Competing risks with uncertain endpoints

[ahdvnd]

Hi there,

I was wondering if you would like to collaborate in adding functionality that includes uncertain endpoints in the competing risks framework. I'm working on a method paper and would appreciate some advice on that.

Best,
Aboozar Hadavand

[raredd]

Hi, Aboozar, that does sound interesting although I am not sure what is meant by uncertain endpoints--for example, an unknown number of competing events? or states which are unknown to be absorbing states? Feel free to comment here or use my .edu address

[ahdvnd]

Hi,
Thanks for getting back so quickly. I couldn't find your .edu email so I'll just answer here. What I mean by an uncertain endpoint is that when you observe an event but not with full confidence. For example, if the event of interest is manic episodes, the doctor may only observe the manic episode with probabilities (not full confidence). In this case, we have uncertain endpoints. The states can be absorbing or non-absorbing. I have written down the formulation (in terms of partial likelihood) so it should be straightforward to implement it.