Prenatal substance use affects many families in the United States and is associated with negative maternal and child outcomes.1,2 Newborn toxicology testing may be ordered for a variety of indications in the context of known or unknown prenatal substance use, including to guide clinical management, formulate safety assessments, or substantiate reports to child protective services (CPS).3 However, toxicology testing is not without harms, potentially deterring patients from seeking care, incurring costs from low-value testing, and carrying a risk of inaccurately interpreted results, including frequent false-positive results.3–5 Inequities in peripartum toxicology testing are common, with clinicians being more likely to order toxicology tests for patients who are Black, both mothers and newborns, and more likely to report Black families to CPS for the same testing results.6–8In this clinical environment, the new Hospital Pediatrics article by Costa et al describes a quality improvement (QI) project aimed at reducing these exact racial disparities in newborn toxicology testing practices.9 Having gathered data demonstrating that newborns of Black parents disproportionately experienced toxicology testing compared with newborns of white parents at their own institution, the study team created a new set of institutional guidelines that were implemented through a QI change package.We applaud the authors for their efforts to improve equity for Black families affected by parental substance use, given the multiple intersecting dimensions of stigmatization and marginalization faced by these families. One strength of their study was their multipronged approach to implementation. In addition to policy creation and clinician education, the team made use of higher-impact workflow change strategies, including use of an electronic health record order set and changes to clinical documentation templates. The team also brought together multidisciplinary perspectives encompassing pediatric and obstetric care, social work, and hospital leadership, which ensured representation of diverse perspectives during policy design and implementation buy-in across multiple settings.Within the year following implementation of their new practice patterns, they observed an increase in the rate of toxicology testing of Black newborns (5.2% to 9.9%) relative to white newborns (3.2% to 3.7%). They concluded, based on this result, that “health care systems that desire to eliminate health care inequity must reach beyond hospital policy changes to identify legislative, community, and other drivers that lead to such inequity in the first place.”9This commendable study was strongly grounded in multidisciplinary team alignment, clear-eyed and clearly stated goals with tangible benefits for families, and a robust QI approach, but the results must have been disappointing to the authors. From our vantage point, a pediatrics resident with clinical and qualitative research expertise in parent communication related to newborn toxicology testing and an internal medicine-pediatrics hospitalist with QI expertise in decreasing rates of and inequities in newborn toxicology testing, we have some ideas about why this project showed an outcome entirely counter to its goals.When considering newborn toxicology testing, having evidence-based and consistent indications for testing is a clear starting point. The authors do not describe in detail why the indications of “no prenatal care” or “prenatal care starting after 20 weeks” were chosen. It is well documented that rates of first-trimester presentation to prenatal care are lower for Black women than white women, because of factors including limited appointment accessibility, insurance or financial barriers, and discriminatory care.10–13 As a result, it can be be anticipated that using late or no prenatal care as an indication for newborn toxicology testing would disproportionately affect Black families. In this article, late prenatal care accounted for 21.9% of tests on Black newborns vs 8.0% of tests on white newborns, a statistically significant disproportionality (Table 1). Given the multifactorial reasons that a patient could present later to prenatal care, prenatal care timing is clinically low yield for detection of prenatal substance use and should not be included as a routine reason for newborn toxicology testing.14 In fact, the authors highlight evidence that reflects that when perinatal toxicology testing for limited prenatal care was eliminated (along with the indication for cannabis use only), it led to improved racial equity in testing and CPS reporting, decreased overall rates of CPS reporting, and did not result in changes in newborn outcomes after delivery.14 This collective evidence can empower providers to not test newborns for indications based on prenatal care timing, particularly if their goal is excellent, equitable care.Another indication for testing that was included for unclear reasons is “known maternal drug use.” The authors highlight ambiguity in the legal requirements in their state for when newborn toxicology testing is required, and that “many health systems operationalize state law by employing institutional policies that call for a newborn toxicology test to confirm PSE prenatal substance exposure.”9 In a review of each state’s minimum legal requirements related to the Child Abuse Prevention and Treatment Act, Michigan does not require ordering toxicology testing for known PSE (only required by 4 states).15 Additionally, as Michigan does not require CPS reports for known use of medications for opioid use disorder (MOUD) during pregnancy, the use in the QI project of MOUD alone as an indication for newborn toxicology testing is also in excess of legal requirements. As such, infants would be observed for development of neonatal opioid withdrawal syndrome regardless of toxicology test results; such testing also does not aid in clinical management during the newborn hospitalization.The tension between clinical use of testing vs mandated reporting responsibilities is particularly highlighted in the authors’ findings regarding prenatal tetrahydrocannabinol (THC) exposure. Of the 55 additional tests performed in the postpolicy period relative to prepolicy levels, only 4 were positive for substances other than THC (Table 2). For Black newborns specifically, 18 of the 19 incremental positive results were positive for only THC. Given the limited evidence on the relationship between prenatal cannabis use and neonatal outcomes, these THC-only positive results are unlikely to change clinical management.16 Perinatal toxicology testing for cannabis use as a sole indication is also unlikely to reveal exposure to clinically actionable substances but carries the risk of false positives, and prenatal exposure to cannabis in isolation is not associated with increased risk of child maltreatment.5,17 Although Costa et al provide the context that their state CPS agency does require reporting for prenatal cannabis use, they thus identified elimination of testing for low-risk cannabis use as an important future target for decreasing low-use tests and decreasing racial disparities in testing.Another driver of racial inequity identified by the study team was differential rates of verbal screening for prenatal substance use, leading to racially disproportionate testing. Current evidence supports that parental history is positive for the majority of substance-exposed newborns, demonstrating that verbal screening can identify prenatal substance use, obviating the need for newborn toxicology testing.3,18 As a result, the study team implemented universal screening for parental substance use by pediatricians in the newborn setting as part of their project. This approach is consistent with current guidance, which recommends that all pediatricians screen for parental substance use, but there is a range of comfort among pediatricians with performing screening and ensuring appropriate referrals and social supports for positive screens.19 Although details on implementation and postpolicy rates of verbal screening, including if they used a validated screening questionnaire, were not included here, this aspect of the project is novel. Future studies describing the implementation of universal screening and qualitative evaluation of its acceptability would be useful for other institutions seeking to implement similar policies.Although this article focuses on universal changes to testing indications and verbal screening, it is also worth noting that some health care systems that have identified equity challenges take an approach of accountability and respond in a differential way to different patient groups.20 Such a differential approach has not yet been tried (to our knowledge) in the area of newborn toxicology testing. One way this might be put into action would be a real-time clinical decision support tool for providers when ordering a toxicology test on a minoritized patient that would provide information on the existing racial disparities in testing.Overall, Costa et al have made a significant contribution to the vitally important topic of improving equity for families affected by substance use. Their work describes the potential unintended consequences of QI efforts in inadvertently increasing racial disparities and, in doing so, has highlighted several potential avenues through which hospitals and health systems can improve newborn toxicology testing policies. Starting with the “why”—and only obtaining toxicology tests when they will answer a clinical question in an actionable way—is a must. Although teams caring for families affected by prenatal substance use must work within the constraints of larger sociolegal contexts, care teams are not powerless. Rather, the article serves as an important reminder of the impact of institutional policy and the importance of gathering data to assess the effect of policy changes on our ability to provide the most equitable and just care for all our patients.We acknowledge the strength of all patients and families affected by substance use and substance use disorder.
Liu et al. (Mon,) studied this question.