Urine drug screening on labor and delivery

Substance use including opioids, methamphetamines, benzodiazepines, and barbiturates during pregnancy is harmful for the pregnant person and the fetus. Routine screening using validated questionnaires is recommended, but often biologic sampling is done instead. There is often bias in urine drug screening on labor and delivery units.

This study aimed to compare characteristics of people who did and did not receive urine drug screening during labor and delivery and to examine the relationship of maternal results to neonatal results.

This was a retrospective chart review examining all people in 2017 who delivered in the labor and delivery unit at our institution. We collected urine drug screening result information, maternal demographic data, follow-up after positive maternal tests, and neonatal test results. Individual characteristics and obstetrical outcomes were analyzed.

Of 6265 deliveries, 297 urine drug screening tests were ordered. People who were tested identified most commonly as Native Hawaiian or Pacific Islander (P<.0001). The most common indications for ordering tests were a history of substance use and insufficient prenatal care (P<.0001). People who tested positive were more likely to self-identify as White (P=.03) and have history of substance use (P<.0001). Among the positive test results, 24 (24%) were caused by a provider-ordered medication. Self-identification as Native Hawaiian or Pacific Islander was not predictive of a positive result. Of the tested people, 36% (108/297) had a positive result on preliminary testing, and 33% (98/295) on confirmatory testing.

Native Hawaiians and Pacific Islanders were more likely to undergo testing, whereas White people were more likely to have a positive result. Maternal results were not reliable for predicting neonatal drug test results and vice versa. With rising rates of substance use disorders in the pregnant and reproductive-age population, standardized unbiased race-neutral guidelines for urine drug screening should be implemented using laboratory test results that include preliminary and reflex confirmatory results.