CODEWITHSUNDEEP
variablesmodelspsscox
Anthony Cutajar
Anthony Cutajar

Reputation: 21

spss query HRs for time dependent covariate at multiple periods of follow-up

Consider the model: h(t, X) = h0(t) exp [clin2 + prison + dose + clin2gt] where the variables are defined as follows: clin2 is time dependent (categorical coded 1/0); survt follow-up time in days; clin2gt is the clin2*time interaction; Prison is categorical coded 1/0; dose is a continuous variable in mg.

I am interested to do an extended Cox PH model in SPSS (Ver.28) that uses one time-dependent variable and the following time periods: 0 to 183 days, 184 to 365 days, 366 to 548 days, 549 to 730 days, and more than 730 days. 2. Obtain the hazard ratio that corresponds to each time period in the model that you created in Question 1. 3. Calculate confidence intervals for each of the hazard ratios calculated in Question 2.

I tried using the following syntax:

TIME PROGRAM.
COMPUTE hv1= (0<=T_ <= 183)* clinic.
COMPUTE hv2= (183<T_ <= 365)* clinic.
COMPUTE hv3= (365<T_ <= 548)* clinic.
COMPUTE hv4= (548<T_ <= 730)* clinic.
COMPUTE hv5= (T_ >730)* clinic.
COXREG
survt2 /STATUS=status(1)
/METHOD=ENTER prison dose hv1 hv2 hv3 hv4 hv5
/CRITERIA=PIN(.05) POUT(.10) ITERATE(20).

Where hv1 to hv5 are the 5 time periods specified and survt2<365=0 and survt2>=365=1. I am not getting the output I need. Any help?

Upvotes: 1

Views: 25

Answers (0)

Related Questions