BackgroundThe substance use crisis continues to progress. Medication for Opioid Use Disorder (MOUD) are prescribed to reduce opioid use and related harms; however, many individuals continue to use substances while on treatment. The objective of this study was to describe the temporal and demographic trends of the agreement between self-reported and urine tested substances. MethodsThe current study is a retrospective secondary analysis of three phases of a prospective cohort study (Pilot 2011, Genetics of opioid addiction (GENOA) 2013–2017, and Pharmacogenetics of opioid substitution treatment (POST)) 2018–2022) spanning 2011–2022. We compared the self-reported substance use data for opioids, benzodiazepines, amphetamine/methamphetamine (AMP/MET), and cocaine with urine drug results. We compared the positive predictive value (PPV), false omission rate (FOR), sensitivity, and specificity between (i) different drugs; (ii) by sex, and (iii) age group at enrollment in each phase of the study using self-reported substance use at baseline and retrospective electronic health record data on urine drug screenings collected over the same time period. ResultsOverall, the average PPV and FOR for any drug across all phases was 80.7 % and 37.9 %, respectively. Sensitivity and specificity were highest for cocaine and lowest for benzodiazepines. We found no specific trend by sex. Lastly, we found a higher sensitivity for opioids and AMP/MET in those under 25 years of age compared to other age groups. PPV increased over time for benzodiazepines, AMP/MET and cocaine and FOR was higher during the pilot and POST phases than the GENOA phase. ConclusionOur study highlights the unique challenges associated with ascertaining substance use behaviour for individuals receiving MOUD, indicating many patients will accurately report substance use while others do not. It is therefore important to consider the context of the patient, and the type of the co-substance used to select patient-centred testing as indicated. Therefore, the answer to the question of do we need urine drug screen is yes in some cases.
Read full abstract