Comments on "parachutes"

I'm not particularly amused by this paper; I think that some of the rapid responses point out the unfairness of the comparison. Robert Temple is spot on, in my opinion, when he writes:

"Smith and Pell are not confronting a genuine problem and I am concerned that their teasing could undermine appropriate attention to study design. Evidence-based medicine is never fully secure from attack by anecdote-based medicine. Their point is, of course, that the effectiveness of some interventions is so obvious that even purists should recognize that no randomized trial is necessary. To make that point, they utilize a straw man: the putative insistence on a controlled trial to evaluate the obvious effectiveness of parachutes. But what seems obvious is not always true and the doubters of the “obvious” are not necessarily hide-bound, rigid, or silly."

I have a particular bee in my bonnet about this because of the issue of cycle helmets. It seems obvious to many people that cycle helmets must have a protective effect, and there is a profusion of anecdotal evidence of the "helmet saved my life" variety. This leads well-meaning people like the BMC to call for cyclists to be compelled to wear helmets, and punished if they don't. All this when the actual evidence is far from convincing and often points the other way.

I understand (and am inclined to agree about cycle helments; I don't wear one myself). I still thought it was funny; I hadn't thought it was really intended to imply that doubters-of-the-obvious are necessarily hide-bound, rigid or silly. I suppose my own confidence that an evidence-based approach is best may make me less sensitive than I should be to the danger of encouraging the partisans of irrationalism.