It would be easy, too. Find all studies on the subject, find correlations between some parameters and the results you want/don’t want, add those parameters in your inclusion/exclusion criteria.

Who even replicates review articles?