Dear All, I would like to calculate the standard error (s.e.) for the Hazard Ratios (HR) involving an interaction term. I have the following model: HR =exp(b1x+b2y+b3z+b4x*y). All variables being continuous. I would like to fix x at 1 (i.e. x=1) while holding all other variables constant. That is, HR=exp(b1+b4y). Then calculate HRs at specific values of y. My problem is calculating the Standard Error for these hazard ratios so I can calculate the confidence intervals. I know how to calculate the variance-covariance matrix for the model but I am not quiet sure how to cal s.e. for the above HRs. Can anyone help? References to journal article would also be helpful. Raphael