Editore"s Note
Tilting at Windmills

Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon Sign up for Free News & Updates

September 5, 2007
By: Shannon Brownlee

HOSPITAL ERROR....Two large studies, published today in the Journal of the American Medical Association, found that cutting the grueling work hours of doctors-in-training had little effect on reducing hospital errors and patient deaths. Surprised? So were the researchers who did the studies.

There are three possible explanations. One, most errors aren't caused by groggy, sleep-deprived, over-worked residents, so giving them more time off won't make any difference in the error rate. Two, the new regulations, which cut residents' typical workweek from 100 hours to 80, didn't reduce their hours enough to make a difference. I mean c'mon, 80 hours a week still doesn't leave much time for eating and sleeping and all those romantic couplings we see on television shows like "ER" and "Gray's Anatomy." Or, three, the number of mistakes made in hospitals is so large, any drop in the errors committed by residents was too small to be measured.

My vote goes to . . . well, let me just offer a couple of statistics. In its seminal 1999 report on the subject, To Err is Human, the Institute of Medicine estimated that as many as 98,000 American patients are killed each year by medical error. Hospitals are such complicated places, the ways that care givers can screw up are almost too numerous to count. A doctor can accidentally perforate a patient's colon during a colonoscopy, leading to infection. Surgeons leave devices or sponges inside wounds and stitch patients up. One intensive care unit that tracked near misses reported 1.7 errors per day per patient, about 30 percent of which could have been serious or fatal.

I know a doctor who was asked by another physician in the ER to check the blood pressure on both arms of a patient wracked with chest pains. When he couldn't get a proper reading from one arm, because the man was writhing in pain, the other doctor assumed the patient's blood pressure was the same in both arms. That meant the patient might be in the throes of a heart attack. In fact, he was suffering from a dissecting aortic aneurysm -- a rupturing abdominal blood vessel -- not a heart attack. Had his blood pressure been obtained from both arms and found to be different, the doctors might have properly diagnosed the dissecting aneurysm in time to prevent the patient from bleeding to death.

Maybe letting residents get some rest will eventually bring down the error rate, but I doubt it. Here's why. Rested residents aren't nearly as important to reducing errors as coordinating the care that everybody in a hospital delivers.

The two studies published today suggest why that might be the case. The studies included 318,000 veterans who were cared for at Veterans Administration Hospitals and another 8.5 million Medicare recipients. It turns out, error rates did go down at VA hospitals, but not at the other hospitals in the study.

When it comes to reducing medical error, the VA health system has three things going for it that most other hospitals don't have. Numero uno, every VA hospital has a fully-functioning electronic medical records system.

This system not only helps physicians and nurses avoid many kinds of errors, like giving a patient the wrong drug, it also allows each hospital to track the treatment of every patient. Hospital safety officers can easily give physicians and nurses feedback on how they're doing when they implement any sort of error-reduction program. Other hospitals are left flailing along, hoping that they are making a dent when they initiate some new plan. VA doctors know -- and in real time.

And finally, VA hospitals do a better job of coordinating all the different people who have a hand in a patient's care. Veterans don't tend to fall through the cracks during hand-offs between one shift and the next, for instance. That's why my colleague Phil Longman's book about the Veterans Health Administration is titled Best Care Anywhere. It's also why a Democratic presidential candidate will probably be pointing to the VA in a speech later this month as one model for improving American health care.

You can find the articles at the JAMA site. They're at the top of the list.

NOTE: Longman's book is based on an article he originally wrote for the Monthly back in 2005. You can read it here.

Shannon Brownlee 1:34 PM Permalink | Trackbacks | Comments (62)

Bookmark and Share
 
Comments

Shannon Brownlee: And finally, VA hospitals do a better job of coordinating all the different people who have a hand in a patient's care.

The proper buzzword is Operations Research. It got its real start in WWII, and has made enormous differences in flight safety, factories, you name it. Hey, it's only been 62 years since the end of the war, maybe medicine needs to give it a shot.

Posted by: alex on September 5, 2007 at 2:04 PM | PERMALINK

I'm guessing that was the descending aorta.

Posted by: Michael7843853 G-O/F in 08! on September 5, 2007 at 2:17 PM | PERMALINK

From your own comment you could add:

4. In combination with the factors above, any decrease in the number of errors committed by residents was offset by the effects of reduced staffing on patient care. A resident is low cost labor and while a resident beween 80 and 100 hours may be prone to mistakes, they still have some capacity to potentially prevent the mistakes of others. If you lower hours without increasing staffing the system has to get by with fewer staff hours.

Posted by: Catch22 on September 5, 2007 at 2:20 PM | PERMALINK

Shannon Brownlee: Two, the new regulations, which cut residents' typical workweek from 100 hours to 80, didn't reduce their hours enough to make a difference.

Or the new regulations only cut on-the-clock hours, not actual hours. Here in NYS residents hours were limited to 80/wk a few years ago, but it was widely reported that residents would sign out and keep working. Best not to make waves when you're in residence.

Hey, in my day, we had to work at least 169 hrs/wk. Kids today have it easy.

Posted by: alex on September 5, 2007 at 2:24 PM | PERMALINK

There must certainly be far more possible explanations than just the 3 offered by Kev.

How about potential reason #19: the studies were flawed? They both used patients' input as their primary data source.

Posted by: wishIwuz2 on September 5, 2007 at 2:30 PM | PERMALINK

Setting upper and lower statistical control limits and then measuring new care against those limits and then resetting the control limits to tighter standards requires a dedicated quality control management system. Doctors do not want to be under the control of such a quality regime.

Posted by: Brojo on September 5, 2007 at 2:40 PM | PERMALINK

Link to the JAMA studies please?

Posted by: Chocolate Thunder on September 5, 2007 at 2:49 PM | PERMALINK

VA medical centers are not the same place as a teaching hospital. VA medical centers to do have big Emergency Departments, do not by OB, do not have pediatrics or neo-natology, do not perform many times of surgeries, do not have sports medicine, and barely have any GYN.

VA Medical Centers look better because they spend a larger portion of their budgets on mental health versus the large teaching hospitals.

Posted by: superdestroyer on September 5, 2007 at 2:50 PM | PERMALINK

No superdestroyer. VA hospitals outperform their private-sector counterparts on measures of clinical quality that can be compared across institutions and over time (such as proper administration of antibiotics for surgical patients). I suggest as recommended reading Longman's original article, Best Care Anywhere.

Posted by: Chocolate Thunder on September 5, 2007 at 2:54 PM | PERMALINK

Rested residents aren't nearly as important to reducing errors as coordinating the care that everybody in a hospital delivers.

A hair raising experience a couple of years ago--while the resident was explaining to me that my partner was suffering from an allegic reaction to a drug they had given him, an attendent came in and hooked up another IV bag--plainly labeled with the name of the drug in question. Good thing I pointed it out--the resident hadn't noticed, and nobody had bothered to change the attendant's orders. Not too many minutes later, my partner was supposed to have a MRI. The attendants there were very rough with him physically, while making remarks about drunks and druggies--apparently unaware that the shakes and hallucinations my partner was undergoing were the product of an adverse reaction to a drug administered by the hospital, not meth or something. Somewhere in there, I had to be restrained . . .

Posted by: rea on September 5, 2007 at 3:00 PM | PERMALINK

I won't rock the boat today. I'm amazed at how poorly the medical profession takes advantage of technology and how little doctors research their patients.

A year ago, my wife was having serious medical problems, so we went to several specialists. Every specialist had us fax records and send films before our initial visit, and then did not look at them. My wife had to tell her story from scratch to each one of them, and they would then spend a few minutes looking at films to make their decisions. If my wife left off one detail from her history during the nth time of telling it, it altered the probable cause of the disease. Each one told us that her case was very complicated, but only one ever said that he was going to have to do some research and consultation. I found it surprising that specialists did not pay any attention to referring doctors and rarely shared information.

My son recently had minor surgery. The assistant nurse took his temperature incorrectly and wrote down 95 F in the chart. She then took his blood pressure incorrectly and did not get any reading. He was then carted off to surgery. Fortunately, there were no complications, so the baseline measurements never had to be consulted. It drove home, however, how little hospitals do to make sure that procedures are done properly.

Most hospitals, even though they use independent doctors, have protocols to deal with various situations. For some reason, it seems like protocols that would lead to more consistent positive outcomes don't get put into place enough.

Posted by: reino on September 5, 2007 at 3:00 PM | PERMALINK

Ah, technology. A friend of mine was having knee surgery on his left knee (having already had a similar procedure on his right one--which should be a lesson to all you joggers to find something better to do to stay fit). While he was waiting on a gurney in some featureless room an orderly came in and asked "Which knee?" Told it was the left one, he took out a magic marker and wrote "NO" in big letters on his right sole.

Three hours later, when my friend comes out of surgery, he sees that someone had shaved his right leg--the one that had already been operated on--until they got to the scar from the previous procedure, which apparently alerted them to the fact that they were prepping the wrong leg. Word to the wise--don't rely on the staff to read what's on your foot, write "NO' yourself on the knee, foot or whatever you want to save.

Posted by: Henry on September 5, 2007 at 3:26 PM | PERMALINK

As a surgeon, I think your comments on these articles were right on target. I lived through the old days, and I dont want to do it again, but a part of the old days (and the fact that led to 100 hour weeks) was the fact that we stayed and cared for our patients during the most critical parts of their hospitalizations. Handoffs during critical periods are extremely high risk, and the patient safety initiatives in these areas only came about AFTER the implementation of the work hour changes. Thus we removed a safety net without replacing it.

Posted by: Jeff on September 5, 2007 at 3:34 PM | PERMALINK

Numero uno, every VA hospital has a fully-functioning electronic medical records system.

And how about establishing a national diagnostic expert system which not only serves as a diagnostic aid, but keeps track of every new constellation of symptoms and diagnosis?

Good gawd, we in the US may have the best med tech in the world, but med mgmt is locked in the blood letting and leaches stage.

Posted by: Disputo on September 5, 2007 at 3:36 PM | PERMALINK

It's also why a Democratic presidential candidate will probably be pointing to the VA in a speech later this month as one model for improving American health care.

I'm inferring you mean a speech by a 'top tier' Dem candidate? But none of them support a VA-type plan -- a single-payer plan.

Of course, one of them might "point to the VA" but that hardly means they'd implement -- or even try to implement -- such a plan once elected.

Only Kucinich and Gravel (and Gore!) are for a truly single-payer plan.

Posted by: Jake on September 5, 2007 at 3:40 PM | PERMALINK

Link to the JAMA studies please?

Sorry about that.

http://jama.ama-assn.org/cgi/content/full/298/9/975

http://jama.ama-assn.org/cgi/content/short/298/9/984

Posted by: Shannon Brownlee on September 5, 2007 at 3:40 PM | PERMALINK

I don't let my undergraduates hand in handwritten papers. Why are illegible medical charts still OK? Is it because we want medical personnel to spend all day trying to read them, or because we don't want them to know the patient's history? Full computerization of medical records has to happen yesterday.

Posted by: calling all toasters on September 5, 2007 at 3:41 PM | PERMALINK

Word to the wise--don't rely on the staff to read what's on your foot, write "NO' yourself on the knee, foot or whatever you want to save.

Similar story. Many moons ago I went in for knee surgery. Got wheeled into the operating room, everyone there but the surgeon. While the anesthesiologist was starting to put me under, an argument arose as to which knee they should prep for operation. The last thing I remember before slipping into unconsciousness was that they finally decided to wait for the surgeon before picking a knee to prep.

Waking up in post op to find that they had fixed the correct knee was quite a relief.

I'd rather take a knife and needle and thread to my own knee then go through that again.

Posted by: Disputo on September 5, 2007 at 3:44 PM | PERMALINK

VA medical centers are not the same place as a teaching hospital.

VA medical centers are the single largest provider of residency training in the U.S. It's true, they don't have very big OB/GYN departments, and not a lot of neonatal intensive care, but they care for a population of patients that on average has worse health habits, higher rates of smoking, drinking, and drug use, and lower levels of education -- and higher rates of mental illness. They have a big job, and they do it well.

Posted by: Shannon Brownlee on September 5, 2007 at 3:45 PM | PERMALINK

Unless you're expecting that the shorter hours would yield no increase in co-ordination-of-care errors, the correct conclusion to draw from no significant change in overall error rate is that fatigue errors did get reduced significantly. (Even if 80 hours isn't a big improvement.)

Moreover, moving errors from the fatigue column to the coordination column is significant, because we know from other professions how to avoid coordination errors. We don't know (other than not having people work fatigued) how to avoid coordination errors.

(Incidentally, reducing resident hours may also have a long-term good effect on patient care. Many of the friends I knew who became doctors reported that negative attitudes toward patients and colleagues became ingrained during the hell that was residency, much as the training for rookie police officers is reported to inculcate the clannish attitudes that make community policing so difficult.)

Posted by: paul on September 5, 2007 at 3:45 PM | PERMALINK

Jeff: Thus we removed a safety net without replacing it.

Are you saying that the patient safety initiatives in these areas are not an effective safety net? Otherwise they constitute the new patient safety net.

the fact that led to 100 hour weeks) was the fact that we stayed and cared for our patients during the most critical parts of their hospitalizations

80 hrs/wk is not enough time to watch a patient through critical periods? How long do these critical periods last? How long can one person stand vigil over a patient? And what about non-teaching hospitals, where there are no residents to work 100 (or even 80) hrs/wk?

Lastly, but most importantly, do you have any statistical data to verify that the residents working 100 hrs/wk was a more effective safety net than newer protocols.

Posted by: alex on September 5, 2007 at 3:50 PM | PERMALINK

Incidentally, reducing resident hours may also have a long-term good effect on patient care. Many of the friends I knew who became doctors reported that negative attitudes toward patients and colleagues became ingrained during the hell that was residency, much as the training for rookie police officers is reported to inculcate the clannish attitudes that make community policing so difficult.

Interesting point. Most of the doctors I've talked to have great attitudes about patients, but they are discouraged by many aspects of their work, especially those in primary care.

Posted by: Shannon Brownlee on September 5, 2007 at 3:59 PM | PERMALINK

Shannon,
Would you please rewrite the 4th paragraph of your post to remove the ambiguity? I can't tell whether you are implying that the failure to obtain both-sides BP contributed to the patient's death. If so, please explain how. Also, please explain what condition(s) might cause BP to differ from one side to the other, and how.
Thanks.

Posted by: Claire on September 5, 2007 at 4:08 PM | PERMALINK

Shannon Brownlee: Two large studies, published today in the Journal of the American Medical Association, found that cutting the grueling work hours of doctors-in-training had little effect on reducing hospital errors and patient deaths.

Not quite. The conclusion of the 2nd paper states: "The ACGME duty hour reform was associated with significant relative improvement in mortality for patients with 4 common medical conditions in more teaching-intensive VA hospitals in postreform year 2. No associations were identified for surgical patients."

So reducing resident hours had a beneficial effect in many cases, particularly noticeable in those hospitals which most rely on residents. Not a panacea, but certainly an improvement.

Posted by: alex on September 5, 2007 at 4:09 PM | PERMALINK

My brother does a lot of work for the government of Estonia. His 11 yr. old daughter went with him on a business trip and got hit in the eye with some gravel from a passing car. As my brother had an Estonian government "smart card" with a microchip embedded with all of his and his families medical records, allergy and medical data, etc., they swiped it once when they went into the government-run clinic. My niece was treated, her eye cleansed and bandaged and they were out within an hour and a half. My brother paid nothing and didn't fill out even one form.

Someday, the United States' health care system may catch up with Estonia. If we elect a Democratic president....

Posted by: The Conservative Deflator on September 5, 2007 at 4:13 PM | PERMALINK

I worked in the medical device industry for many years, and came to appreciate the philosophy, if not always the application, of the FDA’s good design practices (GDP) and good manufacturing practices. Similar to ISO9000, but with much more serious teeth, it is a way of operating that strives to prevent mistakes. In comparison, doctors and hospitals are largely unregulated, and to be blunt, occasionally slap dash. An interesting example was issue of “off label” use of medical devices. As a manufacturer we could only support use of our device for procedures they were approved for, but a physician could take the device and use it for what ever he or she wanted to (the situation is similar for drugs). The only catch was, when they called us up for advice we couldn’t help, as this would be encouraging off label use.

Clinical medicine will never be as controlled as manufacturing, but there is still much modern quality policy that medicine could adapt from industry. I always feel a bit irritated when after an unfortunate death due to medical error some representative comes on the radio and says you can’t prevent human error, when in fact there are all kinds of things you can do to minimize it. The tragic case a few years ago when a transplant surgeon killed a young Guatemalan girl by giving her a incompatible heart was shocking because it showed the level of error possible even in high profile cases.

Posted by: fafner1 on September 5, 2007 at 4:18 PM | PERMALINK

Someday, the United States' health care system may catch up with Estonia. If we elect a Democratic president....

Finally someone admits that the Dems want to replace the greatest healthcare system into the world with communism!

Posted by: wingnut on September 5, 2007 at 4:18 PM | PERMALINK

Off thread, but, while kudos should be given to the VA for their hospital and out-patient care, and well mentioned in this thread, very bad marks should be given to the VA system for their handling of claims by veterans.

In this morning's Oregonian, there is a front page article about the extremely long process of approving claims. In their lead, the reporters write about a Marine from Nam, who is in a race with the grim reaper and the VA approving his nursing care billings. The chalk is with the grim reaper.

In the hospitals, one receives excellent care - In the claims department, there is a myriad of paper work clogging the wheels. DoD loves to pass off claims to the VA system; then it is "Take a number and bring your War and Peace" time.

Posted by: thethirdPaul on September 5, 2007 at 4:19 PM | PERMALINK

Did anyone consider overworked nurses as a possible issue?

Some of you may be too focused just on physicians.

A symptom of not understanding health care?

Posted by: save_the_rustbelt on September 5, 2007 at 4:23 PM | PERMALINK

**

Posted by: mhr on September 5, 2007 at 4:26 PM | PERMALINK

save_the_rustbelt: Some of you may be too focused just on physicians.

Or some of us may have understood that the topic of the thread was the value of reducing residents workloads vs. an operations research approach to health care.

Posted by: alex on September 5, 2007 at 4:53 PM | PERMALINK

though i agree with most of the points made,
my guess is that no matter where you are sitting,
if your aorta is ripping to shreds,you are
pretty much sol.

Posted by: trp on September 5, 2007 at 4:56 PM | PERMALINK

The proper buzzword is Operations Research. It got its real start in WWII, and has made enormous differences in flight safety, factories, you name it. Hey, it's only been 62 years since the end of the war, maybe medicine needs to give it a shot.
Posted by: alex

Ops Research (the study of man-made systems, analog to Physics for natural systems) isn't really the right pick. A better analog is Crew Resource Managment from the aviation industry.

Posted by: SJRSM on September 5, 2007 at 5:12 PM | PERMALINK

"That meant the patient might be in the throes of a heart attack. In fact, he was suffering from a dissecting aortic aneurysm -- a rupturing abdominal blood vessel -- not a heart attack. He bled to death internally before anybody figured it out."

This suggests unfamiliarity with anatomy and the condition being used as an example. The left subclavian artery comes off the descending aorta and is the least satisfactory site for blood pressure to be taken, especially in a case with chest pain. The dissection usually involves that segment of aorta and the subclavian is often reading a pressure far below the real blood pressure because its origin is being compressed by the dissection.. Since disection is usually a consequence of severe high blood pressure, the error is critical.

Electronic medical records are major factors in improving quality as are decision support modules built into the EMR and which can remind everyone of dose levels and allergies. They can also get quite sophisticated in ICU settings. Part of the problem is cost and a lot of it is hospital administrators who do not understand the operations in health care. They are financial types and lawyers.

I also agree with Jeff above about the hand-off problem, especially with surgery residents. As far as private hospitals are concerned, surgeons who do critical care are going to work a lot more than 80 hour weeks. I've worked 40 hour stretches without sleep in private practice. One factor pushing the limitation of hours is students who know the financial rewards are far less than they once were and they will not work those hours without compensation. That is behind a lot of the trends in postgraduate medicine where "lifestyle" specialties are popular and surgery residencies are not filling.

Posted by: Mike K on September 5, 2007 at 5:13 PM | PERMALINK

Claire: I can't tell whether you are implying that the failure to obtain both-sides BP contributed to the patient's death. If so, please explain how.

Hope it's clearer now.


Posted by: Shannon Brownlee on September 5, 2007 at 5:38 PM | PERMALINK

Word to the wise--don't rely on the staff to read what's on your foot, write "NO' yourself on the knee, foot or whatever you want to save.

When I had my ACL replaced last year, they handed me the marker and had me draw an arrow to the correct knee and then initial it.

I also drew one of those universal "no" symbols on the other kneecap. Just in case.

Posted by: Mnemosyne on September 5, 2007 at 5:46 PM | PERMALINK

When I was a kid, one of my great-aunts died because the hospital cut out the wrong kidney.

Posted by: Vlad on September 5, 2007 at 5:46 PM | PERMALINK

"Word to the wise--don't rely on the staff to read what's on your foot, write "NO' yourself on the knee, foot or whatever you want to save."

I went in for ankle surgery about a year ago (torn tendon), and before I left for the hospital I printed "Cut this leg" on the leg that was supposed to be cut and "Wrong leg" on the one that wasn't.

When I got to the hospital, they made me wash the writing off the leg that wasn't supposed to be cut, so that I didn't confuse the doctor. Apparently, they don't actually read the messages (at least, not at that hospital).

Posted by: Vlad on September 5, 2007 at 5:49 PM | PERMALINK

Apparently, they don't actually read the messages (at least, not at that hospital).

Your surgeon was probably one of those illegal immigrants who keep stealing jobs from Americans and don't even read English....

Posted by: Disputo on September 5, 2007 at 5:52 PM | PERMALINK

Mike K: I also agree with Jeff above about the hand-off problem, especially with surgery residents.

See my 3:50. Statistics speak louder than anecdotes, which is a key point of this thread.

As far as private hospitals are concerned, surgeons who do critical care are going to work a lot more than 80 hour weeks.

Non-residents, on a regular basis? Show some hard stats or just gimme a break.

I've worked 40 hour stretches without sleep in private practice.

I'm not in any kind of practice, but I've occasionally done the same thing. How often? Why? If done on a regular basis it is heroic but foolish. How effective were you after 40 hours without sleep?

One factor pushing the limitation of hours is students who know the financial rewards are far less than they once were and they will not work those hours without compensation. That is behind a lot of the trends in postgraduate medicine where "lifestyle" specialties are popular and surgery residencies are not filling.

from http://www.allied-physicians.com/salary_surveys/physician-salaries.htm

U.S. Physician Salaries - Ongoing Salary Survey
Survey includes base salaries, net income or hospital guarantees minus expenses
June, 2003 - Present

SPECIALTY Years 1-2 >3 Max
Surgery - General $226,000 $291,000 $520,000
Surgery - Cardiovascular $336,000 $515,000 $811,000
Surgery - Neurological $354,000 $541,000 $936,000
Surgery - Plastic $237,000 $412,000 $820,000
Surgery - Vascular $270,000 $329,000 $525,000

If you know of any more authoritative sources, please list them.

Posted by: alex on September 5, 2007 at 6:35 PM | PERMALINK

I've worked 40 hour stretches without sleep in private practice.

I stop being an effective lawyer after about 20 straight hours of brief-writing, as I know from experience. Truck drivers have to stop after about 10 hours, if I recall correctly. Evidently, doctoring isn't as demanding as lawyering or truck driving . . .

Posted by: rea on September 5, 2007 at 7:25 PM | PERMALINK

One of the built-in safety features that wasn't mentioned is that residents don't really DO all that much--they write orders for nurses to do things. And if the orders are wrong, nurses get them corrected.

Posted by: Shamhat on September 5, 2007 at 8:47 PM | PERMALINK

Shambat, you are dead wrong. Pun intended. Residents do most of the heavy lifting in most teaching hospitals.

Posted by: wwc on September 5, 2007 at 9:20 PM | PERMALINK

Plenty of snarks, I see. The 40 hour stretches were not common but 24 is. It's what happens when you run a trauma center. My point is not to encourage it but to point out that hours limitations in residency may not prepare one for the responsibilities of practice. As far as the salary figures, a friend of mine who was chief of surgery at a university program near Kevin told me he could hire assistent professors of surgery with followship training in subspecialties for $60k. I don't do surveys but I do talk to medical students every week for the past 10 years. I know where they plan to go and why.

Lots of people who read surveys but not too many with years of experience.

Posted by: Mike K on September 5, 2007 at 10:39 PM | PERMALINK

The aortic dissection anecdote is "amusing". In 12 years of practice as a cardiologist, I've seen about 10 cases, and only a few presented as the classic "tearing sensation in the chest." Each time I've come across a dissection, I nearly shit in my pants, since the quoted mortality rate is 1% per hour! Ironically, since every chest pain that comes through the ER seems to get a (usually unnecessary) CT scan to "rule-out" a pulmonary embolus, most of the dissections are now found coincidently.

So Ms. Brownlee, what specialty were the doctors you speak of? ER physicians? Could you understand why they might order a lot CT scans of the chest, based on their missed diagnosis?

Finally, a lot of the purported innovations made by the VA were developed during the 1990's, while Medicare greatly cut back reimbursements to hospitals. If they are so great, why not distribute them throughout the country free of charge? It has already been paid for by tax payer money, why not use it to benefit everyone's health?

Posted by: gyp on September 5, 2007 at 11:07 PM | PERMALINK

I stop being an effective lawyer after about 20 straight hours of brief-writing, as I know from experience. Truck drivers have to stop after about 10 hours, if I recall correctly. Evidently, doctoring isn't as demanding as lawyering or truck driving . . .

Doctors on call who don't happen to be residents don't have the option of ignoring pages or refusing to deal with an emergency, regardless of how long they've been working.

How residents in the current system will fare in the real world, where they'll have to see new patients after being on call, won't have a guaranteed day off each week, and may sometimes have to work more than 30 hours straight or 80 hours in a week will be interesting. And not in a good way.

Posted by: chinois on September 5, 2007 at 11:29 PM | PERMALINK

Mike K: Plenty of snarks, I see.

Yes, and with good reason. This is not the first time that you've bragged about your heroic working hours. Depending on the definition of "common" and the particular reasons for doing so, commonly working 24 hours, and occasionally 40 hours, at a stretch is a dangerous practice. If pilots, truck drivers, and just about anyone else in a job where errors can kill people have their working hours limited, then, except under extraordinary circumstances the same should be true for doctors. Unless, of course, you can cite a medical reason why doctors aren't subject to the same difficulty concentrating, increased reaction time, etc. that any other sleep deprived person is subject to.

The fact that you have performed that long without sleep and not killed anyone that you're aware of is no excuse. I've driven for 24 hours straight and not killed anyone. It doesn't change the fact that statistically I'm more likely to kill someone that way, and so I avoid it.

It sounds like Hero Syndrome, which can be one of the greatest impediments to a properly functioning system.

My point is not to encourage it but to point out that hours limitations in residency may not prepare one for the responsibilities of practice.

80 hours/week leaves room for three 24 hour shifts per week. That's not enough?

As far as the salary figures, a friend of mine who was chief of surgery at a university program near Kevin told me he could hire assistent professors of surgery with followship training in subspecialties for $60k.

And that $60k is their entire compensation? Uh, huh, and I'm the Queen of England.

I don't do surveys

But do you ever read them? Data is not the plural of anecdote.

I do talk to medical students every week for the past 10 years. I know where they plan to go and why.

Well, if it's a matter of money, then where do they get their information about compensation?

Lots of people who read surveys but not too many with years of experience.

Yes, lots of people who want objective data, and not too many who think that your anecdotes trump it.

The reliance on anecdotes, limited subjective observations, and the belief that your personal "experience" and "judgment" always trump objective, statistical observation is exactly the sort of thing that the original blog post was criticizing. It's all part of the Hero Syndrome, highly unscientific, and has been shown many times in many fields to be inferior to an Operations Research approach. Maybe the VA can help drag medicine from its early 20th century approach into the post-WWII era.

Posted by: alex on September 5, 2007 at 11:37 PM | PERMALINK

Don't normally post here but I'll add my support to Mike K and Jeff's assertion, since it's excluded by Shannon's "three possible explanations" and has drawn rather strong objections from Alex: my wife was in residency in Manhattan at the time they imposed the h/wk limits, and she felt strongly that the limits were a mistake for exactly the reason they state --- namely, that the unbroken supervision of one tired physician is far better for the patient than a risky handoff from one rested physician to another. She wasn't thrilled about the total hours, obviously, but she felt that continuity of those hours served a crucial purpose.

Alex, I wouldn't dream of arguing that anecdote trumps statistical evidence, but I don't think I've seen any statistical evidence *either way* on this question, and in the absence of hard data I find myself persuaded by the opinion of a few intelligent people I know whose argument makes sense to me. Nothing wrong with that, I trust.

Posted by: andrew on September 5, 2007 at 11:43 PM | PERMALINK

It sounds like Hero Syndrome, which can be one of the greatest impediments to a properly functioning system.

Is there a shortage of lawyers or truck drivers? Have you considered what would happen to ER and hospital coverage at nights and on weekends and office wait list times if physician hours were limited by law?

Posted by: chinois on September 5, 2007 at 11:50 PM | PERMALINK

Whoa ... just want to add quickly that I wrote the above before I saw Alex's last post. No intention to escalate here. And surely the VA-style, Operations Research approach is desperately warranted, I don't mean to dispute that for a minute.

But I will add that a friend of my wife's from med school just got hired as a first-year rheumatology attending at a major Manhattan teaching hospital for $65k a year. That's not a lie, just a sick fact.

Posted by: andrew on September 5, 2007 at 11:51 PM | PERMALINK

gyp: a lot of the purported innovations made by the VA were developed during the 1990's, while Medicare greatly cut back reimbursements to hospitals. If they are so great, why not distribute them throughout the country free of charge? It has already been paid for by tax payer money, why not use it to benefit everyone's health?

They are distributed free of charge (including their software IIRC). A better question is why the resistance to adopting them?

Your reference to "purported" innovations suggests that aversion to change may be part of the reason. Also, as has been discussed in previous threads, the VA has strong incentives to reduce lifetime medical costs. Unlike insurance companies they can't dump sick patients, and usually treat them for the rest of their lives. And unlike most doctors and medical institutions they don't get paid on a fee-for-service basis.

Posted by: alex on September 5, 2007 at 11:51 PM | PERMALINK

They are distributed free of charge (including their software IIRC). A better question is why the resistance to adopting them?

And deprive companies of their opportunity to make huge amounts of money on still more fragmentation in the system? Heresy!

My residents rotate at three hospitals, including a VA. It's not unusual for them to spend time at two hospitals in a given day and on rare occasions at all three. This means they have to shift gears between three different electronic health record (EHR) systems, which is, frankly, a recipe for error. It's hard to get into a routine when one screen demands the vital signs be entered in the upper right hand corner and another in the lower in the page. Routine is a good way not to forget to do things.

They also say they like the VA system the best. The lack of interoperability implicit as companies compete for lucrative contracts is absurd. The time wasted as nurses, physical therapists, pharmacists, respiratory therapists, doctors, and everyone else making chart entries have to train on a different system every time they change jobs can't be trivial.

If, for example, the VA system were made the standard, students in nursing, pharmacy, physical therapy, and medical schools all over the country could learn the EHR system in school. Wisely designed, it could even be a teaching tool. No more wasting hours training on a different system with every new job, to say nothing of the inefficiencies that result until one is accustomed to a new system.

Posted by: chinois on September 6, 2007 at 12:09 AM | PERMALINK

andrew: my wife was in residency in Manhattan at the time they imposed the h/wk limits, and she felt strongly that the limits were a mistake for exactly the reason they state --- namely, that the unbroken supervision of one tired physician is far better for the patient than a risky handoff from one rested physician to another.

That argument has always made sense to me, although it should be subject to an operations research analysis. However it seems that 80 hrs/wk should leave time for those cases where a critical patient requires that extreme continuity of care.

I wouldn't dream of arguing that anecdote trumps statistical evidence, but I don't think I've seen any statistical evidence *either way* on this question

See my 4:09 post. The VA study did show a reduction in problems with reduced resident hours. This was true of several medical (non-surgical) categories. The surgical category showed no change, and hence no harm.

a friend of my wife's from med school just got hired as a first-year rheumatology attending at a major Manhattan teaching hospital for $65k a year. That's not a lie, just a sick fact.

I'll take your word for it. Is $65k/yr for full-time work, or does that position allow for "private" patients on the side? If not, then it's certainly way down on the tail.

Posted by: alex on September 6, 2007 at 12:20 AM | PERMALINK

chinois: Is there a shortage of lawyers or truck drivers?

No, which raises the question of why there is a shortage of doctors in the US. Most other countries have no such problem.

The grueling hours and low pay of residents is an enormous barrier-to-entry. As such it limits the number of doctors, and so requires them to work crazier hours. This justifies grueling residency hours in a vicious cycle.

People who've been through the system justify it on the basis of providing superior training, but is that justified from an overall operations point-of-view? Do other countries (with presumably lower residency hours or other barriers-to-entry) have any worse health care? Statistically, no.

That's part of the hero syndrome. That's a term I use snarkily with Mike K because of his habit of bragging about his hours. In reality though it's a common human failing, and one that I've been prey to.

What matters to the patient though is not how heroic his doctor is, but how the overall system works for him. That can only be assessed from an operations research POV, and there are precious few institutions with real incentive to do it (nor can it be done by any one person, especially one busy with clinical practice).

Imagine if we decided to make medical licensing requirements (MCAT's, med school, residency, etc.) twice as tough as they are now. We'd have fewer doctors, but each doctor would be better. I doubt though that we've have a better medical system in this country, as we wouldn't have enough doctors.

So where's the optimum? Other countries are the best thing to study for that. But if the conclusion is that we'd be better off with lower barriers-to-entry (or less stringent requirements if you prefer), I'm sure there would be enormous resistance from existing doctors.

Posted by: alex on September 6, 2007 at 12:39 AM | PERMALINK

alex: I went back and read your 4:09 post and stand corrected, I had missed that in my first read-through of the thread. And that does rather tilt the balance of evidence in a different direction.

re: the $65k/yr thing; yes, it's for full-time work, no patients on the side; I know that he's attached to a research lab (which probably also accounts for some aspects of the low figure) so there may be a possibility for additional grant income there, but what I've heard of the story over the last few months doesn't sound like it. It should be added that this happened basically because he didn't go out and interview at a bunch of places and play hardball about salary --- he wanted to stay with the department where he'd done his fellowship, had agreed to, and didn't expect to get screwed by the department head when it came time to actually sign a contract. This of course points up something else that I find absurdly primitive about operations at many of these hospitals --- namely the degree to which management duties can fall into the hands of certain doctors who, though very good physicians, have almost no managerial aptitude, or are inclined to run their departments like petty fiefdoms rather than functioning businesses ... again this is grounded in nothing more than a string of anecdotes, but some of the stories along these lines I hear from my wife and her friends in medicine baffle me.

Last thing: clearly I'm biased, and clearly the above anecdote is way down on the tail, just as you say; but when I look back at the unbelievable amount of work these guys have put in over the last ten years, and the unbelievable amount of debt most of them still have as a result, the opportunity cost of, say, not buying a house in 1998 or since because of being in med school and because of the huge debt, and the ongoing shaving of physician salaries ... it's tough for me to believe that the fabled payoff is really going to be there for any physicians that aren't doing billable procedures.

Posted by: andrew on September 6, 2007 at 1:03 AM | PERMALINK

Alex, I am skeptical about everything in medicine these days. Frankly, given my experience of working in the West Roxbury/Brockton VA 20 years ago, I just find this stuff about the VA being so great a bit difficult to believe. (after 4 pm and on weekends, the only doctors there were us residents. Don't get me started about the 2 million awning on the outpatient center, or the earthquake proofing of the buildings at the Brockton VA).

As far as aversion to change, my group is currently evaluating EHR's from various vendors. My hospital has an OK EHR, but it has numerous flaws, and frankly, it is inferior to the system I used 20 years ago @ the BI in Boston. Tell me how to get the VA system free of charge and I'll send you a commission, if I use it. OTOH, is it too much to ask to test the VA systems in non- VA hospitals before adopting it for the entire country?

Posted by: gyp on September 6, 2007 at 2:56 AM | PERMALINK

No, which raises the question of why there is a shortage of doctors in the US. Most other countries have no such problem.

Not really.

Since 9/11, tougher immigration barriers have exacerbated physician shortages in US rural areas, which rely more heavily on foreign physicians. More broadly, doctors aren't practicing until they keel over any more. The increasing proportion of women among younger physicians also cuts into physician-hours, since women necessarily take time away to have children.

The grueling hours of residency aren't the only obstacles to entry. It's worth keeping in mind that cutting back those hours would necessarily extend residency training, by years in some disciplines. Right now, a doctor who went straight to medical school and then to residency (and, increasingly, fellowship) is in or her early 30s by the time training is done. For surgeons or the many doctors--the average age of matriculating medical students is higher than 21--who have done other things before starting medical school, they may well be pushing 40 before training is complete.


Posted by: chinois on September 6, 2007 at 9:47 AM | PERMALINK

>>Numero uno, every VA hospital has a fully-functioning electronic medical records system

And, as implied, since this software was developed on behalf of us--the citizens of the United States--it is in the Public Domain. It's free software for anybody who wants it. It was developed mostly by and for the professionals who actually use it daily. I doubt that every part of it has a pretty GUI interface with all the latest bells and whistles, but it works. It's available now.

Posted by: JMoore on September 6, 2007 at 10:11 AM | PERMALINK

Newsflash: THEY'RE NOT ENFORCING THE RULES.

A resident is expected to get all of their work done and if it means going over the 80 hour limit limit to get it done, then the time limit gets ignored.
An unenforced regulation isn't a rule, its a suggestion.

Meanwhile, in the field of commercial aviation, the FAA limits pilots to 30 hours flying a week, requires at least a 9 hour break between flying and gives the pilot a statutory right to cancel (without job repercussions), any flight he feels he's too tired to handle.

Aviation is simply a better run system because when a doctor screws up he gets to go home to his wife, but when a pilot screws up-- he's the first to die. I'd say the incentives are a little different.

Posted by: beowulf on September 6, 2007 at 12:13 PM | PERMALINK

If, for example, the VA system were made the standard, students in nursing, pharmacy, physical therapy, and medical schools all over the country could learn the EHR system in school. Wisely designed, it could even be a teaching tool. No more wasting hours training on a different system with every new job, to say nothing of the inefficiencies that result until one is accustomed to a new system.

It's always nice when somebody else agrees with one's opinion. I argue precisely this point in my book. Make the VA system the standard. Other vendors can sell systems, but they should all be able to interface with the VA system, which is called Vista. And they should probably look a lot like it, too.

Posted by: Shannon Brownlee on September 6, 2007 at 5:47 PM | PERMALINK

I'm always late to the party because of the day job....

VistA is a beautiful system. The VA converted to it during my third year of residency and it was a painless start up compared to other EMR implementations. But -- and there's a big "but" -- VistA is hardly a solution to the EMR problem. It was designed specifically for the VA. It would really not be appropriate for most practices. A single payer, fully integrated, multi-specialty group like Kaiser might use it, but smaller groups would not be able to use some features and would need additional capability added.

Posted by: J Bean on September 7, 2007 at 12:31 AM | PERMALINK

VistA is a beautiful system.

Hello again, J Bean. yes, VistA is a great system, and could probably be adapted fairly quickly to other venues. I think the critical issue here is making all systems compatible.

Posted by: Shannon Brownlee on September 8, 2007 at 10:02 PM | PERMALINK




 

 

Read Jonathan Rowe remembrance and articles
Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon Sign up for Free News & Updates

Advertise in WM



buy from Amazon and
support the Monthly