Reduction of racial disparities in urine drug testing after implementation of a standardized testing policy for pregnant patients

Abstract

Drug use during pregnancy can have implications for maternal and fetal morbidity and mortality and legal ramifications for patients. The American College of Obstetricians and Gynecologists guideline states that drug screening policies during pregnancy should be applied equally to all people and notes that biological screening is not necessary, stating that verbal screening is adequate. Despite this guidance, institutions do not consistently implement urine drug screening policies that reduce biased testing and mitigate legal risks to the patient. This study aimed to evaluate the effects of a standardized urine drug testing policy in labor and delivery on the number of drug tests performed, self-reported racial makeup of those tested, provider-reported testing indications, and neonatal outcomes. This was a retrospective cohort study. A urine drug screening and testing policy was introduced in December 2019. The electronic medical record was queried for the number of urine drug tests performed on patients admitted to the labor and delivery unit from January 1, 2019, to April 30, 2019. The number of urine drug tests performed between January 1, 2019, and April 30, 2019, was compared with the number of urine drug tests performed between January 1, 2020, and April 30, 2020. The primary outcome was the proportion of urine drug tests performed based on race before and after the implementation of a drug testing policy. The secondary outcomes included total number of drug tests, Finnegan scores (a proxy for the neonatal abstinence syndrome), and testing indications. To understand perceived testing indications, pre- and postintervention provider surveys were administered. Chi-square and Fisher exact tests were used to compare categorical variables. The Wilcoxon rank-sum test was used to compare nonparametric data. The Student t test and 1-way analysis of variance were used to compare means. Multivariable logistic regression was used to construct an adjusted model that included covariates. In 2019, Black patients were more likely to undergo urine drug testing than White patients, even after adjusting for insurance status (adjusted odds ratio, 3.4; confidence interval, 1.55-7.32). In 2020, there was no difference in testing based on race after adjusting for insurance status (adjusted odds ratio, 1.3; confidence interval, 0.55-2.95). There was a reduction in the number of drug tests performed between January 2019 and April 2019 compared with between January 2020 and April 2020 (137 vs 71; P<.001). This was not accompanied by a statistically significant change in the incidence of neonatal abstinence syndrome measured by mean Finnegan scores (P=.4). Before the implementation of a drug testing policy, 68% of providers requested patient consent for testing; after the implementation of a drug testing policy, 93% requested patient consent for testing (P=.002). The implementation of a urine drug testing policy improved consent for testing and reduced disparities in testing based on race and the overall rate of drug testing without affecting neonatal outcomes.