Lying with statistics? — The JAMA(Journal of the American Medical Association) report

So the cat’s out of the bag. A recent JAMA report has shown that outcomes are correlated with funding sources.

Let’s throw one benign factor into the mix. I’m not saying that this is the defining characteristic of this apparent assocation shown in the JAMA article, but it’s one I don’t hear discussed much. Perhaps, maybe, just maybe, pharma companies are pretty good at knowing whether their drugs will succeed in a clinical trial. (This can actually be benign or evil — their drug might have a better chance of success if one surrogate marker is used, for example.)

Before we skip merrily to any conclusions let’s dig deep (as I saw in a recent conversation). Real deep. After all, we’re about improving health and quality of life, right?


One Response

  1. Very interesting. Frankly, I was expecting it to be nothing more than a “shock value” headline, but those numbers are significantly different comparing for-profit to non-profit studies. Pretty hard to argue it is bias-free.

    This isn’t exactly the same thing, but in the software biz, there are plenty of analysts that will write favorable white papers about your product, technology, etc. — as long as you pay them for it. Same thing is available in the dietary supplement industry. ConsumerLabs will review a product for about $4000. (We told them thanks, but no thanks.) Conversely, SupplementWatch does NOT take any money from any vendors.

Comments are closed.

%d bloggers like this: