Failures in substance use surveys.

This paper reviews some of the potential sources of failure in surveys concerned with substance use. These failures are both conceptual and methodological in nature. Conceptual failures include nonuse and misuse of survey data, the inability of these data to detect emerging substance use trends and problems, and relevant ethical concerns. Methodological failures include sampling, coverage, nonresponse, and measurement and processing errors, which are indigenous at some level to all epidemiologic surveys. Consideration of these is necessary if survey data are to be appropriately used for their intended purposes.