[istultus]

Ugh, we're going through the cycle of the author of the book going around spreading his little tidbits in all the main sites/podcasts (I heard him on Econtalk).

For us rest-of-the-worlders - note that Peanut allergy is not an invention - there is a small minority of people with peanut allergies, even after exposure to peanuts as infants. However, there is a larger subset of people who developed peanut allergy in the US through full avoidance of peanuts, so that when later in life they came into contact with them, it started an IgE-mediated cascade (according to the model, on the second exposure to peanuts) that could've been avoided had their body been habituated to peanuts earlier in life.

The real thing to learn here is how suggestions become dogma in human society. "We don't know what's causing allergy, maybe just avoid it?" becomes "We [some formal group of pediatric allergologists] suggest to avoid peanuts" becomes "Our guideline is to avoid peanuts".

The point of Evidence-Based Medicine is exactly to avoid this, but note that this is the standard affairs in anything else in life. Iterative improvement is built on doing something until it stops working/starts harming and then doing something else. It's just magnified where human life is concerned - Military, Aviation, Medicine, etc.

Just like in Aviation - the best solution is process - to admit mistakes and force changes and improvements without knee-jerk opprobrium and litigation (here it's obvious that no medical association should suggest something extreme if there's no evidence for it one way or the other, duh) - but creating sensationalist headlines demonizing a group just makes sure that future mistakes will continue to be covered up.