Thursday, June 18, 2009

Strange Policy from Bad Science - Part One

I despise the misuse of statistics. I'm a firm believer in the axiom, "statistics don't lie, but liars use statistics."

It is unequivocally true that you can use statistics, bent, folded, spindled, and manipulated, to prove nearly anything. Time was, however, that most people knew when a statistical abuse was at hand, and dutifully ignored. These days, however, where mathematical education and critical thinking take a clear back seat to whatever politically correct curriculum is in vogue, the ability to recognize and refute such abuses is diminishing daily.

A classic case in point arose just this week, when our beloved Government rode to our rescue and told the masses to stop using Zicam Nasal Gel, because it was "linked" to incidents of loss of smell.

Now that, on the surface, sounds pretty serious. If something out there is causing people to lose their smell, we ought to know about it. But a review of the data beyond the hysterical news headline reveals a much less hysterical reality.

The FDA based its admonition the basis of 199 cases of lost smell, or anosmia tied to the use of Zicam over a ten-year period. Although I could not find hard statistics on the number of actual Zicam users in that period, one news site reported the number to be in the "millions."

Let's say that, for the sake of argument, that establishes at least a minimum number of Zicam users at 2 million. And the FDA is suspending sales of the product based on 199 of those users. Assuming a 100% causal relationship, that means the chances of a Zycam user experiencing anosmia as a result is, at best about 1:10,000.

What the FDA, nor the media, didn't bother to tell us is how many cases of anosmia are found in the general population from all causes. What if we were to discover that the general incidence rate for anosmia was no more than 1:10,000? Or significantly lower? Or significantly higher?

The point is we don't know. Maybe Zicam really is causing a problem. Maybe it isn't. I'm not here to debate the efficacy or risks of Zicam, but to take our media and our government to task for failing to report the entire story, one with the information necessary for the reader to draw their own conclusions - conclusions based on fact and data, not hysteria and innuendo. The omissions are conspicuous, driven by a variety of motivations - some financial, some political.

We're going to explore this topic further next week, with a discussion on how pharmacetical companies (and political pundits) neglect to explain the difference between changes in absolute risk versus relative risk, and leverage that ignorance into a multibillion dollar industry that plays no small part of our current health care problem....

No comments: