0

I have fitted the Cox model with time-varying covariates and the results are here enter image description here

where gdp_time is time-varying covariate.

and here are the values for gdp_time time1 = 0, time2 = 1, time3 = 0.

When I calculate the result with

h(t|Z) = h(t)*exp(-0.1116*priodmonths + 1.7774*rate_com - 0.518*gdp_time)

I got a different result from that R gives me.

Could anyone help me to calculate the interaction term present in this equation?

Bakytbek Begaliev
  • 1
  • 1
  • you are missing an exponential term for the linear combination of coefficients. Should be $\exp(-0.1116 \cdot \text{period months} + 1.7774 \cdot \text{rate_com} - 0.518 \cdot \text{gdp_time})$. Also that term I just gave you is the hazard ratio. If you multiply it by the baseline hazard function (What you call h(t)) you get a hazard function, not a hazard ratio. I don't know how else you estimate the hazard function (and what other way you get the hazard from R). – AdamO Mar 06 '19 at 17:51
  • Sorry, there is exp function. I have edited the hr to h(t|Z) that is cumulative hazard function at a time t. As a gdp_time is time-varying covariate, there must be an interaction between this covariate and others. – Bakytbek Begaliev Mar 07 '19 at 14:13

0 Answers0