Hello!
I have a question regarding the targetSubjects count and comparatorSubjects count on the Shiny Dashboard. From the code, I understand that the shinyDashboard gets their data from getMainResults(), which in turn gets it’s information from cohortMethodResult.
cohortMethodResult shows that for the right targetId - comparatorId - outcomeId - analysisId > targetSubjects is 47,240. This is correct with the shinyDashboard.
However, in the cmOutput folder, there is another file which I thought was the stratified population after Ps matching (StratPop_l1_s1_p1_tXXXX_cXXXX_s1_oXXXX) for the same targetId - comparatorId - outcomeId as above. If I look at the amount of people using the below code:
StratPop_l1_s1_p1_tXXXX_cXXXX_s1_oXXXX %>% group_by(treatment) %>% count()
I get a different amount of people in the target and comparator cohort than in the shiny dashboard.
treatment n
1 0 47263
2 1 47263
It makes sense that they are the same, as the groups have been matched with 1:1 matching, but I can’t find what other analysis gets done after the stratPop is created and why the targetSubjects are suddenly different from the shinyDashboard. I can see these differences for a variety of analyses and whilst they are very small (13 patients here), I wonder where and why they got filtered out? It doesn’t show in the attrition…
@schuemie @Patrick_Ryan or anyone else?
Thanks in advance!