had been received. I was not the only one among my colleagues struggling to stay away. Most of my fellow residents had their faces down in their Styrofoam plates trying to corral the last few stray grains of rice. I was incredulous that the presenter was being given a free pass.
After the last graph was shown, I raised a hand. My question was something to the effect of, “I’m curious to understand more about your calculation of the costs of resident education. I did not see any consideration of the resident tasks that allow attendings to be more efficient, to see more patients, and to bill more.” It was probably more groggily worded, but not hostile in tone.
He took on a new level of haughtiness. He asserted that because resident labor cannot be billed or reimbursed that it ultimately did not count towards the health care product. He went further. He said that our hospital was an attending run hospital and not a resident run hospital. In case this point wasn’t clear, he actually said that if all of the residents in the hospital did not show up for work the next day, he and his fellow staff physicians would be able to do just fine without us. Having residents in our medical center was necessary for the academic mission but was in all other ways a burden. An exchange that I felt should have provoked a work-stoppage walk-out went almost completely unnoticed. My colleagues were too focused on stepping out to answering pages or on scurrying back to our wards to discharge patients to listen or care.
As any medical insider will tell you, beware of the First of July. July 1st is the start of a new year in academic medicine and as a result, is the worse day to be a patient in a hospital or an emergency room. The junior resident on June 30th is suddenly deemed to have the experience and judgment of a senior on July 1st. A newly minted doctor who a few weeks before was traveling around the world on their 4th year medical school junket is now admitting patients and ordering medications. Imagine another industry when nearly everyone is promoted to a new job with increased responsibilities on the same day. From the trainee’s perspective, July 1st is long, hot, and terrifying. July 1st, although disorienting, is indispensable to medical education. A trainee has to be put in new and challenging situations in order to learn. This means making an unfamiliar decision, working through a new clinical scenario, having a conversation with a patient and family unlike any that has come before. The essential tension of July 1st, and the whole mission of medical education, is allowing physicians to grow in competence and confidence while having just the right amount of supervision in order to make sure that patients are safe.
In the golden age or the “bad-old days” of medicine (depending on who you ask), residents worked overnight shifts every other day (in the parlance, call was q2, meaning that physicians were on-call in the hospital every 2nd night). That’s how medical trainees came to be called residents; they literally resided in the hospital. When I talk to senior physicians about the medical education debate, they often tell me that the only thing wrong with the q2 system was that they missed the chance to learn from half of the cases. Like all clubs, medicine has its hazing rituals and boot-camp mentality. I’m always amazed at how hindsight can make people nostalgic for painful experiences and leave them with a desire to make sure that future generations suffer accordingly. I do speak with some senior physicians who will share anecdotes about being in way over their heads as trainees, making decisions and doing procedures for the first time without supervision, support, or back-up.
Since the days of q2 call, academic medicine has been gradually imposing limits on duty hours for medical trainees. It is important to note that this reform effort has been guided by concerns for patient safety and not by a debate on how physicians should ideally be trained. High profile cases of exhausted physicians making errors motivated patient safety groups to pressure for government-level regulation of trainee work hours. A 2010 survey conducted by the Harvard Medical School Work Hours, Health, and Safety Group, led by Children’s Hospital Boston’s own Chris Landrigan and the Brigham’s Charles Czeisler, reveals that potential patients think that doctors work much less than they do and that they do not want to be cared for by exhausted physicians. Those surveyed estimated that trainee doctors work 12.9 hour shifts and 58 hour work weeks. They felt that the maximum shift should be 10.9 hours and that work weeks should be limited at 50 hours. Most intuitively, 81% of respondents felt that patients should be informed if their doctors had been working for more than 24 hours and in this situation, 80% would want another doctor.
My friends who were two years ahead of my original medical school class (they graduated in ’01) were among the last physicians to train in a q3 schedule where they worked 30+ hours shifts every third night. Given the pressures for enhanced patient safety, the Accreditation Council of Graduate Medical Education (ACGME) work hour standards were passed in 2003. The rules limited residents to 80 hour work weeks averaged over a 4 week period and mandated at least four 24 hour days off in the course of a month. The rules did not go further to end overnight shifts of >24 hours and programs could still use schedules with q4 30 hour shifts or even short bursts of q3 call and be in compliance.
The new regulations were implemented with much gnashing of teeth, acrimony, and on the part of some residency programs, foot dragging. There were cases of residents who felt intimidated by their program directors to under-report hours to the ACGME even though their programs were consistently not in compliance. The ACGME investigated and threatened to remove program’s accreditation, the only punishment that seemed to motivate recalcitrant residencies. Assessing the impact of a system-wide intervention such as the duty hour limits is non-trivial. Reviewing the evidence in a 2008 commentary in JAMA, Landrigan and Kevin Volpp concluded that studies could not establish a consistent improvement in patient mortality after the regulation took effect. They identify three potential reasons why after 5 years patients were not safer: 1) residencies continued to flout the rules 2) residents continue to work >24 hours shifts 3) more transfers of care are needed when hours are shortened and transfers are a source of potential medical errors.
In academic medical centers, this July 1st will be different from the last. At the request of Congress, the Institute of Medicine investigated the resident work hours issue and recommended that trainees be limited to shifts of less than 16 hours. The ACGME decided to adopt the standards for interns and the new rules will go into effect for all of the residency programs across the land on Friday (the 1st). The 16 hour limit is based in part on a key study that appeared in the New England Journal of Medicine in 2004 (again, important work by Landrigan and colleagues). Interns staffing intensive care units either worked a traditional schedule which included >24 hours shifts or an “intervention” schedule where a coverage system limited shifts to ~16 hours. The results were striking: interns working in the traditional schedule made 22% more serious errors and 5.6 times as many serious diagnostic errors, arguably the most cognitively challenging task that can be assessed. Although limited to an elite group of interns at the Brigham, this study still seems to be the best of its kind in comparing the performance of the work schedules head to head. The data are hard to ignore and the Institutes of Medicine and the ACGME took notice. For the time being, the 16 hour limit will apply only to interns, the foot soldiers in the trenches of medical wards. Supervising residents can still be asked to work overnight shifts with some restrictions. There is every reason to think that overnight shifts may soon become a thing of the past for all medical trainees.
In recent weeks, residency programs across the country have been scrambling to shift coverage, often to a night float system, in order to eliminate shifts greater than 16 hours. Here in Dallas, Childrens Medical Center’s new pediatric resident schedule has eliminated all overnight call. At many hospitals, fully trained staff physicians, known as hospitalists, will take over a larger share of the inpatient admissions in order to reduce the load on hour-limited residents.
It will take years to determine if the latest work hour interventions will decrease medical errors, which in a recent NEJM study of North Carolina’s improvement-minded hospitals occur at the rate of 25 errors per 100 admissions (well done co-author Andrew Hackbarth!). It will be even harder to assess the effect of work hour limits on medical education. Is the training of today’s young physicians inferior to that of their q2 predecessors? Certainly there is no replacement for hands-on experience and “seeing” things in the hospital. In my view, there is an important difference between looking and seeing. While looking is passive and can be done by the sleep-deprived, seeing is an active process. It combines careful observation with comparison to previous experience, the identification of areas of uncertainty, and the formulation of hypotheses. Seeing requires energy, motivation, and mentoring and this is the activity that medical education must maximize. Is it possible that trainees can work less and see more? The challenge is to seek a sweet spot on the work hours vs. medical errors curve that optimizes meaningful learning opportunities and minimizes situations that are ripe for error. A successful reform effort will require careful empirical study of training regimens and open-mindedness that the best way to train doctors may not be the way it has traditionally been done. Sometimes the best way to walk to school is not up-hill both ways in a blizzard.
For this July 1st, best wishes from Left on Longwood to all who will be stepping up another rung on the ladder of medical training. For everyone else, here’s to a healthy and hospital-free holiday.