Seeking Help with GEE Model

[zulfiqar ali shah]

Hello every one. I am working on analysis of pooled survey data for my project. The hypothesis that I am interested in is exploring is assessing whether the COVID pandemic and related restrictions affected modern contraceptive use among married women of reproductive age in Ethiopia. It is suggested that I use poison link if used as a proportion at cluster level, the cluster level here would be enumeration area (EA_ID). The goal is show whether the trends (slopes) varied between pre-post period. It is suggested to use interaction term as well with year X pre-post and check whether the slopes varied. I need guidance whether in such GEE is the right model with family link poison. The dependent variable here is MCP which is binary while independent variable is year.
. dataex year EA_ID household marital_status mcp

----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. To install: ssc install dataex
clear
input float year long EA_ID int household long marital_status float mcp
2015 2260   1 1 1
2015 2294   3 5 1
2015 2308   1 1 1
2015 2738  55 1 0
2015 2759   2 1 1
2015 2915   8 2 0
2015 2085   1 5 0
2015 2373   6 1 1
2015 2248   2 3 0
2015 2231   1 1 0
2015 2013   1 1 1
2015 2215   2 5 0
2015 2853   2 3 0
2015 2853   1 5 0
2015 2855   1 1 0
2015 2797   1 2 0
2015 2963   1 5 0
2015 2276   1 5 0
2015 2798   1 5 0
2015 2387   1 1 0
2015 2799   1 5 0
2015 2894 145 5 0
2015 2837  98 1 0
2015 2853   1 5 0
2015 2853   1 1 1
2015 2294   8 1 0
2015 2271   3 3 1
2015 2175   1 1 0
2015 2563   2 1 1
2015 2433   1 5 0
2015 2954   1 1 1
2015 2391   1 1 0
2015 2097   1 5 0
2015 2896   1 1 1
2015 2216   1 4 1
2015 2091   1 1 0
2015 2002   1 3 0
2015 2215   1 1 1
2015 2522   1 1 0
2015 2659   1 5 0
2015 2008   1 1 0
2015 2119   1 1 1
2015 2882   2 3 1
2015 2620   1 1 1
2015 2756   4 1 0
2015 2853   1 1 0
2015 2295   2 1 1
2015 2896   6 2 1
2015 2539   1 1 0
2015 2557   1 1 0
2015 2615   1 1 0
2015 2879   1 1 0
2015 2225   1 1 0
2015 2808   1 1 0
2015 2538   8 1 0
2015 2051   3 5 0
2015 2003   8 1 1
2015 2429   1 5 0
2015 2570   1 1 0
2015 2299   1 1 1
2015 2385   1 1 0
2015 2442   4 1 0
2015 2119   1 1 1
2015 2076   1 5 0
2015 2543   1 1 0
2015 2601   1 1 0
2015 2557   1 1 0
2015 2541   1 4 0
2015 2648   1 2 1
2015 2559   2 5 0
2015 2879   1 1 0
2015 2801   2 5 0
2015 2841   1 5 0
2015 2508   1 1 0
2015 2344   1 1 1
2015 2994   1 5 0
2015 2894 112 5 0
2015 2799   1 4 0
2015 2716   1 5 0
2015 2491   1 5 0
2015 2595   1 5 0
2015 2673   1 1 0
2015 2283   1 1 0
2015 2645   1 4 0
2015 2907   2 1 0
2015 2673   1 1 1
2015 2125   1 1 0
2015 2308   1 1 0
2015 2630   1 1 0
2015 2756   1 4 0
2015 2661   1 5 0
2015 2231   5 5 0
2015 2153   2 1 1
2015 2048   1 1 1
2015 2113   1 1 1
2015 2080   1 5 0
2015 2813   1 1 1
2015 2281   1 5 0
2015 2383   3 5 0
2015 2615   2 1 0
end
label values marital_status marital_status_list
label def marital_status_list 1 "1. Currently married", modify
label def marital_status_list 2 "2. Currently living with partner", modify
label def marital_status_list 3 "3. Divorced or separated", modify
label def marital_status_list 4 "4. Widow or widower", modify
label def marital_status_list 5 "5. Never married", modify
label values mcp yes_no_list
label def yes_no_list 0 "0. No", modify
label def yes_no_list 1 "1. Yes", modify

------------------ copy up to and including the previous line ------------------

Listed 100 out of 74111 observations
Use the count() option to list more

.
I availed many options but am nit satisfied with it. would love to take your guidance.