I have to perform a survival analysis using cox model with time varying features.
-
Thromboembolic Event as the Outcome: I am interested in studying whether pregnant women can develop a thromboembolic disease. The follow-up duration starts at the beginning of pregnancy and ends either when the woman experiences the thromboembolic event or 6 months after childbirth.
-
Explanatory Variables:
- Taking Anticoagulant Medication: based time event.
- Type of Birth (Cesarean or Natural): This is where the challenge lies.
Handling Type of Birth: for women who experience the thromboembolic event before childbirth, we don’t know whether they would have had a cesarean or natural birth. This creates a dilemma because we can’t directly include it as a time-fixed variable (since it changes after childbirth) nor as a time-dependent variable (since we lack information for some cases).
Can anyone please help fix this problem? how can I handle this type of variable ?
what I tried:
- Use type of birth cesarean, and type of birth natural as 2 based events features.
todd chavez is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.