This forum is about wrong numbers in science, politics and the media. It respects good science and good English.
We have a problem too with international databases... do different countries tend to have different standard answers where no autopsy performed?
If the databases were culled to only retain data where there has been an autopsy performed and not simply the attending physicians best guess/standard response there'd be virtually no database left and it would be seriously distorted since autopsies would be for atypical deaths anyway.
Admittedly, where there is a recent medical history and the attending physician is the patients own physician, there is a good chance the autopsy is accurate, or so I would have thought. The data given above is worrying. So too when you know cases where the patients own doctor can't add two and two and come up with four.
For example, our neighbour had a brain tumour, was operated on and returned home. A year later she presented with the same symptoms. We all thought she had had a recurrence. The doctor thought otherwise and she died of it.
What should happen is the Government should require autopsies on a percentage of "normal" deaths in order to build a reliable data base.
Of course, that isn't going to happen so the best we can do is recognise there is a serious problem with the data base and ask that nice Michael Mann (or is it James Hansen? I can't keep my vilains straight)if he can "correct it".
I hate smoking. Smokers stink, spread litter everywhere and set fire to buildings. Smoking ought to be legal everywhere outdoors and in pubs and bars. The fact that people such as me stronmgly dislike it is no reason whatever for banning it.
Yet more propaganda has hit the airwaves this morning. I saw this reported on Sky News, and online at the beeb http://news.bbc.co.uk/1/hi/health/8267523.stm
From the links on this page you can get to the study itself here http://content.onlinejacc.org/cgi/content/abstract/54/14/1249
Unless you join the American College of Cardiology you can only read this Abstract. However even from this we can see that it was a meta-study. The numbers quoted are so far beyond the realms of possibility that it must be lies, but without the full text it is again hard to know exactly how they've done it. However there might be a slight clue in the Abstract itself where it says