The Drug Use Forecasting program is an important national data set that estimates the prevalence of psychoactive drug use among recent arrestees. However, because provision of urine samples is voluntary and many arrestees refuse to provide specimens, use of the urinalysis results in estimating prevalence of use may lead to inaccurate estimates. This paper provides a selection model to estimate drug use among those who refuse to take urine tests. The results indicate that those arrestees who refuse urine testing generally have a higher percentage of drug use than those who agree to take the test. The applications of this model allow an overall estimation of the extent of drug usage for all arrestees.
Evaluating the drug use “gateway” theory using cross-national data: Consistency and associations of the order of initiation of drug use among participants in the WHO World Mental Health Surveys