intracluster correlation coefficient after Cox regressin
This post has NOT been accepted by the mailing list yet.
I am a cardiologist with interest in statistics and stata (which makes it all much easier…). I am working with a large hierarchical database (~700 000 observations) and we are interested in different question in regard to prediction of patients outcome e.g. survival.
I would be very grateful if someone could help me with the following question. How to calculate ICC (i.e. intracluster correlation coefficient) for a particular set of data (or model). I have try to google this but was unable to find something appropriate (that I can understand properly). Perhaps I have misunderstood the concept. If we model the survival with e.g. shared frailty Cox-regression (with hospitals as clusters) how do I calculate ICC from the model?