September 9, 2005
Drunk with Fatigue
The physician who ran our Emergency Room when I was a resident used to say, “This is the easiest rotation of your residency. You only have to work half the time.” Said with a huge sarcastic smirk, he found himself to be quite funny. What he meant was that we had to work 12-hour shifts every day of the rotation except, of course, when we were switching from day to night shift. And that was every day; no days off. Ever.
About three weeks into the rotation, I was driving home around 8 o'clock one morning when I awoke to find myself bumping down the median of the street. Luckily it was one of those wide grassy ones without many trees and no bayou. This is one of three distinct times I recall falling asleep while driving home that rotation. I suppose my angels were looking out after me. I came to no harm from it. All the same, it sucked.
Medical training and practice can be personally destructive and socially isolating. There is no need for this in today's world.
Perhaps the greatest failing of the academic medical center is a reluctance to devise better ways to train physicians. No one should be asked to do a forced march for five years or more, working 80-100 hours a week in some sado-masochistic, machismo endurance test. Ultimately there is a choice in medicine: to do it or not. My old professor used to say, “No one is twisting your arm.” This binary all-or-nothing approach is a failure.
Results of a recent study published in the Journal of the American Medical Association suggest sleep-deprived residents react in driving simulation and cognitive tests at the same level of impairment as those with significant alcohol levels. As reported on ScientificAmerican.com:
An 80-hour limit for a resident's workweek was introduced in July 2003 in response to concerns about overwork. In the new study, J. Todd Arnedt of the University of Michigan and his colleagues measured the performance of 34 doctors, who had been on call, on an attention test and in a driving simulator. The volunteers took part in the tests on four different occasions, after working mostly day shifts with only a few overnight calls, or after working intense overnight shifts that added up to about 80 hours in a week. For some of the tests, the doctors were also given alcoholic drinks or nonalcoholic placebos. After a month of difficult work schedules, the doctors exhibited reaction times that were seven percent slower than their responses after working a lighter schedule. In the driving simulator, doctors coming off a month of working nights displayed comparable skills to the subjects who had an easier schedule but had a blood-alcohol level just below the legal driving limit. What is more, the post-call doctors were 30 percent more likely to not maintain a steady speed in the driving simulator compared to well-rested doctors who had been drinking.
The interesting thing is the reported response of the co-author, Judith Owens of Brown Medical School. She is quoted as saying:
We have to continue to educate doctors-in-training, and we should help them develop sleep risk-management strategies. This is particularly important since our study shows that many sleep-starved residents don't recognize that they're impaired.
Develop sleep risk-management strategies? Are you kidding me? I guess that would be sleep, right? This quote would be laughable if it weren't such a serious issue.
It is accepted as an article of faith that physicians cannot be trained except in an abusive system. This view should be rejected and the educational approach re-examined.
Academic medical centers should be held to account. Educational methodology and process get little to no attention and certainly no innovation in these centers. Most educators are amateurs at best. I should know; I was one for many years.
Here is the conundrum: humans learn complex cognitive tasks best by repetition. Hence, the guild-master-apprentice approach to medical training. The structural approach to the academic center could best be described as a feudal one. So, I guess we haven't made much progress for 400-500 years.
Guild-master-apprentice also works ok when there isn't much to know and whether you know it or not doesn't matter because there isn't much to do. I think we are past this phase.
There is some empirical information suggesting that learning is environment specific. So, learning in a fatigued state could be a good thing if you want to perform in a fatigued state. But being alert and happy to be there would be even better.
Finally, there is some empirical information suggesting that one can adequately perform previously mastered tasks even when impaired, but acquiring new information or solving unfamiliar problems remains impaired. When you develop a new and acute medical problem in the middle of the night, who do you want at your bedside? Probably not a fatigued, impaired, inexperienced, un-mastered person.
Why do we expose our patients to such danger?
There is considerable professional resistance to changing training approaches. Physicians are extremely resistant to the idea of shift work. And there is the “culture” of medicine. And let's not forget tradition. And hell, it's my turn to be guild master.
Academic medical centers must develop new ways of training physicians so they are both competent and maximally functional at the point of service care delivery and an integral part of a system that is equally functional 24x7x365.
