The best way to guarantee interval validity in the experimental paradigm is to ensure that randomisation is correct. This makes it possible to validate the hypothesis that any differences observed are due to the factor under evaluation, and not to other prognostic variables.

Table 1 answers this question, showing that of the prognostic variables evaluated, only the volume of prescription (packages prescribed) is different; it is lower in the control group.