Measurement of drug use in a defined population

Abstract

To evaluate the accuracy of the DDD per 1000 inhabitants per unit time as an estimate of the fraction of drug users within a population, DDD figures have been compared with the proportion of apparent drug users, i.e. individuals in the population for who a prescriptions had been dispensed. An individual-based prescription monitoring project provided the necessary data for eight drugs representing continuous long-term medication, short-term medication and medication falling in between those two categories. For digoxin, the long-term drug, the number of DDD/1000 inhabitants/day was about 40% below the proportion of apparent drug users. The DDD figure for antibiotics ranged from 4% below to 28% above the apparent users, and for the remaining drugs it ranged from 17% below (bendroflumethiazide, 1982) to 80% below (naproxen, 1985). The DDD methodology is a valuable first step in overall drug use measurement, but for more precise estimates of drug use it must be supplemented by other techniques.