Just enter your email
I think if you're looking at attitudes and beliefs "grounded theory" tools (http://www.psych-it.com.au/Psychlopedia/article.asp?id=150) integrates theories and overall/common attitudes and beliefs much better than a systematic review/meta-analysis.
A systematic review/meta-analysis based on interviews, qualitative transcripts.. probably won't be statistically valid and will mess up validity/reliability. There wouldn't be much basis for bias appraisals with so many different complete and partial evidence included in the review. Also extracting and standardising diverse variables might be extraordinarily frustrating.
It sounds like lots of "sub-groups" rather than heterogeneity. Too many sub-groups increases false negatives/positives findings. Might be good from onset to determine a definite research question with parameters and a definite quantitative analysis plan. That way those specific conditions can be included in the search. I still think grounded theory tools is more satisfying when investigating attitudes and beliefs.