Symptom-Based Dosing In Neonatal Opioid Withdrawal

Key Takeaways

• In ESC-managed infants, symptom-based dosing was associated with faster medical readiness for discharge than scheduled tapering.
• In the ESC cohort, treatment initiation and overall length of stay did not differ significantly, and the Finnegan cohort showed no significant differences in time to medical readiness for discharge or length of stay, although symptom-based dosing was associated with a higher likelihood of initiating pharmacologic treatment.
• Safety events through 3 months were rare and similar between groups, while a substantial minority of treated ESC infants receiving symptom-based dosing converted to scheduled dosing.

The adjusted mean ratio was 0.79, with a 95% CI of 0.65 to 0.96 and P=.02. This comparison came from a broader cluster crossover trial across hospitals using either ESC or Finnegan-based withdrawal care.

The trial used a multicenter cluster crossover randomized design at 23 US hospitals in the HELP for NOWS Consortium, with run-in periods, 5-month treatment periods, and 3-week washouts. Sites were randomized by sequence across 15 ESC hospitals and 8 Finnegan hospitals within the network. Investigators enrolled 626 infants, with 383 in ESC and 243 in Finnegan; participants were at least 36 weeks’ gestation after documented second- or third-trimester opioid exposure. The primary endpoint was time from birth to medical readiness for discharge, defined as clinical discharge or study readiness at 96 hours and 48 hours after opioid dosing ended. The ESC analysis was powered, whereas the Finnegan analysis was a planned secondary analysis and was not powered to detect efficacy differences.

Within ESC, pharmacologic treatment started at similar rates, 0.40 versus 0.41, with an adjusted risk ratio of 0.99 and a 95% CI of 0.77 to 1.27. Overall stay was 10.91 versus 12.09 days, with an adjusted mean ratio of 0.90 and a 95% CI of 0.72 to 1.13. Among treated ESC infants, discharge readiness was 13.1 versus 17.56 days, with an adjusted mean ratio of 0.75 and a 95% CI of 0.61 to 0.91. In that treated ESC group, 35% required conversion to scheduled dosing because intermittent doses did not control withdrawal, while 65% avoided scheduled dosing. These findings were consistent with less opioid exposure for many treated infants without a significant reduction in overall hospital stay.

In Finnegan-based care, time to medical readiness for discharge was 15.99 versus 17.56 days, with an adjusted mean ratio of 0.91 and a 95% CI of 0.72 to 1.15. Length of stay was 17.38 versus 19.39 days, with an adjusted mean ratio of 0.90 and a 95% CI of 0.69 to 1.16. Receipt of pharmacologic treatment was higher with symptom-based dosing, 0.76 versus 0.63, with an adjusted risk ratio of 1.21 and a 95% CI of 1.03 to 1.42. Under symptom-based dosing, infants received a single opioid dose when a site threshold was met, with repeat doses only if symptoms remained above or returned to threshold. Conversion to scheduled tapering occurred after three doses within 24 hours or after two consecutive short-interval doses. Primary opioids included morphine, buprenorphine, or methadone by site practice, and secondary medications included phenobarbital and clonidine.

Safety outcomes through discharge and a 3-month safety follow-up did not differ significantly between groups, and inpatient critical events were rare. Critical inpatient events were nonaccidental trauma or death, while inpatient safety events included seizures or weight loss greater than 15% from birth weight. Outpatient events included urgent care, emergency department visits, or readmissions by 3 months, although some missing outpatient data were assumed absent.

The Finnegan cohort was not powered to detect efficacy differences, and contamination remained possible despite delayed training access and washout periods. Postdischarge ascertainment relied on records and public searches, and only ESC showed a shorter discharge-readiness interval without a significant reduction in length of stay.