I haven't said much about it, but if you've worked with radiation counters, you know that there are lots of ways they can go wrong. Get a particle or drop of something on the probe. Spikes (the up-and-down kind, not the current misuse of that word) from electrical transients or cosmic rays. High or low humidity. In my experience, the errors tend toward false positives rather than false negatives. Alpha counters are particularly prone to things going wrong, but they should be less used than beta-gamma counters at the Fukushima plant right now.
Today there's a report of a measurement that the plant workers want to re-check. Always a good idea, whether it's an unexpectedly high measurement or unexpectedly low.
Program note: I'll be mostly away from the internets for the next week. There'll be another post this afternoon on how radiation exposure standards are derived that you won't want to miss. Then there will be a few more at BMJ that overlap with what I've posted here already. I'm not ignoring you if I don't respond to comments. I'll try to tweet.