A friend of mine told a story about the way statistics can
be used to “prove” anything. He was in a class, and his professor took them
through a step-by-step statistical analysis that led to the conclusion that
everyone should brush their teeth with sugar.

A few recent articles have led me to throw up my hands when
it comes to medical research and empirical data.

First, there was the Atlantic article that mentioned how
subjective and error-ridden many medical studies really are: “Lies, Damned
Lies, and Medical Science.”

Then another Atlantic piece–“Playing Doctor: How to Spin Pharmaceutical Research“– that describes how those medical
studies get written for medical journals. Turns out that ghostwriters for the
pharmaceutical companies often write the articles that then are “signed” by a
doctor. Who wants the makers of Vioxx to tell me how great Vioxx is?

Finally, the new research that suggests pregnant women
drinking alcohol–in moderation–isn’t really detrimental to the health of the
fetus. In fact, the article I read about this mentioned a study that suggested
smoking crack had little impact on the health of a fetus. I’m worried about
eating goat cheese and I could be smoking crack? (Obviously there are plenty of
other reasons not to smoke crack, but still…)

Of course, perhaps the it’s-okay-to-drink-while-pregnant study
was ghostwritten by Julio-Gallo. Still, I think I’ll have a glass (yes, just one) of champagne
on New Year’s Eve.