A colleague recently sent me the following headline from Yahoo! Health:
http://health.yahoo.net/news/s/hsn/your-gut-bacteria-may-predict-your-obesity-risk (do me a favor and wait on clicking that link until you've read the rest of this post…)
Now, he was just sharing information – he's interested in this field – but I'm teaching evidence-based medicine to the medical students this month, and my skepticism levels are turned to 11 (from their usual level of 9…it's probably not safe to have these levels pegged all the time…).
concepts that apply here….
information for their ability to weed out rat studies, conjecture,
disease-oriented evidence and just plain invalid evidence. If I'm kind to
"Yahoo Health" in this respect, I would say that their mission in reporting health information
is very different from mine in reading health information. And my grandmother always
said, if you have nothing kind to say, it's best not to say anything. Frankly, I barely have enough time to write this blog post, never mind wonder about the validity and findings of a study that's very unlikely to change my practice any time soon. So, I'll wait for a headline that says something like: Oral neomycin (or probiotics or whatever) reduces mortality in obese patients. I won't feel bad about not knowing this preliminary research.
Health for medical information, two words in this headline stand out. "Predict" means very little to me
– it's often the result of a prognostic or harm study. Hardly level one
evidence to base any sort of therapy on. "May" turns me right off. It is a word used by researchers and medical writers as a kind of code that says either "Our results are preliminary, and should not be accepted as the final word on the issue" or "We ALMOST got a p-value of 0.05". The combination of "may predict", I'm afraid, results in me not clicking this link. I am NOT calling researchers weasels…just to be clear…but as critical readers, we should watch for these words and avoid interpreting them as more than they are.