well, to the credit of Bayes, dementia is likely a safe choice (depending on age/etc.) but dementia is largely a diagnosis of exclusion and most doctors, besides being unfamiliar with Bayes, are also just plain lazy and/or dumb and shouldn't immediately jump to the most likely explanation when it's one with the worst prognosis and fewest treatments...

I work in biomed. Every textbook on epidemiology or medical statistics that I've picked up has had a section on Bayes, so I'm not inclined to believe this.