my opinion about margarine

I do not disagree, but the confounding factors in a comparison study such as you describe are complex as hell. There is a crazy old study that suggests bald guys are immune to heart attacks, and you know that's gotta be nuts.

Among other hats I wear around here is radiation safety officer. In the Health physics world there are lots of strange studies. One by Dr. Bernie Cohen found a lower level of lung cancer in the counties with the highest radon levels all over the US, and that was adjusted for other confounding factors like the percentage of smokers in each county. Another involves an apartment complex in Taiwan where the steel in the framework is badly contaminated with Cs-137 (I think that was the isotope but it might be Cobalt), and the residents have a much lower cancer rate than the general population.

It is clear that we do not have a better way to evaluate drugs and treatments than trials and studies. Evidence based medicine rocks! However, one must be cautious about promulgating more sweeping public policy based on such studies. The percentage of a percentage scam, especially where the underlying absolute numbers barely rise to clinical significance, is especially dangerous (if folks taking a drug have 1 heart attack per thousand, but folks not taking it have 2 per 1,000, then that drug has a miraculous 50% improvement claim to advertise).

Syd Posted by: "AnaLog Services, Inc."