Not sure why you'd want to, but you can can dichotomize both variables. Are there sensible places to do this (i.e., not just the medians).

I expect dichotomising will produce different results depending on where you draw the line. You also lose lots of explanatory power.

Having said that - the E-value is a measure answering the question "what if there was an (unmeasured) confounder associated both with the treatment and the outcome?" It is always presented on the risk ratio scale, but can be applied to continuous measures if that is what you are thinking of (see table 2 of the publication). Also, by all means, use the e-value calculator I put in Research Gate Code E-value calculator

Risk ratio is only for dichotomous variables. Continuous variables can be compared with t tests, anova or correlation/regression if normal in distribution, or non parametric tests, such as Wilcoxon, if non parametric