Establishing Neonatal Toxicology Testing Protocols to Prioritize Clinical Utility Reduces Racial and Socioeconomic Bias

Toxicology testing is frequently performed on newborns with known or suspected prenatal substance exposure, even though research shows that these tests often add no new clinically relevant information to that gathered during pregnancy and are rarely clinically actionable. Yale New Haven Children's Hospital lacked a protocol for newborn toxicology testing. The authors suspected that test-ordering patterns demonstrated clinically inappropriate overuse and racial and socioeconomic discrimination, leading to harm to newborns and families, including unnecessary involvement of child protective services. Their hypothesis was that reduced testing, with continued social support, would lead to no change in newborn outcomes or safety. Prior to implementation, they identified wide disproportionality in testing by race and insurance status as a marker of poverty. Compared with all babies born at Yale New Haven Children's Hospital, Black non-Hispanic newborns and newborns with Medicaid insurance were tested at 2 times and 2.6 times the average frequency, respectively. After implementation, the authors evaluated the protocol's impact on the testing volume and racial and socioeconomic distribution of tests ordered, and tracked outcomes in the institution's newborn population. By the time the protocol went live in January 2022, after extensive discussion and education among staff, testing rates had already decreased from 3.2% to 0.3%. The absolute decrease in testing rates for Black newborns was 6.7% to 0.6%, while that for white newborns was 2.5% to 0.2%. The testing rate for newborns with Medicaid insurance decreased from 8.2% to 0.5%, while the rate for newborns with commercial insurance decreased from 0.4% to 0.1%. There were so few tests ordered after the protocol was in place that differences in testing frequency by race and type of insurance could not be statistically evaluated. These reduced testing rates have persisted since that time, with no further interventions. Importantly, there were no identified hospital readmissions for withdrawal symptoms or adverse events for newborns who had no previously identified prenatal substance exposure before or after the protocol was adopted. This work succeeded due to a close evaluation of the clinical utility of newborn toxicology testing, institutional support, interdisciplinary collaboration, a culture accepting of change, and recognition of quality improvement as a health equity tool.