Re: st: RE: Cox model with a segmented time-dependent covariate

[Marta Garcia-Granero <mgarciagranero@gmail.com>]

Wow!

Thanks a lot Kieran. Now I have to peruse the commands carefully until I 
understand completely how stsplit works.

So much to learn still after one year using Stata...

Regards,
Marta

El 15/01/2014 23:32, Kieran McCaul escribió:
> input  patient time status treatment age lbr0 t1 lbr1 t2 lbr2
>    1    281   1   1   46    3.2    47 3.8   184    4.9
>    2    604   0   1   57    3.1    94 2.9   187    3.1
>    3    457   1   1   56    2.2    61 2.8    97    2.9
>    4    384   1   1   65    3.9    92 4.7   194    4.9
>    5    341   0   1   73    2.8    87 2.6   192    2.9
>    6    842   1   1   64    2.4    94 2.3   197    2.8
>    7   1514   1   0   69    2.4    74 2.9   202      3
>    8    182   0   0   62    2.4    90 2.5   182    2.9
>    9   1121   1   0   71    2.5   101 2.5   410    2.7
>   10   1411   0   0   69    2.3   182 2.2   847    2.8
>   11    814   1   0   77    3.8   167 3.9   498    4.3
>   12   1071   1   0   58    3.1   108 2.8   187    3.4
> end
>
> gen lbr = lbr0
>
> stset time, f(status) id(patient)
>
> stsplit one, at(1) after(time=t1)
> stsplit two, at(1) after(time=t2)
>
> replace lbr = lbr1 if one==1
> replace lbr = lbr2 if two==1
>
> list patient time status lbr0 lbr1 lbr2 lbr _st- _t0 , sepby(patient)
>
> stcox lbr treatment age

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/