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.
Share your take
Add a clinician perspective alongside expert commentary.
Cuijpers et al. (2016) studied this question.