Does this unnamed monitoring tool (and undefined query) really think higher percentages are worse or are you mis-communicating?
Forget about the tool used for monitoring, important is to monitor the "Cache hit ratio" is good or we can ignore it?
The query at least would help, since it seems to be measuring something other that "cache hit ratio".
I'm not in a position to judge whether monitoring "cache hit ratio" is something you should be doing in your situation. It provides information and is fairly cheap to capture and store. From the sounds of it you should probably continue capturing the data but turn off the alert. That way if there is a problem the data still exists to be considered. The metric itself does not measure something of critical importance.