Cox regression with time dependent continuous covariate with long followup

[Tiny_Kane]

Hi there,

I am trying to fit a cox regression with time dependent covariate. I am pretty new to this method. I looked related paper and the SAS PHREG guide, however none of them fit my case. the measurement X was repeatedly taken and it is time dependent. I want to estimate the effect of X on the death. However, the follow-up for this study is very long, it can be up to 10 years. The unit of the time to death is days. I cannot use any array function when preparing the dataset  for the PHREG. The array is always out of range. Does any body have any recommendation? Thank you very much in advance. 

Data example. 

ID	X	X_date	death	time to death
1	23	1/1/2003	1	1880
1	15	2/1/2007	1	1880
1	27	…	1	1880
1	28	…	1	1880
1	20	…	1	1880
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
2	…	…	0	2990
…	…	…	…	…

[SteveDenham]

How large are the arrays?  If you are getting "out of range" errors, test your code in a DATA step, and if it still occurs, share the pertinent section of the log.  I don't know what the maximum dimension of an array can be, but I don't think it is less than 3650 (ten years counting by day).  Therefore, I suspect that there is something unusual in your situation that is causing the out of range error.

SteveDenham.

[Tiny_Kane]

Please see the error log below. Thank you. 

[SteveDenham]

Here is a link to someone who had a very similar problem, and the answers they received:

https://communities.sas.com/t5/SAS-Procedures/ERROR-Array-subscript-out-of-range-at-line/td-p/478927 

SteveDenham