24 September 2009

Annoying Bias

I like Ezra Klein.  He's serious, and studious and laudably progressive.  But his approach to physicians is becoming lamentably more adversarial.  From his commentary on the NYTimes article on Medical Malpractice:
But the best way to reduce malpractice costs would be to reduce malpractice. The problem isn't in courtrooms so much as on the operating table. But because it's doctors who are angry about malpractice suits, most of the fixes are from their perspective. What we need is malpractice reform from the patient's perspective. That wouldn't be the system we have now, or mere caps on damages: It would be serious work and investment in better practices.  (Emphasis added.)
This is not the first time that Ezra has voiced this perspective, and it's just wrong.  It's also offensive, but that's beside the point.  I can see where Ezra gets this -- he's a policy wonk and heavily into the quality buzzword lingo.  It demonstrates a problem with the organizational quality initiatives which focus on patient safety and medical errors: they generates a fixed belief (especially among non-clinicians) that all medical errors are malpractice/negligent and that they are preventable through best practice measures and innovation in the delivery of health care.  But the Pollyanna-ish belief that "if doctors were just better, we wouldn't have these lawsuits" betrays a lack of real-world experience actually delivering health care. It betrays a fundamental confusion on the distinction between Quality Improvement and Negligence.  It's distressing that even some of the better health care policy people bring this sort of ill-informed understanding to the debate.

I admit that I was a skeptic at first when the "pin-headed bureaucrats" started pushing these quality standards down on us.  Part of the problem was that the recommendations seemed too trivial (hand-washing? really?), and part was that some were actually wrong on the science (antibiotics within 4 hours for pneumonia).  But I've come around over the years.  I've come to realize that there are things that we do over and over that can be done better, that can be standardized, and that can have a huge impact on cost and on human lives.   If you haven't read Atul Gawande's seminal piece on checklists, you should.   I'm now an enthusiastic supporter of finding the ways we can improve quality, and developing real best practices that work.

This, however, has little to do with malpractice.  There is an ill-concieved proposal floating out there that doctors who follow to-be-established "best practices" should be protected from liability, that there should be a "rebuttable presumption" that there was no negligence if the defense can show quality guidelines were met.  Spiffy.

Consider the Emergency Room: the greatest liability cost for ER docs in gross dollars is Chest Pain, most typically relating to missed MI or missed PE or missed TAD.  And rightly so.  It's a common complaint, and it can have huge costs, both economic and human.  This ought to be a perfect substrate for best practice standards, right? 

But it doesn't work well in real life.  There are too many variables for a safe harbor to pragmatically be applied and be defensible. 

Let's consider a fairly famous case.  A 54-year old man presents to the ER with chest pain and syncope.  His ECG shows evidence of an acute MI.  The doctor follows the "best practices" approach and administers beta-blockers, blood thinners, activates the cath lab and the patient goes off for urgent revascularization.  Woo-hoo! All Quality Standards Met!  Unfortunately, the patient's name is John Ritter and he is found to have an Aortic Dissection mimicking an Acute MI and he dies.  

Under the utopian "Quality Shield" the defendant doctor is going to claim that the standards were followed, but the plaintiff's attorney is going to rebut that presumption with ease.  There are enough inconvenient details (some relevant, some not) that the "Best Practice" defense will be torn to shreds, and this will be the pattern in each and every case that is brought.  Plaintiffs' lawyers will become skilled in the argument that the Best Practices were followed but not in a timely manner, or that they were followed incompletely, or that they were incorrectly applied.  It sounds great on paper, but it is going to be an utterly toothless protection in application.

And that's just for the disease which lend themselves to standards and best practices.  Consider the most common lawsuit against an ER doc: wound care problems, most commonly a wound infection or retained foreign body.  These cases are usually not lethal, so the dollar value is low, but the frequency is high.  The variability of lacerations is so high that I can't even get my head around "best practice" protocols beyond no-brainers such as irrigating and exploration of the wound.   And once again, the "best practices" will offer little real life protection, because the plaintiffs will simply allege that they were not, in fact, followed.

Of course, in the aggregate, malpractice liability for ER docs is clumped into "other stuff" -- the bizarre and uncommon things that we see (and occasionally miss).  The funny and inexplicable neurologic symptoms.  The subtle fracture not easily visible on x-ray.  The atypical appendicitis.   It's hard to see how reasonable safe harbors could be developed to have any impact on these pathologies, and the significant liability associated with them.

So Ezra, keep on arguing for healthcare delivery system improvement; I'm with you on that.  But I'd rethink your stance on medical malpractice reform, because it's seriously misguided.


  1. I think you missed the point of Ezra's remarks; he's not advocating the 'best practices' proposal you take down. Instead, he's arguing for changes that promote transparency, to better encourage errors to be recognized, remedied, and prevented. Thus these sentences, which you skipped in your quote:

    "The sad irony of the malpractice system is that it has led doctors and hospitals to be much less transparent with their data, which has made it harder to find out when things go wrong, which has made it harder to put in place systems that keep things from going wrong."

    The problem is that doctors now have a massive incentive to hide their mistakes, which means it's harder for them to address those mistakes. And if they can't address those mistakes, then they'll still make the same mistakes over and over.

    And when doctors make expensive mistakes, somebody has to bear the cost. Ezra is right: the best way to reduce malpractice costs is to reduce malpractice. Most of the reforms proposed by doctors and conservatives simply shift the costs of malpractice onto patients. But if a physician incurs a million dollars worth of damages, that doesn't magically go away because the law says the patient can only recover $250,000; the balance is born by the patient, his insurer, or society. So Ezra is exactly correct: the best way to reduce malpractice costs is to reduce malpractice. But that doesn't mean Ezra is advocating for a "rebuttable presumption".

  2. I do think Ezra's enthusiasm for reducing malpractice costs is both commendable and well-warranted. However, it's very easy to dish out a glib solution to a problem ("Just Reduce Malpractice"), without really grasping the complexity of medical care and recognizing that there is a certain baseline level of uncertainty in medical diagnosis and treatment [cue statistics on sensitivity and specificity]. Perhaps a day or two shadowing in the ED would prove to be beneficial.

    Thanks Shadowfax for such an excellent post.

  3. Thanks for the link to the Atul Gawande piece; very impressive. I agree with Duncan and DM, transparency would go a long way in restoring trust in the patient/physician relationship, and reduce the costs of repetitive and preventable errors.

  4. Also there were two important omissions in the Ritter case, no CXR and no bilateral BPs. Either of those might have caught the aneurysm.

    So it may have been malpractice. Best practices weren't followed.

    The trouble with malpractice is that the jury system is too fallible. Do we want the same people that found OJ innocent second guessing our medical decisions.

  5. What is the infallible system you would recommend?

  6. pdx,

    It would be good to see on the chart if I were going to court and defending the case but CXR and bilat BP's are completely insensitive to rule it out or rule it in.

  7. Part of the problem is that the 'clinical guidelines and best practices' approach to ensuring high quality care (i.e., error free and good outcomes) is grounded in a paradigm that is not easily transported from production engineering, where it originates, to clinical medicine. The issues are certainty and homogeneity.

    When manufacturing an electronic device, it is appropriate and effective to carefully define each step in the manufacturing process of each component and in the assembly process. Each part X is the same.

    Clinical medicine works with orders of magnitude more variability. The disease process is less predictable. Each patient has her/his own genetic makeup, health status, life style, and personal preferences. And guidelines are consistently drawn from clinical studies that (necessarily) focus on a single issue and specifically exclude co-morbidities.

    I'd suggest Designing Care by Bohmer for some nice discussion of this.

  8. I'm always astounded how naive these health policy figures seem to be, and how tenous their grasp of real-world health care appears to be. I'm particularly amused by Duncan Cross' posts, which reminds me of Glenn Beck -- sounds great, but completely wrong.

    There are specific mechanisms in place to promote transparency in medical care, and certain unavoidable checks and balances that ensure that medical errors are documented and examined. And yet, here we have it again -- "The problem is that doctors now have a massive incentive to hide their mistakes" -- I'd give a 'A' for paranoia and demogagry, but an 'F' for failure to understanding what you are talking about.

  9. Anonymous: I only know, understand, and write about the real-world health care I've lived through. If there are mechanisms to ensure that errors are examined and documented, they've failed every time I've encountered a serious error in my care - which is at least three times, and probably more. You can call me naive, but I doubt that anybody with my sort of experience - that is, any patient - has a clue what you're talking about. Keep telling yourself "mechanisms" exist, but it doesn't matter if we never see them.

  10. Well said. I ... er ... agree. Wait a moment ... that can't be right ... must read post again.

  11. Medicine is not an exact science. Conditions can mimick other conditions, and mistakes are sometimes made.

    You shouldn't be able to sue someone just because things didn't go your way.

  12. "You shouldn't be able to sue someone just because things didn't go your way."

    You can't sue just because things didn't go your way. Well, you can file, but you can also file because the moon didn't come up. The clerk who takes your money doesn't have the authority to reject it. But your case will be dismissed if the sole basis is things didn't go your way.

    Although even if the doctor is negligent but you suffer no damage, you can't sue.

  13. We must all realize that doctors are also humans and are prone to commit mistakes. But if it causes harm to human life they should never be allowed to run free.

  14. We must all realize that doctors are also humans and are prone to commit mistakes. But if it causes harm to human life they should never be allowed to run free.