The approach is analogous to a MA with two arms. You've to calculate an effect size - in your case a raw mean difference or SMD between pre- and post. However, please note that pre- and post- values are correlated and you have to include this in calculation of the estimate variance/SE (Example: http://handbook-5-1.cochrane.org/chapter_16/16_4_6_4_example.htm). But if you have access to raw data,then both mean difference and its variance can be calculated from it.