Data Sources: Problems and Issues

Abstract

Several nationwide data systems have been specifically designed to assemble data on prevalence of drug use. Additional information is available from other systems that are oriented to different purposes, but also collect drug-related data. National prevalence estimation efforts are variously based on one or more of these prominent data sources, which are briefly reviewed in this article. Also discussed is the utility of these existing data in terms of sampling and coverage, nature and validity of measures, methods of enumeration, and consistency of data collection over time and across systems. To determine the nature of drug use at the local level, local data should be examined. Regardless of their geographic scope, all data systems should improve sampling and coverage of high-risk groups and results should be assessed by occasional small-scale validity studies. Most important, comparability among data systems using comparable measures is needed to enable integration of information, leading to more comprehensive and accurate estimates of the prevalence of drug use.