Why We Need More Evidence-Based Medicine

As mentioned here previously, the stimulus package passed in February includes funds to encourage evidence-based medicine. Some uninformed critics will claim that this is some big government conspiracy to exert socialized control over private medicine. But, truly, encouraging a firmer empirical basis in all aspects of medicine--through more studies, government guidelines, and just improved common practice--is a very desirable outcome.

A post by David Newman at The New York Time's Well blog lays out a variety of examples of why this is so (with links to original studies!). Also, Hugh Pickens has a nice summary of this post at Slashdot. Newman lists a variety of clinical interventions that are commonly used, despite ample evidence indicating they're ineffective, or even harmful: early administration of beta-blockers, general use cough remedies, various applications of antibiotics, and certain surgeries. Check out his post for more.

Beyond the fact that doctors must strive to give the best, most effective, and most efficient treatments for their patients, and even beyond the fact that some of these needless interventions have more far-reaching consequences (i.e. the spread of antibiotic resistant bacteria), this status quo is just very wasteful (to the tune of billions of dollars). Newman's main point, though, is about ideology trumping empirical evidence:

The administration's plan for reform includes identifying health care measures that work, and those that don't. To place evidence above ideology, researchers and analysts must be trained in critical analysis, have no conflicts of interest and be a diverse group.

Perhaps most importantly, we as doctors and patients must be open to evidence. Pills and surgery are potent symbols of healing power, but our faith in these symbols has often blinded us to truths. Somewhere along the line, theory trumped reality. Administering a medicine or performing a surgery became more important than its effect.

Scientists--and professionals in related fields, such as medicine--are often driven by the desire to understand and explain nature. Still, a good scientist knows that evidence always trumps theory, regardless of how elegant your hypothesis is. Of course, those most guilty of letting ideology trump evidence are the pseudoscientists. The creationists who won't accept evolution because despite all of the evidence it fundamentally disagrees with their worldview. The peddlers of new age treatments that have never been demonstrated to have any real medical value but fit with some personal idea of how the body is supposed to work. The anti-vaccinationists who latch on to one thoroughly discredited study and a variety of anecdotal evidence at the expense of all other studies on the topic--and at the expense of the health of their children and the children they come into contact with.

The list goes on, partly because such thinking is human nature. We're drawn to patterns and to neat ideas. That's why we must continually go out of our way to avoid unempirical thinking, particularly in such an important field as human medicine.

Hat tip to Ben Goldacre.


More like this

Excellent post - I'm a med student. From what I've seen its very easy for a doctor to slip into using experience from their clinical practice, even when it contradicts the evidence. Sometimes it's legit - not all patient populations are the same - but mostly its chance that their experience is the way it is, so they're practicing "Voodoo medicine"

The list goes on, partly because such thinking is human nature. We're drawn to patterns and to neat ideas.

Yeah, it's pretty seductive. My usual strategy is rest and a slight change of diet (Chicken soup, orange juice and tea with lemon instead of coffee). I usually recover in a cople of days, with a cough persisting for the following week. This time I took and OTC cough medicine, which seemed to eliminate the week-long cough.

That was enough to convince me to use that it was of use, despite being a single anecdotal story. It could well be that the particular cold I caught was simply less severe, or the placebo effect was effective in my case. Unfortunately, it seems that as long as there is no effective treatment, there will always be a market for placebos.

I agree with the overall premise - there are common practices in medicine that are either unproven or disproven and should generally be stopped.

However, Newman's post is a lousy example of the evidence-based practice he claims to support. He laments the use of beta-blockers, saying they don't reduce deaths and they do increase heart failure. But the link he provides actually paints a more mixed picture. It does say what Newman says, but it also says metoprolol significantly reduces re-infarction and atrial fibrillation (p = 0.001 in each case vs. placebo). Newman conveniently omits any mention of that.

Regarding cough medicines, he again misleads. He claims that "no cough remedies have ever been proven better than placebo," but his link only indicates that over-the-counter cough remedies don't work.

Next, he claims "Patients with ear infections are more likely to be harmed by antibiotics than helped." Not one of his links actually supports that claim. The first only shows that 80% of ear infection symptoms improve within 2-3 days without antibiotics. It doesn't conclude that antibiotics are completely ineffective, and it doesn't conclude that they're harmful on the balance. The links on bronchitis, sinusitus, and sore throat aren't even less supportive of his claim. Each one concludes that antibiotics actually do help somewhat for some patients. To be sure, they also caution that the magnitude of such benefits should be weighed against the potential adverse effects, but they do NOT conclude that such patients are "more likely to be harmed than helped."

Newman's overall point is a good one. But if he wants to make a convincing case that those are anti-scientific practices, he shouldn't resort to anti-scientific distortions of his own.

Since I think that both CAM and Creationism are slight variations of the same theme (that is, faith-based pseudoscientific woo), I hope that the future of scientific research will be supported by the new Obama administration. I'm tired of stories such as the one about how the Bush administration forced the FDA to follow religious dogma.

BTW, I found this blog while searching for some good reading on my new Kindle. Don't spend the $0.99 that I paid on anything extravagant.

Evidence based medicine helps doctors but if the doctors fail to use their common sense and knowledge, evidence based medicine is useless. If evidence based approach is like "The Bible" in medicine, then I think our computers can probably do our job better.

In 1980s I noticed babies brought into A&E because they stopped breathing at 3 am were the once who were given cough suppressants and some sort of aroma (Vicks, eucalyptus or karval applied on chest). Babies are obligatory nose breathers when excessive secretion is produced due to irritants (aroma) and their cough reflux is suppressed, these babies will drown in their own secretions. I tried to get my observation published but no journal would accept it, now its headline news.

Advances in medicine in the past century were because doctors used their knowledge, skill, observed and not based on evidence. This is called âThe Clinical Hunchâ.

Similarly, red ears diagonised as infections and treated with antibiotics. Not many know we have to check the movement of eardrums to diagnose otitis media (infection). This is done by using a pneumatic pump or ask patient to perform Valsalva manure. Most auroscope have a a small tube near the ear piecel, but most doctors do not have a clue about this and now we have nurses who are dishing out antibiotics based on this clinical evidence telling them âred ears is infectionâ. I have seen read ears soon after a child cries; redness is due to increased blood circulation, similar to red check when you blush.

The downfall of medicine will be because we allowed this so-called evidence based medicine to take over and forgot medicine is an art based on clinical examinations. If we do not apply our knowledge to help diagnose and treat why do we need doctors?

Our teachers (student of Chamberline) thought us "Your eyes cannot see what your mind can't think". To read more, please check out my website and pass on your comments:(www.medifix.co.uk),"Parents/Symptoms"

I'm disappointed. I thought you'd have a vigorous, forthright, 5000-word rebuttal from Orac by now!

I like this article, especially the first part.
I hope people eventually become conscious about what they do to themselves. 'Drugholics' treat colorful pills like candies. Even if some of them are said not to have any influence on their body still can be really dangerous, as taking few kinds of 'safe' medicine can appear to have unpredictable results (http://www.scienceblog.com/cms/aspirin-and-similar-drugs-may-be-associa…). On the other side some doctors believe only in medicine instead of human's body ability to cure itself with just a dose of help. I guess it's high time to changa that poin of view.