Concordance between Self-Reported Drug Use and Urinalysis in a Sample of Black American Women

Abstract

Background

Obtaining accurate drug use data is important in the field of substance use research. Urinalysis, considered gold standard, can be costly or infeasible, whereas self-report is quick and easy, but susceptible to imperfect recall or misrepresentation. It is important to determine the concordance between self-report and urinalysis, and better understand the contexts and participant characteristics that influence self-report accuracy. The current study aims to assess this concordance for marijuana and cocaine in a sample of Black American women, some with criminal justice exposure, and to investigate predictors of non-concordance.

Methods

In this longitudinal study, a sample of Black American women were recruited from community, prison, and probation settings. Self-report drug use and urine drug screens were obtained at 6-, 12-, and 18-month follow-ups, allowing for the calculation of concordance. Generalized linear mixed models were used to assess participant characteristics that predicted non-concordance (both false positives and false negatives).

Results

In general, there was agreement between self-report and urinalysis results for both marijuana and cocaine. Baseline drug use status was the most consistent predictor of non-concordance. Individuals recruited while on probation were more likely to have false negative results and less likely to have false positive results. Additionally, concordance rates for marijuana increased over the follow-up period.

Conclusions

Self-reported marijuana and cocaine use are accurate measures of actual drug consumption in a sample of Black American women with a variety of criminal justice interactions.

Keywords:

• Self-report
• urinalysis
• concordance
• substance use
• criminal justice
• generalized linear mixed models

Acknowledgments

We thank the women who participated in the project. Also, we are grateful for the Kentucky Department of Correction’s (DOC) support, but it should be noted that the Department disclaims approval or endorsement of the findings.

Additional information

Funding

This work was supported by grants from the National Institute on Drug Abuse (NIDA) to CO (R01DA022967).