Hi! I'm trying to do an meta-analysis and i have a few questions!

The subject is depression, where i've included studies using two different scales (hamilton and MADRS). What i would like to do, is to calculate the effect of exercise vs treatment as usual, based on the change in the depression rating scales.

I've plotted in the final scores mean (SD) of each study into RevMan. But, two of the studies report the mean change from baseline to posttreatment + standard error.

I've followed the cochrane handbook to use the confidence interval / standard error to calculate the SD. So now i have the two studies with mean change + SD.

My question is. According to cochrane handbook; "final value and change scores should not be combined together as standardized mean differences "

How should i include these two studies in my meta-analysis, when 5 of the studies report final score, mean SD?