10 February 2006

Defensive Medicine -- what is the cost?

An issue that raised its head briefly in the last presidential campaign was defensive medicine, in relation to the malpractice crisis and tort reform. The president's argument was that the US could save ~$100 Billion per year in defensive medicine expenses if tort reform were passed. I don't know about that -- I think he may have misplaced a decimal there. The culture of defensive medicine is so engrained that I suspect even sweeping malpractice reform might not change practice very much, and the data from states that have passed tort reform seem to back that up.

But it's a little maddening to see otherwise excellent health policy boggers such as Ezra Klein and Kate Steadman claim that defensive medicine does not exist. Hard to measure? Damn straight. But every day of my working life, every patient I see, every diagnostic/therapeutic decision I make, the specter of "what if" haunts me. The decision-making for every single patient encounter is driven by three factors:

  1. What does my gut tell me this person likely has wrong with them?
  2. What do I need to do to reasonably establish the diagnostic impression?
  3. What do I need to do to cover my ass in case I am wrong?
Number one is by far the most important factor. A good ER doc will know the diagnosis five minutes after meeting the patient, most of the time. Number three probably accounts for 2/3 of the CTs and Ultrasounds ordered in the ER. Defensive medicine is endemic -- especially in the ER, where the standard of care seems to be that any missed diagnosis, no matter how obscure or unlikely, constitutes negligence.

For example, the "Standard of care" in ruling out an aneurysm in someone with the worst headache of their life is a CT scan and a lumbar puncture (spinal tap). This is something every ER doctor will agree on to near-unanimity. However, with modern high-resolution CT scans, the negative predictive value of a CT scan is something like 99.8% -- yet every single ER doctor will still do that LP, because if you don't and you are unlucky enough to have that one person in 500 who had the false-negative CT any ER doc will testify that you fucked up. That's defensive medicine. What's the cost? Well, a spinal tap costs maybe $300-$500, and you are doing 499 unnecessary spinal taps for each aneurysm you diagnose by the tap -- that adds up (in dollars as well as pain inflicted on patients who didn't need the LP).

There are examples of this kind of defensive practice for most any presenting complaint in the ER. Like chest pain. Good lord, how many people with atypical pain and no risk factors wind up staying in the ER six to eight hours (or admitted) for cardiac enzymes and a rule out? And it's because the acceptable percent of patients discharged with myocardial infarction is zero.

Don't just take my word for it; there was a lovely study recently published in AEM:

Emergency Physicians' Fear of Malpractice in Evaluating Patients With Possible Acute Cardiac Ischemia, Annals of Emergency Medicine, 46,6, 525-533 (abstract only)

Which documents that the more fearful a doc is of malpractice the more likely s/he is to admit a patient with chest pain, and the more likely s/he is to order extra tests or admit the patient to the ICU. And, as imporantly the diagnostic yield of the extra admissions/tests was nil. So this is not "better medicine" being practiced, it is CYA medicine.

Those of us who work in the pits call this "proving the inherently obvious."

So it makes me a little crazy when it's claimed that doctors aren't motivated by fear of lawsuits -- we are. Now if you want to claim that the overall fraction of healthcare dollars spent on defensive practice is low -- 1-2% of all spending -- I might agree with that. But bear in mind that the healthcare expenses in the US are in the trillions of dollars, so we are talking many billions spent on unnecessary care. If you want to argue that tort reform won't change doctor's practice patterns, I would be open to the possibility -- doctors don't want to be in the NPDB whether it's for a million dollars or for fifty thousand. But don't tell me that defensive medicine doesn't exist.


  1. This is great stuff. I really don't have anything to add other than to say that I find this stuff very interesting and enjoy getting this perspective from a doctor that is a friend that I trust.

  2. Thanks. I try to mix the personal/professional stuff to keep some variety and to appeal to professional and non-professional readers. This is a good place to rant a little, and you know I am a politics wonk. At least with medical stuff, there's always the "war stories" that are interesting to medical and non-medical folks alike.

  3. I think this gets back to some of your earliest blogging discussions/posts about feeling like a blog needed a them to be good. I think you have a nice personal/professional/political mix going. From the perspective of an old friend that is horrible at frequency in most forms of communication, I can tell you I think this is great.

  4. Again, thanks. Praise from Ceasar, indeed, because I have been equally impressed by your efforts (both in the box and at Caringbridge). I feel like if I just write about the stuff that interests me it may stay interesting to some readers. If nothing else, it'll be interesting to me! But I am getting a better handle on my "mission."

  5. I have written many entries like this in the past. DB has as well and so has gruntdoc.

    I do agree that it's so become the standard within the medical profession now that even reform will unlikely change practice at this point. we don't even want to be in court, even if we are right.

    so, great point.

    and BTW, add me to your blogroll. I am plannign on adding yours ;-}

  6. Will do. After this week's grand rounds, there are a lot of new medblogs I have gotten familiar with, so the blogroll is going to expand dramatically.


  7. I think it goes beyond what you suggest. Often there is a referal to a specialist who then does a lot of CYA tests, so the cost is compounded. And the Society position papers usually don't help. For example, the International Headache society guidelines for migraine state and MRI is indicated for a new onset headache or a change in the established pattern of headache. That is soft enough around the edges to include just about anyone.

  8. But don't you want to detect that 1-in-500 aneurysm? You don't want to just send that guy home, do you? One in 500 isn't that rare, and an aneurysm is something you don't want to overlook.

    Why do a spinal tap on the guy with the positive CT scan, though?

  9. I'd like to read your opinion on the push for special "Health Courts" backed by the bipartisan legal reform group Common Good.
    Is this a step in the "right" direction?


  11. This is so true. My son was 20 years old is away at college in Washington DC. One evening after returning from gym, he got wet and developed chest pain. He was concerned and decided to go to hospital. He was admitted and they scared him with every diagnosis in the book from Cardiomyopathy to pneumonia and pleural effusion. While he called me, I reassured him and told him he has nothing more than muscle pain. After over 24 hours at hospital and thousands of dollars paid by insurance, that is what it turned out to be - myalgia chest.

  12. All tests and procedures are defensive. When I went to medical school we did test and procedures because they had been verified by good statistical studies. Now we order many more test and procedures because we know of docs who didn't order them and got sued. The fear of a malpractice is a stronger motivation to do test and procedures than well done studies.


Note: Only a member of this blog may post a comment.