“It is Simply No Longer Possible to Believe”

This piece by Marcia Angell in the New York Review of Books, while very good, mostly consists of stuff that would be familiar and unsurprising to OB readers.  But I was somewhat surprised that she went so far as to say this:

The problems I've discussed are not limited to psychiatry, although they reach their most florid form there.  Similar conflicts of interest and biases exist in virtually every field of medicine, particularly those that rely heavily on drugs or devices.  It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgment of trusted physicians or authoritative medical guidelines.  I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as an editor of The New England Journal of Medicine.

That's pretty strong stuff for someone who is enough of an establishment figure to become the editor of the NEJM.  It's worth pointing out, though, that most of the biases that she is talking about are the product of plain old financial corruption, not the subtle cognitive biases that we mostly worry about here (though those undoubtedly play a role in allowing physicians to delude themselves into believing that they are not being swayed by the money).  So these kinds of problems could probably be mostly eliminated by a conceptually simple (though of course politically very difficult) change in the rules of the game.  Getting rid of problems like physician overconfidence would be much harder.
GD Star Rating
loading...
Tagged as: ,
Trackback URL:
  • sonic

    Greed easier to get rid of than overconfidence?
    Well, we are talking about doctors…

  • http://drchip.wordpress.com/ retired urologist

    David: The later years of my practice were filled with anger and frustration because of my insider experience with the system. I know of no source at all for accurate official information in my field, and I suspect a similar situation in the more “important” health fields. Those seeking to overcome bias readily realize, however, that overconfidence is a blind mistress, hardly faithful to doctors alone, but sharing her charms with AI researchers who would save/destroy the world, economists who hold the secrets to unbiased living, other economists who hold the secrets to everything, short-sighted bankers, and a plethora of OB reader students/researchers/philosophers who never seem to be at loss for an expert insight. Perhaps we’re describing the human condition.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    This is indeed a pretty damning admission.

  • delta

    And what do you suggest, if you can no longer believe? Do you suggest that we give up medicine and just die? Do you think we should take up the herbal secrets of the ancient and lost traditions of Antarctica? Or would you rather suggest less corruption, and more scientific thinking? I suspect it’s the latter that you suggest. But then why do you employ the kind of rhetoric that will enable the quacks to abuse it?

  • David J. Balan

    Retired, you’re certainly right that doctors are not the only ones with problems of this sort. They may not even be the worst offenders, though I suspect they’re up there. But the article I cite is just one more bit of evidence that bias-related problems in medicine are extremely severe. This fact makes it a worthwhile project to try to fix the problem. The “human condition” is relevant only to the extent that you believe that these problems are so tightly woven into human nature that nothing can be done to mitigate them. But this is clearly not the case: legal bans on certain corrupt financial practices would solve a large portion of the problem right away, and changes in the way that doctors are educated would fix a lot more of it in the medium term.

    delta, You are right that harshly criticizing modern medicine runs the risk that people will respond with a flight to quackery, and I would hate like poison to contribute to that happening. On the other hand, I also don’t want to tone down the evidence that a lot of medical expert opinion really is untrustworthy. I’m not really sure the best way to handle this. Maybe criticisms of mainstream modern medicine should be accompanied by something like the following:

    “It is not obvious how a rational person should respond to the serious problems that have been found with modern medicine. After all, if I can’t trust the medical establishment, who can I trust? Myself? The internet? The only thing I can really do is support necessary reforms, and in the meantime muddle through as best I can: probe physicians for signs of rationality and listen to the rational ones more, upgrade some the weight placed on my own analysis (tricky!), have a modest thumb on the scale against drugs and surgery, and so on. What I should obviously *not* do is abandon modern medicine, which has a great many triumphs to its credit but has revealed itself to have a lot more flaws that most people think, in favor of meritless nonsense of this or that variety.”

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    The alternative to medicine is only quackery if we must do something. The core bias is the do-something bias.

  • frelkins

    @Robin & David

    quackery

    As a believer in the scientific method overall, I think it is fair to say that some respected government peer-reviewed studies have begun to show benefits in certain areas for alternative medicine, such as acupuncture, whereas the case for homeopathy seems currently unclear, due to poor study design.

    To use the term “quackery” seems biased prima facie, as it assumes a result we do not yet actually know. There is controversy now as to whether the “placebo effect” exists or not and how strong it may be.

    Nonetheless if the lactose pills of the homeopath were to be effective, even based on placebo, would that be so bad? As long as we are honest, as long as patients are not misled, and as long it leads to better health outcomes, why not?

    Homeopathy and acupuncture are certainly cheaper than conventional medicine. If their true mechanism is just that people get better when authority figures pay individual attention to them, isn’t it even better?

  • David J. Balan

    Robin, One conclusion that I think can be reasonably drawn from available evidence is that when conventional medical advice says it’s a close call whether to do something or to do nothing, it’s probably a good idea to do nothing; and when it says that it’s a close call whether to do something more invasive or something less invasive, it’s probably a good idea to do the less invasive thing. But there are an awful lot of infra-marginal cases where “nothing” isn’t really on the menu.

    frelkins, Here is another disclaimer that should probably be made whenever you’re talking about quackery:

    “In principle, medicine should be open to anything that can be scientifically shown to work. If there is evidence that herbs or needles or whatever *actually work*, then they deserve to be a part of conventional medicine every bit as much as a drug developed in a lab. Furthermore, it is fair to point out that for-profit medicine has pretty strong reasons to ignore some possibly effective treatments: if it turns out that an apple a day really does keep the doctor away, drug companies would have every reason to suppress or ignore that fact. This should have the effect of: (i) making you support research funded by the government or by non-profit institutions; and (ii) lowering somewhat the evidentiary threshold an “alternative” treatment would have to clear before you bought into it. But this does *not* mean that you should just start believing any nonsense: there is still a lot of real quackery out there. And you should be highly skeptical of alternative medicine that markets itself mostly on the basis of citing quotes like the one in this post, and not on actual evidence that their treatments work.”

  • frelkins

    @David

    actually work

    Ok, let’s take. . .gosh, I dunno. . .ayurveda, so popular among rich ladies in New York right now. What does “actually work” mean in this context? You might argue that reviewing studies on cholesterol reduction, to see if say, an ayurvedic medicine with garlic or whatever reduces cholesterol would prove “working.”

    But then we still run into the above “authority figure pays attention to me” issue. The ayurvedic doctor has an amazing office, all decorated in Indian silks, with beautiful Indian assistants serving chai or dosha(constitution)-appropriate teas.

    He wears a fantastic white silk “spiritual” outfit and will listen to you talk for an hour. You will take a computerized test to determine your “individual” dosha, you will be prescribed a “unique” diet to follow, with a “unique” set of yoga exercises, then he will pour oil over your forehead, etc. etc. etc.

    Finally, he will give you an herbal mix or pleasant tasting herbal-flavored ghee-candy paste (rasayana) to eat 2 times a day. And you have to come back in 6 months.

    So maybe the garlic-guggul-arjuna butter-candy mix “actually” reduces cholesterol. I guess we could give it to people via mail order and see if their numbers go down. Or maybe all the lavish impressive “spiritual” attention and time with the yoga instructor does it.

    Heck, maybe the power yoga alone does it, by increasing exercise. Maybe the diet does it alone, by weight loss.

    The point is, does it matter in terms of health outcome which component of the ritual actually works?

    I guarantee the giant tub of fennel-flavored sugared-butter with garlic, guggul and arjuna tastes great (for good patient compliance), and costs a mere US$150 a year, which is certainly cheaper than US$1,400 a year for Lipitor + US$400 for the cardiologist, even when you count in the US$200 initial office visit and the US$35 yoga video.

    I think this is really important to figure out soon, in light of the coming US universal health care plan ($$$,$$$,$$$,$$$). It could be a useful component perhaps of Robin’s suggestion to cut medicine in half, while increasing people’s signal sense of “being cared for” by an expert of their choice.

  • David J. Balan

    frelkins, I agree that “things that work” is a category with somewhat fuzzy boundaries. It’s worth asking whether, say, it makes sense for treatments that are known to work only b/c of placebo effects or “someone paid attention to me” effects to be regarded as legitimate medicine. Does it matter whether they need to be based on deception? That is, does it matter whether they only work if the patient doesn’t know that anything else would have worked just as well? These are hard questions that I have no particular insight into. But I think the broader points above remain. There is plenty of stuff that simply doesn’t work at all.

  • http://drchip.wordpress.com/ retired urologist

    frelkins: Robin Hanson spends many words describing the dollars spent on American health-care without any corresponding improvement (or perhaps some actual decline) in health status. There is much suspicion at present that the isolated achievement of lower cholesterol has no specific health benefit. On the other hand, Lipitor has been shown (hopefully not in studies that fall into Dr. Angell’s categories of fraud) to improve cardiovascular outcomes, not by its intended purpose of lowering cholesterol, but by its anti-inflammatory effects. This should lead to the use of cheaper anti-inflammatory treatments with safe profiles. Among the dangers of abandoning rationalism for “things that sometimes work” are the loss of reproducibility and the distraction from something that might provide an actual health benefit, as opposed to modifying a lab test number. Perhaps we will witness a spate of happy, relaxed rich ladies in New York with wonderful cholesterol levels, looking beautiful at their society funerals. I “heartily” suggest this post and discussion on skepticblog.

  • frelkins

    @retired urologist

    I can’t judge either way about acupuncture, ayurveda, or homeopathy. We lack evidence, NIH says. (If any of them are likely to work, I vote for the homeopathy, simply based on the fact that apparently many of the “medicines” are plant abstracts in brandy — brandy is always good for whatever ails ya.)

    Why not just take a group of Park Avenue ladies, putting 1/4 on ayurveda, 1/4 on Lipitor, leaving 1/4 to spend time with their celebrity hairdresser (the ultimate New York authority figure who gives you individual attention), and abandoning the last quarter to their normal charity events? At the end of 1 year see which has the lower cholesterol.

    Voila. There’s your NIH study right there – but seriously, why not a comparative survey of the treatments? If ayurveda gives you a “good enough” number as the Lipitor, who cares about the “distraction?”

    Why protect the mystique of conventional medicine and doctor authority if I can get a similar number on the cheap? Especially if you’re gonna make us pay for their expensive Lipitor with our meager public purse.

  • http://drchip.wordpress.com/ retired urologist

    frelkins: You ignored the fact that lower cholesterol is not the goal for better cardiovascular health. It’s just an achievable end.

  • frelkins

    @retired urologist

    lower cholesterol is not the goal

    You will please forgive me if I find your statement literally incredible.

    As everyone knows, I bought an insurance policy this fall – all they cared about was that number. I went to the doctor and the first thing he asked me was if I knew that number. Almost every gentleman of my acquaintance over 50 is obsessed with that number.

    Many reputable health articles you read tell you to care about that number, and the American Heart Association is happy to tell you what it should be. Doctors care so much about that number they are now considering statins for children and suggest some cholesterol screening at the age of 2. If this number isn’t the goal, why is everyone fixated on it?

    The only person who questions this fixation seems to be Gary Taubes, whose book has been discussed here. And he’s considered a lunatic by the medical establishment, apparently.

  • delta

    David J. Balan,

    as you can see, you need to be careful with words.

  • http://drchip.wordpress.com/ retired urologist

    frelkins, re: Taubes: he’s considered a lunatic by the medical establishment.

    The “medical establishment” is the subject of Balan’s post, or rather, their dishonesty, inaccuracy, immorality, and overconfidence. If you prefer to go by the guidelines of the AHA and the pharmaceutical cartel, you live in a free country that is perfectly suited to accommodate you. Every discovery is a departure from previously held notions, and is typically resisted by the “establishment”.

  • Aaron

    retire urologist:

    There’s a researcher in town at Johns Hopkins who is developing a test on blood stickiness, which he and apparently some other researchers feel is a better indicator than cholesterol level. Have you heard anything about it? I was pretty interested in it (from a layman’s “I’d like to stay alive” standpoint), but Angell’s article has dashed the hopes of a lot of recent innovations I’ve read about. But at least there is some pockets of criticism of accepted protocol.

    Aaron