Evidence Based Medicine
When my son was being born, toward the end, during the pushing part of labor, my doctor told me to hold my breath so I could push better. I'd been doing my reading - pregnancy is a great time for terror-driven research - and I knew that holding your breath is no longer recommended. It doesn't actually help with the pushing, and it can spike your blood pressure dangerously. But the doctor hadn't had the same motivation for research, and she kept demanding that I do it. So, as my baby boy came into this world, I was pushing, holding my breath, and shouting "this is not evidence-based."
That anecdote was my little way of letting you know that I am completely obsessed with evidence-based medicine (EBM). The guiding principle of evidence based medicine is that for a health intervention to be worthwhile, it needs to be proven to work. One would expect that all medicine is evidence based, but in fact it's not.
Evidence, in this context, consists of scientific research that demonstrates that medical intervention A causes good health outcome B. A double-blind study is the gold standard. Meta-analysis of less rigorous studies is often used (the most prestigious source of meta-analysis is the Cochrane collaboration). How many more depends on the nature and rigor of the research. Anecdotes, theories, or epidemiology alone are not evidence. EBM looks at the risks and benefits of medical treatment to see if it is worthwhile.
We have evidence that breastfeeding reduces infant ear infections and diarrhea, and reduces infant mortality in the developing world. We suspect it also helps improve infant eyesight and raise IQ, but we don't have evidence for that. It probably also reduces infant mortality wealthy countries, but we don't have evidence for that either. However, our research on ear infections, diarrhea, and infant mortality in lower income countries is so solid that it's really all we need to promote breastfeeding.
One would expect that all of medicine would be evidence-based care, but that's just not true. The less depressing reason is that there is a lot of research we can't ethically do on human beings. We can't do placebo surgery for a double-blind study, for example. And we can't just stop providing supplemental food to Darfurian refugees to check the before and after impact. For things like that, we have to look for natural experiments, such as when the government of Darfur suddenly kicks out aid agencies.
A somewhat depressing reason that medicine doesn't tend to be evidence-based is that medicine tends to be based on theory. If it follows logically, and physiologically that something must work, then it is done. There is a limited supply of research funding; no one may ever test and see if it really does work. Examples include lowering
The much more depressing reason that medicine doesn't tend to be evidence based is that physicians love to provide medical care, and drug companies need to sell their drugs. A physician's entire training structure gears them toward taking care of people. Leaving well enough alone just doesn't come naturally to doctors. They are eager to try new treatments - new treatments drug companies constantly develop -and slow to abandon them. If a doctor isn't constantly keeping up on her reading, she may not even know a treatment isn't evidence-based.
For more information
Wikipedia has a surprisingly excellent article about evidence-based medicine. The Cochrane collaboration home page has a nice introduction, and this is a nice discussion of EBM and setting policy. Ezra Klein has a grim blog post on why EBM isn't dominant in the United States.
COMMENTS (1)