Finally, the limitations of the databases used should be acknowledged. The epidemiological data used in these studies have typically been gathered for other purposes. For example, ims collects data on total sales of drugs in Canada. Drugs sold are not necessarily ingested, thus drug utilization data provide only crude estimates of the extent of drug use and do not indicate the appropriateness of such use. Furthermore, the ddd is a technical unit of measurement used to permit comparisons which may not accurately reflect the average prescribed daily dose in Canada.
Linkage between morbidity (eg, abuse, dependence) and use is only indi-ect since patient-specific da ta bases are pres ently gath ered primar ily for sur -veillance and have not necessar ily been developed to test rigorous hypotheses. The kinetic parameters used in the study are population-based averages and may not reflect the large variations in pharmacokinetics among individuals .
Despite the above-mentioned limitations of the databases, it is in-erest-ng to observe that our find-ngs us-ng available epidemiological data closely support rigorously conducted experimental studies. Defining concepts and improving methodologies in this area may be the drug epidemiologist’s contribution to comparative abuse and dependence liability testing.