The advice I give my students is to organize the Results section from this kind of comparative work by describing the themes that the groups had in common, and then move on to the ways in which they differed. To the extent that there is meaningful diversity in the important themes, the goal is to account for the sources of those differences.

Your comparison should focus on quality rather than quantity, I don't think the same number of themes comparison would do the job. As David rightly said, identify important themes that could address your research questions and/or purpose of your study and conduct your analysis along that. I hope this helps. All the best!

Have you looked at Charmaz's guidance on constant comparative analysis in grounded theory? Although you might not be using grounded theory methods, the kinds of questions offered by Charmaz that you could put to your data might be useful. 

I completely agree with David and Adekunle.  If you just need to indicate the number of topics covered in common, doing a coding task, perhaps your work relates more to content analysis than to discourse analysis. In this last case, I think the key would be to observe the diffrent dimensions and views each group presents (every group formed according to certain distinctive features of interest)  by treating the same theme. Good luck! :)

The use of matrices could come in handy. This would show you in which group these themes are more pronounced than others or was more reported than others (qualitatively). Where more dominant then you could follow up to find out why and where less dominant you would find out why not. You could read Miles and Huberman (2002)

Marianna, You've already gotten excellent advice from others here, but I'll offer one maybe more pragmatic thought.  As I read this, I found myself asking, are you doing this for an academic assignment, or for 'real life' i.e. policy analysis context? In translating myself ;'] from graduate school (don't ask when!) to consulting and then Government, I've found that sometimes part of the deal is figuring out, and then reminding yourself, what really you are trying to accomplish, grounding yourself as well as your study. So, in this case, is your interest more about demonstrating your technical skills (though of course that still extends to more than counting) or more, seeking out information that may inform policies for the field, or for your organization as they and you define priorities. I'm a big fan by the way of looking for unintended consequences and comments that reveal aspects of (whatever) that we'd never glean sitting in our respective towers.

And yes, don't get so much hung up on numbers as patterns, differences (and similarities) with consideration of why woven in.  I hope this helps a little...Not a morning person...