Articles

Resident’s Working Hours: Should We Let Sleeping Docs Lie?

The Accreditation Council for Graduate Medical Education (ACGME) has released proposed new standards to limit working hours for medical residents. Bus drivers are allowed to drive for 10 hours and then are required to have 8 hours off duty. Airline pilots can be scheduled for up to 16 hours on duty — being at work, ready to fly — and up to eight hours of actual flight time in a 24-hour period, with a minimum of eight hours for rest between shifts. Physicians in residency training work 80 hours or more a week (compared to 75 hours a month for airline pilots) and are regularly on duty for more than 24 hours at a time. If adequate rest is an important safety measure for drivers and pilots, isn’t it important for doctors too?

When I was an intern and resident, my hours were a little better than some. Instead of every other night, I was on call every third night. I had to work from about 7 AM one day to 5 PM the following day (34 consecutive hours). I stayed in the hospital: there was a call room with a bed, but if we got to lie down it was never for very long. When I got off duty, my sleep-deprived body demanded that I go home and crash. It was only every third day when I worked “only” a 10 hour shift, that I could devote an evening to all the other activities of my life like laundry, grocery shopping, and trying to read medical journals. One memorable weekend I worked from Saturday morning to Monday evening and only got to lie down for about 20 minutes. I don’t think I made any fatigue-induced mistakes that hurt patients, but by Monday afternoon I was groping my way through brain fog and running on fumes.

The arguments for the brutal schedule seemed to boil down to these:

  1. We had to do it in our training, so we’re going to make you suffer too. (Were they arguing for tradition or revenge?)
  2. You need to spend all night at the hospital to have the experience of continuity of care for your patients, to see how illnesses evolve over 24+ hours.
  3. Someone has to be there to take care of the patients at night, and there are only so many residents.
  4. It teaches you that when you feel like you are so fatigued that you can’t possibly carry on any longer, you actually can. This is a valuable lesson to learn before you are confronted with a situation in your practice or in a disaster where your sleep is interrupted.
  5. It’ll grow hair on your chest.

I thought:

  1. Since you had to do it yourselves, you should have some compassion.
  2. Is there any evidence that getting the experience of continuity in this way actually produces better doctors? If so, is it possible that we could find a more humane way to get that kind of experience?
  3. So hire more residents or put staff physicians into the roster.
  4. Couldn’t you wait to learn that same lesson naturally during an actual emergency situation? Aren’t we smart enough to learn this lesson from one experience? Does it really require repetition every 3 days for several years?
  5. No thank you, I’m a girl.

Libby Zion and Repercussions

We bitched a lot, but nothing happened until a young woman named Libby Zion died in 1984. The ensuing accusations of murder and an official investigation and civil suit attracted a lot of media attention. Libby’s case was complicated, but her death was attributed to an injudicious combination of medications, and it came out that the prescribing doctor was an intern 8 months out of medical school who was making middle-of-the-night decisions on her own during a long shift (well over 24 hours), was responsible for 40 patients, and was supervised by an attending physician who was at home and available only by phone. In 1989, New York State passed the Libby Zion Law restricting residents to an 80 hour work week and 24 hour shifts and requiring attending physicians to be physically present in the hospital. (It is far from clear that Libby’s outcome would have been different if those restrictions had been in place at the time.)

In 2003 the ACGME followed New York State’s example and enacted similar standards for all accredited residency programs in the US.The 24-hour shift limit can be extended by up to 6 additional hours to allow residents to participate in didactic activities, transfer care of patients, conduct outpatient clinics, and maintain continuity of medical and surgical care. “Adequate supervision” by qualified faculty is required but not clearly defined.

Did it do any good? We’re not entirely sure. This study showed no increase in re-admissions or deaths following the 2003 reforms. These results were questioned: hours worked were self-reported; the actual hours worked may not have decreased. Another study showed that interns made substantially more serious medical errors when they worked longer shifts of 24 hours or more.

When a resident’s shift is over, responsibility for patients is transferred to another resident who didn’t do the admission or workup, knows the patients only on paper, and may easily miss crucial details. Will patient care suffer as much from lack of continuity as it does from physician fatigue? We don’t know.

New Recommendations

A task force of the ACGME recently reported:

Over the past year, the task force conducted a thorough examination of patient safety, duty hours, resident supervision, educational outcomes and training standards that included hearing testimony from more than 100 individuals, receiving written presentations from 100 medical organizations and commissioning three independent reviews of the literature on sleep issues, patient safety and resident training.

In response to the recommendations of its own task force and to an independent report by the Institute of Medicine (IOM) the Accreditation Council for Graduate Medical Education (ACGME) has issued enhanced standards scheduled to take effect in July 2011. Programs will have to follow them or lose their accreditation. Shifts will be limited to 16 hours for first year residents and 24 hours for subsequent years, with 4 more hours for handing off patients or remaining with an acutely ill patient. First-year residents will get more supervision.

Anticipated Adverse Impact on Family Medicine Residencies

In a letter to the ACGME, the AAFP and five other organizations have raised concerns.They are worried that the new requirements

  • will decrease the overall educational time and clinical experiences for family medicine residents.
  • will impair many programs’ ability to meet the required continuity patient care visit thresholds in the ambulatory setting, thus compromising a key educational component of family physician training.
  • are likely to promote a “shift work” approach to practice that is not consistent with efforts to move toward more patient-centered care.
  • could result in the need to extend family medicine training to 48 months, thus increasing costs and potentially hurting current efforts to recruit medical students to choose careers in primary care.
  • have the potential to cripple small, community-based residency programs because of insufficient personnel to provide full patient care coverage.Worse, according to results of a July 2010 survey of family medicine residency programs, the adoption of the proposed standards would threaten the existence of nearly 40 percent of programs with fewer than 22 residents.

Among other recommendations they ask the ACGME to :

  • develop and implement pilot studies in which different duty hour requirements are measured against medical errors and patient safety guidelines;
  • publicly acknowledge that the new duty hour restrictions will require an increase in training program faculty, with associated increases in program costs;

Other Dissenting Opinions

A letter to the editor in the New England Journal of Medicine (NEJM) argued that the new guidelines redistribute the responsibility for care.

This forced change diminishes a critical experience that previously contributed to defining a physician as having a profession rather than just a job. Personally, we would rather be cared for by a fatigued professional who feels responsible for our care than by a well-rested shift worker who does not.

An article in the same issue raised other concerns. The IOM report stressed that additional funding would be needed. They estimated that approximately $1.7 billion would be required to hire other medical providers just to bring programs into compliance with the 2003 ACGME standards.

The American College of Surgeons expressed grave concerns, predicting “a negative impact on patient safety and continuity of care unless there is a substantial increase in human resources to replace the residents.”

A Duty Hours Congress convened by the ACGME voiced opposition to the recommendation for extended sleep periods, and the AMA adopted a policy opposing it.

Conclusion

The proposed new guidelines will be kinder to residents (more sleep, more time with family). They can be expected to reduce physician fatigue, reduce fatigue-induced errors, and improve supervision. They can also be expected to increase the cost of medical education, result in the closing of some training programs, reduce hours of education, reduce direct responsibility and continuity of care, and possibly increase errors in patient care due to lack of continuity. Whether there will be a net benefit to patient care or physician education remains to be seen.

Posted in: Politics and Regulation

Leave a Comment (53) ↓

53 thoughts on “Resident’s Working Hours: Should We Let Sleeping Docs Lie?

  1. Marta says:

    I had kind of a gut reaction to this: last summer a close friend was struck and nearly killed by a resident who had just gotten off a long shift and fell asleep at the wheel. It’s not just their performance on the job that is of concern.

  2. weing says:

    I’m all for the shift work mentality. Sleep deprivation has been shown to have the same effect on cognitive functioning as being drunk. Are the detractors of the shift work mentality OK with a surgeon operating under the influence?

  3. David Gorski says:

    One problem with the workshift mentality: Patient hand-offs. The more “shifts” you have, the more patient hand-offs there are, each of which introduces the possibility of error and, at the very least, guarantees that physicians who don’t know the patient very well end up covering.

    One thing that I find very interesting is that none of the extravagant claims made for work hour limitations before they were instituted around 7 or 8 years ago have come to pass. While there is no good evidence that work hour limitations have compromised patient safety (which detractors warned about), there is also no good evidence that, as it was originally claimed when work hour limitations were instituted, that work hour limitations have improved patient safety either. My guess is that the potential problems from hand-offs have probably canceled out any potential benefit from work hour limitations. The other claim made for work hour limitations is that they would result in better performance of residents on their board examinations and other measurements of training achievement. That hasn’t happened either. To be fair, on the other hand, detractors were wrong, too, in that work hour limitations have not resulted in detectable declines in board pass rates, etc. There is conflicting evidence that they might have resulted in decreased numbers of surgical cases, but the evidence isn’t particularly convincing; most studies I’ve seen show no significant change attributable to such limitations.

    The problem with work hour limitations, of course, is that we are completely unwilling to pay for them. To implement them properly, there have to be adequate support personnel, sufficient numbers of doctors, etc. There are not, nor are there ever likely to be in this cost-constrained health care environment. Consequently, every time there are new work hour limitations, the burden to make up for the resulting personnel shortfall falls onto those who are not work hour restricted. An excellent example will begin in July, when interns will be forbidden from working more than 16 hours straight. What will happen is that this workload will be shifted upwards to 2nd and 3rd year residents, who will be getting totally screwed, to put it mildly. This sort of thing will continue unless we as a society are willing to bear the cost of these restrictions. Unfortunately, I see no sign that we are or will be any time soon.

    Bottom line, the problem with work hour limitations is that they do not apply to every physician. They only apply to residents, not to fellows or attendings. What that means is that, whenever there are more work hour restrictions, the work load shifts upward to fellows and attendings, many of whom already work 70 or 80 hours a week. If residents, who presumably have backup from higher level residents, fellows, and attendings, allegedly can’t function if they’re working too many hours, what about the attendings, who are the ones who are ultimately responsible, both legally and morally, for patient care decisions?

  4. Mhops says:

    In my experience, one of the more commonly quoted arguments in favor of this was the need to be prepared for what it is like after residency. I never understood how “practicing” staying up for 30 hours makes you any better at doing it.

    Secondly, depending upon where you choose to work or what you choose to do, you may only rarely be called upon to work 30-40 hours straight (everyone I know says residency was harder than anything they are doing now). Finally, once you are an attending physician/surgeon, you likely have the power to cancel a clinic or a surgical case if you feel impaired – a power you never had as a resident.

  5. David Gorski says:

    Cancel a clinic or a surgical case? Yes, attendings have the power to do that, but in all practicality there is a lot of pressure on them not to do so, particularly if they are in private practice. In private practice, it’s a matter of money and patient satisfaction; if they cancel a clinic, they lose a lot of money and piss off a lot of patients, even risking losing one or two. When you have staff to support and razor-thin profit margins due to ever-declining third party payor reimbursement, canceling a clinic is usually not an option other than in an emergency situation. Then there’s the issue that, if you do it too often, referring physicians hear about it from their ticked off patients, making them less likely to refer.

    As for surgery, there are those pressures plus additional factors that, depending on the surgery, may come into play. First, patients really get pissed off when their surgery is canceled. Often they have rearranged their lives for it. Second, sometimes it’s not feasible. In breast surgery, for example, often a wire is inserted the morning of surgery to guide us to the lesion we’re removing. Unless we make the go/no go decision by very early in the morning, often the patient will already have the wire in place. We can’t send them home with the wire; so the choice boils down to doing the case or canceling and pulling the wire out, thus wasting a procedure to place it and resulting in the patient being billed.

  6. Peter Lipson says:

    Although we do not yet have data that confirm that improved work hours lead to better care, it is certainly plausible. We do have evidence, much of it indirect, that excessive work hours lead to errors. My take…

  7. Cowy1 says:

    You guys are looking at this from the wrong angle. Think about it from a human-rights abuse perspective and you will see what it really is; no-one should be required to work for 30+ hours straight for any reason really.

    The bogus excuses given for working residents 80+ hours a week are just that, bogus. Hospitals and physician groups are just protecting their (financial) interests by clouding the issue with “Patient Care”.

  8. I think the doctor/pilot comparison is relevant here as well. When the choice was made to limit flight hours some of the same concerns could have been raised by the airline. If you limit the flight hours, we do not have adequate staff to support our flight demand. The cost of flying will increase beyond an acceptable level. If flight hours are limited then pilots will take longer to proceed through training, which will be a disincentive to become a pilot, etc.

    In fact, I recall the discussion years ago, when after a fatal hospital helicopter crash it was determined that the pilot had been flying well over the prescribed guidelines for commercial pilots (which legally don’t apply to EMS helicopter pilots, I guess). The defense of the extended hours used some of these same arguments.

    That said, in order to implement a solution for a problem it’s important to consider the consequence of a proposed solution. Once you know the potential drawbacks you can perhaps alter the solution in the attempt to mitigate the consequences.

    I am with weing, It seems the evidence is good that sleep deprivation impairs cognitive function*. I wonder if there are strategies that might mitigate the potential problems in support shortages, expense and lower training hours?

    *Curious that we recruit the brightest individuals possible to become doctors, then set about forcing them to work in a way that has a negative impact on their cognitive function. Perhaps we should recruit more people with more average cognitive abilities then force them to work in an environment that positively impacts cognitive function. :)

  9. I forgot to include patient hand-off in the potential problems of limiting hours. It would be another thing that might need new strategies to mitigate the negative effects.

  10. I’m not sure I understand why we can’t pay more doctors.
    1) If remuneration is per act, the number of acts shouldn’t change with the number of doctors.
    2) If we don’t require heroic sacrifices of doctors, many of them might be willing to settle for a little less pay in exchange for a better life.

    If the issue is not doctors but residents, it still doesn’t make sense. If we need more doctors because they’re requiring a better work/life balance than they used to, then we also need more residents to supply them. That need exists with or without humane work hours.

  11. overshoot says:

    I never understood how “practicing” staying up for 30 hours makes you any better at doing it.

    How it works in medicine is a matter for research, but how it works in other fields is similar to how it works with habitual alcohol abuse: as you acquire experience at performing while impaired (e.g. driving drunk) you build a baseline of “normal” which includes your impaired behavior as perceived by your impaired judgment. Since the ability to recognize poor performance is degraded before the ability to actually perform, especially for tasks for which you have trained extensively, your self-perceived performance may actually improve. Thus the common opinion that “I actually drive better after a few drinks.”

  12. I might add to Alison Cummins comment. That I have heard of (and spoken to) very capable, smart women with excellent academic records who are interested in medicine, choosing another line of work because they don’t feel they can manage both the grueling hours of residency, fellowship, etc and have a family. I’d guess the same applies to men, (who I have not had the opportunity to quiz on the topic) who may be interested in a better work/non work balance.

    It is a bit of a vicious cycle. We don’t have enough residents so they have to work grueling hours. We can’t get more residents because they are required to work grueling hours. :)

  13. tuck says:

    @Dr. Gorski: It may be that reducing the time awake to 24 hours didn’t show benefits because 24 hours is still too long.

    Unfortunately all your points about the economics make perfect sense, absent a showing of lower incident rates.

    The military has done a lot of research on sleep deprivation and performance, and degradation occurs within 24 hours: “This finding is consistent with laboratory studies of impaired performance in 24 hours of sustained, continuous work (Mullaney, Kripke and Fleck, 1981). Angus and colleagues also reported performance reductions of 30 per cent during the first night and 60 per cent during the second night when sleep deprivation was combined with continuous cognitive work (Angus, Heslegrave and Myles, 1985).”

    http://faculty.nps.edu/nlmiller/docs/ch12.pdf

    This is a great post. I’ve never understood the motivation for keeping doctors up for so long. Obviously there are some circumstances where it’s unavoidable (combat or natural disaster), but as an ongoing practice it seems pretty indefensible in light of clear evidence that it dramatically impedes performance.

    Obviously there’s a trade-off between having an alert doctor who’s not familiar with your case, and a tired doctor who’s familiar but working at a 30% cognitave deficit. Based on my own experience working long hours, I’d take the alert doctor who can read the chart correctly.

  14. Peter Lipson says:

    @Alison

    Resident salaries are largely paid for by Medicare, meaning all of us, and specialty/training choice is often driven by the large debt load of US med school grads. It is a huge systemic problem.

  15. David Gorski says:

    Although we do not yet have data that confirm that improved work hours lead to better care, it is certainly plausible. We do have evidence, much of it indirect, that excessive work hours lead to errors.

    Correct. However, as Harriet correctly points out, we don’t have good evidence whether, in aggregate, when all other factors and tradeoffs are factored in, work hour restrictions lead to fewer medical errors and better patient outcomes. Remember, that was one of the major claims, that work hour restrictions would lead to decreased medical errors, particularly in the wake of the Libby Zion case. Personally, I think that people (not you) tend to blow off concerns about continuity of care a bit too blithely. Hand-offs are not a minor deal; the problem is, they are all too often treated in a very perfunctory fashion. What is often given to the covering doc as the primary doc leaves is a brief list with a one-line description and, if applicable, a list of things that need to be done or checked. Sometimes not even that much detail is given.

    Work hour restrictions, assuming that well-rested physicians make fewer mistakes than overly fatigued physicians (a reasonable assumption), will not lead to decreased medical errors in and of themselves. Other changes in the system need to be made in concert with them, for example, improving the quality of information exchange during hand-offs. My point is not to dismiss work hour restrictions; it is to point out that we as a society want to have our cake and eat it too. We want work hour restrictions, believing that they will decrease medical errors and improve the quality of care (which may or may not be true), but we are clearly not willing to pay the cost for them (more residents and/or physician extenders and, potentially, with longer residencies). My other point is that claims for benefits in terms of decreased medical errors due to work hour restrictions have not yet, at least in the current system, been borne out by the evidence, nor have the claims for better resident education.

    I have no problem with making the argument on a pure humanitarian basis that it is brutal to subject doctors in training to 100+ hour work weeks of the sort that I endured for five years. I’m not one of those doctors who claim that I liked working those hours (I most definitely did not and even came close to quitting surgery residency on more than one occasion) or that they are necessary to train competent doctors. Hell, in general surgery, work hour restrictions have arguably helped to save the specialty by leading medical students who want to become surgeons but don’t want to work 100+ hours a week to consider surgery programs, where before they would have rejected them before on lifestyle considerations alone. Before work hour restrictions, a lot of general surgery programs (and very good ones at that) were having trouble filling their slots. Such is not the case now, and there is some evidence that this is more than just correlation with the introduction of work hour restrictions. I do, however, see a problem with the evidence for whether these work hour restrictions provide all the benefits originally claimed; namely, it’s just not very convincing.

    We should be careful in arguing for work hour restrictions that we don’t promise what can’t be supported by evidence.

  16. overshoot says:

    Once upon a time, resident hours were much worse. Then again, weren’t those also the times when “resident” was literal rather than a title? Didn’t teaching hospitals actually provide, like, residences to the residents, so that there wasn’t a commute-time issue? So that some of the “more than work and sleep” activities could be delegated (a patently cost-effective allocation of resources) etc?

    Times change, expectations change, but I wonder if the current topic is in part an unintended consequence of an otherwise unrelated shift in practice.

  17. Peter Lipson says:

    “Back in the day”, most residents were young, unmarried men. This has changed significantly.

  18. overshoot says:

    “Back in the day”, most residents were young, unmarried men.

    Yup — and that certainly made it easier to get them to live in dormitories and work insane hours. Just like the Army.

    Perversely, though, the desire for more time with e.g. family has resulted in rearranging things so that in addition to time with family a resident also spends time with traffic, time with a vacuum cleaner, time with …

    Somehow, though, there are still 168 hours in a week.

  19. weing says:

    @Marta,
    I agree completely. I still remember driving home at night after being on call the previous night, windows open, slapping myself to stay awake till I got home.

  20. David Gorski says:

    Oh, I did even worse than that. There were a couple of times where I drove the three hours from Cleveland to Detroit after having been on call because I didn’t want to waste any vacation time that might be spent visiting my family and came close to falling asleep at the wheel on more than one occasion. Stupid, I now know (and probably knew even then). I should have at least taken a nap or waited a day before heading out.

  21. I would guess that continuity of care is a pretty big deal.

    Even in IT, abruptly handing off an issue or task to someone else often results in mistakes, outright disasters, or unnecessary duplication of efforts, but in IT, people rarely die when the email server goes down.

  22. Mhops says:

    I understand that the idea of canceling cases or clinics has, itself, plenty of negative ramifications. However, when it comes down to patient safety, it needs to at least be on the table.

    The idea that is often presented – that “this isn’t an option, so get used to it” (paraphrasing former attendings) – is a flawed premise used to justify the need to “practice working 2 days straight”. I know there are other arguments for the long shifts, but this was the one I kept hearing. Personally, I haven’t yet made any cancellations for this reason, but then I also chose a specialty and work environment in which this is rarely an issue.

  23. daedalus2u says:

    I have mixed feelings about this. On the one hand it is a human rights violation akin to slavery to compel and exploit individuals to work under conditions that are so dangerous and self-damaging. On the other hand, if you want people to be prepared to handle a “real” emergency, they have to practice at doing things that are emergency-like when it isn’t an emergency.

    I think the analogy to war-fighting is appropriate. The objective in fighting a war is to minimize the sum total of all casualties, casualties in training and casualties in battle. Training that is brutally severe enough to kill a few people in training might save many more lives in a battle. This is the unfortunate calculation that I am glad I don’t have to make.

    The objective in training a physician is (or should be) to minimize the total mistakes the physician makes over his/her lifetime. The same cost constraints that compel resident hours to be long also compel medical resources to be insufficient for foreseeable emergencies. Then there are the unforeseeable emergencies.

    http://en.wikipedia.org/wiki/Coconut_Grove_fire

    The calculation that I do, is that in return for being patients to residents in training to “practice on”, society has a right to expect that residents will become MDs who “give back”. There is no such quid pro quo with insurance companies. I see that as the problem with insufficient allocation of funds to training residents. It will always be cheaper to not train the next generation of physicians and/or exploit them simply because they can be exploited. The question is (or should be), who gets the “value” of that exploitation? Does it accrue to insurance company stockholders in the form of higher profits? Does it accrue to society in the form of greater care for those without insurance? Does it accrue to society in the form of increased abilities of MDs in the event of a “real” emergency?

    I don’t think the exploitation premium should simply go to those who are willing to do the most exploitation.

  24. “The calculation that I do, is that in return for being patients to residents in training to “practice on”, society has a right to expect that residents will become MDs who “give back”.”

    Eeww, I don’t like that calculation at all. I figure that in return for having access to practicing MDs with appropriate training, I willingly accept residents in training “practicing on” me.

    (Also, that in return for being available at crazy hours, in whatever community they are in doctors are entitled to a level of remuneration that will allow them to hire permanent, reliable help at home.)

    That’s it. I don’t feel entitled to have doctors sacrificing themselves for me.

    I’m currently reading The Great Influenza* http://amzn.to/gueqK8 about the 1918 flu, and one of the themes addressed is professionalism. Professionals worked. Nurses and doctors went where they were needed and worked until exhausted or dead from flu. Lay people… not so much. They were afraid of being killed by the flu and stayed away from the sick. I’m guessing it’s some sort of bystander effect. In an emergency, if you’re hurt in public, it’s much less effective to say “Help, help!” than to say “You there in the yellow sweater. Call 911. And you in the propeller beanie, can you see where the blood is coming from?” Professionals feel like they have been individually called and have a personal commitment, whereas non-pros kind of lurk in the crowd and feel it’s not their problem.

    If that’s what’s going on, I think that makes professionals open to exploitation. I don’t think professionals owe society any more than I do. (In 1918 it wasn’t just a question of specialist knowledge. People died because there was nobody to give them food and water.)

    *I keep thinking that Harriet Hall should review it, if she hasn’t already.

  25. qetzal says:

    Is this an issue in all developed countries, or mainly in the US? If the latter, how do other countries avoid these problems, and can we use similar approaches here?

  26. David Gorski says:

    The idea that is often presented – that “this isn’t an option, so get used to it” (paraphrasing former attendings) – is a flawed premise used to justify the need to “practice working 2 days straight”. I know there are other arguments for the long shifts, but this was the one I kept hearing.

    So what do you do when you’re the only general surgeon for 100 miles, have been up all night operating on an emergency, and you get called to the emergency room for a patient who’s got a lower GI bleed that just won’t stop and needs surgery now? Do you tell the ER doc, “Sorry, I can’t do it” and get him call the helicopter to fly the patient to the nearest hospital that can handle him, which will result in a delay, possibly a fatal delay, in getting the patient to the operating room when you could be in the OR with the patient within the hour? This is not a hypothetical situation. I’ve met a number of surgeons who have similar emergency consults after they’ve been up most of the night on a near weekly basis.]

    Those of us who live in an urban or suburban area rarely, if ever, have to face this sort of situation, true. There’s almost always some backup. However, there are more parts of the country than you might think that are underserved. Obviously, it would be better to get more doctors out there, so that this isn’t as much of an issue, but what do you do in the meantime if you’re a doctor out there?

    Being able to tap deep into oneself and do the case, even if you’re not at 100%, is professionalism.

  27. I think if the goal is to train doctor’s to tap into their inner resources in order to overcome fatigue when needed in an emergency situation, then the question would be; how many experiences of fatigue does the student need to learn the lesson? Could this be accomplished through drill sessions that students are randomly called to participate in once, twice, four, six times a year or does it have to be a grueling weekly routine?

  28. aaronupnorth says:

    I too think this is a human rights issue rather than a patient safety issue. There are certainly ways to make hand-overs safe though we in medicine have shown great impotence in pursuing any of these options.
    As a resident I too (like Dr. Gorski) fell asleep at the wheel while driving home after a call shift. I was on a freeway traveling 70mph and was saved only by a very wide paved shoulder and rumble strips. A fellow resident at my hospital was permanently disabled after a similar accident. The majority of my cohort who began as 30-something married people ended up divorced. Most safety sensitive industries now have work hours restrictions to deal with both client and worker health, but medicine is a laggard in this regard.
    Even if you wish to approach this issue in an evidence driven way the most relevant evidence is the effect of 24+ hours of consecutive work and 80+ weekly hours of work on the workers.

  29. ConspicuousCarl says:

    David Gorskion 30 Nov 2010 at 9:15 am
    The more “shifts” you have, the more patient hand-offs there are, each of which introduces the possibility of error and, at the very least, guarantees that physicians who don’t know the patient very well end up covering.

    How many patients have a short enough stay that 24-hour shifts make a difference in how much total time is spent with a specific doctor? Or much difference in how many total hand-offs there are? I am not a doctor and have never worked in a hospital. Please humor and/or correct me…

    It seems like there are three general situations:

    1. A patient leaves the hospital after 1 doctor-shift-length or less. This seems ideal, though some patients may be unfortunate enough to have to switch doctors in the middle.

    2. A patient has to stay in the hospital for more than 1 doctor-shift-length, but less than 2. Hopefully, this is “good enough” if not perfect, with one doctor being there “most” of the time.

    3. A patient has to be in the hospital for 2, 3, or more doctor-shift-lengths. This sucks, even if a particular doctor is able to stick with that patient as much as possible as allowed by the rules.

    I found one source that said the average hospital stay is 4.8 days, which is way longer than I thought. I assume 4.8 is a mean. If the median is much less (e.g., most people are out in a day, but a few are stuck there for months), then a long minimum shift length makes sense.

    But if the median stay is long enough, not only does the system have to be arranged such that multiple hand-offs can be done safely (whether the average is 5 or 7 hand-offs, at that rate you have to figure out a reliable method). Also, having more than one or two handoffs means the patient is unlikely to be under the care of the same doctor for much more than about 50% of the time.

    So “1″ is probably impossible for most patients, unless you want to have the average doctor work for 5 days straight or longer. “2″ may range from good to neutral depending on exactly how long the median or mode hospital stay is. Unless the range of hospital stays is heavily weighted to the shorter end, “3″ is probably most common and not very sensitive to the whether a doctor works 24 or 34 hours straight.

  30. I think there probably is something character-building about discovering that you can do something you didn’t think you could. Surgery seems to be particularly gruelling that way.

    For tales of a lonely surgeon who is regularly the patient’s only hope, see Other Things Amanzi. This post describes one of his early character-building exercises.

    http://other-things-amanzi.blogspot.com/2009/02/alone.html

  31. Geekoid says:

    @david G – regarding patient hand off.

    What you describe indicates to me that patient hand off process should be improved, not that the short shifts aren’t working. If the hand off is worse then having a doctor who hasn’t slept in 24 or more hours, is seems to be broken.
    What am I missing?

    When I worked with doctors, I has asked many times way the need to work so many hours. There answers seemed to boil down to ’cause’.

    Of course I also wonder with a demand for entrance so high, why medical schools can’t expand over time to have more students. It seem that while only 10% of the applications are great candidates, only 1% or so get into the school.

  32. Wolfy says:

    per dr gorski:

    “Bottom line, the problem with work hour limitations is that they do not apply to every physician. They only apply to residents, not to fellows or attendings. What that means is that, whenever there are more work hour restrictions, the work load shifts upward to fellows and attendings, many of whom already work 70 or 80 hours a week.”

    This is an excellent point. I would add that in the area of “resident work hour restrictions” fellows frequently end up as “glorified residents” helping residents out with their work (admissions, daily notes, discharges) instead of focusing on fellow level work.

  33. Mark says:

    Great discussion. I would like to weigh in as a current internal medicine program director. The problems we are facing in academic medicine are vast, and this duty hours issue is a major challenge. I too trained in the pre-2003 era, and did not particularly enjoy being dog tired all the time after sometimes spending 36+ hours in the hospital. 80 hours per week and 30 hour shifts are reasonable limits. Sadly studies have shown reduced hours in the hospital have not increased resident sleep time. They spend time with family and friends and likely end up just as fatigued. Cutting hospital time further will likely have a similar result.

    some of the comments above though deserve some rebuttal:

    “1) If remuneration is per act, the number of acts shouldn’t change with the number of doctors.” – simple answer – remuneration for resident services is by annual salary, not by act or by hour. The number of acts is the same, but far more docs are needed if fewer hours are worked per doc. Increasing the numbers of residents in a program is generally not an option as there are Medicare imposed caps on funded resident training slots. I would happily add more residents if I had the money to pay them. Another dirty little secret is that soon there will be more graduating US medical students than there will be residency slots to train them in. Additionally, their funding is only good for 3 years. If they are not competent at the end of that time, the cost of any additional training is eaten by the program. Some programs will thus terminate underperforming residents that may eventually become competent physicians with enough time, rather than continue attempts to remediate at a significant cost to their program.

    “I too think this is a human rights issue rather than a patient safety issue.” I could not disagree more. This is not an “Arbeit macht frei” situation. We as program directors strive to ensure that patients are cared for safely, residents are well treated, and that we graduate residents that are competent, with training fixed to a maximum 3 year time frame. These are often mutually conflicting goals. I wish 40 hours a week would do the trick and residents could see and experience enough to become competent. That is simply not the case. These residents are not merely cheap labor. They are recieving a tremendous amount of education and mentorship, usually from some of the best docs to be found.

    Simply pushing the workload onto attending physicians is a recipe for disaster. They will have less time to teach, conduct research, and will see essentially no increase in compensation. Teaching attendings forgo a significant amount of potential income becuase they enjoy working with and teaching residents. Increasing their work burden further will alienate many talented physicians and make recruitment of teaching faculty all the more difficult.

    I wish I knew the answer to this. I can forsee the need in the future for some period of supervised practice after initial residency training or the need to prolong residency training. The problem is that there is simply no funding. Medicare is broke, and only congressional shenanigans keep the books fudged enough to hide the true deficits in the system. The ongoing saga of the so called “doc-fix” is an example.

  34. Stuartg says:

    *Is this an issue in all developed countries, or mainly in the US? If the latter, how do other countries avoid these problems, and can we use similar approaches here?*

    It probably is a worldwide problem. I’ve done my time and even withdrawn from a resident programme because I had more respect for my family than the hours that the programme demanded.

    In New Zealand the approach to limiting residents hours arguably happened earlier than in the USA, initially in the late 1980s, but now is lagging. It wasn’t done by state or governmental statute but by the action of unions, the resident medical officers managing to gain increasingly punitive pay rates the more hours they were required to work.

    Senior medical staff are well behind the residents, being mainly salaried and expected to finish the job however long it takes. 24 hour call is still common for senior medical staff but no longer for the residents.

    And no, I don’t have an answer, sorry.

  35. David Gorski says:

    Simply pushing the workload onto attending physicians is a recipe for disaster. They will have less time to teach, conduct research, and will see essentially no increase in compensation. Teaching attendings forgo a significant amount of potential income becuase they enjoy working with and teaching residents. Increasing their work burden further will alienate many talented physicians and make recruitment of teaching faculty all the more difficult.

    This is another point that I didn’t get to. As more and more of the work that residents normally do gets pushed upward to attendings with no additional compensation, at the attending level academic medicine will start to look more and more like private practice, only not as well paid and with a lot of additional academic responsibilities added to a private practice job. And if you’re going to do the same work as private practice plus all of the other duties that come with academic medicine, private practice is going to start to look more and more attractive to more and more docs.

  36. Mark P says:

    When I left High School I had the grades to get into Medical School directly (in New Zealand).

    I discounted becoming a doctor because of the whole residency thing. I did not value the money at the other end enough to go through that. I earn far less than if I were a doctor, but I’m not sorry with my choice.

    Perhaps if doctors were paid less some would not take up the profession. Others would take their place. People who might value money rather less, true, but would that make them worse doctors?

    From my perspective, residency is a way for doctors to justify extremely large earnings. They did it the hard way, and so are entitled to anything they take home. Take away those residency years and the sacrifice becomes less, so the pay seems less justified.

  37. wertys says:

    Resident hours has been a big issue in Victoria (Australia) for years, and the average hours worked by junior doctors has reduced from around 60 per week in my day (mid-90s) to 45 per week. This has been achieved by training more doctors and by using IMGs (International Medical Graduates), ie stealing them off other countries who can’t afford to lose good doctors. As part of the package of demographic shift, the junior doctors are also taking far more time off regular salaried postions in hospitals to do locum work, often at the same place but for more pay and fewer hours. The crisis is now not about overcommitted work-crazed junior docs, but undercommitted, me-first, work-shy doctors who don’t work the long hours we used to, but also have far more of a casual attitude to their professional responsibilities.

    Clearly generalizations are not invariably correct, but as someone who has worked over the last decade at the same hospital at both junior and now senior levels, I believe it is fairly accurate. I have no idea which approach is safer for patients, and no accurate data because the new risk-management systems in our hospital have by themselves made severe medical errors much less common.

  38. Mark Crislip says:

    The solution is to become a naturopath etc. No long residency training to worry about.

  39. JMB says:

    I was told the rational for long hours in training was; (1) experience as many cases as possible in a short period of time, (2) learning how to function in case of a disaster scenario. As a first year resident I remember working with medical students who were in a much more protective environment than the school I attended. They were very limited in the number of patients they would work up and follow, and never took overnight call. Even the brightest students of that program seemed ill prepared for residency training.

    I remember that certain residents, even if they were fatigued, would make better decisions than other residents who were not as sharp. Fatigue is a factor in medical errors, but so is lack of experience. I think I would put more faith in the decisions of an experienced resident than the well rested resident. Given the importance of experience in the functioning of a physician, I have to wonder what will happen as more primary care is provided by non-physician practitioners. If physicians accumulate experience through 60 to 80 hour work weeks in the last two years of medical school, and 3 years (or more) of residency, how does that compare to the clinical experience of non-physician primary care providers?

    Was the major contributor to the medical errors in the Libby Zion case the fatigue of the intern, or the lack of supervision of the intern? If we really wanted to correct such medical errors, wouldn’t expansion of the role of the hospitalist be more important than limitation of work hours of residents?

    Whether limiting hours of residents will result in a reduction of medical errors is an assumption that should be tested. The indisputable toll of long hours is the high divorce rate, high drug abuse rate, and high suicide rate of doctors in training.

    ****************

    So many in the USA think the supply of training spots is arbitrary. The training is very expensive. Monetary support for training comes from the government, which ultimately limits the number of doctors trained. Want more doctor training spots? Write your congress person. Many doctors wondered why healthcare reform didn’t increase the number of training slots. There is a major looming shortage of all specialties, not just primary care specialties. If everybody gets health insurance, the most obvious shortage will initially be in the primary care specialties.

  40. JMB says:

    @Mark P

    Others would take their place.

    In the late 80′s in the USA, the interest in the medical profession as a career dropped. Instead of an average of 15 applicants for each spot, we were getting about 8 applicants. Consequently, the grade point average of the entering medical school class dropped. At the same time, the percentage of students failing the national board tests increased. Students with a B- average in college had a significant failure rate on the national boards. If you drop the economic incentive, you will drop the qualifications of applicants to medical school. You will have to make it easier to become a doctor, and the doctors will be less qualified. Fewer will be willing to work long hours, weekends, and holidays for less pay, so you will exacerbate the doctor shortage. Still, that is an easy path to reigning in healthcare costs, drop the requirements for physician licensure to 4 years of training after graduating high school, and make the national boards much easier, and we will have lots of cheap doctors to choose from. I’d bet we wouldn’t even see much change in our average life expectancy, that sensitive measure of quality of care.

  41. Hmm, there’s a lot of projections of what might happen to prices, demand, labor supply in this thread. It’s beginning to look like you need an economist with an interest in health care on this board. :)

  42. overshoot says:

    RE: remuneration

    Those suggesting that part of the problem is the high cost of MD remuneration should consider the staggering burden of paying off medical education. The out-of-pocket bills are bad enough, but don’t overlook the opportunity costs.

    Blogger WhiteCoatTales is a good example: BS in engineering with grades good enough to get him into medical school. Right there, he’s looking at a starting salary of $80K/year to start. By the time he finishes residency, many of his engineering classmates will be making over $110K/year. Call it a total forgone income of $700K or thereabouts. Add the bills from med school itself (don’t double-count living expenses) and subtract any stipends received prior to completion of residency. The total cost (actual and opportunity) easily exceeds a million dollars, and it has to be amortized over a career.

    The interest alone on that tab is going to run north of $50/year [1], and you (again) have to add that to the pay of his engineering classmates. Which means that his break-even figure for when he completes residency would be around $200K [2]. A more realistic amortization time would front-load the payoff, which suggests that his income when he hits 50 (and is in the stage where many of us are shifting our priorities towards passing on what we have learned to another generation) will appear rather generous. It will certainly make for more expensive training for residents and medical students.

    All of this, remember, is to maintain parity with what he could reasonably expect from just a BS in engineering. There’s no value assigned to the work he put in himself, just the financial investment.

    [1] I’m using pre-tax figures for all of this, since most of it is income-based except for the out-of-pocket medical school bills. That makes the results somewhat conservative.
    [2] Uunless you assume negative amortization early in his career, which increases the total lifetime break-even cost.

  43. Dawn says:

    To be honest, I have to say that one group of help is being overlooked. Nurses.

    I am NOT saying that nurses can replace residents. However, when I graduated from nursing school, I worked in hospitals where the nurses took care of the routine issues (start IVs, push most medications, etc).

    Then I went to work at a hospital where the nurses were not allowed to do any of those things on anyone! So, instead of starting a new IV in anyone, I had to call the resident, wake him/her up, and wait for them to arrive. EVERY time. It was frustrating for me and, I felt, useless for the resident who could have used the solid sleep time.

    When I transferred to Labor and Delivery, I ignored the restrictions and began to start my own IVs, drawing the admission labwork at the same time (with the lab’s OK). My patients were thrilled they only had to be stuck once, the resident got to sleep a little more since they were not required to do anything to admit private patients (they only cared for the clinic patients unless the attending requested their help or the woman delivered before the attending arrived). They trusted the nurses to call them in emergencies but to take care of routine things.

    In hospitals where the nurses are trusted to handle simple, routine issues, the residents get more rest, I think, because the nurses expect the residents will come when they are really needed. I worked with mostly great residents and we had a great relationship. If they could get a good night’s sleep while on call in Labor and Delivery, I was more than willing to let them do so, as long as they came when I really needed them.

  44. Fifi says:

    “Fewer will be willing to work long hours, weekends, and holidays for less pay, so you will exacerbate the doctor shortage.”

    This just isn’t true – certainly not in Canada. Many female doctors (and increasingly male ones) choose to work shorter hours for less money because they value having a personal/family life.

    http://www.cbc.ca/health/story/2009/11/26/doctor-supply-canada.html

    The whole internship system is more like a hazing than it is any real form of learning/teaching (sleep deprived people don’t learn well and obviously they also make more mistakes) – it’s also at least partially about hospital hierarchies/pecking orders.

  45. Mhops says:

    Dr. Gorski – I certainly agree with probably sucking it up and doing that case in your example. That is indeed professionalism — unless you are so impaired that you may as well be a little drunk, then we might use other words to describe it. My overall point was that such examples, while they do exist, are the exceptions to daily practice and never made for a very compelling argument to justify extended shifts. For some reason I more often heard these macho arguments presented over the others.

  46. pulsetsar says:

    I wonder how the cognitive science research on the magic “10,000 hours” rule for development of expertise plays into this. Certainly for surgeons, but likely for many other subspecialties, the new policies will require extending the number of years spent in residency in order to create equivalent levels of expertise to what we currently achieve. Although, focusing on didactics more may help.

  47. BillyJoe says:

    “In hospitals where the nurses are trusted to handle simple, routine issues, the residents get more rest, I think, because the nurses expect the residents will come when they are really needed. ”

    That’s a refreshing attitude.

    …in contrast to the local experience:
    Australia has seen recently the introduction of nurse clinics. The nurses working in these clinics are demanding pay equal to that of doctors working in general practice. This is despite the fact that they refer onto these doctors the cases they cannot handle. One Hospital emergency dept that happens to be close to such a nurse clinic has seen an increase rather than the anticipated decrease in attendances for non-emergency conditions because patients have attended the nurse clinic instead of their GP who would have been capable of treating these cases himself (or herself :)).
    There are also nurse clinics set up by pharmacists right inside their own pharmacies. Conflict of interest anyone?

  48. pmoran says:

    Dr. Gorski – I certainly agree with probably sucking it up and doing that case in your example. That is indeed professionalism — unless you are so impaired that you may as well be a little drunk, then we might use other words to describe it. My overall point was that such examples, while they do exist, are the exceptions to daily practice and never made for a very compelling argument to justify extended shifts.

    In my day (before Dr Gorski’s) the long hours were definitely considered a rite of passage into medicine. We surgical trainees also didn’t doubt for a minute that the more time we spent in hospitals the better clinicians and surgeons we would be. The decision whether an abdominal emergency required surgery, for example, required highly sophisticated clinical skills which could only be developed through prolonged contact with very many patients.

    When you are young and enthusiastic I think you can get away with it most of the time. Nevertheless I recall two occasions when I was aware of major impairment. On one occasion my consultant happened to also be in the hospital at 3.00am and dealt with the matter. On another horror weekend the administration took the extreme step of closing the RNSH to further admissions. The entire medical staff had had no sleep for two nights and needed some rest before turning up for work on Monday.

    One point is that young residents/interns are probably more likely to make errors through inexperience or inadequate knowledge, so that making sure that they have adequate overall supervision and backup is at least as vital a matter as the hours they work. Medication errors should eventually be largely eliminated by computerised systems.

    Technology may have reduced the need for some of the old skills.

  49. JMB says:

    @Fifi

    I was examining the hypothesis that making the training of doctors easier (for example, completing training in only 4 years) could be a solution to the expense and availability of healthcare. I was suggesting that by lowering the requirements of training, we would also see a decrease in the percentage of physicians willing to work nights, weekends, and holidays. Canada has substantially similar requirements for licensure to the USA. I don’t know how the two systems compare in the percentage of trained physicians that choose to work less than full time, but that comparison does not reflect on the hypothesis that we could reduce expense and increase availability of healthcare by reducing the requirements for training (since the two systems have little differences in education of physicians, at least as far as I know).

    The bottom line is that there is an incredible amount of knowledge required to be a functional physician. Much of the knowledge can only be gained by experience. Part of the reason for the rise of specialties is that it is easier to master the amount of knowledge to be a functional physician in 8 to 10 years following a college degree by restricting your field of practice. I think I remember reading once that medical training added about 20000 words to the vocabulary of the college graduate. That is just one measure of the amount of knowledge required to be a functional physician.

  50. Dawn “Then I went to work at a hospital where the nurses were not allowed to do any of those things on anyone! So, instead of starting a new IV in anyone, I had to call the resident, wake him/her up, and wait for them to arrive. EVERY time. It was frustrating for me and, I felt, useless for the resident who could have used the solid sleep time.”

    Wow, this must vary quite a bit by hosptial? I don’t believe I’ve ever had a resident start an IV (with me or my son). It seems to me it’s always been the nurse in pre-op or the anesthetist in one case where the nurse had trouble.

    BillyJoe – Your story on the nurses in Australia is interesting. In my experience, here in Michigan, the NP work within practices with GPs. A NP might handle the more routine office/sick visits or in some cases more specialized counseling that is in high demand or more time consuming that the typical 15 min sick visit. For instance our pediatrician’s office GP handles alternate yearly physicals for children without complex health concerns, sick visits where she is not shy about consulting one of the pediatricians on a thorny diagnostic issue and she is a very good advisor on potty training, constipation and stomach pain for the younger set.

    I believe one of our local pediatric emergency room also has a NP (along with Pediatric ER docs) who handles things like nursemaid’s elbow…but I’m not sure about that.

    Sorry off topic.

  51. Charon says:

    I’m trying to decide if doctors have completely overinflated egos and warped senses of reality, or if their profession is really so different from mine. I think it’s a little of both.

    If you’re learning complicated material, you can’t do this tired. Limiting to 24 hours instead of 36 or 48 is completely pointless, that’s still way too long. I say this as someone who’s spent his whole life in academia, and both taken and taught lots of math, physics, and astronomy. But it’s true that so much of medicine is wrote memorization without much thought (as anyone who’s tried to teach physics to pre-meds has discovered… great at memorization, thinking not so much), and maybe this can be learned and regurgitated when extremely tired.

    That said, I have to agree with the comment above that noted that if patient hand-off is such a problem, that’s because you’re doing it wrong. I would much rather have two well-rested doctors take care of me than one bleary-eyed automaton, and if patient hand-off is done correctly, there shouldn’t be much of a problem with this.

  52. Charon says:

    Also, people who are thinking-impaired tend not to be aware of this. This sounds kind of obvious, but I don’t know if a lot of people appreciate this. Extreme-altitude climbers who are hypoxic have horribly slow and bad thinking processes, but most of them subjectively experience their thought process as fine. This is why you always get someone else to check your knots, even if you’ve tied them thousands of times before, and why very experienced climbers have died when they didn’t do this. This is also why we teach people in wilderness first aid what signs of HACE to look for in other people. Because the person with HACE is the last to know about it.

    So, all you doctors who think you’re doing just fine after 36 hours on the job, or (even more subjection to distortion) thought you did just fine as residents years ago, I don’t believe you. This is SBM. No anecdotes.

  53. Harriet Hall says:

    @charon:

    Rote memorization is not an issue in residency training. Doctors are actively caring for patients and gaining experience in applying their book-learning from medical school to practical situations. One of my biggest frustrations in residency was that I couldn’t read textbooks and medical journals because I was too tired to concentrate and didn’t have time.

    “all you doctors who think you’re doing just fine after 36 hours on the job” What doctors are those? Did anyone say they think they’re doing just fine? I think it is exactly the opposite. Doctors realize they’re not able to function optimally when fatigued, and the proposed regulations are a response to that realization.

Comments are closed.