calibration plots after competing risk regression


Less is more. Stay pure. Stay poor.
Not off hand but I have a competing risk model I need to run in a couple of weeks, so would be interested.

Do you believe there is a difference in selecting a calibration method in competing risks? Also, also when you say calibration are you thinking something like the hosmer-lemeshow?
Last edited:
Thank you for your response. I am thinking about something similar to observed /predicted curves based on Cox regression and KM. I try to recreate this based on competing risk approach but no way to do it in Stata to my knowledge.
I believe there is a difference in calibration methods since they need to manage an additional set of data those with competing risk.


Less is more. Stay pure. Stay poor.
I will be getting two datasets in the next 2 weeks (both time to event), one which may have competing events. But much like the quote, "once you know multilevel models everything seems like a multilevel model, competing risks seems comparable, that many things should be modeled with competing events. I only run Cox about every couple of years, so over the next month I will be freshing up on the topic and will keep you apprised if I find anything.

I will also, likely post general links into this thread, so in the future I will have an archive of references, so I don't have to look stuff up again. I may also post in questions I may have on Cox's reg.