Identifying dataset dependencies - which fields / measures are used where

If a dataset changes, online reports can break

Is there a way to see which fields and measures from each online dataset are used, and in which reports (and visuals)? Since everything is online and organised in a structured way, all the information is available within the platform to do this.

...or is it not possible, and everyone else also goes through the regular routine of "what broke" and quickly fixing broken reports - it's a very unreliable way of pushing out "improvements" so I have the impression I'm missing something obvious

How do people in mid to larged sized enterprises update datasets without causing multiple reports to break?