I've run a parametric survival analysis on the number of days to an event, where there is an interaction between sex and another variable. I would like to report the odds ratios for this variable for each sex. How can I calculate these manually using the parameter estimates for the variable and the interaction term? The proportional hazard fit will report odds ratios for single variables (i.e. sex, variable 1, variable 2), but only gives a warning for the variable involved in an interaction (same as with logistic regressions).