Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to run survival analysis where the total count of different types of events matter?

    Hi Stata Community,

    I'm currently trying to conduct survival analysis where events are not just the binary variables of 1/0 (i.e., did the event happen or not). However, I'm not quite sure how to proceed in this case. My "events" are integers representing the count of non-mutually exclusive events. For instance, a value would be 3 if three kinds of events occurred (in the context of car insurance data, the three different events would be cracked axle, a smashed rear windshield, and a blown head gasket). It would be 2 if any two of these events happened, 1 if only one occurred, and 0 if none occurred

    Currently, I'm interpreting any integer count greater than 0 as an event occurring (a conventional 1 or 0). However, I would like to capture the nuances between the different integer counts (3 versus 2 versus 1 versus 0). Should I apply a standard Cox Proportional Hazards model separately for each count level? That is to say, define an event as 1 when all three kinds of events occurred, define an event as 1 if more than two kinds events occurred, etc.

    Thank you so much!
Working...
X