This is analogous to reasoning that low dose radiation causes cancer even though an epidemiological study doesn't discern a difference between a control group and the exposed group.
When drugs are developed they are tested in clinical trials for efficacy as well as unexpected side effects. If there is no efficacy demonstrated (usually at a 90+% statistical certainty level), between those who get the drug and those who don't, the drug flops.
But just because the drug flops, doesn't mean it didn't help anyone who was given the drug.
n-of-1 studies are about understanding why a drug may have helped someone, or why it appears that it did though the cause may not be the drug.
No comments:
Post a Comment