Drug Use Measurement

Is drug use declining in American households, schools, and prisons? Is heroin use on the rise? Are Americans still using marijuana, a "gateway" drug, in large numbers? National drug control policy demands answers to these questions. But, in obtaining responses, it must also be asked: How valid are the data? GAO concludes that three leading federally funded annual surveys of national drug use patterns and trends yielded questionable results because of various methodological weaknesses, such as excluding groups at high risk for drug use and relying on self-reported data. Promising new methodologies, such as hair sample analysis, deserve attention as ways to validate self-reports and determine drug use over an extended period. Expanding the subsamples of current surveys and conducting new studies aimed at hard-to-reach, high-risk groups should improve the coverage of underrepresented target populations.

GAO found that: (1) the rate of cocaine use among arrestees in 1990 was 22 times higher than the rate for high school seniors and 53 times higher than the rate for households; (2) arrestees' cocaine use remained stable from 1987 to 1990; (3) all groups showed a decrease in the use of marijuana; (4) between 1979 and 1990, general household drug use decreased by 60 percent and high school senior drug use decreased by 62 percent; (5) between 1987 and 1990, arrestee marijuana use decreased by 48 percent; (6) the National Household Survey on Drug Abuse (NHSDA) employs a highly developed research design that emphasizes a national multistage probability sampling procedure; (7) the High School Senior Survey (HSSS) uses a national multistage probability sampling procedure for collecting drug use data on approximately 15,000 to 19,000 high school seniors each year; (8) arrestees underreported cocaine use by more than 50 percent in 20 sites, raising questions about the utility of the self-report procedure for this population group; (9) survey findings do not support the annual collection of data on the various forms of substance abuse among high school students and the general population; (10) adequate state-level drug use data are necessary to assist local policymakers in the planning and evaluation process; and (11) proposals to expand NHSDA to the state level are costly and potentially duplicative of other funded studies.

Recommendations

Our recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.

Director: Team: Phone: