Limits of so-called 'evidence based medicine'
I just came across an interesting study on the limits of so called "evidence based medicine" which has relevancy here. Note that "evidence based medicine" is quite different than merely relying on scientific evidence; it has it's own set of rules and standards. A summary of the case:
A patient is prescribed an antibiotic after getting some dental work done. A pharmacist makes an error, and instead of dispensing the amoxicillin, gave the patient the antidepressant doxepin. The man starts taking the doxepin at a dose that is four times the maximum daily dose (which would have been a normal dose of antibiotic). After taking the overdose of the wrong medicine for several days, the toxic effects of the antidepressant are so bad that the patient stops taking it, and the error is discovered. The patient develops glaucoma and neurological damage, which is blamed on the medicine.
There were no studies showing the effects on humans of a massive overdose of the antidepressant (probably hard to recruit people for such a study). The doctors representing the injured man looked to a variety of scientific studies about doxepin toxicity, and presented enough evidence linking doxepin to glaucoma so that the man wins a lawsuit against the pharmacy.
Of course, the pharmacy appealed the case - they admitted that they goofed on the prescription, but though it was unfair that they pay for damages related to the glaucoma. Their argument was that there were no "evidence based medicine" to prove that doxepin can cause glaucoma - that is, there were not multiple, large controlled studies where humans had been given large doses and then developed glaucoma. Without such studies, the pharmacy company said, there was only conjecture. The fact that a drug might cause glaucoma in animals was dismissed as irrelevant. The fact that the drug in humans has an effect on the optic nerve that is consistent with developing glaucoma is interesting, but again, not proof of anything. The fact that the warning that came with the drug mentioned an increase in eye pressure as a possible temporary side effect was not even seen as proof of anything - the fact that normal dose of doxepin causes a temporary increase in eye pressure is not proof that an overdose of doxepin would cause the pressure to go so high that it caused glaucoma. The pharmacy argued that it would be wrong to put all of these fragments of evidence together and claim that it proved that an overdose of doxepin causes glaucoma.
The judge agrees with this argument, reverses the decision, and orders a new trial. In the new trial, evidence that is not 'evidence-based' is strictly limited.
One obvious problem here is that for certain type of questions, we cannot look to 'evidence based medicine' for an answer, because that type of study hasn't, won't, or can't be done! And while questions that get asked a lot (do statins lower cholesterol) are likely to be answered quite well by such "evidence based" methods,...