Gravar-mail: On the covariate-adjusted estimation for an overall treatment difference with data from a randomized comparative clinical trial