Posted by Lindenblüte on October 26, 2006, at 10:32:09
In reply to Evidence and civility, posted by Squiggles on October 25, 2006, at 19:09:56
> It would be nice if there were
> resources on this board, where a
> medical falsehood posted could be caught
> and reprimanded as uncivil -- much
> like being impolite, or making
> racist remarks.what's a falsehood?
1) in the literature there are often conflicting studies. One study finds a positive result of X drug on depression. One study failes to find result of X drug.
2) what does it mean when X drug and placebo both cause reduction in symptoms? I'd rather take X drug than nothing (my doctor won't prescribe me placebo)
3) sometimes a "truth" is nothing more than a small study that has been over-interpreted.
4) sometimes there is a big complicated study, and a poster (or even another researcher in the field) takes a tiny snippet of their results or conclusions and then extrapolates. This happens when I read fashion magazine articles trying to give me medical advice- "drink red wine- you'll live 7 years longer!" whatever.
5) sometimes we know a treatment works, but there is rampant speculation about the mechanism involved. one paper publishes in vitro studies of mouse hippocampal cells and reports that drug X binds to these receptors. Another study presents PET scans of live humans doing spatial tasks and says that drug X impairs map reading, but not mental rotation- and then finds a relation between map reading impairment and drug X dosage, and wow- there is little evidence that drug X even binds to the hippocampus of humans.
I could go on and on...
My point, squiggles, is that science is incremental. Every study builds on what we know, but no study is really so perfect that it can tell us anything with certainty.
I am guilty of having a bad memory. I don't trust my own recollection of mechanisms, results, studies, for the most part. And I take other peoples' interpretations with a grain of skeptic salt.
I don't think this is uncivil though, it's just advancing science and understanding. increments. 1 step forward. .0003 steps back. margin of error. confidence interval. variability.
there is no "truth" in the biomedical sciences. There are only some theories that have strong empirical support (like evolution perhaps, or the central dogma of molecular biology DNA-> RNA-> protein) and other theories that have weaker or antithetic empirical support.
-Li
civility has nothing to do with it- unless you think that one poster is *deliberately* trying to POISON another one! well, that's why better talk to a doctor before following advice you get on pbabble.
poster:Lindenblüte
thread:697727
URL: http://www.dr-bob.org/babble/20061020/msgs/697883.html