Scientists don't make snide remarks about "convincing" people, they simply publish their data, and are able to substantiate it to a reasonable degree. You have my data. I would like to see yours, so I can reach objective conclusions. My team are not salesmen, we are engineers and scientists. We work with facts and numbers.

Scientific studies should be objective. You state here your result and ask for data to "prove it". You mention that your result is different from mine, so you must have a different number for either:

- The number of rebreather accidents (and you have a list of 162 I sent you, after your scrutiny just two were uncertain and there was also two unallocated that were very likely the rebreather in question). So an error that might be a few percent in mortality numbers in your check.

or
- Sales figures. You have my sales figures, which are within 10% of the figure the manufacturers have stated publicly in almost every case.

or
- You are blending the figures from different types of rebreather for your study (eCCRs, mCCRs and SCRs have very different mortality figures).

or a combination.

Your 23/10,000 figure gives a mortality rate per unit for a rebreather life of 7 years, of 160/10,000 (i.e. a probability of surviving of (1-23/10000)^7). This is one in 62.

Take for example, the biggest selling sports eCCR. The sales figures should not have to be debated much because the MD stated them on oath ("few thousand", which I take to be 3,000 and you said on NBD that you understand to mean 3,000 to 6,000 - and our number for the sales was 3,300 and now is 2,700 - pretty close to the MD's statement). The accidents are known within about one or two percent after your own review of these. How come your number is out by a factor of 2 when you have such accurate source data? I think I know ...

I suspect the reason our numbers "differ", is that our 1 in 39 mortality figure per unit sold is for a specific eCCR. Our mortality number per unit on mCCRs is about 1 in 900. Our mortality number per unit on SCRs is about 1 in 600. By blending these data sources our team assembled, of course you will end up with a figure of around 1 in 60 or 70. Which is pretty close to your 1 in 62 number. So why are you trying to denigrate me publicly with this falsehood you are circulating? So you can hide what you know to be a high eCCR accident rate, by blending it with the mCCR and SCR figures? As your conclusion seems no different from ours, if you take all rebreathers, what are you trying to prove by the statement I quote?

Finally, the problem with your study is that it is not scientific. You need to fix the population from which you take your samples, and the samples need to be random. If you want divers' health data, you would need to offer free medicals or something. You are a GP who does diving medicals, so you should get a good cross section of divers. That is, good samples. If you ask them if you can use their data for research purposes, you will get a decent body of data, instead of appealing for those who have health problems on an internet forum.

It would be good to have answers for the diver health questions you raise, but the data has to be scientific otherwise it is worthless. That means either comprehensive research to get at the numbers, or in this case using a random selection of the population.

Alex

NB: I had a look around your web site. Seems to have taken over from NBD. Could you put a link to the data please on which your figure is based? I have my data fully in the open.