New developments in biological measures of drug prevalence.

Drug use among different populations such as household members, students, and arrestees vary substantially and the accuracy of their self-reports may be questionable. The accuracy of prevalence estimates based on self-report data can be monitored by chemical drug testing of biological specimens such as urine, saliva, sweat, and hair. Each biological specimen is unique and offers a somewhat different pattern of information regarding drug use over time. Also, each specimen has unique strengths and weaknesses regarding the type of information obtained from drug testing. The performance characteristics of the assay methodology may also be important. The validation of self-report data by drug testing must be performed with careful consideration of the limitations imposed by the testing methodology and the biological specimen.