I have need help to make a graph to show a Spline Cox proportional with log
hazard versus a continuous variable. I like to appropriate knots on the
spline curves. I have 2 cohort of patients with similar follow up time. I am
not able to adjust for age gender other biochemical parameters by the
following codes. I am also unable to make a graph with log hazard ratio with
95% confidence interval
stset obstime censor
stpm opg, scale(odds) knots(%20 40 60 80)
predict s0, survival zero
predict surv3, survival time(3)
predict median, centile(50)
predict semedian, centile(50) stdp
predict sem, centile(50) stdp
gen lci = exp(ln(median)-1.96*sem)
gen uci = exp(ln(median)+1.96*sem)
predict h0, hazard zero
predict h1, hazard zero at(opg 1)
predict lnH, cumhazard at(opg 1)
gen hr = h1/h0
predict bz, tvc(opg)
predict sbz, tvc(opg) stdp
twoway scatter lnH opg
I appreciate all help
Thanks
Tony Qureshi
*
* 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/
