Hello,
I have a retrospective cohort of 145000 subjects. My exposed group has 526 subjects, and I would like to have two unexposed groups by age, sex, and time in follow-up.
The exposed group is followed up from the ending date of the sickness episode until event or end of follow-up.
The first unexposed group should start the follow-up at the same time-point (and have same age and sex). Hence, I want to match them by FREQUENCY MATCHING method. This unexposed group is selected by the criteria of NOT having any sickness episode. However, I have several registers of them in the form of episodes of employment. So my idea was to use the starting date of the employment episode of the unexposed group closest to the ending date of the sickness episode of the exposed to match. This way, I would start the follow up on similar time points and follow them up until event.
The second unexposed group is selected by the criteria of HAVING a sickness episode with a different diagnosis than the exposed group. Same as the first unexposed group the follow-up should start on he same time-point, and they should have same age and sex. In this case, my idea is to match by the ending date of the sickness episode of the exposed with the ending date of the sickness episode of the unexposed.
I have read two discussions about frequency matching on the forums:
https://www.statalist.org/forums/for...age-and-gender
https://www.statalist.org/forums/for...uency-matching
And applied the codes recomended but:
1. I am not quite sure on how to interpret the outputs.
2. It doesn´t give me 1:1 match, the resulting database has half of the exposed.
If anyone could help me with this I would be very grateful.
Thanks!
Related Posts with Frequency Matching in a retrospective cohort: output and matching for follow up time
extract variable labels for new variable namesHi there I try to automate my programming as much as possible and one challenge I've come up agains…
Count number of cases if dates are within a certain range (a la statsby)Greetings all, I have single line per observation survival data (4 million lines). Here is a simpli…
Generate balance tableHi, I'm trying to replicate this balance table (as in the picture) using some of the example dataset…
Multinomial logit with sample selectionDear everyone, I am looking for something similar to Heckman selection model/svysemlog with a modi…
gsem covstructHello. I am running LPA analysis using gsem command. I ve run the analysis in R using the mclust com…
Subscribe to:
Post Comments (Atom)
0 Response to Frequency Matching in a retrospective cohort: output and matching for follow up time
Post a Comment