Fatigued Physicians Make Mistakes and Harm Patients

Fatigued drivers cause accidents. In response to this obvious fact, we limit bus and taxi drivers to a maximum of 10 hours of driving after 8 consecutive hours off duty. Yet when it comes to physicians, the current standard is significant more lax; first-year residents are restricted to 16-hour shifts! That already is nuts. I often teach a night class, 7:20-10 pm and I always try to teach the more difficult material early because by 9pm I am not at the top of my game. Needless to say, medical residents are far more stressed and fatigued than teachers. Moreover, while first year residents can work up to 16 hours, second year residents can work up to 24 hours straight and even up to 30! Isn’t it amazing how one year of residency can teach physicians how to function without sleep?

The current standards, which strike me as absurdly low, are actually due to restrictions put in place in 2003 and 2011–restrictions which are now being lifted. The new plan is to allow longer hours for first year residents:

Rookie doctors can work up to 24 hours straight under new work limits taking effect this summer — a move supporters say will enhance training and foes maintain will do just the opposite.

A Chicago-based group that establishes work standards for U.S. medical school graduates has voted to eliminate a 16-hour cap for first-year residents. The Accreditation Council for Graduate Medical Education announced the move Friday as part of revisions that include reinstating the longer limit for rookies — the same maximum allowed for advanced residents.

An 80-hour per week limit for residents at all levels remains in place under the new rules.

Studies have found that physicians who work longer hours are much more likely to get into auto accidents on the way home. Physicians and nurses who work longer hours also make more medical mistakes.

The main argument in favor of long hours is that the 2003 and 2011 restrictions do not seem to have greatly improved patient safety. That is surprising but the micro and experiential evidence that fatigue makes for mistakes is so strong that the lesson to be drawn isn’t that longer hours don’t lead to mistakes–the lesson is either that the restrictions were routinely ignored (as the National Academy of Science study found), that the studies done to date are misleading for a statistical or design reason or that there is another constraint in the system that needs to be examined. One possibility for another constraint is that handoffs of patients between physicians aren’t handled well. But that means that poor handoffs are killing as many people as fatigue!

In no other field do we tolerate error as much as we do in medical care. Why does the government regulate driving hours more than medical hours? It’s not just the government. It’s amazing that in a society where McDonald’s can be sued for making people fat that the tort system hasn’t shut down absurdly long residency hours (there have been a few cases). Medical care is a peculiar field (cue Robin Hanson).

Aside from Hanson-type factors, a key factor that explains what is going on is that residents are a huge profit source for the hospitals. Much like student athletes, residents are underpaid. As a result, hospitals want to use residents as much as possible so they lobby for longer hours even at the expense of patient safety.


Doctors are the third rail of the American medical system. Discuss.

Economist Alex thinks physicians need even more government regulation & finely tuned rules. Apparently government licensed medical professionals are too dumb or malicious to get adequate sleep for their work ... if so, what does that fundamentally reveal about government medical regulation and the current heavily regulated medical profession?

Do government regulators get enough sleep to perform their duties well?
(how 'bout economists?)

I relly love the Fundamentalist answer: "reality is of no concern, sleep is a government invention".

Kimball, see how it sounds with a substitution: "Apparently hockey players are too dumb or malicious to know when they should wear helmets to protect themselves...." This is a very good analogy in terms of the forces that are at work in the minds of rationally self-interested decision-making doctors. It would be a better analogy if, in addition, a large subset of coaches strongly implied (or even came right out and told you behind closed doors) that wearing a helmet would show you're not really committed to the game and would decrease your playing time.

Something something asymmetric information something. iow, caveat emptor.

Seems like more branding in medical services might help. Let the branding company decide who can do what.

"Much like student athletes, residents are underpaid. As a result, hospitals want to use residents as much as possible so they lobby for longer hours even at the expense of patient safety."

If greed is good, everything is allowed. Even to kill one's brothers to make more money.

Greed is like gravity--whether we think either one good or bad, we'd better take them into account when designing institutions and buildings, or our plans will collapse into rubble.

Yet there is a big difference between acknowledging greed - or gravity or cancer - , planning accordingly, and celebrating it as Americans do. In the USA, man has become a wolf to man.

It depends on your definition of greed. The meaning has changed, by the modern definition greed is never good. (a selfish and excessive desire for more of something (as money) than is needed motivated by naked ambition and greed). At one time, it meant desiring more than you need, not so bad.

My more exact definition: Willing to do immoral things to get what you desire.

As a medical student Myself and fellow students were regularly forced to stay up during rotations for over 36 hours at a time. Some kids got into car accidents. Some fell over while walking home alone side roads. Patient Medical care and student learning was hampered. SUCH CONDITIONS SHOULD BE ILLEGAL!

Does it frighten you that medical administrators are too stupid to figure out safe coverage and hand offs in a humane way? American exceptionalism!


I think it is like an initiation ritual, like those that fraternities do. Some say I went through it and you should have to do it too. It holds the club together. Perhaps, we need to close the club (AMA).

It's more complicated than that. While physicians worked long and grueling hours back in the day, the work itself was more educational - it contained a much higher ration of hands-on medicine and learning to scut and paperwork. Now the latter has risen to the point where it's swamping the former. If you've been out of residency for forty years, it looks like modern residents are trying to avoid learning medicine. If you're in the trenches now, you realize that work hours restrictions don't leave enough time in the day for residents to get their paperwork done _and_ learn their jobs. The best solution is to cut down on the paperwork required of physicians - but do you really see that happening?

Boy are you wrong- the scut level was much higher in the past. We had to do our own microscopy & staining of sputum & at night spun our own HCT. And I am talking the 80s. Furthermore, hours were much, much longer. Many did every other night call for months- 80 hours was a slow week for an intern. But the patients weren't as sick. Prior to the 70s & 80s patients spent a lot of time in the hospital simply recuperating-not now. The procedure rate is much higher as is the patient "churn rate". BTW NO ONE is cutting down physician paperwork- EMRS are actually increasing it & it is of poor quality- a lot of cut & paste boilerplate in the record & when reviewed it is real hard to separate the wheat from the chafe- i.e. anesthesia records from EMR are often 20 pages of size 9 font with most of it shear BS- and it is very difficult to find salient intraoap events. The training of physicians is a difficult process & it is clear that severe sleep deprivation is bad and if we limit hours we probably have to lengthen residency to ensure residents get the training they need.

Why don't residency programs open up to 2x students? Would costs be so much higher?

Interns and residents need a lot of handholding and supervision, especially in the first couple of years. Double the number of interns/residents, and you'd need to double the number of supervisors and seniors.

Ah, but that is holding the length of program constant. Even worse, if overwork reduces retention of lessons, it locks in repetitive non-learning. And extended length of overwork to make up for fatigue.

Maybe the submariner's "6 on, 12 off," gets more through the program, with better learning, and the same student-teacher ratio.

It strikes me that the whole thing has not been studied and optimized, it has just become a cultural tradition and rite of passage.

I wonder if older docs would be better at giving up disproven therapies if they had been better rested when first learning them. Sleep deprivation, after all, is among often used brain-washing technique.

I don't know that the training programs need to be longer for efficiency of learning, especially since the issue really isn't learning at all. No one is saying doctors don't learn enough during internship/residency.

And if programs were longer (which would meet tremendous resistance from the people who would have to attend these longer programs), while program size was held constant, you'd still end up with more interns/residents and more attending physicians. Unless I'm missing something, your solution dumps the problem onto doctors in training (yes, you get to eat and sleep, but that time sleeping and eating is going to delay the time you get paid FMV for your work!), while also increasing hospital costs.

Doctors and hospitals refuse to pay to train their own employees. They insist that all residencies be paid for by the federal government. They also explain that it is impossible to increase the number of doctors because there are not enough residencies.

Residents have to have a certain amount of exposure to patients and cases. For example, a surgery resident has to keep a case log of the operations that participate in by number and type, and submit it when they sit for boards. This means that a program can only have so many residents. If your resident to patient ratio is too high, the residents don't see enough cases to be properly trained. Hospitals are allotted training spots based on patient volumes, procedure types and caseload.

The cynic would look at the long hours pulled by residents as a "narrow gate" similar to medieval to keep competition low.

I understand that doctors need to know how being sleep deprived impacts judgment, just like nuclear engineers need to know. But it seems like the same message could be sent by depriving them of sleep and then having them play doubles ping-pong or taking tests. No need to put patients at risk to send the message.

If only this were the case. The gate is kept narrow in that the number of residencies are essentially fixed and med students (who must have attended a med school where seats are currently outnumbered by applicants 2:1) must match into a residency via a giant cartel. There is an explicit carve out of monopoly law to keep the gate nicely narrow without the hours.

The truth is, this is pretty much all about the money. An average ER doc makes around 300K. A resident makes around 60K. After the itern year, a resident can legally do most all of the things an attending can do, but still is overseen by an attending. Cutting back resident hours leaves us with just a few options:
1. Hire more attendings. This is very pricey and will quickly get into bidding wars, particularly as noted above the supply of med students and new attendings are already limited by regulations.
2. Make residencies longer. This is already happening. Pretty much every specialty is watching the length of residency creep higher. The amount learned drops off heavily, but in theory your doctors are better trained and hospitals can access the cheap labor longer. The down shot is that medicine becomes even less attractive when you are looking at almost two decades of additional education.
3. Push more work down from the MD level. Physician extenders (e.g. nurse practitioners) are already massive, but given the need for physician oversight, this sort of care directly trades cheaper labor costs for more overhead. I also worry that for areas of care where good, sustained doctor-patient relationships are vital (like primary care, cardiology, etc.) this will lead to ever more rationing of doctor-patient time and erode the relationships further.

Medicine is very much a guild, but the guild is not based around the insane hours expected of physicians.

"who must have attended a med school where seats are currently outnumbered by applicants 2:1"
So what? In Brazil's med schools, it is much higher than 50:1. It does not prove there are few slots, it proves there are many applicants.

"The gate is kept narrow in that the number of residencies are essentially fixed"

No it isn't!

Any insurance company can fund teaching hospitals to have more residencies!

Any business group, say the Chamber of Commerce can fund more residencies in the State or region they serve.

Any conservative think tank foundation arguing the government funds too few residencies can fund an equal number of residencies as CMS does.

The Koch brothers and Thiel, and the millionaire Republicans in Congress, can easily fund 5000 more residency positions privately each year.

Hey, the big hospital corporations, one of which had a founder in Congress seeking tax cuts, can self fund two, three, four times the residencies as the government funds.

The conservative State governments can fund more residencies in their State instead of submitting to the actions of distant unresponsive Washington because each State can still tax and spend, and it's the conservative States that tend to have the doctor shortages.

Medicare law funded residencies to meet the criticism that funding more medical care would drive up costs because there were too few doctors, and because Medicare pays doctors too little given the crushing debt from their paying for their med school and residency.

The funding by Congress of residencies was to increase the number of residencies by adding to the funding of residencies.

Unless you want more tax and spend by Congress, you want fewer doctors in the US because the private sector has failed to pay for increasing the number of residencies even when the opponents of taxes have billions in wealth they could use to pay for hundreds or thousands of added residencies each year.

Beautiful concise comment. Conservatives will not admit that they actually believe the poor do not deserve medical care:

This would be true, except that is wrong about most everything.

Currently we spend about $12 billion funding resident training in this country. Of that $9.5 billion is coming from Medicare. This obviously leaves around $2.5 billion coming from other sources. Many of them are indeed from state government, private sources, and other areas.

So why not just pay out of pocket? Well in part, doing so makes your overhead figures look worse, this in turn can have impacts on reimbursement rate negotiations. Expanding the pool of residents at your institution without federal funds is accounted differently, this can result in your hospital taking it on the chin from major insurers. Bumping up your resident count risks lowering the payout you get. It has been a while since I have seen our financials, but I think adding residents risked lowering both our DGME and IGME funding. Or in other words, we at least were afraid of losing federal funding if we paid in too much ourselves.

Now other hospitals are in different settings (e.g. they have local monopoly) and some of those have been funded privately. However when we talked with DeVry about them purchasing residency slots, it was thought to be highly risky for the reputation of our program and again risking an order of magnitude more money on other reimbursements.

In theory, the number of residents is limited at each facility based on case loads - residents who fail to see sufficient breadth of practice not being sufficiently trained to practice safely unsupervised. Admitting that we could go higher has the downside risk that Congress will cut IGME.

Long story short, sure you can fund a residency however you want. A few poor residents have to pay their own salaries when their department closes. However, due to the billions being dumped in by the federal government, every funding decision is impacted. This leads to a lot of perverse incentives because reimbursement and even IGME/DGME are screwy.

Ultimately this is no different than the EMR debacle. It is net negative for many hospitals to follow EMR guidelines, but if you buck the federal government too much you can be certain that when time comes for to write new regulations or for discretionary grants being given, hospitals that do not play along will not have seats at the table. Shockingly organizations will sacrifice long term benefits (like more residencies with external funding) in order to avoid risking larger short term assets (e.g. program reputation, grants, reimbursements).

The medical profession is no more. What is left is a medical cartel/syndicate that makes its money by rent-seeking

I am not in any way connected to the medical industry, but my understanding is that there are significant and collusive limits on the number of medical student slots allowed in the system. What that means is that we end up largely with a group of physicians that is more competitively selected than if the system were more like, for example, lawyers. If we crack down on resident hours and frankly physician hours, then we might end up with medical errors due to short staffing, rather than sleep deficiency. Should that trade-off now include a lowering of average academic talent via an increase in available med school slots? I would be interested to hear from those in the industry.

There are currently two medical school applicants for every slot at a medical school in the US (DO and MD). Of those applicants fewer than 25% are not academically qualified for medical school. However, med school is a grind and many people are not really prepared for how much BS they will have to put up with and who hard they will have to work. The goal of medical school admissions in the lower tier medical schools is mostly to find students who will be willing to endure the grind rather than give it up and go do something else.

However, medschool slots are not what limit residencies. Residencies are paid for by Medicare and unless that changes, we will continue to have limits on the number of residents out there. Because residencies have a carve out from monopoly law, you will not see a healthy functioning balance between supply and demand. One possible way out is to let residents pay for their own "training", but mostly that will mean debt financing and either impoverished doctors or (much more likely) the costs of the interest will be pushed into reimbursement later.

The biggest thing that would actually make the numbers work is cutting out a lot of the BS - like paperwork and pointless EMR work - which consumer about half of physician hours. This will make it harder to combat fraud, easier for physicians to coast on minimal effort, and harder for insurance companies/the government to ration care through hassle and annoyance, but we could have physicians easily work more sane hours if we dumped a bunch of the oversight. Of course this would tend back towards the God-complex days so it will never happen.

Why are residencies are paid for by Medicare? Why wouldn't a hospital pay the $60k/yr for a resident out of their own pocket to get such cheap labor?

Because of political horse trading in the 1960s.

Why is it still that way? Because your reimbursement rates from insurance (often the single biggest pot of money) depend on things like overhead rates. If you self-fund residencies that mucks around with the accounting. At my hospital we were afraid that we would see the local insurance companies demand lower reimbursement, as physician costs are far less than 10% of our costs, this is a huge gamble we could not take. Likewise, if you rock the boat by not following the status quo setup, you end up losing clout when it comes to grants, legislation, and regulation.

Hospitals are fantastically risk averse, so going out on your own is generally discouraged. Some places do it, but they are atypical.

I don't know of a single MD program with a long term attrition rate above 10%, and last I had heard 4 year graduation rates were around 85%. So quitting isn't really a significant factor.

It is by far a bigger factor than academic inability. Typically if you can pass the MCAT, you can manage to pass your courses and the USMLE if you are willing to put in the work. It is also a larger factor for lower tier schools than for upper tier schools.

Because, as you note, pass rates are already high small changes from a single student quitting can have outsized impacts on rankings, particularly at the lower end. When I worked with admissions, avoiding attrition was bar nothing the most important concern.

This has nothing to do with quality of patient care. If it did, we would keep better statistics on medical errors ( here is a nice list of links to various data points: https://www.justice.org/what-we-do/advocate-civil-justice-system/issue-advocacy/medical-errors ).

This has to do with incentives. Here are two big ones I see:

1. Doctors are expensive. First year rookie doctors are much less expensive.

2. First year rookie doctors likely have lots of big debts to pay.

I believe just these two incentives explain much of the reason for this change and why it is so lax to begin with.

The crews on nuclear submarines work 12 hour shifts (two crews, each working 12 hours). Every day. And nuclear submarines rarely surface. 12 hour shifts, in confined space, without sunlight for months on end. And these submarines carry weapons that could destroy most of humanity. A fatigued doctor can harm only one patient at a time. A fatigued submarine crew can kill millions in an instant. Is this a good idea?

Bad idea for docs, patients, sub crews and humanity. @rayward How long until Armageddon?

Another thing: the crew (other than the captain and navigator) don't know where they are. Think about how disorienting that would be: submerged for months, working 12 hour shifts, in a confined space, and not having a point of reference as to where you are. Talking about vertigo!

@rayward: the US Navy has another story. It's good to know the people in charge of missiles is well-rested: "Subs operate on an 18-hour schedule – and here’s how a typical day breaks down: a) 6 hours spent on watch (actively operating assigned equipment), b) 12 hours spent off watch (this time is divided between eating, studying, training, qualifying and free time – which includes sleep time)" https://www.navy.com/navy-life/life-on-a-sub#sub-life-underway

Ps, A typical submarine deployment would be: 3–6 months for a smaller fast attack submarine (SSN), 3 months for a larger ballistic missile submarine (SSBN)......the medical resident just knows that after a year of medical residence follows another another year of residence.

Should I believe the two crewman I met from a nuclear submarine based at the nearby nuclear submarine base or the Navy recruiter? 12 hour shifts makes perfect sense, especially given that each crew member shares a bunk with another crew member on the opposite shift. How could they share a bunk when they have overlapping shifts?

Off watch doesn't mean sleeping. You can obviously hot bunk with overlapping watches.

How long ago did you meet them? 50 years ago? Were they 80 years old?

Radiation oncologists work with very small radioactive warheads whatcha they fire into cancer cells which are made up of millions of living organisms. Ever living cell has its own soul and so we have a moral duty to let radiation oncologists get the same shift structure as nuclear submarine technicians because we must protect souls.

Ages ago, when I was a Computer Science grad student, I did lots of all nighters programming. It's amazing how you think a lack of sleep is not impacting you that much, and then you clip your shoulder on a door frame walking through a completely normal open doorway - something you'd never do when you've gotten enough sleep. I can't believe this is even a close call.

It's easy to work alone. Try doing an all nighters while not cursing nurses or patients. You have the freedom to be a robot but other people does not have to suffer you bad mood. I learn this by the difficult way with my wife. I could work all I want but I have to leave a fraction of my daily energy/patience level to deal with things at home.......and that's dealing with people you love. Imagine having to be patient and interested in people you don't care about, that's the origin of burnout.

Yep....compassion is the first casualty. It is all too common when working abusive hours to come to see yourself as different than other people and deserving of money money money; driving a more mercenary attitude.

I was stationed on such a submarine, and I confess I'd sometimes doze off and almost hit the red "Nuke 'em high" button by accident. But if a crew member does hit the red "Nuke 'em high" button by mistake he gets a stern talking to, so there is incentive to stay awake.

It is good to know the person who will destroy my family and me by mistake will be sternly lectured and maybe even have to make a report. Anything less and he should have joined the Air Force.

Here is a surprising statistic: older patients do not make up the bulk of patients in the hospital. Instead, three age brackets (18-44, 45-64, and 65-84) constitute roughly equal shares (about 25% each) of hospital patients. And I thought we tolerated medical mistakes in hospitals because most of the patients were nearly dead anyway. Here's a question rarely asked: is the hospital the best place to train doctors? After all, most people who visit a doctor aren't sick or injured, but doctors seem to find something wrong with them anyway, most likely because they have been trained in hospitals where everyone is sick or injured.

Source: https://www.hcup-us.ahrq.gov/reports/statbriefs/sb180-Hospitalizations-United-States-2012.pdf

Thanks for the source. The stays per 1000 were not close though . (18-44, 45-64, and 65-84) (78,108,260).

I was surprised that the 18-44 cohort was that high, but makes sense: 22.2 percent were maternal or neonatal.

Also, Medicare paid for the largest number of hospitalizations (14.3 million stays), followed by private
insurance (11.2 million stays) and Medicaid (7.6 million stays). About 2 million hospital stays were for
patients without insurance. Private insurance is down to 32% of stays.

is the hospital the best place to train doctors?

It is one place we need training, but still too heavily emphasized in most programs.

Dr. Watson doen't sleep.

"the state of New Jersey has recently amended its vehicular-homicide statute to add to the definition of reckless driving “driving after having been without sleep for a period in excess of 24 consecutive hours,” a revision that explicitly subjects drivers in that state to a conviction of criminal homicide under such circumstances.24 Similar legislation is pending in New York, Massachusetts, and Michigan. Moreover, appeals courts in two states have ruled that an employer's responsibility for fatigue-related crashes can continue even after an employee has left work, similar in concept to the liability incurred by people who serve alcohol to drivers who are subsequently involved in alcohol-related motor vehicle crashes."(http://www.nejm.org/doi/full/10.1056/NEJMoa041401#t=article)

Summary of current legislative efforts:

This post is very un-libertarian of you Alex. If you don't like this I would suggest not looking into the schedules of airline pilots.

From the FAA:

Regulations limiting flight time and pilot rest have been in place since the 1940s. The rules for domestic flights do not explicitly address the amount of time a pilot can be on duty. Rather, the rules address flight time limitations and required rest periods. Current FAA regulations for domestic flights generally limit pilots to eight hours of flight time during a 24-hour period. This limit may be extended provided the pilot receives additional rest at the end of the flight. However, a pilot is not allowed to accept, nor is an airline allowed to assign, a flight if the pilot has not has at least eight continuous 
hours of rest during the 24-hour period. In other words, the pilot needs to be able to look back in any preceding 24-hour period and find that he/she has had an opportunity for at least eight hours of rest.

Thank you for googling that Liam. Out of curiosity are you a pilot?

It's freight pilots that we should be concerned about. They aren't paid much and work very long hours to make up for the low pay. And they share airspace and airports with the rest of us. Years ago I knew a pilot who told me about long flights carrying freight across Canada on extremely dark nights (I suppose nights are darker up there), the combination of fatigue and darkness often resulting in vertigo. He said he would have to fight the onset of vertigo while maintaining his focus on the instruments (IFR for you pilots). I sometimes travel, by car, at night so as not to interfere with the work day - I work and reside in different places. One night I reached my house at about 2-3 am and as I drove down my driveway, there they were, two monkeys walking across my driveway, my headlights shining on them. Monkeys! I stopped the car and watched them walk across the driveway and into the marsh. They didn't run, they walked.

"It’s freight pilots that we should be concerned about. They aren’t paid much and work very long hours to make up for the low pay"

Here's some data:

"Cargo pilots are allowed to fly up to 8 hours (as opposed to 9 hours for passenger carriers under their rules) then legally must have a rest period. In a situation where there are three crew members or more, cargo pilots may fly up to 12 hours. While, cargo pilots may be on duty for 16 hours, under no circumstance do they ever fly 16 hours without rest. "

This is the FAA, I don't have any idea what the rules are in Canada.

October 4, 1984: The day medicine changed forever. Forever seems to be getting shorter and shorter... some kind of relativistic effect, I guess - http://www.conciergemedicinemd.com/blog/2013/11/07/october-4-1984-libby-zion-the-day-medicine-changed-forever/

The comments (by doctors and one pharmacist) in the linked article give support for two claims:

1. The long hours for residents and interns allows the hospitals to save minor amounts, by letting doctors do hands on work that could be handled by nurses and lab technicians. I.e. the long hours are side effect of MBA thinking.

2. Six doctors were willing to lie on the witness stand to protect one of their own.

HA- plaintiff attorneys can find plenty of doctors to lie for the expert witness fee- i have heard 2 plaintiffs attorneys discuss this fact in an airport lounge- they stated the could get Dr X (whom I know personally) to lie to just about anything on the stand for them. I don't know what was more surprising- that Dr X was so sleazy, or the attorneys would discuss that in a public venue

This post really understates the impact of poor hand offs on medical error.

A lot of posts here focus on the anti-libertarian aspects of Alex's post. The real issue is that it's not restricting doctors in general, it's restricting doctors/hospitals to abuse a small, captive population of workers who will very quickly progress out of the situation of being abused. You have to do an internship to become a physician. Be a complainer, and the program can boot you. You have zero power. Then you have to do a residency to become board certified and make a decent living/get insurance/etc. Again, if you want to get the recommendations and connections to get a good practice situation, you play the game. And these are short-lived situations. A 24-karat carrot is always dangling in front of you, unless you're the caring type who wants to do peds or FM or psych.

Furthermore, working 90-110 hours a weeks, which I went through during internship, is so exhausting that you don't have the energy to organize and try to change the system. And at each step of the process, as you progress through residency, things get better. Going from insane hours as an intern, to only 70-80 hours as a first year resident, feels like a big improvement. Then the next year it gets a bit better, and unless you're doing a long surgical residency, the end is in sight.

There's one other factor about why hospitals love interns/residents that I don't think anyone here has mentioned. The hours are elastic, though the pay is locked. When the ER gets 'slammed,' the team on call may have their patient load swell. You may literally not sleep for 36 hours straight, and for the next week or two you might we working 100+ hours/week. But then there are times when few patients come in, and you have a fair number of transfers/discharges, and your patient load drops. You work 10-hour days, have time to eat actual meals in the hospital, and get 7-8 hours sleep (you lose the ability to sleep in, even if given the opportunity). If hospitals accepted 50% or 100% more interns/residents, then during peak times things would be much better, but during lull times the hospital would be paying some of them (and their attending physicians) for doing little or nothing.

In other words, "attendings" aka instructors, are incompetent and can't find things to teach "residents" aka students when not slammed with work that basically prevents any teaching, but ideally prevents disaster by the instructor stopping his student from killing or maiming a patient? But otherwise covers up the screw ups....

Why don't instructors go over the entire system the students will be working in as doctors? Spending time following through on the billing and payment collection for the care that was provided on instant reactive decision Making?

Conservatives argue patients should purchase medical care like they purchase things on Amazon or like they buy a car which might cost less than what they agree to based on a doctor advising them of options. When patients do ask about costs, they find most doctors have no clue what their very expensive recommendation will cost. Primary care docs generally have a good idea what their recommendations will cost and what insurance will and will not pay for. But not ER docs, nor specialists.

One might argue doctors should not know about costs because that would color their approach to a patient, but given bills must be paid, the solution is to create non-medical gatekeepers blocking access to doctors. The poor and uninsured can't see a doctor because they can't get an appointment.

First, attendings are not instructors. They're experienced doctors who have their own patient loads and responsibilities. They're more mentors and supervisors. Interns are mostly instructed by first or second year residents. First year residents are mostly instructed by second and third year residents, and so on. Contact with supervisors is usually limited, maybe half an hour or an hour a day, at best. They review cases, they rarely instruct.

Second, residents are not students. They're physicians who are 'resident' in that facility as they get specialized experience and training. The emphasis is overwhelmingly on experience, not training (put another way, most of the training is experiential).

And yes, it would be helpful if there was some instruction in billing and payment collection, but the insurance companies have worked hard to make this an ever changing and impossibly complex system, one that almost no physician handles on their own anymore. Every single insurer has their own system, all apparently designed to delay and minimize their pay outs to providers.

I agree that medical care is unlike any other service or product that we purchase. The nonsense that gets spouted about driving down costs with "more choice" etc. is smoke and mirrors. And most doctors have a pretty good idea what their treatments and recommendations cost. But doctors are not trained to be financial advisors, they are trained to treat illness and disease. Most patients seem to prefer that, for better or worse. Try telling a patient that you're recommending a treatment that is "mostly as good" as another treatment that costs 5x as much, and that statistically they're unlikely to notice the difference. Most patients react badly to that suggestion.

Finally, the creation of a massive population of non-medical gatekeepers and case managers is the primary reason I stopped practicing medicine. This is, frankly, a bigger issue than overworked, under-rested interns and residents.

Daylight Savings Time kicked in today. Studies report a 17% increase in car accidents on the Monday after DST. If a potential loss ( I stayed in bed about the same duration this morning) of one hour can cause a measurable rise in adverse events, we should expect that a shift length increase from 16 to 24 hours will certainly result in adverse events. I would like to see an economist use the tools of his profession to analyze who will benefit materially from this change, what the costs of the inevitable adverse events are, and how the winners will compensate the losers.

Physicians are no longer the only ones privy to the secrets of health and disease. They're technicians who order tests performed by other technicians. The results of those tests are subjected to computer analysis and interpreted by other technicians. Treatment is based on more computer analysis by technicians, ultimately approved by the physician. In other words, doctors are just one group of health industry technicians, not a fraternity of magicians. They are over-educated and overpaid.

Citing a non-preregistered study afflicted by all the p-hacking, multiple comparisons and garden-of-forking-paths self deceptions whose data was drawn from a "web-based survey" and published in the NEJM which had been advocating for shorter work hours for its guildsmen - Sad!

You should read up on the reproducibility crisis and then read through some of the studies citing this silly paper. Then you'll understand the problem. The meta analyses are all corrupted by the pollution of the literature with similar likely false positives all published at the same time while efforts to reproduce the findings by looking at say real patient outcomings (and not what they remember years later for an internet poll) refute your claim. E.g. "The implementation of duty-hour restrictions correlated with an increased risk of postoperative complications for patients undergoing brain tumor and cerebrovascular neurosurgical procedures. Duty-hour reform may therefore be associated with worse patient outcomes, contrary to its intended purpose. Due to the critical condition of many neurosurgical patients, this patient population is most sensitive and likely to be negatively affected by proposed future increased restrictions." - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4527330/

Alex, you've fallen for the tricks of the NHST racket (RIP PE Meehl). Time for you to be weened off the confirmation bias teat.

Your are arguing that it's worse to have no doctor around than to have a sleep deprived doctor making critical decisions because 50% bad decision quickly, corrected 50% of the time 20 minutes later by a non-sleep deprived doctor, is better than a decision 5 minutes later by a non-sleep deprived doctor with experience rushing in and taking 5 minutes to assess the situation?

After all, the budget for everything is always fixed and can never be increased. Ie, you can only buy a $50,000 luxury car made by factor workers working 24 hour shifts because limiting their shifts to 16 hours max would result in a bad hand off of cars on the assembly line and deflects caused by sleep deprived workers never being fixed?

The big issue is the private sector considers government funding to increase the spending on an activity to be a mandate to drastically cut private funding of that activity and turn the added government funding into a cap on total spending.

I'm old enough to remember the debate in the 60s about too few doctors to handle Medicare and thus the funding of more doctors by funding residencies through Medicare out of CMS.

I am not old enough to know if doctors/residents worked 24 to 36 hours back in the 50s and early 60s before Medicare. And I don't recall debates about doctors having shifty jobs because they worked days without sleep seven days a week while factory workers rarely worker more than 8-10 shifts only 5 days a week. In fact, in those days, the doctor was in the elite who got to go golfing multiple times per week.

But so very few of the decisions made in an ER doctor are serious. Sprained ankles, runny noses, mild foodborne illness, Mexicans screaming unintelligibly, too much to drink, an Arab woman with a mysterious black eye, someone freaking out over something about their genitals. These can't be seriously screwed up in most cases.

I'm sure if something real comes in the adrenaline pumps and they're in peak state.

the hours were VERY long in the old days- and in fact many residencies were UNPAID. In fact I remember reading an interview with the director of resident training at Vanderbilt stating he opposed paying residents b/c they got better residents when they didn't pay them. Residents got free housing (an apartment on hospital grounds) and free cafeteria privileges- which worked out b/c you had so little free time you basically lived in the hospital. But the work was less stressful on a per hour basis.

Do you want to work in a profession where 80 hours a week is normal, and where people basically expect you to drop anything you're doing to help people any time ... or you're the useless piece of crap doctor who didn't?

American doctors are overpaid. But that's because the US health care system is an overpriced uncompetitive and poorly organized system that enables it to happen. Not because the doctors want to have a bit of life for themselves in the process of providing the huge services to society that they do.

So .,. how did efforts to get more for less in educatoin work, by the strategy of throwing mud at and generally insulting those most useful of experts who you want to extract more from, for less?

If you want doctors to give more for less, how about propose a system that doesn't force them to turn away someone in need for the fact of not being insured. Many doctors require a significant premium, often to the tune of hundreds of thousands a year, to avoid the situation of being banned from providing care to someone who does not have money in pocket to pay.

Huh? I don't understand how doctors can be requiring a significant premium to avoid being banned from providing care for people who can't pay. I mean, couldn't any doctor who wanted to provide such care do so at a community clinic or something? Or just take a lot of medicaid patients whose re-embursement rates are pretty low? I am pretty sure doctors working in practices with mostly medicaid patients aren't paid *better*.

Do doctors working for Kaiser[1] get paid a lot better than doctors in ordinary private practice?

[1] Basically a private version of socialized medicine--the doctors, hospitals, radiology clinics, labs, etc., are all part of the same organization and the patients don't generally get bills other than their copays.

Say, you're working in some place caring for people in your community.

And someone ways "why don't you go work in the USA where you can take home 300k instead of 150k"?

To which they may respond "yeah, but there's a test to take. And anyways, I couldn't look myself in the mirror to work for a system that would turn away someone in need.

Do you know how mnay Americans go bankfupt every year due to a common type of accident? I wonder how that affects there forward progress and potential.

Great post Alex, but I disagree with your final paragraph, which points out that hospitals are incentivized to get more hours out of their semi-captive workers. Longer shifts and longer hours are related, but independent. You could still have residents work 80 hour weeks without subjecting them to 28-hour shifts.

Start suing for malpractice.

It is all the threats of malpractice that drive the working of residents 24 to 36 hour shifts because they need to do unnecessarily tests which take lots of extra time to prevent being sued for malpractice. ;-)

Tradition can be a way of capturing, storing, and transmitting collective wisdom. Sometimes it is just plain stupid. Long shifts for interns and resident pbysicians-in-training are just plain stupid. The medical industry buries it's mistakes.

Doctors are high status, truckers are low status. This is why the government regulates the sleep of one and not the other.

See also: ADD medication versus smoking.

Everyone knew this was a serious problem even 30 years ago. Like the nonsensical academic model that leads to absurd tuition rates, residency is another example of how institutional inertia dominates our society.

Aside from Hanson-type factors, a key factor that explains what is going on is that residents are a huge profit source for the hospitals. Much like student athletes, residents are underpaid. As a result, hospitals want to use residents as much as possible so they lobby for longer hours even at the expense of patient safety.

This factor is underrated, and I write a lot about it in "Why you should become a nurse or physicians assistant instead of a doctor: the underrated perils of medical school." The essay is based on living with someone who went through the whole, and wholly mad, residency process.

Repetitions matter. As a surgeon, the process sucked and I never would want to do it again. But I am glad it was hard, because when you are alone at night doing an emergency case, I know I can do it because I don't have think about it. I just do it. You only get that with repetitions. There's a saying, the key to medicine/surgery is good judgement. You get good judgement from experience. You get experience from the accumulated knowledge of bad experiences from yourself and others. Pay attention to every case and every nugget of wisdom someone shares.

True statement: worst kind of Doctor is part time. If true then the converse... best kind of doctor works full time. Who decides the sweet spot? Personally as a physician when you work long hours you get in a groove. As a fellow MD once said "bad going in, bad coming out, not bad while your there"

There's a lot in medical education that is flatly very stupid. Ridiculous shifts for residents is part of it, but I don't think the effects are overly serious.

The AAMC holding the number of seats far below the number of qualified applicants is another story. The AOA responded to market demand and was growing tremendously in influence, but the AAMC has pretty well put an end to that with the residency merger. This is very serious and bluntly is rent-seeking behavior.

The premed ciriculum is another archaic and largely irrelevant reason to restrict market supply. I very seriously doubt doctors regularly apply information they learned in chemistry, organic, or physics, yet the AAMC is expanding the premed ciriculum substantially by adding biochemistry and psychology to the MCAT.

Finally diversity preferences have grown to absolutely gigantic proportions; there are large number of sub 3.0 black matriculants each year in midst of the AAMC bragging about declining white matriculants.

I thought each place independently decides on its pre-med content.

For example, aren't some pre-med programs tilting away from loads of pure science and promoting more diversified knowledge and thinking skills?

Of course most countries have very similar systems, so blame for this system can't be placed on just bad US regulation or poor incentives for hospital administrators in the US. I think in reality people have found this is the best way to train doctors, very intensive training at an early stage. Chalk and talk just isn't an effective way to train people, there needs to be lots and lots of hands on experience. Of course patients suffer in the short term but the overall system benefits. I don't know if the trade offs are worthwhile, but given the rest of the world sticks with this system it suggest it does.

Been there and done that.

Any physician trained in a system where you work 36 hours straight knows what happens under the blinding fluorescent lights all night long. They also know that these shifts are complexity unnecessary in the real world. All teaching hospitals should be staffed by hospitalist and intensivist staff physicians in addition to residents. With that kind of coverage, there is no need for a resident to work a 36 hour shift or even a 24 hour shift.

A related issue is the 7 days on and 7 days off schedule of staff physicians. I think that even with 10-12 hour shifts a day - you are not nearly as cognitively efficient on day 6 or 7.

I doubt that these hours have anything to do with training. I would not be shocked to find out that there are some institutions out there who still need residents for the work force.

Lot of opinions on how doctors should divvy up their hours. Not much perspective on what it's like to take care of patients.

Here's my take.

First, I did q4 call as an intern. The next year the new rule went into effect. So I know firsthand it is a lot easier to do one 30 hour call every 4 days but have a light pre-call day and every other weekend off, than it is to do 6 straight days of 12 hour shifts (which can easily become 13-14 hour days when a patient decompensates right at the end of the day and requires the care of the doctor who knows him best). I also know it's easier to see friends and family outside of the hospital with the q4 schedule. Believe it or not, having a life outside the hospital is important for mental health, and this reflects on patient care. So the grind of 6 days in a row with no golden weekends was actually tougher than q4 call. Especially when those 6 days in a row are from the hours of 7pm to 7am! When the rule change went into effect, everyone missed the post call and pre- call days, and people felt a lot more burnout with the new schedule.

Second,you guys don't understand the role of an intern. Interns are the least experienced doctors in the hospital. Their job is primarily to handle small issues and document a patient's history. Major medical decisions in the middle of the night are made by the more senior supervising resident or fellow, who have 1-6 more years of experience, and who don't have to deal with the nonsense of charting or writing tylenol orders all throughout the night. So these more experienced cots can actually sleep when things are quiet. Furthermore, in a true emergency, it's the attending physician who makes the final call -- and that doctor is at home sleeping but available by page. Improving an intern's sleep probably cuts down on the orders incorrectly entered into a computer EMR, or on typos in an H and P note, but any decent hospital pharmacist or nurse can minimize the damage caused there. Focusing on intern fatigue and not resident or attending fatigue is not going to alter clinical outcomes much. However, minimizing intern fatigue at the expense of resident or attending fatigue is definitely not going to help patients. THAT'S WHAT THIS RULE DID.

Third, handoffs truly are difficult. No one likes them. Patients are complex. A brief signout on why a patient is admitted and what to look out for never substitutes for continuous care. Furthermore, you can't handoff personal relationships with a patient. An overnight doctor not only needs to learn a patient's medical issues, but also needs to establish a rapport at some godforsaken hour. For patients, they have to meet an entirely new doctor and may not feel comfortable putting their trust in this new physician, especially since this doctor didn't spend an hour in the room for morning rounds with a team of other doctors and the attending physician. For the doctor, losing the personal connection with a patient also causes a loss of ownership. The line "I'm just covering, overnight" starts popping up when nurses page for help. When families arrive after hours, these doctors have a tough time defending or explaining plans crafted during the days, because they simply weren't privy to the conversations or lines of thinking discussed during the day. And from a simple human level, Having an established relationship with a patient is important and any rule that diminishes it is bad for patient care!

Last, both doctors and hospitals preferred the q4 system. I've tried to outline my own feelings for why I preferred that system, but it wasn't because I wanted to be exploited by the hospital. This regulation only made things worse. I knew the outcomes for these recent studies would be negative because it was obvious that patient care was not being improved with the new rules. The only thing that changed was increased resident fatigue, increased burn out with a 6 day 12 hour schedule, and increased number of medical decisions being made by a covering "night team".

So I agree with the commenter above. An economist arguing for more regulation for a field he understands little should look closely in the mirror.

I agree with Cardiologist, who commented above. I believe that the new 6 day 12 hour schedules creates much more fatigue than the previous system. It only makes sense that this would be the case because work half the day every day, with only one day per week to rest does not leave much room for the residencies to recover mentally, let alone physically which would in turn hinder their performance in the work place. Though these interns or residencies are probably overworked because they get paid less, it allows the more experienced doctors to rest so that when there is a real emergency that are able to operate to the best of their abilities.
Furthermore, even though the hours on interns, doctors, residencies, and nurses are atrocious they are necessary. For instance some surgeons have to perform surgeries that can last over 24 hours straight. So overworking the interns and residencies is harsh and clearly fatigues them, it is necessary. This way when they become the more experienced doctors in the hospital and become the "shot callers" they are ready to handle each emergency and would, by then, have become used to excessive work hours. Personally, I'd much rather have a doctor, or surgeon, operate or take care of me for 36 hours straight if I knew they had pulled hours like that in the past. It would provide me with a sense of ease knowing that my care taker had done this before, and would allow me to think that his work performance would not worsen because of his lack of sleep.

Inequality is everywhere. From who goes first at restaurants to who gets paid a bigger salary in job fields. But this article points out another places that inequality comes into place, the medical field and other professions. It is crazy to think that someone just beginning their medical residency can work 24 hour shifts and 80-hour work weeks. When pulling an all-nighter to get homework done, it is all I can do the next morning to function, let alone make crucial decisions about someone’s medical standards.

Society loves “free” labor, and that is what is happening when first year residency students work long and frequent shifts. These people are still in school so they are not getting paid for any of these shifts. When they make mistakes because of fatigue or any other reason hospitals deem it as a mistake because they are still learning. But what are they still learning, how to make better decisions? Or how to make them even though they are working their third 24 hour shift that week?
When we put more restrictions of the amount of hours one can drive instead of how many hours a medical professional can work, then we are increasing our chance of error every time we go to the doctors as well as every time one goes into surgery. As a society we need to worry less about free labor and more about the health and well-being of ourselves and our medical professionals.

I've been reading MR since before I went to med school, and am now a practicing hospitalist. I'll try to make an analogy for the economists:

Sometime in the future, we realize that college students (patients) learn best (feel better), if you let them pick the times of their courses, even if they decide on short notice (randomly show up at the ER). So you, as an econ prof, end up teaching at all hours. In fact, about 40% of your classes (patient encounters) occur between 5pm and 7am. This is hell for your personal life, but you love the long-term relationships you're building with the students in your favorite class, How To Be Dismal 101.

Unfortunately, these hours make you tired, and this leads to objective mistakes. The administration decides to fix this. You suggest that they give you more help - maybe you could have a teaching assistant (another healthcare provider) to help catch these mistakes and allow you to continue teaching, but for various logistic (money) reasons, this isn't possible. You agree to the next best solution: you and your colleagues instead are each assigned a time slot rather than a class. This makes your life predictable, but on the other hand, you teach different students and different classes each day. You've lost the ability to develop the types of relationships that promote meaningful learning in your students (improved health). You can't prove this, since no one knows how to accurately measure it. And since fewer of the easily measured mistakes are occurring, the issue seems closed. So now you sleep more, but you have to prepare more lesson plans (patient notes) for the various classes, and you have to hand off the day-to-day details of your classes to your colleagues, since you never know who will next teach your classes.

It's easy to imagine an ideal way to do a handoff - you simply through each student and his/her learning needs with your colleague. But this takes hours, and your shifts would have to overlap for this to work. There isn't time for this, so you have to choose between teaching the students who are in front of you, and spending time handing off their teaching points.

Doctors know that better handoffs mean less mistakes. But when the changes that you mention made contact with the realities of our ridiculous system, they lead to more work with less time to do it in.

Comments for this post are closed