Not necessarily, it reduces false positives. It just doesn’t do anything for false negatives (arguably makes the problem worse).

If you just want to see if there is incidence of valid differences, this seems fine. But I wouldn’t use it as a guarantee.