Geoff: "Once you demonstrate a link, you must next show it is causal." No, you actually don't. You just need to show the closeness and possible relationship of the link. Demonstrating causality is difficult, and desirable, but not a necessary condition for public health policy.

"Typically you must rule out alternative causes and a dose-response is useful. e.g., that people with the highest exposure to the suspect substance had the highest level of problem."

Err, that is what Wertelecki is doing in the Ukraine, is it not?

"People need to do the tough science of working out what is going on at Fallujah, and pretending to know the answers when you have absolutely zero evidence is a horrible thing to do"

Actually, "people" don't need to do any such thing, scientists need to do the tough science. If the scientists working on the problem were pretending that radiation was the only cause, then yes, I would agree with you, but I see no evidence that they are. So you can stick that little piece of fake indignation back in your bottle. "People" have every right to discuss the possibilities and public health implications, and it's not a horrible thing to do. OTOH, is pretending that that radioactive particulates do not cause any birth defects, or that the probability is so low, we can effectively ignore it or not discuss it is not a horrible thing to do?