[iebaltab]: Allow continuous treatment

[bbdaniels]

This should be quite simple as it should only result in the removal of the column for summary statistics of further groups. Allowing continuous() rather than grpvar() should request only one column of summary statistics, and then the estimate of the continuous effect of the treatment variable as the difference column.

This is needed for dose-response situations where the treatment is only something that can be varied in intensity across the population, but is not a binary or grouped thing.