test to assess proportional difference over time between groups

I am trying to determine the appropriate test to run for a community-based intervention. There are 2 groups (intervention and control) and 2 time points - pre & post (many, but not all, of the same people will be surveyed at the 2 time points). The dependent variable is dichotomous (in remission y/n). The proportion in remission is significantly different at the pre-test between the two groups, so I am interested in testing the hypothesis that the increase in proportion of patients in remission will be greater in the intervention group than in the control group from the pre to the post test. What would be the correct test to use? Thanks!


Omega Contributor
Different options, perhaps cochrane mantel haenzel. Need to examine your loss to followup to ensure there was no systematic bias and they are missing at random.