st: Survival analysis: Time interactions for hazards that cross multipletimes

[Maria Khan <maria_khan@unc.edu>]

Hi,

In survival analysis we may need to create an interaction with 
continuous time or categorical time. If we see a continual increase in 
the HR for men versus the referent women, it would suffice to create an 
interaction with continuous time with our gender variable ("gender"). To 
create an interaction with continuous time, the code would be:
stcox gender, efron nohr tvc(gender)

Say we must create an interaction with categorical time in the case that 
there is a non-monotonic departure from proportionality of the hazards 
of the two strata of the variable of interest. Say before 96 months of 
follow-up, we observe one HR for males v females, and after time 96 
months, we observe a drastically different HR for males v females. The 
STATA code would then be: stcox gender, tvc(gender) texp(survtime>96) nohr

But say the HR changes three times. For example, at 96 months and again 
at 120months. There would be three different HRs for the effect of 
gender, depending on the time period.

Does anyone know code for this??

Thanks-
Maria Khan

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/