There are some simple tricks that allow you to measure correlation between separate sets of evidence.

This you can then adjust for.