Assessing the clinical utility of toxicology testing in the peripartum period.

BACKGROUND: Toxicology testing is frequently used as a means of gathering objective data about substance use in pregnancy, but little is known about the clinical utility of testing in the peripartum setting. OBJECTIVE: This study aimed to characterize the utility of obtaining maternal-neonatal dyad toxicology testing at the time of delivery. STUDY DESIGN: We performed a retrospective chart review of all deliveries in a single healthcare system in Massachusetts between 2016 and 2020, and identified deliveries with either maternal or neonatal toxicology testing at delivery. An unexpected result was defined as a positive test for a nonprescribed substance that was not known on the basis of clinical history, self-report, or previous toxicology testing within a week of delivery, excluding results for cannabis. We evaluated the characteristics of maternal-infant dyads with unexpected positive results, unexpected positive results by rationale for testing, changes in clinical management after an unexpected positive test, and maternal outcomes in the year after delivery using descriptive statistics. RESULTS: Of the 2036 maternal-infant dyads with toxicology tests performed during the study period, there were 80 (3.9%) with an unexpected positive result. Diagnosis of substance use disorder with active use in the last 2 years was the clinical rationale for testing that yielded the greatest number of unexpected positive results (10.7% of total tests ordered for this rationale). Inadequate prenatal care (5.8%), maternal use of medication for opioid use disorder (3.8%), maternal medical indications such as hypertension or placental abruption (2.3%), history of substance use disorder in remission (1.7%), or maternal cannabis use (1.6%) yielded lower rates of unexpected results compared with a recent substance use disorder (within the last 2 years). Solely on the basis of findings from unexpected test results, 42% of dyads were referred to child protective services, 30% of dyads had no documentation of maternal counseling during delivery hospitalization, and 31% did not receive breastfeeding counseling after an unexpected test; 22.8% had monitoring for neonatal opioid withdrawal syndrome. Postpartum, 26 (32.5%) were referred to substance use disorder treatment, 31 (38.8%) attended a postpartum mental health visit, and only 26 (32.5%) attended a postpartum visit. Fifteen individuals (18.8%) were readmitted in the year after delivery, all for substance-related medical complications. CONCLUSION: Unexpected positive toxicology results at delivery were uncommon, particularly when tests were sent for frequently used clinical rationales for testing, suggesting a need to revisit guidelines surrounding appropriateness of indications for toxicology testing. The poor maternal outcomes in this cohort highlight a missed opportunity for maternal connection to counseling and treatment in the peripartum period.

Incidence of and Risk Factors for Failed Induction of Labor Using a Contemporary Definition.

OBJECTIVE: To estimate the incidence of failed induction of labor and the associated patient risk factors. METHODS: We performed a case-control study from a cohort of nulliparous women who delivered between 39 and 41 weeks of gestation after an induction of labor in one of seven hospitals. Cases of failed induction were defined using the Obstetric Care Consensus criteria (ie, cesarean delivery performed in early labor [less than 6 cm dilatation] after at least 12 hours of oxytocin administration from membrane rupture). For each case of failed induction, the next four women who did not meet the criteria for failed induction of labor were selected for the control group, matched by hospital. We identified characteristics associated with failed induction of labor using a multivariable conditional logistic regression that was constructed with backward stepwise method for variable selection. RESULTS: Across the hospitals, 4,123 of 10,175 nulliparous women were induced (40.5%), of whom 82 had a failed induction of labor (2.0%). A total of 328 women were selected for the matched control group. Baseline characteristics were similar between the groups. Compared with women in the control group, women with a failed induction were more likely to have a delivery body mass index (BMI) of 40 or higher (28.0 vs 8.2%, P<.001), shorter height (mean 63.9 vs 64.8 inches, P=.01), and closed cervix on admission (41.5 vs 24.1%, P=.002). Factors significantly associated with induction failure in the multivariable model included: 1) delivery BMI (adjusted odds ratio [aOR] 7.93, 95% CI 3.48-18.09, for BMI 40 or higher relative to BMI lower than 30, 2) height in inches (aOR 0.89, 95% CI 0.80-0.98), and 3) number of centimeters dilated on admission 2 or more (aOR 0.30, 95% CI 0.14-0.65). CONCLUSION: Failed induction of labor occurs infrequently. Risk factors include shorter height, BMI 40 or higher, and cervical dilatation of less than 2 cm on admission. Even so, most women with these risk factors will not experience failed labor induction.