Try to go on some kind of short course, it will be well worth the time invested. Meta analyses are incredibly complex (as you are finding out!) and there are many potential pitfalls. Find yourself a good statistician to advise you, your Uni should have a source of stats support for PhD students. Good luck.