Factor Analysis of Drug Supply Time Series at Pharmacies

Abstract

The stochastic properties of the daily variations in drug sales at a community pharmacy in the season of 2004/2005 are studied by a factor analysis. The four factors which can closely be related to chronic diseases, common cold, influenza and allergies are extracted from the data of 31 drugs. For most of the drugs, the sales time series can approximately be represented by the four factors alone. Similar results are obtained from the same pharmacy in the 2003/2004 season and also from two other pharmacies. Interestingly, the classification of drugs according to the factor analysis corresponds to their medicinal actions.