Could you please tell in what situation the maentl hanszel test is more appropriate than using chi square. I believe maentl hanszel is used when a confounding variable is present thus allowing for adjustment. But when the confounding variable is taken into account after randomization of patients how is stratification performed?