How can we deal with differences between the intervention and comparison groups before the intervention in a non-equivalent control group quasi-experimental design?

You are welcome 

Thank you.

Generally speaking, if the design includes the pretest and posttest, the difference between the intervention and the comparison groups is not a major threat to the validity. However, if only the posttest is included, we should minimize the difference by statistical methods (e.g., balancing or matching) as much as possible.