Estimating daily dose for pharmacoepidemiologic studies: Alprazolam as an example

Pharmacoepidemiologic research utilizing large automated prescription databases requires measures of daily drug dose. The objectives of this report were to develop a measure of the average daily dose from such a database and to apply it to a specific drug (alprazolam) to determine if the average daily dose had changed over time. An estimate of the average daily dose was developed by calculating total exposure (in mg) over time and dividing it by the total number of days of exposure over the same time period. The total number of days of exposure was calculated with an algorithm linking prescriptions into periods of use if certain conditions were met. Applying the measure, the average daily dose of alprazolam increased after it was approved for use at higher doses, and we observed no seizure disorders among any users with highest average daily doses. Since the data fields to develop the measure appeared to be valid and are common to most automated prescription systems, and given an assumption that the receipt of multiple dispensings in a consistent temporal sequence indicates that the drug is consumed, average daily doses can be estimated from large automated prescription databases.