The Only Way to Win Is Not to Play

If this article from CBS News is to be believed, medical error is the third leading cause of death in the United States:

Medical errors, including wrong diagnoses, botched surgeries and medication mistakes, are the third leading cause of death in the United States, a new study suggests.

Scientists from Johns Hopkins found that more than 250,000 Americans die due to medical mishaps every year, greater than the toll from any major medical condition except heart disease or cancer.

The findings, published in The BMJ, come from an analysis of death rate records spanning eight years.

I have considerable sympathy for the medical professionals on this score. When you intervene, some of the patients will die. That’s just the way it is. And sometimes you make mistakes. To err is human.

There are any number of ways to reduce the rate of medical error. The most obvious is to curb the predisposition to intervene.

Another is something that the medical profession has been fighting a vain, inevitably losing battle against for decades: automation. Computers don’t tire, don’t rush so they can make it to their daughters’ dance recitals, and don’t need vacations. Increasingly, computers are looking over physicians’ shoulders and that’s something that will only increase. And that will reduce the rate of medical error.

10 comments… add one
  • Modulo Myself Link

    Medical errors can take a number of different forms, including diagnostic errors — missing the correct diagnosis due to substandard evaluation of a patient — and drug mishaps. Unnecessary surgery, not calling in a specialist when one is needed, and missing life-threatening conditions such as septic shock also feed into the problem. Most such instances reflect system-wide problems, such as poorly coordinated care and inconsistencies in insurance coverage, among other issues, and cannot simply be blamed on “bad” doctors.

    It doesn’t sound like this is a problem of unnecessary interventions.

  • PD Shaw Link

    Computers can also be used to track their own errors.

  • The number of interventions matters because it’s a tally. Assume that the rate of medical error per million procedures is 1. On average if 100,000 procedures are performed there will be no errors, if 1 million are performed there will be 1 error, and if 100 million are performed there will be 100 errors.


    Computers can also be used to track their own errors.

    and they don’t have to face each other at cocktail parties.

  • Modulo Myself Link

    I’m not sure how many accidental deaths take place in the segment of patients where the initial intervention is debatable.

    I actually had surgery last week to remove a benign nerve sheath tumor from my right brachial plexus. The surgery went well. I’m not in any substantial pain and the pain that was there before is completely gone. Interestingly, though, they gave me a Percocet scrip just in case Tylenol wouldn’t cut it. When they filled it, I ended up with 120 Percocet. The other drugs, a steroid, an antacid, and an anti-constipation drug for the side-effects of the Percocet–these were all rationed out diligently. But for some reason they gave me a briefcase filled with the only substance that might be fun.

    It’s like they want people to have serious opiate problems.

  • PD Shaw Link

    . . . or its a sting operation to catch you selling your suitcase in an alley to a conservative radio talk show host. Its Obama’s fault.

  • Modulo Myself Link

    My defense being that I needed the money to pay for my Obamacare premium…

    Actually, I have a good Obamacare plan with no deductible and it’s been great, relative to my other pre-Obamacare health care, one of which failed to authorize an MRI two years ago after x-rays suggested I had tendinitis in my right shoulder. It was the MRI that showed the tumor; otherwise I would be dealing with physical therapy for an old swimming injury.

    The surgery plus the biopsy seems to have cost around 80K so far.

  • steve Link

    Very complex issue. I have never had a chance to look at the methods on these studies. I do know that there is a lot of difference between practicing high risk medicine in real time and reviewing cases after the fact. That said, we docs could be a lot better. As a whole, we have been pretty resistant about checklists and protocols. A lot of worries about cookbook medicine. Yes, they will occasionally lead to bad results, but the large majority of time they will produce much better outcomes. A good deal of my time over the last 10 years has been spent getting docs to accept these. So, at our larger facilities where people now follow them pretty wholeheartedly, mortality rates and costs are 20%-40% lower than at our other facilities. Now we need to get these other places to comply, where we have much less authority. (These places are full of the “I am a doctor and no one is going to tell me what to do”.) Will be a battle.

    Also, Dave’s point is excellent. One of our big ongoing projects is a death panel. (Ok, that is what conservatives would call it if they were running for office.) What we are doing is more testing pre-operatively, especially cognitive tests. We are finding that the risks for some procedures are so high they should probably not be done. In the past, the proceduralist would get paid regardless of outcome, so they had incentives to go ahead. We are trying to change that. Next, we need to stop doing procedures that have been shown to not improve outcomes.


  • Next, we need to stop doing procedures that have been shown to not improve outcomes.

    I think a lot of patients would be astonished how frequently that’s the case.

  • Andy Link

    I would agree with Steve about checklists and protocols. There’s a lot of evidence of resistance to that in the medical profession.

    Ironically, I had an experience with multiple medical professionals curbing the predisposition to intervene. My 91yo Dad fell down the stairs at his house. Fortunately, he’s is better physical health than most of his peers and did not break any bones. But he did go to the hospital (he had some significant bruising) and the only thing the doctors and nurses seemed to care about was getting his wife and my brother to sign a do not resuscitate order. Other than that, they were interested in discharging him as quickly as possible.

  • Gray Shambler Link

    Drs. do the best they can with what they have. They don’t have to care, they want to succeed, not lose patients, especially in the E R.
    Family members can help by revealing information the ill patient may not. Some years ago. my wife began to pass out by degrees and we could not rouse her. We went to the ER and there she quickly went completely under, blood pressure dropping rapidly, oxygen levels falling. The staff was fully alert and out of time to run tests, and they were losing her. They turned to me for anything that might help them, I called to consult with our adult daughter out of state. She was very upset but as a local ambulance volunteer she had some training and knew her mom better than I did, I guess. Years of back pain had caused her to rely more and more on increasing amounts of Tylenol. All I can say is I knew this but it was normal for her.
    At any rate, Liver failure due to Tylenol overdose. This excited the staff because they had a treatment via injection that neutralizes Tylenol.
    All worked out well but wouldn’t without our daughters input.

Leave a Comment