Risk of premature birth after appendectomy compared to reference population

[rollanders]

I have data om 5.000.000 pregnancies of which 7.000 had appendectomy during the pregnancy. I want to analyse the impact of appendectomy on the risk of premature birth and stillbirth (among other things). The impact of appendectomy is strongly related to the length of pregnancy when it happens. I have date of start of pregnancy, date of appendectomy and date of birth plus a some covariates. I need help what method to use to analyse the impact of appendectomy. I have been thinking of stsplit as an alternative, splitting the dataset on the date of appendectomy, and analyse the risk of birth pre/post appendectomy. I have also thought of using a matched dataset matching on the start of pregnancy. May I also split the followuptime for the matched (nonoerated) pregnancies to get a more easily communincable illustratation of the impact on appendectomy on the risk of prematuiurity?