When I was a medical student, I remember wondering what my attendings did when they weren’t on the wards. As an attending now myself, trying to cram three half-month ward blocks into my hyperscheduled life each year, I find that sentiment charmingly naïve. I – like most of my faculty colleagues – am awfully busy these days, both on and off the wards.
But one thing that makes the wards doable is that there is a certain rhythm to the experience. Check that: made and was. Until last year, the ward team (consisting of a resident, a couple of interns, and a med student or two) admitted new patients all day and all night, every fourth day. As their attending, I might have seen a couple of new admissions during the on-call day and chatted with the resident overnight about the most complex cases. But the team was mostly on its own for the first 24 hours, particularly at night. The data dump – from the residents to the legally responsible physician (in this case, me) – occurred on the post-call morning, when I sat down with my bleary-eyed team to hear about all the new admissions – sometimes as many as ten. As we raced from one case presentation to the next, we desperately sought time to accomplish any number of tasks: listen to the third-year student’s cinematic depiction of his patient’s family history, run to radiology to look at some films, talk to a couple of patients and families, and – channeling my inner Abraham Verghese – demonstrate some key physical findings.
Oh yes, all the while trying to ensure that the patients received the best possible care. For the attending, the post-call day was a hellish fire drill, but the work got done, most patients did fine, and the rhythm was predictable, lending a certain comfort to the whole exercise. Moreover, the non-post-call days offered relatively tranquil interludes of 60-90 minutes each morning for attending teaching rounds; in a half-month block, I could count on about 6 “good days” when all the residents and students were around and I could cover some of my favorite topics (diagnostic reasoning, patient safety, decision-making about end of life care, etc.). And – because they were left alone for so much of the on-call day and night – the housestaff learned to be autonomous clinical thinkers.
I just finished my first ward stint since the new ACGME duty hours regulations took effect last July. These regulations, you’ll recall, now prohibit interns from working shifts of more than 16 hours. Like programs everywhere, UCSF was forced to jettison the time-honored 24-hour call system and replace it with some version of shift-work. In our new model, on-call teams now admit from 8 a.m. to 6 p.m. The interns go home at about 9 p.m.; the resident stays overnight to mop up on the new patients and to cross-cover them. But new admissions between 6 p.m. and 8 a.m. are handled by a separate night crew, consisting of housestaff and hospitalists. All these patients have to go somewhere, of course, so the daytime on-call teams receive several new patients who were worked up overnight during an early morning handoff.
You may think that this is the time when the old geezer begins crooning about the “Days of the Giants”: how overnight call made us Strong-Like-Bull, how fraught handoffs are, how today’s housestaff are acquiring a shift-work mentality. I do have concerns. In fact, in this month’s issue of AHRQ WebM&M, we highlight many of them with a thoughtful article by Dine and Myers and my interview with Larry Smith, the founding dean of the new Hofstra/North Shore-LIJ med school and a former residency program director. Both pieces are worth a look.
But, recalling the times I fell asleep driving home after 36-hour intern shifts, I think the new model is an improvement. Even if it isn’t, we should get over ourselves and work on making lemonade. The public finds the idea that we can provide high quality, safe care in our 32th consecutive hour impossible, particularly when good research shows that after such a period of prolonged wakefulness, our cognitive skills degrade to a state that mirrors that of someone with a blood alcohol level of 0.1 – legally drunk in every state. Even though studies on the impact of duty hour limits generally don’t prove that safety is improved, the limits have led to better rested housestaff, less burnout, and greater public confidence in our system. So the duty hour limits are here to stay (by the way, the max in Europe is 48 hours per week), and those of us who run educational programs must accept our new challenges: how to preserve the core values of professionalism, intellectual autonomy, and medical education when the wards have a very different rhythm than in the past.
It’s not an easy task.
The new system has several advantages over the old. For the attending, rather than trying to digest a mountain of information about many new patients on the post-call morning, the task has been shifted one day forward and spread around. Last month, I eventually settled into a routine in which I rounded with my team twice on their on-call days: once at about 10 a.m. to hear about all our old patients and the new admits that they received 3 hours earlier from the night teams (usually 3-4 patients), and again at 6 p.m. to hear about all the additional patients admitted throughout the day as well as any major status changes (remember that they’re done admitting new patients at 6 p.m.). The beat was completely new to me: I saw patients throughout the on-call day and evening, generally leaving the building around 11 p.m. It was an exhausting day, particularly for an Old Timer, but it also felt great knowing that I had tucked in all the new patients by bedtime. The dread I used to feel driving into work on the post-call day was replaced by Zen-like calm; it was now a smooth, predictable day.
I also like the fact that my input into the cases was much more timely. In the old days, for a patient admitted at 6 p.m., I might have heard a thumbnail sketch by phone five hours later, a full presentation late the next morning, and not gotten around to meeting the patient until that afternoon, nearly a day after admission. Now, sick patients admitted at night are “staffed” by one of my colleagues, a nighttime hospitalist (“nocturnist”). I usually saw all of them by noon. And I saw those admitted during the day within six hours of admission, sooner if they were acutely ill.
So far so good. What’s the bad news?
First, this just-in-time attending involvement carries a potential cost. You may recall me sharing some of Dr. Cindy Fenton’s observations when she returned to academic medicine after a decade’s absence. Perhaps the most memorable were her comments about housestaff autonomy:
…The more involved role of the attending was striking. My resident, who was superb, asked for and seemed to appreciate my involvement in holding family meetings and moving care along, especially on post-call days. She wanted me to round with the team in the ICU daily at 8:30 am. I realized I had no idea how to function as an attending on work rounds (this was something I pretty much never did in the “old days”), so I had to develop these skills.
That was written before the new rules make things worse – or better, depending on your perspective. On call days, I was much more of a presence, and offered my input much sooner, than in the past. On the post-call day, the resident (who is allowed to stay overnight but can work for only 28 consecutive hours) had to leave by 11 a.m. When she left, I morphed into the team’s res-attending. My interns were terrific, as was my resident, but, with each look to the attending for guidance on decisions that residents used to make completely independently, I could see the team’s autonomy slipping away.
Is that good or bad? Both. (And not just from the attending perspective – this superb essay, written by one of our 3rd year residents, Chris Moriates, and published in JGIM, offers a residents’-eye view of the new system.) Learning from one’s mistakes is fundamentally unethical when you have a human life in your hands. But an environment in which the housestaff are trained to read the attending’s body language before making a tough call can’t be right either, particularly when our third-year residents morph from resident to attending on June 30th each year. I often had to push myself to say, “Tell me what you’d like to do,” but it isn’t easy when you’re so busy, the duty hours clock is ticking, and the quickest path to the right answer is to offer it up yourself. As midnight draws near, Cinderella doesn’t have time to discuss the pathophysiology of pumpkins. She just needs to know where the staircase is.
A second worry is the relative dearth of patients being followed by a single resident from admission to denouement. Our teams inherited nearly half their patients as handoffs from night admitters. While some trainees forced themselves to rethink their patients’ problems and actively ward off anchoring bias, others didn’t, accepting what they were told as gospel and never coming to know the handed-off patients as well as those they admitted themselves. So many emergency admissions traverse a trajectory in which an early assessment is followed by a period of data gathering (tests, consults), followed by an initial patient response, which is evaluated in context. In a system in which half the patients are cared for by two sets of doctors during these crucial stages, neither group fully sees this arc play out, and their education suffers. While some handoffs are inevitable, I wonder if there is some way to drop the handoff percentage to more like one-third of all admissions. This feels like an intricate math problem – tough but soluble.
The other thing that worries me about the new schedules is the palpably limited time available for education. In the 16 days I spent as attending in January, I recall only two in which the entire team was available for our traditional hour-long teaching rounds. Sometimes I felt that my resident believed part of her job was to protect the rest of the team from the attending so that the interns could get their work done. I completely understand why that would be, but this is a terrible position for residents – and attendings – to be put in. We simply have to figure out how to bake in structured teaching rounds – maybe at 6 p.m. rather than 10 a.m. – lest we lose something very precious. And perhaps the hour-long chalk talk is no longer the optimal teaching forum for the Twitter generation; today’s trainees may need to learn through multiple small feedings rather than 500cc boluses. If so, that’s fine. But I’m confident that if we don’t hard wire teaching times into the daily schedule, then formal teaching will sink to the bottom of the priority list in the name of efficiency and duty hours. That would be a terrible loss.
Let’s not romanticize the old days. My schedule during internship was inhuman and dangerous. While the old rhythm had the virtues of continuity and more time for teaching, the costs in patient safety, housestaff burnout, and the credibility of the enterprise to a skeptical public were too high. A new system is welcome and needed. But those of us trusted to teach the next generation need to seek ways to mitigate its harms. If we don’t, we will have traded off some old problems for some powerful new ones – and the latter outcome could pay itself forward for generations to come.