Exploring the utility of an estimation procedure to reveal drug use among arrestees: implications for service delivery.

One of the most persistent questions plaguing researchers and service providers is how to best estimate the extent of targeted behaviors in relevant populations. One problem of particular importance is the prevalence of drug use in justice-involved populations. Data have been collected through such methods as self-report and analysis of biological specimens, although both have notable limitations when used alone. As a means of drawing on the strengths of both methods, such data can be used in a confirmatory manner or, alternatively, may be summed to estimate prevalence. However, this latter approach is not without difficulty as different sources lack substantial agreement. The focus of this study is to employ a methodology that utilizes multiple data sources and adjusts for nonreporting from either source. Compared to more commonly employed techniques, the results indicate that the alternative method yields higher estimates of marijuana and cocaine use among a sample of arrestees. These findings, in turn, suggest that current behavioral health interventions and policies may be based on underestimates of drug use.