Meanwhile the "old way" is interesting it is useless for any prediction.
None of the 3rd party data providers' report are exact since they do not get all data (from all data point) but the coverage of the universe is high, over 80/85% (IIRC) … the data (script#) are accurate, no - significant - under or overreporting.
"old way" / script#: it shows how many patients took V during the relevant week
"new way" (normalized): it it shows how many V was took by the patients during the relevant week
e.g. during week ending 05/04 42,988 patients took 55,197 bottles V … if patients took 1-month or 3-months supply (no other x-months) it means - 36,898 patients took 1-month - 6,100 patients took 3-months