Key points are not available for this paper at this time.
We conclude that pre-post SMDs should be avoided in meta-analyses as using them probably results in biased outcomes.
Cuijpers et al. (Fri,) studied this question.