-
“Sorry, I thought it was 19 deaths over a 10 year period or 1.9 people per year. Where did 33 come from? If the numbers don't work you can't just double them!”
Simon,
The data set used in Beliaev’s study was of 19 documented deaths over and beyond the norm in only 4 hospitals with trauma services out of 14 such hospitals in 2 of 4 regions of New Zealand.
- 2 of these 4 hospitals (Auckland and Waikato) are advanced trauma service. New Zealand has 5 hospitals with advanced trauma services. Data was collected from only 2 of these 5 hospitals for the Beliaev study.
- 1 of these 4 hospitals (Middlemore) is a district trauma services. New Zealand has 16 hospitals with district trauma services. Data was collected from only 1 of these 16 hospitals for the Beliaev study.
- 1 of these 4 hospitals (Northshore) is a basic trauma service. New Zealand has 8 hospitals with basic trauma services. Data was collected from only 1 of these 4 hospitals for the Beliaev study.
Hence Beliaev’s study never examined documents from any of the other trauma hospitals in New Zealand, including others with advanced trauma services. Most notably this did not include Wellington Hospital in the Central Region though this hospital is an advanced trauma service in a region that is one of 2 most likely to treat major trauma patients. The other region is Northern and 3 of its hospitals were included in Beliaev’s study.
What does this mean?
It means in order to match a New Zealand sample of JWs to the additional deaths (19) in Beliaev’s study an adjustment must be made to correlate population with source (of the data set).
There are at least 2 ways to accomplish this. 1 way is to multiple the total New Zealand population of JWs by the percentage of total New Zealand population living in the 2 regions. A 2nd way is to adjust the hard number (19) by prorating it to match New Zealand’s total population based on population of the 2 regions with hospitals in the study. I used the latter though mathematically either will give the same result. This is explained in my blog article on the subject. (Here: http://marvinshilmer.blogspot.com/2012/02/more-than-50000-dead.html
Making this adjustment we end up with a statistical number of 33 deaths beyond the norm over the period of 1998 to 2007. This adjustment is based on a hard number of documented deaths over and beyond the expected, and not on a mortality rate.
I did not “just double” numbers to make something work and do not understand your insinuation otherwise. To the contrary, I let the numbers speak for themselves.
My recommendation is that you learn the math of this thing and stop with the insults. If you have questions feel free to ask.
Marvin Shilmer