Estimating baseline values of the variable of intervention in a clinical trial