c daily dose (ADD) of the considered drug used in this population, the latter being estimated from prescription panels or other sources. By default, the defined daily dose (DDD) or the recommended daily dose (RDD) can be used as proxy.
Example: 780 000 packages of 20 capsules have been sold in a one-year period, the used daily dose is 2.1 capsules. This corresponds to the quantity necessary for a cumulative duration of treatment of: (780 000 x 20)/2.1 = 2 666 667 days, or 87 719 months. In a more epidemiological parlance, the exposure level in the source-population is 87 719 person-months.
As for incidence density calculations, this total probably sums individual exposure periods which are extremely different. Moreover, because of its ecological character, this approach precludes any risk analysis based on the duration of exposure.
To estimate the number of treatments or the number of subjects treated would require knowing the average duration of a treatment (ADT) with the considered drug. In the previous example, if the ADT was 23 days, the number of treatments for the considered period would be: 2 666 667/23 = 115 942.
However, in the absence of direct information from a health insurance database, the use of measurements made on panels or relatively small samples, both for the average daily dose and duration of treatment, will greatly increase the statistical instability of the estimate. In the previous example, if the 95% CIs were [1.6; 2.7] and [16; 31] for the ADD and ADT, respectively, then the CI for the number of treatments would range from 31 860 to 104 167. For this reason, it is often preferable to keep person-time estimates for further calculations.
Was this article helpful?