In some countries, the size N and characteristics of the exposed population and its conditions of exposure can be precisely derived from health insurance databases. In this case, except for the poor quality of case collection (i.e. under-reporting), SR approaches the cohort design.
Unfortunately, in most cases, it is necessary to estimate these parameters from sales statistics and/or drug prescription on drug utilization panels (Begaud et al., 1993). The use of such agregated data precludes any possibility of considering some individual or sub-group characteristics in the analysis.
The necessary "ingredients" for computation are: the number of exposure units, e.g. tablets, capsules, injection doses sold in the territory during the relevant period of time, and the average c
Was this article helpful?