Old joke: Severity scale for lies: lies, damn lies, and statistics. In my areas of professional endeavor, when I execute comparative studies I'll use commonly excepted statistical methods like 2-sample t-tests, which in turn require certain technical prerequisites like, is the population's behavior I'm studying normal (you know, that old college friend of ours, the bell curve)? Am I comparing populations which can be sorted by the variables I've selected to partition them by? Of course as the health care reform continues to simmer, the usual suspects (Our editorial policy does not affect our news editing functions. No really. Stop laughing.) appear with their helpful information to educate the public with. Stuff does get technical, but out and out misrepresentation doesn't help. Here's a data set which, if it was normal, would look like a diagonal line:
What does this mean? Well, most statistical methods assuming a "normal" probability distribution will probably give you a questionable conclusions on questions like, "Does treatment "A" effect my process the same way as treatment "B"?", or "What is the probability of making what we call a Type I error (the thing I have said is true is actually false)?". More importantly, what does this have to do with "The Paper of Record". Notice the Times is talking to physicians who utilize manual surgical methods for relieving prostrate restrictions of the male urethra and the recent popularity of Intuitive Surgical's DaVinci system with urologist patients (1. Study participants should not bias your conclusions). Further, the Times produces data saying clinical outcomes between manual methods and more instrumented methods are not statistically different. The implication is "idiot" patients are being sold by clever marketing campaigns which consume more healthcare dollars, while delivering few real benefits. What the Times buries in the article is the fact the Medicare data used confounds robot surgical systems and endoscopic surgery. They both use instruments, and they both work with small incisions, but the robotic system is much more precise. Does this matter? It needs to be validated by actual clinical outcome studies (which take time), but I suspect the answer from the anecdotal record is yes (not to mention the reduced postop recovery time).
The interesting thing about the article is it seems to unjustifiably conclude we don't need this "expensive", "unproven" medical device. Does this article seem to prepare the ground for the Times support of health care reform (which focuses on basic medical services we want to get for "free")? I'll let you reach your own conclusions, but I'm a cynic with most media reports these days.
And if my prostrate doubles or triples in size, I'll probably look for a urologist who has an extensive track record with the DaVinci machine.
No comments:
Post a Comment