I stopped reading at this point--they're not counting preventable, they're counting cases where something *might* have been able to be done. That doesn't mean success and it means their 80% figure is an attempt to make the data fit what they want, not legitimate research.