(1) Racial Disparities in Urine Toxicology Screening for Labor and Delivery Admissions

Background & Introduction: Urine toxicology screening (UTS) is one tool to help identify perinatal substance use, which can impact perinatal outcomes. However, with unclear guidance on when and whom to test, UTS practices are prone to bias, especially in racially diverse populations. Non-Hispanic Black or African American patients are disproportionately subjected to UTS, often without corresponding increases in positive results, indicating systemic bias rather than clinical need. These disparities raise concerns about equity in perinatal healthcare and the social and legal consequences for minoritized populations. This study evaluated disparities based on race and ethnicity in UTS orders during labor and delivery (L&D) admissions across multiple health systems, controlling for clinical, demographic, and institutional factors.

Methods: This retrospective cross-sectional study analyzed electronic health records (EHR) from L&D admissions between July 1, 2018, and June 30, 2022, at three Upper Midwest health systems. Patients with multiple pregnancies during this time period were included and admissions were nested within patients. Admissions resulting in live births or fetal demise among patients aged ≥12 years were included. Patients opting out of research were excluded. Extracted data included demographics, comorbid diagnoses, pregnancy and admission characteristics, and hospital characteristics. Medical conditions that may warrant a UTS order were captured using ICD-10 codes. The primary outcome was the presence of a UTS order; results of a completed UTS for each substance was recorded. In secondary analyses, we evaluated whether a UTS order resulted in a positive test (presence of a substance in urine sample). Positive results were analyzed in two models: (1) positive test result for any substance and (2) positive test result for any substance excluding cannabis/THC. Analyses were conducted in SAS 9.4. Multivariable logistic regression assessed associations between race/ethnicity and UTS orders, adjusting for potential confounders. The study procedures were reviewed by each institution’s Institutional Review Board and approved as exempt non-human subjects research.

Results: Among 81,999 admissions for 71,341 patients, UTS was ordered in 14.2% (n=10,176) of the admissions. Most patients identified as White (66.2%), followed by Black or African American (14.8%), Asian (9.5%), unknown or not reported race (4.9%), multiracial (2.5%), other race (1.4%), and American Indian or Alaska Native (0.7%). Most patients identified as non-Hispanic/Latinx (78.2%), with 6.1% identifying as Hispanic/Latinx and 15.7% having unknown ethnicity. When controlling for sociodemographic characteristics, pregnancy characteristics, medical diagnoses, and time and place factors, adjusted analyses showed that Black or African American (aOR=2.09, 95% CI=1.93, 2.25), American Indian or Alaska Native (aOR=2.50, 95% CI=2.00, 3.13), and multiracial (aOR=2.04, 95% CI=1.77, 2.36) patients were significantly more likely to have a UTS ordered than White patients. Adjusting for all other factors, patients who were Hispanic/Latinx (aOR=1.16, 95% CI=1.04, 1.30) or unknown ethnicity (aOR=1.39, 95% CI=1.29, 1.49) were significantly more likely to have a UTS than non-Hispanic/Latinx patients. In addition to race and ethnicity, the variables with the largest associations with UTS orders were having a comorbid mental health or substance use disorder, being a current smoker, having a fetal demise, and having a medical indication for UTS. Adjusted positive UTS rates, excluding cannabis/THC, were comparable across racial groups (7.5%). Including cannabis/THC increased positive rates to 21.3%, with significantly higher rates among Black or African American and multiracial patients after adjusting for other factors.

Conclusion & Discussion: This study highlights racial bias in UTS ordering among patients admitted to L&D, unsupported by clinical indications or substance use prevalence, and suggests that non-clinical factors drive testing disparities. Standardized, evidence-based UTS protocols are urgently needed to ensure equitable practices. Emphasizing clinical indications and informed consent to protect patient autonomy and reduce harm is also necessary to optimize care. It is important to note that although UTS can be a helpful ancillary tool in care, both pregnancy care and addiction medicine organizations support universal written or verbal screening as the primary tool for substance use screening in pregnancy.

Study strengths include large sample size, data extracted from multiple health systems, and controlling for multiple covariates. Limitations to consider are inclusion of one geographic region, EHR data did not include clinical notes, and who ordered the UTS was not examined.

Future research should explore institutional and clinician-level contributors to these disparities and evaluate the effects of standardized protocols on perinatal health outcomes. Addressing these inequities is essential for achieving fair and effective perinatal healthcare for all patients.

References:

1. Forray A. Substance use during pregnancy. F1000Res. 2016;5(F1000 Faculty Rev):887.

2. Ecker J, et al. Substance use disorders in pregnancy: ethical imperatives. Am J Obstet Gynecol. 2019;221(1):B5-B28.

3. Pflugeisen BM, Mou J, Drennan KJ, Straub HL. Demographic discrepancies in prenatal urine drug screening in Washington State surrounding recreational marijuana legalization and accessibility. Matern Child Health J. 2020;24(12):1505-1514.

4. Jarlenski M, Shroff J, Terplan M, Roberts SCM, Brown-Podgorski B, Krans EE. Association of race with urine toxicology testing among pregnant patients during labor and delivery. JAMA Health Forum. 2023;4(4):e230441.

5. Roberts SC, Nuru-Jeter A. Universal alcohol/drug screening in prenatal care: questioning assumptions. Matern Child Health J. 2011;15(8):1127-1134.