Surveys with drag use indicators may have different objectives: to describe current treatment practices; to compare the performance of individual facilities or prescribers; to monitor or supervise specific drug use behaviours; or to assess the impact of an intervention. This chapter discusses some important study design and sampling issues.
The best design for a particular study depends not only on statistical theory but also on the objective(s) of the study and on the practical aspects of collecting the data. Guidelines for the sample size for each of the different types of study are therefore included and are based on extensive field testing. They are summarized in Table 4. By following these recommendations the results of the drug use studies will be valid and comparable.
The methodological issues are discussed in full in a separate document, in which theoretical and empirical support for the recommendations made in this section are presented. If it is essential that study results be highly accurate and reliable, for example when the effects of an expensive intervention have to be tested, it is recommended that a sampling expert be consulted before the study is undertaken.