08-01-2011 05:35 PM
I have a collegue who is insisting that we implement Lin's regression (Biostatistics (2000), 1, 1, pp. 35–47) for modeling costs with censored data. I've not heard of this before, nor have I been able to find applications that actually have coded this. Anyone been able to get this to work in SAS? It should be similar to PHREG I would think....
Thanks for your help!
08-01-2011 05:49 PM
I'm not familiar with it, but it may be amongst the macros provided at:
Also, a macro for obtaining Lin's correlation coefficient is provided in the mccc macro at:
08-02-2011 01:56 PM
Wrong Lin. Sue is referring to DY Lin (not X Lin). I missed it too the first time.
Here is the paper
and a nice summary
Lin's Regression is parametric, so I is probably not a variation on PHREG. However, I have not found any SAS software for it. There are several follow-on articles as well. You may want to contact DY directly, though my guess is that his software will be in R.
08-03-2011 09:39 AM
Thanks. I hadn't seen anything on how to do it in SAS either but wasn't sure if anyone else had any experience using it. R would be fine (since it's open source) so I'll contact him directly and see what else I can gather. I have to admit tho, it doesn't leave me with a "warm-fuzzy" feeling when I can't find anything on how to program the method. Almost makes me hesitant to use it at all...but alas, that's not necessarily my call to make.
Thanks for the info!