DISCUSSION
Our study had limitation and the results obtained
have to be considered with these limitations. The
randomization of community pharmacies to the
study and control group were done due to avoid
unintended increase in quality of standard
pharmaceutical services, but it might have
influenced the number and characteristic of patients
enrolled into the study and control group.
Assignment to the study group could be the reason
for resignation from the study of a large part of
community pharmacies. Despite this limitation we
did not observe the differences in characteristic of
patients included in study and control group.