770 - Risk-Based Criteria for Umbilical Cord Toxicology Testing: A Retrospective Cohort Study

Substance use has a major impact on the health of pregnant patients and newborns, leading to short and long-term consequences. Universal toxicology testing is not recommended by any major medical entity. Validated risk-based toxicology testing criteria have not been published. 

Objective: The objective was to determine the association between umbilical toxicology test results and the risk-based criteria used to obtain them.

Design/Methods:

This was a retrospective cohort study on umbilical toxicology tests obtained at a single tertiary care birthing center from 2016-2018. The study was approved by COMIRB.  Umbilical toxicology test results were abstracted from the medical chart. Data abstracted included indication for testing, maternal demographic characteristics, and clinical characteristics. The data collected was entered into a Red-Cap database.

Demographic and clinical characteristics were compared using two-sample t-tests, Wilcoxon Rank-Sum tests, Fisher’s exact tests and chi-squared tests for continuous, non-normally distributed continuous, and categorical variables respectively. The Cochran Armitage test was used to test for trend in positivity rate by the number of reasons the umbilical cord toxicology test was obtained. Analyses were conducted using SAS v 9.4. A p-value of < 0.05 was determined a priori as statistically significant.

Results:

Scant/late prenatal care was reported in 56% of mothers, however no association with positive tests was found. There was no statistical association between birthing persons with no prenatal care and a positive umbilical test result. Few (13%) birthing persons who verbally admitted to use of a substance had an umbilical toxicology test result that was positive for a different/additional substance. Isolated late prenatal care was not associated with a positive umbilical toxicology test result, but did represent a larger percentage of non-English speaking patients (18% language barrier patients in the entire cohort, 67% of non-English speaking birthing persons represented in the late prenatal care group, compared to only 26% of English-speaking birthing persons in the same group).  No single risk-based criterion for toxicology testing was associated with a positive toxicology test result. 

Conclusion(s):

Toxicology testing for birthing persons/newborns based upon isolated risk-based criteria are not predictive of positive test results, thus poorly identifying families affected by substance use. Certain risk-based criteria disproportionately affect patients of marginalized populations and likely introduce bias into the selection of patients for toxicology testing.