Abstract
IMPORTANCE: The direct addition of buprenorphine to urine drug test specimens to mimic results suggestive of adherence is a clinically significant result, yet little is known about the phenomenon. OBJECTIVE: To characterize factors associated with the direct addition of buprenorphine to urine specimens among patients prescribed buprenorphine for opioid use disorder. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study of urine drug test specimens was conducted from January 1, 2017, to April 30, 2022, using a national database of urine drug test specimens ordered by clinicians from primary care, behavioral health, and substance use disorder treatment clinics. Urine specimens with quantitative norbuprenorphine and buprenorphine concentrations from patients with opioid use disorder currently prescribed buprenorphine were analyzed. EXPOSURES: Nonprescribed opioid or stimulant co-positive, clinical setting, collection year, census division, patient age, patient sex, and payor. MAIN OUTCOMES AND MEASURES: Norbuprenorphine to buprenorphine ratio less than 0.02 identified direct addition of buprenorphine. Unadjusted trends in co-positivity for stimulants and opioids were compared between specimens consistent with the direct addition of buprenorphine. Factors associated with the direct addition of buprenorphine were examined with generalized estimating equations. RESULTS: This study included 507 735 urine specimens from 58 476 patients. Of all specimens, 261 210 (51.4%) were obtained from male individuals, and 137 254 (37.7%) were from patients aged 25 to 34 years. Overall, 9546 (1.9%) specimens from 4550 (7.6%) patients were suggestive of the direct addition of buprenorphine. The annual prevalence decreased from 2.4% in 2017 to 1.2% in 2020. Opioid-positive with (adjusted odds ratio [aOR], 2.01; 95% CI, 1.85-2.18) and without (aOR, 2.02; 95% CI, 1.81-2.26) stimulant-positive specimens were associated with the direct addition of buprenorphine to specimens, while opioid-negative/stimulant-positive specimens were negatively associated (aOR, 0.78; 95% CI, 0.71-0.85). Specimens from patients aged 35 to 44 years (aOR, 1.59; 95% CI, 1.34-1.90) and primary care (aOR, 1.60; 95% CI, 1.44-1.79) were associated with the direct addition of buprenorphine. Differences by treatment setting decreased over time. Specimens from the South Atlantic census region had the highest association (aOR, 1.4; 95% CI, 1.25-1.56) and New England had the lowest association (aOR, 0.54; 95% CI, 0.46-0.65) with the direct addition of buprenorphine. CONCLUSIONS AND RELEVANCE: In this cross-sectional study, the direct addition of buprenorphine to urine specimens was associated with other opioid positivity and being collected in primary care settings. The direct addition of buprenorphine to urine specimens is a clinically significant finding, and best practices specific for this phenomenon are needed.