Risk / 33,279,000

The Star (1 June 2008) - Dr. Paul Dorian, a cardiologist at St. Michael's Hospital and the author of the most recent Taser study, would never say it's impossible to meet a deadly fate at the end of a taser. But it's rare. According to his research, an analysis of the existing literature coupled with his own observations and study, the odds of a "death by taser" are the same as dying from, say, living next to power lines, getting breast implants, or drinking from a bisphenol-leaching plastic water bottle. [LINK]

Nice sound-bite, but meaningless. It seems as if Dr. Dorian is denominator washing to the highest possible degree. He seems to be including the entire population as the denominator. It also sounds a bit like Kroll's "Safer than Tylenol" stupidity.

What any reasonable and informed person would be interested in is the worst case death rate from a full-on X26 taser to the chest with the barb(s) in the worst possible location. This is the crux of the argument. Taser and Kroll claim that the tasers are perfectly safe (cardiac-wise) even in the worst possible case. Most critics would disagree (based on the real world news).

Once we can agree on the worst case (even within the same order of magnitude), then we can subsequently and explicitly adjust for external factors such as barb placement, misfiring, missing the target completely, all the way to ...not actually ever being involved with the police in the first place (in case you want to include the entire population in your denominator).

I've already made a rough estimate of the risk-of-interest.

2007 British Columbia: 496 taser incidents. Perhaps 25 full-on X26 taserings across the chest. One or two deaths. Result: 1 or 2 deaths divided by roughly 25 full-on X26 deployments = about a 6% death rate. Might be 1% (maybe). Might be 10% (maybe). Might be a bit higher. Might be a bit lower.

See previous post titled: Death rate estimated - feel free to provide rebuttal [LINK]

If Dr. Dorian has anything to back-up his outlandish claims then I'd like to see it.

Beware denominator washing (a topic of many previous posts including [LINK]).