Articles

Medicine’s Beautiful Idea

For most of human history, doctors have killed their patients more often than they have saved them. An excellent new book, Taking the Medicine: A Short History of Medicine’s Beautiful Idea, and Our Difficulty Swallowing It, by Druin Burch, MD, describes medicine’s bleak past, how better ways of thinking led to modern successes, and how failure to adopt those better ways of thinking continues to impede medical progress.

The moral is not that doctors once did foolish things. The moral is that even the best of people let themselves down when they rely on untested theories and that these failures kill people and stain history. Bleeding and mercury have gone out of fashion, untested certainties and overconfidence have not.

Burch’s conversation with his rowing coach epitomizes the problem:

“I want you to keep your heart rate at 85% of max for the next hour and a half.”
“Why?”
“Because it’s the best way to improve your fitness.”
“How do you know?”
“Because I’ve done it before and it worked. Because that’s what the people who win the Olympics do. I know, I’ve trained some of them.”
“But has anyone actually done an experiment?”
“What on earth are you talking about?”

This book is Burch’s answer to his coach’s question. Medicine’s “beautiful idea” is that we should test all hypotheses and beliefs using the kind of tests that are reliable for determining the truth. Instead of going by tradition, authority, theory, common sense, or personal experience, we now have effective tools to find out for sure whether a treatment really works.

The scientific method developed slowly and there were a lot of hiccups on the way. Researchers frequently misunderstood what constituted evidence.

In an early Chinese experiment, two people were asked to run together. One was given ginseng; the other, who didn’t get ginseng, developed shortness of breath. They thought that was sufficient evidence to prove that ginseng prevented shortness of breath.

Galen gave one of his potions to a lot of patients: some recovered, some died. He thought that was evidence that the potion worked, because

All who drink of this treatment recover in a short time, except those whom it does not help, who all die. It is obvious, therefore, that it fails only in incurable cases.

Galen’s fallacious reasoning is easy to spot, but a 20th century doctor committed a similar error. He gave all his patients aspirin and asserted it was 100% effective in preventing heart attacks. Some of them did have heart attacks, but he didn’t count them because on close questioning he found that they had omitted doses or otherwise didn’t strictly follow the aspirin protocol (which was probably equally true of all his patients).

Even after the importance of randomization was recognized, there were errors in applying the principle. In early trials, randomization was by alternate allocation, where the first subject to enroll is put in group A, the second in group B, the third in group A, etc. But doctors tended to bend the rules to put certain patients in the treatment group. True randomization had to be forced on doctors who thought they knew what was best for their patients and who didn’t even realize they were cheating.

Humility is required of those who have theories rather than evidence. If they design experiments simply to confirm their prejudices, they are in danger of designing bad ones or misinterpreting results. The more researchers want to prove that the results were due to their favored treatment, the more exhaustive should be their search for alternative and equally reasonable explanations.

Burch’s book is a history of medicine with many intriguing stories about people, personalities, penicillin, opium, thalidomide, and the other usual subjects of medical history; but it is also an explanation of the scientific method and a commentary on modern medicine’s failure to rigorously and consistently apply that method.

Despite our increasing acceptance of the scientific method, the term evidence-based medicine (EBM) didn’t appear in the medical literature until 1991. Critics of scientific medicine have unfairly claimed that less than 10% of treatments are EBM. Burch points out that evidence doesn’t just consist of randomized controlled trials (RCTs), and that we have good evidence that parachutes save lives without having to do an RCT on parachutes. The 10% figure is way too low: a recent study estimated that 80% of current treatments are based on evidence.

Testing and experiment have failed to protect us from deluded cures and poisonous remedies. They can’t be relied upon unless they are carried out with method and rigor. Understanding previous mistakes helps us to avoid them.

Burch has some harsh things to say about current medical research and the processes of drug approval. Many treatments accepted as EBM are actually based on poor quality studies. 62% of studies change the definition of what they are studying between ethical approval and publication. Some studies are stopped prematurely because of apparently clear benefits or risks to patients: this is usually a mistake that diminishes the quality of data. It might be better to finish the study as planned and harm a few patients today than to harm thousands of patients later because of a false conclusion.

People worry about withholding new drugs from needy patients while they undergo testing. They worry about the ethics of offering placebos to patients when a new drug offers an apparently effective treatment. But history has shown that the new drugs in these trials are just as likely to harm as to help.

A drug’s effects, even if they are moderately large, can almost never be reliably figured out on the basis of personal experience.

Doctors are still reluctant to trust science when it goes against their prejudices. He tells how cardiologists strongly supported the first Coronary Care Units (CCUs). A study was done comparing CCU treatment to home treatment for heart attacks. The researchers told the cardiologists that there were fewer deaths in the CCU but that the difference didn’t reach statistical significance. The cardiologists all thought this trend was a strong enough reason to insist on CCUs. Then the researchers admitted they had lied: the numbers were correct but reversed. The trend had actually favored home care. Based on the same quality of evidence, the cardiologists now did not consider the data a strong enough reason to insist on home care!

Medicine is becoming more scientific and more evidence-based every day, but we can and should do better.

What is needed is a culture, regulatory and intellectual, where every attempt is made to ensure new medical interventions are used solely in randomized trials. Only when their effects have been determined should they become available for use outside a trial setting. Until then there is a moral obligation on doctors to use unknown drugs and treatments only in such a way as to come to an understanding of them, and a moral obligation on patients to demand treatments that are either supported by sound evidence or only given as part of a trial which will uncover some.

This is good advice for mainstream medicine, and it is even more important for alternative medicine, which Burch doesn’t address. Since by definition “alternative” medicine is medicine that has not been proven effective, following these guidelines would eliminate any use of alternative medicine outside of a clinical trial. I know, the money isn’t there and it would be difficult to implement, but the principle is irrefutably sound. (That’s assuming that we want to avoid using placebos and find out what really works; but I don’t think the general public wants that. I suspect they would resist and prefer to cling to untested beliefs.)  

Here’s a sampling of some of Burch’s quotable words of wisdom:

There is a bitter joke in modern medicine: the violence with which someone makes an argument is inversely proportional to the amount of evidence they have backing it up.

Trials can be full of statistics; difficult to understand and laborious to undertake. They have a loveliness to them all the same, and it comes from their power to uncover parts of the reality we live in.

[It is] our nature to prefer credulity to doubt, confidence to skepticism. We share a tendency to simplify and confuse things, to slip into mental habits that let us down.

The idea that even the most reasonable-sounding theories should be subjected to tests probably has more potential to make the world a better place than all the drugs that doctors possess. Economics, politics, social care and education are full of policies that are based on beliefs held as a matter of principle rather than because they are supported by objective tests. Humility, even more than pills, is the healthiest thing that doctors have to offer.

I highly recommend this book. It’s well-written, entertaining, and provides much food for thought. It’s a great way to learn about fascinating incidents in the history of medicine and a great way to learn what constitutes truly science-based medicine and how to avoid the errors of the past, the errors in thinking that we flawed humans are all susceptible to.

Posted in: Book & movie reviews, History, Science and Medicine

Leave a Comment (18) ↓

18 thoughts on “Medicine’s Beautiful Idea

  1. hat_eater says:

    Thank you, I’ll be adding this book to my short wish list.
    Is the researcher in the story of cardiologists Archie Cochrane? I recall reading about it somewhere… ah, here it is:
    http://www.badscience.net/2010/04/righteous-mischief-from-archie-cochrane/

  2. daijiyobu says:

    Per: “critics of scientific medicine have unfairly claimed that less than 10% of treatments are EBM,”

    I was thinking the other day that this is along the lines of a ‘but everybody was doing it’ or ‘standards are really merely rituals’ kind of excuse by such for allowing anything, without scrutiny.

    -r.c.

  3. zeno says:

    An excellent review of an excellent book, Harriet!

    I finished it a few weeks ago, but want to re-read it. As you say, it tells us a lot about where doctors went wrong and where lessons have still to be learned.

    BTW, the story about the CCU trial, the researcher who lied to the cardiologists about the results to see their reaction was one Archie Cochrane.

  4. cervantes says:

    I would say that the greater challenge right now is not so much to increase the evidence base for medicine and find new and better treatments, although that should certainly continue; but rather to implement what is already known. This is particularly a problem in the United States where we have no equivalent of the UK’s NICE and it is, in fact, politically toxic even to suggest such a thing.

    I’m looking right now at Bridget Kuhn’s article from the April 28 JAMA about inappropriate prescribing of antipsychotic drugs. In spite of a black box warning, doctors keep giving these to elderly people with dementia. She notes that in an analysis of VA data, 60% of people who got a scrip for an AP in 2007 had no record of a diagnostic indication. These drugs have potentially terrible side effects and shorten people’s lives. She quotes Douglas Leslie of Penn State College of Medicine as saying that docs are prescribing these drugs off label because they have heard “anecdotal stories of benefit.”

    Of course it’s because of illegal marketing by the drug companies, and they’re getting away with it.

  5. Versus says:

    Dr. Hall,
    Could you please post the citation for this:

    “The 10% figure is way too low: a recent study estimated that 80% of current treatments are based on evidence.”

    I’d like to add it to my arsenal for when I am attacked by alt med types.

    Unfortunately, as I write this, Florida Gov. Charlie Christ has before him a bill which would mandate an ultrasound prior to having an abortion and make a woman sign a form if she doesn’t want to review the results. This will in effect force doctors to practice non-EBM if Crist signs the bill. As he is now an Independent and not a Republican, maybe we have a better chance of a veto.

    Along with mandating ultrasounds, the bill also would require women to sign forms if they don’t want to view the results.

  6. Versus says:

    OOPS! Ignore that repetitive last sentence.

  7. Harriet Hall says:

    Versus,

    The 80% figure is from this 1995 study in Lancet: http://www.ncbi.nlm.nih.gov/pubmed/7623571
    It was based on inpatient general medicine and actually estimated 82%.

  8. wales says:

    Harriett, this sounds like an interesting book. I have a question, could you please provide a citation for the study you mention here “The 10% figure is way too low: a recent study estimated that 80% of current treatments are based on evidence.”

    A few lines down you say something somewhat contradictory “Many treatments accepted as EBM are actually based on poor quality studies.”

    Thanks.

  9. wales says:

    Thanks for the citation, though one should keep in mind that this small study (109 patients) was published in the Lancet 15 years ago while the BMJ’s “Clinical Evidence” currently suggests that only 12% of clinical treatments are proven to be beneficial.

  10. Ray Greek MD says:

    Hi Harriet
    Nice article! I am a big fan of EBM but do want to point out one, not exception really, but rather situation that should also be considered in clinical practice. The following is modified from a table in JAMA 1998; 280:1618-19. The article is discussing evidence and ranks the following in order of importance:

    Multiple randomized, prospective, double blind, multi-institutional studies.
    One randomized, prospective, double blind, multi-institutional study.
    One randomized study.
    One well designed, controlled trial without randomization.
    One well designed cohort or case-control analytic study.
    Evidence from multiple time series.
    Dramatic results from an uncontrolled study.
    Opinions of respected authorities based on clinical experience, descriptive studies, or reports of committees.

    My only point in bringing this up is that single observations (anecdotes) or “dramatic results from an uncontrolled study” be that study informal and even with a very small number of patients, have historically alerted the clinician to anomalies that proved very important. McBride and Thalidomide, Iressa and the small number of patients that improved, and Flemming and the topical administration of penicillin are some examples.
    I look forward to reading “Taking the Medicine”.
    Thanks!
    Ray

  11. weing says:

    “BMJ’s “Clinical Evidence” currently suggests that only 12% of clinical treatments are proven to be beneficial.”

    Did you ever find out if the universe of clinical treatments they reviewed include homeopathy, acupuncture, and other alternative modalities?

  12. BobbyG says:

    Great stuff. I am an ongoing student of what I would call “the cognitive liabilities of expertise” (e.g., “how doctors think,” “how lawyers think” etc). This post speaks to a good bit of that.

  13. pmoran says:

    “BMJ’s “Clinical Evidence” currently suggests that only 12% of clinical treatments are proven to be beneficial.”

    While not stated, those figures appear to refer to the questions that this group has thus far selected for systematic review, rather than all treatments used.

    There will be a bias towards controversial subjects, and obviously effective treatments are not likely to be considered.

  14. David Gorski says:

    My good buddy Orac has discussed the nonsense that is the overall estimate for treatment efficacy by BMJ Clinical Evidence in the context of homeopath Dana Ullman’s pointing to it as “proof” that SBM is not better than homeopathy:

    http://scienceblogs.com/insolence/2010/04/the_return_of_dana_ullman_2010.php

  15. tmac57 says:

    The quote that really caught my eye was “[It is] our nature to prefer credulity to doubt, confidence to skepticism. We share a tendency to simplify and confuse things, to slip into mental habits that let us down.”
    It should humble everyone that despite the awesome accomplishments of humans in engineering and mathematics, that led to sending astronauts to the moon, and rovers to Mars, that understanding the human body is still such a challenge. That is why it is so important to follow the evidence and do the hard work to overcome our innate tendency to take mental shortcuts in what is arguably one of the most complex challenges that science has taken on.
    We like to keep our understanding of the of the world manageable, and within our cognitive grasp. We strive for control over our environment as a matter of self preservation. But this can come at a cost of misunderstanding reality, or turning our executive function over to someone else, who in turn makes the same mistake either accidentally, or intentionally.
    This sounds like an excellent book for skeptics, and anyone who takes a more circumspect view of the world.

  16. halincoh says:

    The book is now on my list to read. Thank you Harriet.

    But my favorite post on this thread belongs to David Gorski. It suggests that if Superman had created a blog, Clark Kent himself would indeed have read it. :)

  17. It’s funny how the “10%” myth continues to be recycled. The best article debunking it, which I briefly discussed just a few mos. ago, was written by our own David Ramey (look to your right).

  18. JMB says:

    @BobbyG
    I checked out your website. Of course, I always wondered why the government didn’t make more use of its data, it is a goldmine for scientific study. The VA system is exemplary in the way it can make population data available for analysis. They can serve as a model for such use of data (as well as many foreign countries socialized medical systems).

    You might also want to check out some of what is available from Medical Decision Making. SBM focuses on how practicing doctors can be skeptical through scientific principles of empiricism, of all the information and claims we are bombarded with. MDM was focused on how to improve the decisions of clinicians through both the study of the decision process, and the development of aids to incorporate medical scientific evidence into the clinical decision process. If you are tasked with the induction of evidence from population data to guide clinical decisions, you will probably get a harsh lesson in the requirement of skepticism. The principles you learn from these articles on SBM will give you a headstart.

Comments are closed.