Cox regression modelling describes the first time to an event, but later events are discarded. I need a model to describe the impact of recurrent events on disease progression. Preferably applicable in SPSS.
I suppose multiple regression analysis -multiple risk factors, Odds Ratio, etc.
@ Marianne Levon Shahsuvaryan thanks for your respons.
But multivariate analysis is for multiple risk factors, but I want to know what the impact is of recurrent events by the same factor, namely virus infections on the disease progression
There is a class of multistate models.
For example, look at book:
Competing Risks and Multistate Models with R
Authors: Beyersmann, Jan, Allignol, Arthur, Schumacher, Martin
In R there are some packages dealing with multistate models.
The answers is yes. You are correct that the Cox model was designed to model the occurrence of a single event. To solve this you can use
1.The Andersen–Gill model which basically extends the Cox model
2.The Prentice, Williams and Patterson (conditional risk set) model
3.The Wei, Lin, and Weissfeld (marginal risk set) model, and
4. Multi-state models.
I have done a review on this, see one of the papers on my pr-prints titled;
Modelling and forecasting recurrent recovery events on consumer loans
@Richard Chamboko, thanks !
But I need a model for SPSS, as far as I know PWP model is not applicable in SPSS.
In further search, do you think linear mixed modelling is suitable?
To my knowledge, outside what i shared there are no other methods to deal with that.
Competing risk analysis and multistate models should be appropriate.
To my knowledge, you can run competing risk with SAS and R. You can run Multistate models with R (mstate package).
Have a look on SPSS Help inside the survival analysis procedures for Competing Risk.
I suggest these readings
Competing Risks and Multistate Models with R
SAS Online Help for SAS
I agree with Francesca, you could make a multistate model in which each recurrence is represented by a new state (healthy, first episode, second episode, etc).
I have done something similar without any prior experience in R. I managed to do it with the mstate package combined with the book by Mills called Introducing Survival and Event History Analysis.
Good luck!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling