I think the clinical folks would take over at that point and research a common database of trial results for any relevance to potassium levels, age, and any factors which would have been input.
At that point, they may use anecdotal reports regarding pets and not input information.
They might not be able to label every criterion as meaningful, but would you agree it's a good place to start?
Otherwise, it's rather like looking for a black cat, in your basement, with no windows, and a power outage, at midnight, imo.
News 
Market Data 
Discover 
