> The statistics don't support the idea that people disable their smoke detectors permanently over nuisance alarms.

So when people die, it's rarely ever the case that the alarms had been disabled for some reason?