If the meta-analysis didn't actually look at the original data to validate the claims of the original papers, what good is it? Studies should be assumed false

If the meta-analysis didn't actually look at the original data to validate the claims of the original papers, what good is it? Studies should be assumed false until proven otherwise.