Greetings,
I am conducting a case-control study to assess the effect of ethnicity (six main ethnic groups) on suicide risk over a long period (from 2001 to end of 2019).
I am using conditional logistic regression to report odds ratio. Now I need to assess if the risk (odds ratio) has changed over the study period across the ethnic groups, but I am unsure of how to perform that- a sensitivity analysis.
My thought is to do the primary analysis, then run the model over different study periods (eg. from 2001-2010, and from 2011-2019), then "compare" odds ratios of the primary analysis with that of the pre-defined study periods.
Is that the correct way? do I have to assess the % of change in odds ratio?