. 2026 Apr 25;335(24):2114–2125. doi: 10.1001/jama.2026.5782

Symptom-Based Dosing for Neonatal Opioid Withdrawal

The OPTimize NOW Randomized Clinical Trial

Lori A Devlin ^1,^✉, Denise C Babineau ², Stephanie L Merhar ^3,⁴, Sara B DeMauro ^5,⁶, Walter K Kraft ⁷, Scott A Lorch ^5,⁶, Abhik Das ², Scott A McDonald ², Evan Rhodes ², Augusto F Schmidt ⁸, Lillian Trochinski ², Margaret Crawford ², Thitinart Sithisarn ⁹, Lawrence Leeman ¹⁰, Kelley Zagol Kovatis ¹¹, Namasivayam Ambalavanan ¹², Ryan W Smith ¹³, Sucheta Telang ¹⁴, Jennifer A Tioseco ¹⁵, Jennifer M McAllister ¹⁶, Scott L Wexelblatt ¹⁶, Bhanu Muniyappa ¹⁷, Patricia K Williams ¹⁸, Susan C Adeniyi-Jones ¹⁹, Crystal D Hill ^20,²¹, Tanner Wright ²², Gregory M Sokol ²³, Lynette Johnson ²⁴, Richard W Hall ²⁵, Scott D Duncan ²⁶, Karen Puopolo ^5,⁶, Krishna Dummula ^27,²⁸, Ann Anderson-Berry ²⁹, Jonathan M Davis ³⁰, Brenda Poindexter ³¹, Leslie W Young ³², for the HEAL Evaluation of Limited Pharmacotherapies for Neonatal Opioid Withdrawal Syndrome (HELP for NOWS) Consortium

¹Norton Children’s Neonatology, affiliated with the University of Louisville School of Medicine, Louisville, Kentucky

²Data Analytics and Digital Solutions, RTI International, Research Triangle Park, North Carolina

³Department of Pediatrics, University of Cincinnati College of Medicine and Perinatal Institute, Division of Neonatology, Cincinnati Children’s Hospital Medical Center, Good Samaritan Hospital, Cincinnati, Ohio

⁴St Elizabeth Healthcare Edgewood Hospital, Edgewood, Kentucky

⁵Division of Neonatology, Children’s Hospital of Philadelphia, Pennsylvania

⁶Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia

⁷Department of Pharmacology, Physiology & Cancer Biology, Thomas Jefferson University, Philadelphia, Pennsylvania

⁸Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland

⁹Department of Pediatrics, Golisano Children’s at University of Kentucky, Lexington

¹⁰Department of Family and Community Medicine, The University of New Mexico School of Medicine, Albuquerque

¹¹Division of Neonatology, Department of Pediatrics, ChristianaCare, Newark, Delaware

¹²Division of Neonatology, Department of Pediatrics, The University of Alabama at Birmingham

¹³Department of Pediatrics, University of Louisville School of Medicine/Norton Women’s and Children’s Hospital, Louisville, Kentucky

¹⁴Department of Pediatrics, University of Louisville School of Medicine/Norton Children’s Hospital, Louisville, Kentucky

¹⁵Children’s Hospital of Philadelphia Newborn Care, AtlantiCare Regional Medical Center, Pomona, New Jersey

¹⁶Department of Pediatrics, University of Cincinnati College of Medicine and Perinatal Institute, Division of Neonatology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

¹⁷Department of Pediatrics, Division of General Pediatrics, University of Utah School of Medicine, Salt Lake City

¹⁸Section of Neonatology, Department of Pediatrics, Oklahoma Children’s Hospital, OU Health, University of Oklahoma Health Center, Oklahoma City

¹⁹Department of Pediatrics, Sidney Kimmel Medical College, Thomas Jefferson University/Nemours Children’s Health, Philadelphia, Pennsylvania

²⁰Kettering Health, Kettering, Ohio

²¹Perinatal Institute, Division of Neonatology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

²²Department of Pediatrics, University of South Florida, Tampa

²³Department of Pediatrics, Indiana University School of Medicine, Indianapolis

²⁴University of Rochester School of Medicine & Dentistry, Rochester, New York

²⁵Department of Pediatrics/Neonatology, University of Arkansas for Medical Sciences, Little Rock

²⁶Department of Pediatrics, University of Louisville School of Medicine/University of Louisville Hospital, Louisville, Kentucky

²⁷Department of Pediatrics, University of Kansas Medical Center, Kansas City

²⁸Children’s Mercy Hospital, Kansas City, Missouri

²⁹Department of Pediatrics, University of Nebraska Medical Center, Omaha

³⁰Department of Pediatrics and Tufts Clinical and Translational Science Institute, Tufts Medical Center, Boston, Massachusetts

³¹Division of Neonatology, Emory University School of Medicine, Atlanta, Georgia

³²Larner College of Medicine at the University of Vermont, Burlington

Group Information: The members of the HEAL Evaluation of Limited Pharmacotherapies for Neonatal Opioid Withdrawal Syndrome (HELP for NOWS) Consortium appear in Supplement 4.

Accepted for Publication: April 6, 2026.

Published Online: April 25, 2026. doi:10.1001/jama.2026.5782

^✉

Corresponding Author: Lori A. Devlin, DO, Department of Pediatrics, University of Louisville School of Medicine, 571 S Floyd St, Ste 342, Louisville, KY 40202 (lori.devlinphinney@louisville.edu).

Author Contributions: Dr Babineau and Mr McDonald had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Devlin, Babineau, Merhar, DeMauro, Kraft, Lorch, Trochinski, Smith, McAllister, Wexelblatt, Adeniyi-Jones, Sokol, Hall, Davis, Poindexter, Young.

Acquisition, analysis, or interpretation of data: Devlin, Babineau, Merhar, DeMauro, Kraft, Lorch, Das, McDonald, Rhodes, Schmidt, Crawford, Sithisarn, Leeman, Zagol Kovatis, Ambalavanan, Telang, Tioseco, Muniyappa, Williams, Hill, Wright, Sokol, Johnson, Hall, Duncan, Puopolo, Dummula, Anderson Berry, Young.

Drafting of the manuscript: Devlin, Babineau, Kraft, Rhodes, Davis, Young.

Critical review of the manuscript for important intellectual content: Babineau, Merhar, DeMauro, Kraft, Lorch, Das, McDonald, Schmidt, Trochinski, Crawford, Sithisarn, Leeman, Zagol Kovatis, Ambalavanan, Smith, Telang, Tioseco, McAllister, Wexelblatt, Muniyappa, Williams, Adeniyi-Jones, Hill, Wright, Sokol, Johnson, Hall, Duncan, Puopolo, Dummula, Anderson Berry, Davis, Poindexter, Young.

Statistical analysis: Devlin, Babineau, Das, McDonald, Rhodes.

Obtained funding: Devlin, Merhar, DeMauro, Kraft, Das, Ambalavanan, Telang, Adeniyi-Jones, Wright, Sokol, Young.

Administrative, technical, or material support: Babineau, Kraft, Schmidt, Trochinski, Crawford, Sithisarn, Ambalavanan, Telang, Williams, Adeniyi-Jones, Hill, Wright, Hall, Duncan, Dummula, Davis, Poindexter.

Supervision: Babineau, Merhar, Lorch, Das, Trochinski, Leeman, Zagol Kovatis, Smith, Telang, Wexelblatt, Adeniyi-Jones, Hill, Wright, Sokol, Johnson, Hall, Duncan, Puopolo, Davis.

Conflict of Interest Disclosures: Dr Devlin reported receiving grants from the National Institutes of Health (NIH) during the conduct of the study. Dr Babineau reported receiving grants from NIH during the conduct of the study. Dr Merhar reported receiving grants from NIH during the conduct of the study. Dr DeMauro reported receiving grants from NIH during the conduct of the study and outside the submitted work. Dr Kraft reported receiving grants from Thomas Jefferson University (UG1HD107628) during the conduct of the study; grants from Merck, Pfizer, and Rhoshan; personal fees from AbbVie, Johnson & Johnson, and Sun Pharma outside the submitted work; and having a patent pending (US-20200330455-A1). Dr Lorch reported receiving grants from NIH during the conduct of the study. Dr Das reported receiving grant funding from NIH to support this work. Mr McDonald reported receiving grant funding from NIH to support this work. Mr Rhodes reported receiving grant funding from NIH to support this work. Ms Crawford reported receiving grants from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) during the conduct of the study. Dr Sithisarn reported receiving grants from NIH during the conduct of the study; grants from NIH Non-Opiate Treatment After Prenatal Opiate Exposure to Prevent Postnatal Injury to the Young Brain (No-POPPY), National Institute on Drug Abuse POPI: Placenta, Opioids and Perinatal Implications study, and Research Triangle Institute/NIH Treat NOW Study – Target Trial Emulation for Pharmacologic Treatment of Neonatal Opioid Withdrawal Syndrome outside the submitted work. Dr Ambalavanan reported receiving grants from NIH during the conduct of the study and having a patent pending for vital signs analysis for identification of opioid exposure in infants. Dr Smith reported receiving grants from NIH during the conduct of the study. Dr Wexelblatt reported receiving personal fees from Sanofi and Abbott for serving on the speakers bureau outside the submitted work. Dr Williams reported receiving grants from NIH during the conduct of the study. Dr Adeniyi-Jones reported receiving grants from NIH during the conduct of the study. Dr Hill reported receiving grants from NIH during the conduct of the study. Dr Wright reported receiving grants from NIH during the conduct of the study and grants from US Centers for Disease Control and Prevention, Hillsborough County, and NIH Helping to End Addiction Long-Term (HEAL) outside the submitted work. Dr Sokol reported receiving grants from NIH during the conduct of the study. Dr Hall reported receiving grants from NIH during the conduct of the study. Dr Duncan reported receiving grants from Medtronic, Pfizer, and Oak Hill Bio outside the submitted work. Dr Puopolo reported receiving grants from NIH during the conduct of the study. Dr Dummula reported receiving grants from NIH during the conduct of the study. Dr Davis reported receiving grants from NIH to his institution during the conduct of the study and consulting fees from Inflection Medicine outside the submitted work. Dr Young reported receiving grants from NIH during the conduct of the study. No other disclosures were reported.

Funding/Support: The NICHD and the NIH HEAL Initiative provided funding via cooperative agreements (U24 HD107621, UG1 HD107580, UG1 HD107616, UG1 HD107627, UG1 HD107628, UG1 HD107631, UG1 HD107649, UG1 HD107650, UG1 HD107653).

Role of the Funder/Sponsor: NICHD staff had input into the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and/or approval of the manuscript; and decision to submit the manuscript for publication.

Group Information: The HEAL Evaluation of Limited Pharmacotherapies for Neonatal Opioid Withdrawal Syndrome (HELP for NOWS) Consortium members are listed in Supplement 4.

Disclaimer: While NICHD staff had input, the comments and views of the authors do not necessarily represent the views of NICHD, the NIH, the US Department of Health and Human Services, or the US government.

Meeting Presentation: This article was presented at the 2026 Meeting of the Pediatric Academic Societies; April 25, 2026; Boston, Massachusetts.

Data Sharing Statement: See Supplement 5.

Additional Contributions: We extend our deepest gratitude to the infants and caregivers/parents who participated in this study, without whom the study and resulting implications for the medical community would not have been possible. We also extend our gratitude to the HELP for NOWS Patient and Community Engagement Committee, which provided feedback on the study design and assisted in the development of community-facing study materials. We are indebted to our medical and nursing colleagues who agreed to take part in the study. The following investigators and research coordinators, in addition to those listed as authors, participated in this study: Crystal Bass, MD, Elizabeth Hendricks, DNP (AtlantiCare Hospital); David Paul, MD, Jennifer Chambers, BSN, RN, CCRC, Amy Mackley, MSN, CNS-BC, CCRC (ChristianaCare); Ward Rice, MD, Cathy Grisby, BSN, CCRC, David Russell, JD, Lisa Tully, MA, Traci Beiersdorfer, BSN, RN, RNC-NIC; Carla King, BSN, RN, Cynthia Reid, BSN, RN (Cincinnati Children’s Hospital Medical Center, University of Cincinnati Medical Center, Good Samaritan Hospital, and St Elizabeth Healthcare [UG1 HD107616]); Stephanie Wilson Archer, MA (NICHD); Henrietta Bada, MD, MPH, Mina Hanna, MD, MSc, Amanda Wilburn, CCRP, Susan DeGraaff, CCRP, CPC-A, Beth McKinney-Whitlock, CCRP, Carrie Hobbs, RN, MSN, CCRP, Crystal Wilson, LPN (Golisano Children’s at University of Kentucky); Anna E. Thomas, MD, Hannah Rakow, BSN, RN, MBA-HM, Jessica Mendoza, BSN, RN, Betsy Tudor, CCMA (Indiana University, Sidney & Lois Eskenazi Hospital [UG1 HD107653]); Loren Nallu, MD, Mary Connolly, PhD, ACRP-CP, Allison Dymacek, MACPR, BSN, RN, Karen Herzing, RN (Kettering Health); Natalie Goodman, BA, Erin Bohon, LPN, Aprill Shockley, BSN, RN (University of Oklahoma Health Center); Emily Smith, MPH, Jenny Auman, BS, Anna Mazur, BS, Shannon Barnes, MSN, CCRP, Brittany Thomas, BS, Eugene Turner, BS, James Pickett, BS, David Leblond, BS, Savannah Moore (RTI International [U24 HD107621]); Courtenay Fulmor, RN, Lisa Pullaro, Emily J. Murray, RN, Megan Lennon Kundrat, RN, Angela Pallotto, BSN, RN, Melette Pearson, RN (Thomas Jefferson University [UG1 HD107628]); Myriam Peralta-Carcelen, MD, MPH, Waldemar A. Carlo, MD, Samuel Gentle, MD, Sandra Turner, BSN, Rachel Benz, MSN, RN, Sharon Owen, RN (University of Alabama at Birmingham Health System and Children’s of Alabama [UG1 HD107580]); Vallon Williams, DNP, APRN, AGNP-C, Allyson Cheathem, BSN, RN, CRS, Alperdis Keyes, BSN, RN, CRS (University of Arkansas for Medical Sciences [UG1 HD107650]); Kristina Foster, RN, MS, APRN-BC, CCRP, Allison Scott, RN, CCRC (University of Kansas Medical Center); Madison Jessie, BSN, RN, Sarah Morris, BSN, RN, CCRC, Jennifer Nason, BSN, RN, CCRC, Audrey Miller, BSN, RN, Rachel Welch-Miles, BSN, RN, Molly Hemmerle, BSN, RN (University of Louisville, Norton Children’s Research Institute affiliated with the University of Louisville School of Medicine [UG1 HD107649]); Betty Oberle, RN, Matt VanOrmer, PhD, Sarah Newman, NNP, APRN, DNP, Emilee Kendall, BS, Kym Abraham, RN (University of Nebraska Medical Center); Sofia Markee, MD, Nicole Yonke, MD, Jessie Maxwell, MD, MBA, Sandra Beauman, MSN, RNC-NIC, Conra Lacy, RN, Ashley Palmer, BSN, RN, Elizabeth Kuan, BSN, RN (University of New Mexico [UG1 HD107627]); Megan Dhawan, MSN, CRNP, Mary Catherine Gambacorta, BSN, RN, Jon Snyder, RN, Melanie Risch, BSN, RN, Hanna Rosewood, BA (University of Pennsylvania and Children’s Hospital of Philadelphia [UG1 HD107631]); Ronnie Guillet, MD, PhD, Julie Riccio, MD, Melissa Moreland, MPH, Rachel Jones (University of Rochester Medical Center, Golisano Children’s Hospital); Marcia Kneusel, RNC, CCRC, Courtney Casey, RN (University of South Florida); and Robin K. Ohls, MD, Camille Fung, MD, Carrie A. Rau, RN, CCRC, BSN, Kathie Coleman, RN, Kandace M. McGrath, BS, Manndi C. Loertscher, BS, CCRP (University of Utah Medical Center). All study sites not specifically associated with an NIH grant received compensation to conduct the study via a subcontract with RTI International under their U24 grant.

^✉

Corresponding author.

PMCID: PMC13110413 NIHMSID: NIHMS2188718 PMID: 42033722

Key Points

Question

Does pharmacologic treatment with a symptom-based dosing approach decrease time to medical readiness for discharge compared with a scheduled opioid taper for infants with opioid withdrawal cared for with the Eat, Sleep, Console approach?

Findings

In this cluster, crossover randomized clinical trial including 383 infants cared for with the Eat, Sleep, Console approach, time to medical readiness for discharge was 2.3 days shorter for infants treated with symptom-based dosing compared with a scheduled opioid taper. No difference was noted in safety outcomes through 3 months of age.

Meaning

Symptom-based dosing decreased time to medical readiness for discharge for infants with opioid withdrawal cared for with the Eat, Sleep, Console approach.

Abstract

Importance

Infants with neonatal opioid withdrawal syndrome (NOWS) who receive pharmacologic treatment are traditionally treated with a scheduled opioid taper. An alternate approach, symptom-based dosing, may better align treatment with withdrawal severity.

Objective

To compare time from birth to medical readiness for discharge for infants with moderate to severe withdrawal treated with either a symptom-based dosing or scheduled opioid taper approach.

Design, Setting, and Participants

In this cluster, crossover randomized clinical trial with run-in period, 23 US hospitals cared for infants using the Eat, Sleep, Console approach (ESC) or Finnegan-based care (a comprehensive scoring system to quantify severity of symptoms; 15 ESC and 8 Finnegan hospitals) and their preferred primary opioid. Opioid dosing was guided by study-approved, site-specific algorithms. Infants with NOWS with a gestational age at birth of at least 36 weeks and at risk for pharmacologic treatment were enrolled between March 25, 2024, and April 9, 2025, with the last 3-month assessment on July 15, 2025. Sample size analyses were conducted between August 1, 2024, and September 23, 2024.

Intervention

Sites were randomized to 1 of 2 sequences: (1) symptom-based dosing followed by scheduled opioid taper or (2) scheduled opioid taper followed by symptom-based dosing.

Main Outcome and Measure

Time from birth to medical readiness for discharge.

Results

Of the 626 enrolled infants (mean [SD] gestational age, 38 [1] weeks; 49% male), 383 were cared for with ESC (primary outcome cohort). The mean time to medical readiness for discharge was significantly shorter in the symptom-based dosing group compared with the scheduled opioid taper group (9.18 vs 11.61 days; adjusted mean ratio [aMR], 0.79 [95% CI, 0.65-0.96]). There was no difference in the risk for initiation of pharmacologic treatment (0.4 vs 0.41; adjusted risk ratio, 0.99 [95% CI, 0.77-1.27]) or length of stay (10.91 vs 12.09 days; aMR, 0.9 [95% CI, 0.72-1.13]) between groups. For infants in the symptom-based group, 35% (95% CI, 25%-45%) required scheduled opioid dosing due to withdrawal severity that was not controlled with intermittent dosing. In the Finnegan cohort (n = 243; planned secondary outcome), there were no significant differences in time to medical readiness for discharge (15.99 vs 17.56 days; aMR, 0.91 [95% CI, 0.72-1.15]) or length of stay (17.38 vs 19.39 days; aMR, 0.9 [95% CI, 0.69-1.16]). The inpatient composite safety outcome occurred rarely (in the ESC cohort, 3 of 188 in the symptom-based dosing vs 2 of 195 in the scheduled opioid taper groups).

Conclusions and Relevance

Symptom-based dosing decreased time to medical readiness for discharge compared with a scheduled opioid taper approach among infants cared for with ESC.

Trial Registration

ClinicalTrials.gov Identifier: NCT05980260

This crossover randomized clinical trial examines the effect of symptom-based dosing vs scheduled opioid taper on time to medical readiness for discharge in infants with neonatal opioid withdrawal syndrome treated with the Eat, Sleep, Console approach.

Introduction

As the opioid epidemic persists, opioid use during pregnancy remains a substantial public health concern. In the US, 1 infant is diagnosed with neonatal opioid withdrawal syndrome (NOWS) approximately every 27 minutes.¹ Acute withdrawal typically occurs during the first week after birth and is assessed and managed with either the Eat, Sleep, Console approach (ESC)^2,3 or Finnegan-based care.⁴ Among other aspects of care, the Finnegan-based approach focuses on detailed scoring of the signs of withdrawal, whereas ESC emphasizes the functional well-being (eating, sleeping, and ability to be consoled) of the infant. Although both care models are currently used in clinical practice, results from a recent multicenter randomized clinical trial demonstrated a substantial decrease in the length of hospital stay and receipt of pharmacologic treatment when infants were cared for with ESC, supporting the use of this approach.^3,5,6

Nonpharmacologic care is the first line of treatment for infants with NOWS.^3,7 However, if infants do not adequately respond to these interventions, pharmacologic treatment is indicated. The 2 most frequently used approaches for the pharmacologic treatment of NOWS are the traditional scheduled opioid taper approach and a newer symptom-based dosing approach. With the scheduled taper approach, low-dose opioids are initiated for infants meeting a severity threshold and escalated until the signs of withdrawal improve. Opioid dosing is then slowly weaned while withdrawal severity is monitored. With the symptom-based dosing approach, pharmacologic treatment is provided using a single opioid dose when a specified withdrawal threshold is met. Monitoring of withdrawal severity is continued, and additional opioid doses are administered only if the infant’s withdrawal severity remains above or escalates back to this treatment threshold. With consistent application, symptom-based dosing may decrease postnatal opioid exposure and truncate the duration of hospitalization for infants whose withdrawal is adequately controlled with intermittent opioid dosing.

To date, there are limited data to support the use of 1 dosing approach over the other. Retrospective studies and quality improvement initiatives from single centers and small regional collaboratives have evaluated symptom-based dosing, but only a limited number of these initiatives have assessed adoption of symptom-based dosing independent of other changes in clinical practice.^{8,9,10,11,12,13,14} Symptom-based dosing has been associated with improvement in hospital-based outcomes without an increase in the rate of hospital readmission. Although these findings are promising, the efficacy, safety, and generalizability of a symptom-based dosing approach has yet to be rigorously studied.

To address this knowledge gap, the Optimizing Pharmacologic Treatment for Neonatal Opioid Withdrawal Syndrome (OPTimize NOW): A Symptom-Based Dosing Approach study (NCT05980260) was developed. This trial was conducted by the National Institutes of Health Helping to End Addiction Long-Term (HEAL) Evaluation of Limited Pharmacotherapies for Neonatal Opioid Withdrawal Syndrome (HELP for NOWS) Consortium to evaluate the safety, efficacy, and generalizability of a symptom-based dosing approach compared with a traditional scheduled opioid taper approach for infants with NOWS assessed as at risk for pharmacologic treatment.

Methods

Trial Design and Oversight

This multicenter, stratified, cluster randomized clinical trial with 1 cluster (Finnegan) subordinate to the other (ESC) was conducted at 23 US sites in the HELP for NOWS Consortium between March 25, 2024, and April 9, 2025 (last 3-month assessment on July 15, 2025). Advarra served as the single institutional review board, with reliance agreements for all sites. The study was conducted with a waiver of informed consent approved by the institutional review board in accordance with the Code of Federal Regulations (45 CFR 46.116).¹⁵ CONSORT reporting guidelines were followed. An independent data and safety monitoring board, appointed by the director of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, provided oversight for this trial.

Patients and Participating Sites

To enhance study generalizability, we selected geographically and demographically diverse sites that included academic and community hospitals (eFigure in Supplement 1). Sites cared for a varying number of infants with opioid withdrawal in several different hospital care settings (eg, neonatal intensive care units [open bay and single room], newborn nurseries, and pediatric floors). To reflect current clinical practice while acknowledging the increased use of ESC, we included hospitals that used either ESC or Finnegan-based care for the management of NOWS but powered the primary analysis on the ESC cohort.

We enrolled infants who were at least 36 weeks’ gestation at birth, had documented opioid exposure during the second and/or third trimester of pregnancy, were assessed for opioid withdrawal at a participating study site, and were determined to be at risk for pharmacologic treatment (ie, at least 1 Finnegan score ≥8 or at least 1 “yes” for difficulty with eating, sleeping, or consoling on ESC assessment). Complete eligibility criteria have been previously published.¹⁵ All outcomes are reported separately for the ESC and Finnegan cohorts.

Randomization

Sites were randomized 1:1 to 1 of 2 sequences: (1) the symptom-based dosing approach (5 months) followed by a scheduled opioid taper approach (5 months) or (2) a scheduled opioid taper approach (5 months) followed by the symptom-based dosing approach (5 months). Each sequence began with a 5-week run-in period and had a 3-week washout prior to transitioning dosing approaches. Randomization was stratified by the assessment approach used at sites (ie, ESC or Finnegan). Randomization included 16 ESC and 8 Finnegan sites.

Intervention

During the scheduled opioid taper approach, pharmacologic management was per the site’s routine scheduled taper algorithm, including the threshold used for initiation of pharmacologic treatment (using either the ESC or Finnegan-based assessments); the strategy for weaning; and monitoring for worsening signs of withdrawal prior to hospital discharge. Site-specific algorithms were reviewed by the study team and assessed as being within currently accepted practice.

During the symptom-based dosing approach (Figure 1) the same site-specific treatment thresholds were used; however, rather than initiating scheduled dosing when this threshold was met, infants were given a single opioid dose, and withdrawal severity was monitored. The dosage provided for this as-needed dose was the same as the initiation dose used by the site for their scheduled opioid taper. Subsequently, infants did not receive additional opioid doses unless their signs of withdrawal again met the site threshold (either remaining at or escalated back to) for pharmacologic treatment. Each time the threshold for treatment was met, an as-needed dose was given as long as the infant had not already received 3 opioid doses in the preceding 24-hour period or 2 consecutive short-interval doses (morphine dosed every 2 hours, buprenorphine every 3 hours, or methadone every 4 hours). If the infant had already received 3 opioid doses in the preceding 24 hours, or 2 consecutive short-interval doses, the infant’s withdrawal severity warranted scheduled dosing. In these cases, the site’s scheduled opioid taper was initiated, starting with the first escalation dose on the treatment algorithm, and this scheduled opioid taper was followed for the duration of treatment.

Throughout this trial, sites used a consistent assessment method (ESC or Finnegan) and their preferred primary opioid (morphine, buprenorphine, or methadone) for pharmacologic treatment. Opioid dosing was guided by study-approved, site-specific algorithms. All sites administered primary opioid and secondary medication (phenobarbital and clonidine) doses that were within ranges commonly used in clinical practice and agreed on by the protocol team, as detailed in the trial protocol (Supplement 2).¹⁵

Outcomes

The primary outcome was time from birth to medical readiness for discharge among infants cared for with the ESC approach (ESC cohort). Infants were considered to be medically ready for discharge at the time they were discharged by a medical team or when they met the study criteria for medical readiness (ie, at least 96 hours of age and 48 hours from their last dose of opioid medication), whichever occurred first. This definition of medical readiness for discharge was informed by our previous work³ and the American Academy of Pediatrics 2020 published standards.⁷

Secondary outcomes included receipt of pharmacologic treatment and length of hospital stay for all infants. Among infants who received pharmacologic treatment, time to medical readiness for discharge, number of primary opioid doses, days of primary opioid administration, length of opioid treatment, and receipt of secondary medications were evaluated. An important secondary outcome specific to the symptom-based dosing group was the proportion of infants who required initiation of scheduled opioid dosing due to persistently elevated signs of withdrawal, despite intermittent dosing. Similarly, in the scheduled opioid taper group, the proportion of infants who required discontinuation of opioids due to oversedation was assessed. Time to medical readiness for discharge in the Finnegan cohort was also a secondary study outcome.

Safety outcomes were assessed for all enrolled infants and for those who received pharmacologic treatment. Safety outcomes included a composite inpatient safety outcome (seizures or weight loss >15% from birth weight); an outpatient composite safety outcome (acute/urgent care or emergency department visits or hospital readmissions) assessed at 3 months of age; and a composite critical safety outcome (nonaccidental trauma or death), which was measured at hospital discharge and at 3 months of age. Outcomes after hospital discharge were assessed through a review of electronic medical records and through a search of public records (eg, news reports, obituaries, and registries).³

Statistical Considerations

Sample size analyses were conducted between August 1, 2024, and September 23, 2024. The sample size was calculated with data from the INFORM NOW: Informing NOWS Research Through Retrospective Data Collection study, a multicenter chart review study designed to inform clinical trial development in the HELP for NOWS Consortium, using the approach of Hemming.¹⁶ This resulted in a sample size of 320 infants to detect a 21% difference in the time to medical readiness for discharge between the intervention groups, with greater than 80% power, using a 2-sided type I error rate of .05.

All efficacy outcomes were analyzed separately for the ESC and Finnegan cohorts, using an intention-to-treat approach according to the site’s randomized dosing sequence. Safety outcomes were also analyzed separately, using an as-treated approach.

Maternal and infant characteristics were summarized for each dosing group, with frequencies and percentages for categorical variables and means (SDs) or medians (IQRs) for continuous variables, and were compared using Pearson χ² tests, Fisher exact tests, or Wilcoxon 2-sample tests, as appropriate.

Study design–adjusted analyses were performed for each outcome, using a generalized linear mixed-effects model with a negative binomial (for count outcomes) or Poisson distribution (for binary outcomes). Each model included fixed effects for the dosing approach and study period, as well as a random effect for site. Small sample corrections using robust standard errors accounted for the small number of sites.^17,18 An interaction term between dosing approach and study period was also added to determine if the effect of dosing approach varied by study period. Models of count outcomes were used to estimate the study design–adjusted means (95% CIs) in each dosing approach as well as adjusted mean ratios (aMRs; 95% CI) and mean differences (95% CI) between the 2 dosing approaches. Models of binary outcomes were used to estimate the study design–adjusted risks (95% CIs) in each dosing approach as well as adjusted risk differences (95% CIs) between the 2 dosing approaches. To estimate adjusted mean and risk differences and the variances thereof, 1000 bootstrap samples were used, with replacement within sites.

Covariate-adjusted analyses were conducted to determine the impact of prespecified baseline characteristics on the effect size for each outcome by adding the following fixed effects to the study design–adjusted models: adequate prenatal care,¹⁹ use of medication for opioid use disorder during pregnancy,²⁰ concurrent nonopioid psychotropic substance use during pregnancy,¹⁹ infant sex,²¹ and birth weight.²² Additionally, a prespecified sensitivity analysis was performed for the primary outcome that excluded infants transferred prior to discharge, outlined in the statistical analysis plan (Supplement 3). A number needed to treat was calculated to highlight the clinical relevance of study findings.

A prospectively planned interim efficacy analysis was performed when approximately 50% of enrolled infants had met the primary outcome, resulting in a 2-sided significance level of P = .045 for the final analysis to ensure an overall type I error rate of .05 (Supplement 1). Secondary outcomes were not adjusted for multiple comparisons. All analyses were completed using SAS version 9.4 (SAS Institute).

Results

Patients

Of the 626 enrolled infants, 383 were cared for with ESC (ESC cohort) and included in the primary analysis and 243 were cared for with Finnegan-based care (Finnegan cohort) and included in secondary analyses (Figure 2). The ESC cohort was composed of 189 infants in the symptom-based dosing group and 194 infants in the scheduled opioid taper group. The Finnegan cohort included 119 infants in the symptom-based dosing group and 124 infants in the scheduled opioid taper group. Maternal and infant characteristics for the ESC and Finnegan cohorts were similar between intervention groups, with the exception of nicotine exposure in the Finnegan cohort (Table 1).

Figure 2. — ^aIncluded major birth defects (n = 8 [6 ESC; 2 Finnegan]), neonatal encephalopathy (n = 2 [ESC]), postnatal opioid exposure prior to neonatal opioid withdrawal syndrome treatment (n = 9 [3 ESC; 6 Finnegan]), receipt of respiratory support at 48 hours (n = 60 [42 ESC; 18 Finnegan]), or outborn and received pharmacologic treatment at transferring hospital (n = 15 [2 ESC; 13 Finnegan]); multiple exclusion criteria (n = 6 [5 ESC; 1 Finnegan]).

^bIncluded no antenatal exposure identified (n = 4 [ESC]), did not meet the study definition of at risk for pharmacologic treatment (n = 11 [6 ESC; 5 Finnegan]), and gestational age at birth less than 36 weeks (n = 27 [19 ESC; 8 Finnegan]).

^cThe critical safety outcome (nonaccidental trauma and death) was assessed through electronic medical record review and public record search for all infants. Data for the outpatient safety outcome (acute/urgent care or emergency department visits and hospital readmissions) were not recorded for 153 infants. These infants were evenly distributed between dosing groups in the ESC cohort (symptom-based dosing, 50 [26%]; scheduled opioid taper, 49 [25%]) and Finnegan cohort (symptom-based dosing, 27 [23%]; scheduled opioid taper, 27 [22%]). If data were not recorded, the study assumption was that the event did not occur.

ESC indicates Eat, Sleep, Console approach.

Table 1. Baseline Characteristics of the Eat, Sleep, Console (ESC) and Finnegan Cohorts.

Characteristic	ESC cohort, No. (%)		Finnegan cohort, No. (%)
Hospital characteristics^a
Total No.	15		8
Type of site
Academic	12 (80)		6 (75)
Nonacademic	3 (20)		2 (25)
Volume of infants contributed to study
High (≥30)	5 (33)		3 (38)
Medium (20-29)	6 (40)		2 (25)
Low (<20)	4 (27)		3 (38)
Level of NICU
4	5 (33)		2 (25)
3	10 (67)		6 (75)
	Symptom-based (n = 189)	Scheduled taper (n = 194)	Symptom-based (n = 119)	Scheduled taper (n = 124)
Maternal characteristics
Age, median (IQR), y	32 (29-36)	32 (29-35)	32 (28-35)	31 (29-35)
Race^b
Black	24 (13)	32 (16)	18 (15)	19 (15)
White	148 (78)	150 (77)	93 (78)	97 (78)
Other^c	17 (9)	12 (6)	8 (7)	8 (6)
Hispanic or Latino ethnicity, No./total No. (%)^b^,^d	34/184 (18)	35/188 (19)	2/110 (2)	8/119 (7)
RUCA status metropolitan^e	174 (92)	175 (90)	78 (66)	86 (69)
Married, No./total No. (%)^d	22/179 (12)	26/193 (13)	21/110 (19)	23/115 (20)
Public insurance	162 (86)	169 (87)	97 (82)	91 (73)
Adequate prenatal care, No./total No. (%)^d	118/178 (66)	119/185 (64)	69/100 (69)	71/108 (66)
Prescribed medications for treatment of opioid dependency, No./total No. (%)^d	151/187 (81)	144/192 (75)	96/115 (83)	103/121 (85)
Buprenorphine	58/151 (38)	53/144 (37)	39/96 (41)	42/103 (41)
Methadone	65/151 (43)	55/144 (38)	32/96 (33)	24/103 (23)
Buprenorphine and naloxone	28/151 (19)	35/144 (24)	25/96 (26)	37/103 (36)
Other^f	0/151	1/144 (1)	0/96	0/103
Polysubstance use in the second or third trimester	140 (74)	154 (79)	96 (81)	94 (76)
No. of substances used, median (IQR)	1 (0-2)	1 (1-2)	1 (1-2)	1 (1-2)
Nicotine exposure in the second or third trimester^g	118 (62)	123 (63)	90 (76)	79 (64)
Infant characteristics
Inborn	184 (97)	187 (96)	94 (79)	96 (77)
Gestational age at delivery, mean (SD), wk	38 (1)	38 (1)	38 (1)	38 (1)
Sex
Male	87 (46)	98 (51)	59 (50)	62 (50)
Female	102 (54)	96 (49)	60 (50)	62 (50)
Birth weight, mean (SD), kg	3.0 (0.5)	3.1 (0.5)	3.0 (0.4)	3.0 (0.5)
Head circumference at birth, mean (SD), cm	33.4 (1.7)	33.4 (1.5)	33.2 (1.6)	33.5 (1.8)
Length at birth, mean (SD), cm	48.8 (2.8)	48.9 (2.7)	48.3 (2.5)	48.2 (3.0)
Breastfed during initial hospitalization, No./total No. (%)^d^,^h	18/79 (23)	19/81 (23)	31/89 (35)	24/80 (30)
Neonatal toxicology screening	144 (76)	148 (76)	112 (94)	110 (89)
Positive toxicology results, No./total No. (%)^d	130/144 (90)	132/148 (89)	103/112 (92)	105/110 (95)
Positive toxicology results, excluding opioids, No./total No. (%)^d	71/144 (49)	80/148 (54)	38/112 (34)	49/110 (45)

Open in a new tab

Abbreviations: NICU, neonatal intensive care unit; RUCA, Rural-Urban Commuting Area Code.

^{^a}

One ESC site withdrew after randomization and prior to enrollment.

^{^b}

Maternal race and ethnicity were obtained from the electronic medical record. Race and ethnicity were included because they have been associated with neonatal opioid withdrawal syndrome severity. The National Institutes of Health definition of race and ethnicity was used to categorize these characteristics.

^{^c}

In ESC, included American Indian or Alaska Native (n = 4), multiracial (n = 6), and unknown (n = 19). In Finnegan, included Asian (n = 1), multiracial (n = 4), Native Hawaiian or Other Pacific Islander (n = 1), and unknown (n = 10).

^{^d}

Denominators are provided for variables with missing data. For analyses using these variables, missing data were grouped with “no” responses.

^{^e}

Defined as RUCA codes 1 to 3.

^{^f}

Included 1 infant who received both buprenorphine and methadone, but duration prior to delivery was unclear.

^{^g}

P < .05 in the Finnegan cohort.

^{^h}

Breastfeeding data were only collected in infants who were pharmacologically treated.

Primary Outcome

The mean length of time to medical readiness for discharge was significantly shorter for infants who received the symptom-based dosing approach compared with those who received the scheduled opioid taper approach (9.18 vs 11.61 days; aMR, 0.79 [95% CI, 0.65-0.96]; P = .02) (Table 2). Results were similar when adjusted for baseline covariates. The effect size was similar when transferred infants were excluded from the model for the primary outcome and for other planned sensitivity analyses (Supplement 1).

Table 2. Primary and Secondary Outcomes for Infants Assessed and Cared for With the Eat, Sleep, Console Approach (ESC).

Outcome	Adjusted analysis (95% CI)
	Study design^a				Covariate^b
	Symptom-based mean or risk	Scheduled taper mean or risk	Mean or risk ratio	Mean or risk difference	Symptom-based mean or risk	Scheduled taper mean or risk	Mean or risk ratio	Mean or risk difference
Primary outcome
Time to medical readiness for discharge, d^c	9.18 (7.51 to 11.23)	11.61 (10.03 to 13.43)	0.79 (0.65 to 0.96)	−2.3 (−3.96 to −0.62)	9.2 (7.88 to 10.75)	11.41 (10.1 to 12.89)	0.81 (0.69 to 0.94)	−2.12 (−3.61 to −0.57)
Secondary outcomes
Length of stay, d^c	10.91 (8.98 to 13.27)	12.09 (9.98 to 14.63)	0.9 (0.72 to 1.13)	−1.1 (−3.30 to 1.26)	10.85 (9.19 to 12.81)	12.01 (10.03 to 14.37)	0.9 (0.74 to 1.1)	−1.09 (−3.13 to 0.95)
Receipt of pharmacologic treatment^d	0.4 (0.3 to 0.54)	0.41 (0.33 to 0.51)	0.99 (0.77 to 1.27)	−0.01 (−0.10 to 0.08)	0.42 (0.31 to 0.58)	0.41 (0.32 to 0.53)	1.04 (0.86 to 1.26)	0.01 (−0.08 to 0.10)
Secondary outcomes in pharmacologically treated infants
Time to medical readiness for discharge, d^c	13.1 (9.72 to 17.65)	17.56 (14.14 to 21.8)	0.75 (0.61 to 0.91)	−4.54 (−6.79 to −2.25)	12.03 (9.25 to 15.64)	16.45 (13.27 to 20.4)	0.73 (0.6 to 0.9)	−4.55 (−6.57 to −2.22)
Length of stay, d^c	18.62 (13.01 to 26.64)	19.47 (14.01 to 27.07)	0.96 (0.78 to 1.17)	−1.06 (−4.63 to 2.93)	15.34 (11 to 21.39)	16.96 (12.35 to 23.29)	0.9 (0.76 to 1.08)	−1.88 (−4.74 to 1.12)
No. of primary opioid treatment d^c	7.63 (4.21 to 13.85)	11.1 (7.19 to 17.12)	0.69 (0.44 to 1.06)	−3.49 (−5.84 to −1.28)	6.55 (3.69 to 11.61)	9.89 (6.26 to 15.61)	0.66 (0.43 to 1.02)	−3.44 (−5.47 to −1.43)
Length of primary opioid treatment, d^c	8.05 (4.54 to 14.28)	11.43 (7.51 to 17.41)	0.7 (0.47 to 1.05)	−3.42 (−5.81 to −1.12)	6.9 (3.96 to 12.02)	10.16 (6.51 to 15.87)	0.68 (0.46 to 1.01)	−3.38 (−5.44 to −1.34)
No. of doses of primary opioid^c	29.55 (12.12 to 72.02)	45.49 (21.74 to 95.16)	0.65 (0.39 to 1.09)	−16.1 (−26.9 to −6.29)	21.7 (9.86 to 47.77)	35.29 (16.89 to 73.74)	0.61 (0.38 to 1.01)	−13.9 (−22.8 to −5.48)
Receipt of secondary medications^d	0.11 (0.04 to 0.31)	0.25 (0.12 to 0.49)	0.46 (0.19 to 1.09)	−0.13 (−0.23 to −0.03)	0.1 (0.03 to 0.35)	0.22 (0.09 to 0.53)	0.46 (0.17 to 1.19)	−0.11 (−0.21 to −0.02)

Open in a new tab

^{^a}

Conducted in an intention-to-treat population. Adjusted for design effects only (study intervention and study period). There were no missing data for the outcomes presented.

^{^b}

Conducted in an intention-to-treat population. Adjusted for study design effects and baseline demographic characteristics that have been associated with neonatal opioid withdrawal syndrome severity (adequate prenatal care, medications for opioid use disorder during pregnancy, number of psychotropic substances used during pregnancy, sex, birth weight, and positive neonatal toxicity screen for psychotropic substances other than opioids [ie, amphetamines/methamphetamines, barbiturates, benzodiazepines, cocaine, gabapentin, marijuana, medetomidine, methaqualone, phencyclidine, serotonin reuptake inhibitors, tianeptine, and xylazine]). In pharmacologically treated infants, also adjusted for primary opioid medication.

^{^c}

Estimates for this outcome corresponded to adjusted means in each study intervention as well as adjusted mean ratios and adjusted mean differences between symptom-based dosing and scheduled opioid taper dosing.

^{^d}

Estimates for this binary outcome corresponded to adjusted risks in each study intervention as well as adjusted risk ratios and adjusted risk differences between symptom-based dosing and scheduled opioid taper dosing.

Secondary and Safety Outcomes

For the ESC cohort, the likelihood of initiating pharmacologic treatment (0.4 vs 0.41; aMR, 0.99 [95% CI, 0.77-1.27]) and the length of hospital stay (10.9 vs 12.1 days; aMR, 0.9 [95% CI, 0.72-1.13]) were not different between dosing strategies (Table 2). Among infants who received pharmacologic treatment in the ESC cohort, the mean time to medical readiness for discharge was also shorter in the symptom-based dosing group (13.1 vs 17.56 days; aMR, 0.75 [95% CI, 0.61-0.91]) compared with the scheduled opioid taper group (Table 2). Additionally, 35% (95% CI, 25%-45%) of pharmacologically treated infants in the symptom-based dosing group required scheduled opioid dosing due to persistently elevated signs of withdrawal, which equates to a number needed to treat of 1.5 (95% CI, 1.3-1.8) to prevent 1 infant from receiving scheduled opioids. There were no significant differences in the mean number of primary opioid doses administered between intervention groups, the number of opioid treatment days, the length of treatment, or the receipt of secondary medications (Table 2). Only 1 infant in the scheduled opioid taper group required cessation of opioid treatment due to excessive sedation.

In the Finnegan cohort, there were no significant differences in time to medical readiness for discharge (15.99 vs 17.56 days; aMR, 0.91 [95% CI, 0.72-1.15]) or length of stay (17.38 vs 19.39 days; aMR, 0.9 [95% CI, 0.69-1.16]). Additionally, there was no significant difference in any of the secondary outcomes, except for the risk of initiating pharmacologic treatment, which was higher in the symptom-based dosing group (0.76 vs 0.63; adjusted risk ratio, 1.21 [95% CI, 1.03-1.42]). The effect size across most outcomes, while not significant, was similar to the ESC cohort (Table 3).

Table 3. Secondary Outcomes for Infants in the Finnegan Cohort.

Outcome	Adjusted analysis (95% CI)
	Study design^a				Covariate^b
	Symptom-based mean or risk	Scheduled taper mean or risk	Mean or risk ratio	Mean or risk difference	Symptom-based mean or risk	Scheduled taper mean or risk	Mean or risk ratio	Mean or risk difference
Secondary outcome
Time to medical readiness for discharge, d^c	15.99 (10.48 to 24.4)	17.56 (11.31 to 27.25)	0.91 (0.72 to 1.15)	−1.56 (−4.65 to 1.42)	15.26 (10.01 to 23.26)	17.32 (11.22 to 26.74)	0.88 (0.67 to 1.16)	−2.07 (−4.98 to 0.93)
Length of stay, d^c	17.38 (11.64 to 25.95)	19.39 (11.86 to 31.7)	0.9 (0.69 to 1.16)	−1.96 (−5.48 to 1.70)	16.46 (11.27 to 24.05)	19.38 (12.34 to 30.45)	0.85 (0.62 to 1.16)	−2.75 (−5.95 to 0.96)
Receipt of pharmacologic treatment^d	0.76 (0.62 to 0.94)	0.63 (0.49 to 0.82)	1.21 (1.03 to 1.42)	0.13 (0.01 to 0.25)	0.73 (0.64 to 0.84)	0.63 (0.48 to 0.81)	1.16 (1.01 to 1.34)	0.1 (−0.02 to 0.23)
Secondary outcomes in pharmacologically treated infants
Time to medical readiness for discharge, d^c	17.94 (11.93 to 26.99)	22.95 (15.83 to 33.27)	0.78 (0.58 to 1.05)	−5.03 (−8.25 to −1.91)	16.41 (12.25 to 21.99)	21.39 (15.92 to 28.74)	0.77 (0.56 to 1.05)	−5 (−8.05 to −1.93)
Length of stay, d^c	19.73 (13.63 to 28.56)	25.37 (16.61 to 38.74)	0.78 (0.59 to 1.03)	−5.63 (−9.17 to −2.14)	18.17 (14.18 to 23.27)	23.89 (17.44 to 32.72)	0.76 (0.55 to 1.04)	−5.61 (−9.17 to −2.20)
No. of primary opioid treatment d^c	13.45 (8.20 to 22.07)	18.24 (11.34 to 29.35)	0.74 (0.54 to 1.00)	−4.78 (−7.99 to −1.75)	11.98 (8.88 to 16.16)	16.71 (11.9 to 23.47)	0.72 (0.51 to 1.01)	−4.65 (−7.73 to −1.75)
Length of primary opioid treatment, d^c	13.96 (8.71 to 22.39)	18.69 (11.67 to 29.95)	0.75 (0.55 to 1.02)	−4.71 (−8.11 to −1.59)	12.42 (9.41 to 16.39)	17.09 (12.16 to 24.03)	0.73 (0.51 to 1.03)	−4.6 (−7.80 to −1.62)
No. of doses of primary opioid^c	59.98 (22 to 163.55)	79.41 (27.99 to 225.30)	0.76 (0.48 to 1.19)	−19.5 (−36.5 to −4.87)	49.72 (24.14 to 102.42)	67.51 (31.89 to 142.93)	0.74 (0.48 to 1.12)	−17 (−31.5 to −3.78)
Receipt of secondary medications^d	0.2 (0.08 to 0.53)	0.3 (0.1 to 0.89)	0.68 (0.46 to 1.01)	−0.09 (−0.18 to 0)	0.18 (0.04 to 0.83)	0.27 (0.05 to 1.32)	0.67 (0.41 to 1.1)	−0.08 (−0.17 to 0.01)

Open in a new tab

^{^a}

Conducted in an intention-to-treat population for all outcomes. This analysis adjusted for design effects only (study intervention and the study period). There were no missing data for the outcomes presented.

^{^b}

Conducted in an intention-to-treat population. Adjusted for study design effects, nicotine (unbalanced between groups), and baseline demographic characteristics that have been associated with neonatal opioid withdrawal syndrome severity (adequate prenatal care, medications for opioid use disorder during pregnancy, number of psychotropic substances used during pregnancy, sex, birth weight, and positive neonatal toxicity screen for psychotropic substances other than opioids [ie, amphetamines/methamphetamines, barbiturates, benzodiazepines, cocaine, gabapentin, marijuana, medetomidine, methaqualone, phencyclidine, serotonin reuptake inhibitors, tianeptine, and xylazine]). In pharmacologically treated infants, also adjusted for primary opioid medication.

^{^c}

^{^d}

Inpatient safety events were rare in all study groups. There were no significant differences in inpatient (3 of 188 in the symptom-based dosing vs 2 of 195 in the scheduled opioid taper groups [ESC cohort only]) or outpatient safety outcomes between dosing groups for either the ESC or Finnegan cohorts (Table 4). Serious adverse events during the study were also rare (eTable in Supplement 1).

Table 4. Summary Safety Outcomes for Infants in the Eat, Sleep, Console (ESC) and Finnegan Cohorts.

Safety outcome^a	ESC cohort, No. (%)			Finnegan cohort, No. (%)
	Symptom-based (n = 188)	Scheduled taper (n = 195)	RR (95% CI)	Symptom-based (n = 118)	Scheduled taper (n = 125)	RR (95% CI)
Infants at risk for pharmacologic treatment^b
Inpatient
Composite critical safety outcome (nonaccidental trauma or death)	0	0	NA^c	0	0	NA^c
Composite safety outcome (seizures or weight loss >15% from birth weight)^d	3 (2)	2 (1)	1.54 (0.23-10.59)	1 (1)	6 (5)	0.11 (0-2.6)
Outpatient
Composite critical safety outcome (nonaccidental trauma or death)^e	0	3 (2)	NA^c	0	0	NA^c
Composite safety outcome (acute/urgent care or emergency department visits or hospital readmissions)	31 (16)	46 (24)	0.7 (0.37-1.33)	25 (21)	15 (12)	1.6 (0.49-5.19)
Pharmacologically treated infants	Symptom-based (n = 78)	Scheduled taper (n = 82)	RR (95% CI)	Symptom-based (n = 88)	Scheduled taper (n = 81)	RR (95% CI)
Inpatient
Composite critical safety outcome (nonaccidental trauma or death)	0	0	NA^c	0	0	NA^c
Composite safety outcome (seizures or weight loss >15% from birth weight)^d	1 (1)	0	NA^c	0	3 (4)	NA^c
Outpatient
Composite critical safety outcome (nonaccidental trauma or death)^e	0	1 (1)	NA^c	0	0	NA^c
Composite safety outcome (acute/urgent care or emergency department visits or hospital readmissions)	8 (10)	16 (20)	0.51 (0.25-1.05)	20 (23)	9 (11)	1.92 (0.46-8.08)

Open in a new tab

Abbreviations: NA, not applicable; RR, risk ratio.

^{^a}

Safety events were rare throughout the study and were reported as more than 0 if there was 1 or more composite events. The only composite safety outcome for which there was an adequate population to conduct inferential statistics was the outpatient composite safety outcome, and no significant differences were seen between treatment groups in both the at-risk and pharmacologically treated study populations.

^{^b}

Analyses conducted in an as-treated population. Numbers of infants not treated per protocol in the ESC cohort: symptom-based dosing, 9; scheduled opioid taper, 9. Numbers of infants not treated per protocol in the Finnegan cohort: symptom-based dosing, 10; scheduled opioid taper, 3. Infants with missing data for the 3-month assessment were considered to not have the outcome. Missing data were evenly distributed between dosing groups in both cohorts.

^{^c}

Events were rare and an RR could not be calculated when the denominator was 0 or when the event rate was too small to allow for inferential statistical analyses.

^{^d}

No infants were reported to have seizures during their inpatient stay.

^{^e}

One infant died after hospital discharge due to sudden unexplained infant death that was deemed to be unrelated to the study intervention.

Discussion

Among infants with opioid withdrawal who were at risk for pharmacologic treatment and cared for with ESC in this cluster, crossover design randomized clinical trial, use of the symptom-based dosing approach led to a significantly shorter mean time to medical readiness for discharge compared with a scheduled opioid taper approach.

Symptom-based dosing gives clinicians a framework to provide pharmacologic support when withdrawal thresholds are met, without committing infants to scheduled opioid dosing and a prolonged opioid taper when their withdrawal severity oscillates around the threshold for pharmacologic treatment. In this study, 65% of pharmacologically treated infants in the symptom-based dosing group were successfully treated without the use of scheduled opioid dosing. These results support the premise that opioid withdrawal follows a severity continuum to which pharmacologic treatment should be matched and provides the evidence to move beyond the all or nothing approach inherent in the use of a scheduled opioid taper. Additionally, the symptom-based dosing approach used in this study provides the means to readily identify and appropriately support infants who require scheduled opioids, thus minimizing the likelihood of undertreatment.

The study population was defined to include infants at risk for pharmacologic treatment rather than only infants who were actually treated to maintain comparability between groups and counteract the potential that the option to administer a single opioid dose in the symptom-based dosing group might lower a clinician’s prescribing threshold. This choice provided for assessment of withdrawal severity between groups and increased the likelihood that differences observed between dosing approaches were truly reflective of the intervention. However, a difference in the proportion of infants who were started on postnatal opioids was not found between the intervention groups in the ESC cohort, which suggests that symptom-based dosing did not alter prescribing thresholds.

Inpatient safety events were rare in all study groups. There were no material differences in safety outcomes identified through 3 months of age between dosing approaches. Longer-term follow-up is needed to further inform the safety profile for symptom-based dosing.

This study did not have power to detect differences in outcomes between dosing approaches in the Finnegan cohort. Therefore, findings in this cohort were not unexpected, although similarities in the effect sizes for most outcomes in the Finnegan and ESC cohorts suggest that further study may be warranted.

Strengths and Limitations

Strengths of this study include the large study population, inclusion of study sites across geographically diverse regions, and the use of a waiver of consent, all of which increase the generalizability of the results. Additionally, the primary outcome of time to medical readiness for discharge avoids the impact of non-NOWS–related factors on the duration of hospital stay and provides a more accurate reflection of health care utilization specifically due to opioid withdrawal.

This study has limitations. First, assessment of postdischarge safety outcomes was limited by reliance on electronic medical records from the enrolling hospital, linked medical records when available, and public record search. Although this strategy may not have captured all outcomes, it has been previously used,³ missing data were balanced between intervention groups, and this pragmatic approach allows for assessment of short-term outcomes under a waiver of consent. Second, contamination between groups was a potential risk, but was tempered by limiting access to training materials for the symptom-based dosing approach until just prior to each site’s transition and inclusion of a washout period. Third, the results could have been affected by temporal trends, but the analysis adjusted for such trends. Fourth, all studies are limited by the quality of data abstracted from the medical record; this was addressed through the use of a data quality framework that has been shown to decrease abstraction error during the trial.^23,24

Conclusions

In this randomized clinical trial, use of the symptom-based dosing approach decreased time to medical readiness for discharge and resulted in avoidance of scheduled opioid dosing for 65% of infants cared for with ESC. For infants with NOWS who are cared for with the ESC approach, symptom-based dosing should be considered an effective, evidence-based approach for pharmacologic treatment.

Supplement 1.

eFigure. Maps of Study Sites

eTable. Serious Adverse Events (SAEs)

jama-e265782-s001.pdf^{(363.5KB, pdf)}

Supplement 2.

Trial Protocol

jama-e265782-s002.pdf^{(4.1MB, pdf)}

Supplement 3.

Statistical Analysis Plan

jama-e265782-s003.pdf^{(2.2MB, pdf)}

Supplement 4.

Nonauthor Collaborators

jama-e265782-s004.pdf^{(286.7KB, pdf)}

Supplement 5.

Data Sharing Statement

jama-e265782-s005.pdf^{(15.4KB, pdf)}

References

1.Agency for Healthcare Research and Quality . Healthcare Cost and Utilization Project (HCUP) fast stats, neonatal abstinence syndrome among newborn hospitalizations in 2022. Updated July 9, 2025. Accessed February 13, 2026. https://datatools.ahrq.gov/hcup-fast-stats?tab=special-emphasis&dash=79
2.Grossman MR, Berkwitt AK, Osborn RR, et al. An initiative to improve the quality of care of infants with neonatal abstinence syndrome. Pediatrics. 2017;139(6):e20163360. doi: 10.1542/peds.2016-3360 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Young LW, Ounpraseuth ST, Merhar SL, et al. ; ACT NOW Collaborative . Eat, Sleep, Console approach or usual care for neonatal opioid withdrawal. N Engl J Med. 2023;388(25):2326-2337. doi: 10.1056/NEJMoa2214470 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Finnegan LP, Connaughton JF Jr, Kron RE, Emich JP. Neonatal abstinence syndrome: assessment and management. Addict Dis. 1975;2(1-2):141-158. [PubMed] [Google Scholar]
5.Devlin LA, Hu Z, Merhar SL, et al. ; Eunice Kennedy Shriver NICHD Neonatal Research Network and NIH Environmental Influences on Child Health Outcomes (ECHO) Program Institutional Development Award States Pediatric Clinical Trials Network . Influence of Eat, Sleep, and Console on infants pharmacologically treated for opioid withdrawal: a post hoc subgroup analysis of the ESC-NOW randomized clinical trial. JAMA Pediatr. 2024;178(6):525-532. doi: 10.1001/jamapediatrics.2024.0544 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Merhar SL, Hu Z, Devlin LA, et al. ; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network National Institutes of Health Environmental Influences on Child Health Outcomes Program Institutional Development Award States Pediatric Clinical Trials Network . Infant feeding and weight trajectories in the Eat, Sleep, Console trial: a secondary analysis of a randomized clinical trial. JAMA Pediatr. 2024;178(10):976-984. doi: 10.1001/jamapediatrics.2024.2578 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Patrick SW, Barfield WD, Poindexter BB; Committee on Fetus and Newborn, Committee on Substance Use and Prevention . Neonatal opioid withdrawal syndrome. Pediatrics. 2020;146(5):e2020029074. doi: 10.1542/peds.2020-029074 [DOI] [PubMed] [Google Scholar]
8.Blount T, Painter A, Freeman E, Grossman M, Sutton AG. Reduction in length of stay and morphine use for NAS with the “Eat, Sleep, Console” method. Hosp Pediatr. 2019;9(8):615-623. doi: 10.1542/hpeds.2018-0238 [DOI] [PubMed] [Google Scholar]
9.McMorrow TJ, Byrnes K, Gates M, et al. Quality improvement targeting non-pharmacologic care and as-needed morphine improves outcomes in neonatal abstinence syndrome. Pediatr Qual Saf. 2022;7(6):e612. doi: 10.1097/pq9.0000000000000612 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Morrison TM, MacMillan KDL, Melvin P, et al. Neonatal opioid withdrawal syndrome: a comparison of as-needed pharmacotherapy. Hosp Pediatr. 2022;12(5):530-538. doi: 10.1542/hpeds.2021-006301 [DOI] [PubMed] [Google Scholar]
11.Wachman EM, Minear S, Hirashima M, et al. Standard fixed-schedule methadone taper versus symptom-triggered methadone approach for treatment of neonatal opioid withdrawal syndrome. Hosp Pediatr. 2019;9(8):576-584. doi: 10.1542/hpeds.2018-0165 [DOI] [PubMed] [Google Scholar]
12.Parlaman J, Deodhar P, Sanders V, Jerome J, McDaniel C. Improving care for infants with neonatal abstinence syndrome: a multicenter, community hospital-based study. Hosp Pediatr. 2019;9(8):608-614. doi: 10.1542/hpeds.2019-0083 [DOI] [PubMed] [Google Scholar]
13.Townsend SF, Hodapp CD, Weikel B, Hwang SS. Shifting the care paradigm for opioid-exposed newborns in southern Colorado. J Perinatol. 2021;41(6):1372-1380. doi: 10.1038/s41372-020-00900-y [DOI] [PubMed] [Google Scholar]
14.Achilles JS, Castaneda-Lovato J. A quality improvement initiative to improve the care of infants born exposed to opioids by implementing the Eat, Sleep, Console assessment tool. Hosp Pediatr. 2019;9(8):624-631. doi: 10.1542/hpeds.2019-0144 [DOI] [PubMed] [Google Scholar]
15.Young LW, Babineau DC, Das A, et al. ; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research and the National Institutes of Health (NIH) Helping to End Addiction Long-term (HEAL) Initiative . Optimizing pharmacologic treatment for neonatal opioid withdrawal syndrome (OPTimize NOW): a symptom-based dosing approach study protocol for a multi-center, cluster crossover design randomized controlled trial. Trials. 2025;26(1):317. doi: 10.1186/s13063-025-09035-x [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Hemming K. The shiny CRT calculator: power and sample size for cluster randomised trials. Accessed September 1, 2024. https://clusterrcts.shinyapps.io/rshinyapp/
17.Mancl LA, DeRouen TA. A covariance estimator for GEE with improved small-sample properties. Biometrics. 2001;57(1):126-134. doi: 10.1111/j.0006-341X.2001.00126.x [DOI] [PubMed] [Google Scholar]
18.Schluchter MD, Elashoff JT. Small-sample adjustments to tests with unbalanced repeated measures assuming several covariance structures. J Stat Comput Simul. 1990;37(1-2):69-87. doi: 10.1080/00949659008811295 [DOI] [Google Scholar]
19.Devlin LA, Hu Z, Ounpraseuth S, et al. ; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network and the NIH Environmental influences on Child Health Outcomes (ECHO) Program Institutional Development Awards States Pediatric Clinical Trials Network . The influence of mediators on the relationship between antenatal opioid agonist exposure and the severity of neonatal opioid withdrawal syndrome. Matern Child Health J. 2023;27(6):1030-1042. doi: 10.1007/s10995-022-03521-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Suarez EA, Huybrechts KF, Straub L, et al. Buprenorphine versus methadone for opioid use disorder in pregnancy. N Engl J Med. 2022;387(22):2033-2044. doi: 10.1056/NEJMoa2203318 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Yen E, Kaneko-Tarui T, Ruthazer R, Harvey-Wilkes K, Hassaneen M, Maron JL. Sex-dependent gene expression in infants with neonatal opioid withdrawal syndrome. J Pediatr. 2019;214:60-65.e2. doi: 10.1016/j.jpeds.2019.07.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Amiri S, Nair J. Gestational age alters assessment of neonatal abstinence syndrome. Pediatr Rep. 2022;14(1):50-57. doi: 10.3390/pediatric14010009 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Zozus MN, Young LW, Simon AE, et al. Training as an intervention to decrease medical record abstraction errors multicenter studies. Stud Health Technol Inform. 2019;257:526-539. [PMC free article] [PubMed] [Google Scholar]
24.Garza MY, Williams T, Myneni S, et al. Measuring and controlling medical record abstraction (MRA) error rates in an observational study. BMC Med Res Methodol. 2022;22(1):227. doi: 10.1186/s12874-022-01705-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eFigure. Maps of Study Sites

eTable. Serious Adverse Events (SAEs)

jama-e265782-s001.pdf^{(363.5KB, pdf)}

Supplement 2.

Trial Protocol

jama-e265782-s002.pdf^{(4.1MB, pdf)}

Supplement 3.

Statistical Analysis Plan

jama-e265782-s003.pdf^{(2.2MB, pdf)}

Supplement 4.

Nonauthor Collaborators

jama-e265782-s004.pdf^{(286.7KB, pdf)}

Supplement 5.

Data Sharing Statement

jama-e265782-s005.pdf^{(15.4KB, pdf)}

PERMALINK

Symptom-Based Dosing for Neonatal Opioid Withdrawal

Lori A Devlin, DO

Denise C Babineau, PhD

Stephanie L Merhar, MD

Sara B DeMauro, MD, MSCE

Walter K Kraft, MD

Scott A Lorch, MD, MSCE

Abhik Das, PhD

Scott A McDonald, BS

Evan Rhodes, BSPH

Augusto F Schmidt, MD

Lillian Trochinski, BSN

Margaret Crawford, BS

Thitinart Sithisarn, MD, PhD

Lawrence Leeman, MD

Kelley Zagol Kovatis, MD

Namasivayam Ambalavanan, MD

Ryan W Smith, MD

Sucheta Telang, MBBS

Jennifer A Tioseco, MD

Jennifer M McAllister, MD

Scott L Wexelblatt, MD

Bhanu Muniyappa, MD

Patricia K Williams, MD

Susan C Adeniyi-Jones, MD

Crystal D Hill, MD

Tanner Wright, MD

Gregory M Sokol, MD

Lynette Johnson, DO

Richard W Hall, MD

Scott D Duncan, MD

Karen Puopolo, MD

Krishna Dummula, MD

Ann Anderson-Berry, MD

Jonathan M Davis, MD

Brenda Poindexter, MD

Leslie W Young, MD

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Intervention

Main Outcome and Measure

Results

Conclusions and Relevance

Trial Registration

Introduction

Methods

Trial Design and Oversight

Patients and Participating Sites

Randomization

Intervention

Figure 1. Treatment Algorithm for Trial Intervention Groups.

Outcomes

Statistical Considerations

Results

Patients

Figure 2. Flowchart of Recruitment, Randomization, and Follow-Up Assessment for the Optimizing Pharmacologic Treatment for Neonatal Opioid Withdrawal Syndrome (OPTimize NOW) Trial.

Table 1. Baseline Characteristics of the Eat, Sleep, Console (ESC) and Finnegan Cohorts.

Primary Outcome

Table 2. Primary and Secondary Outcomes for Infants Assessed and Cared for With the Eat, Sleep, Console Approach (ESC).

Secondary and Safety Outcomes

Table 3. Secondary Outcomes for Infants in the Finnegan Cohort.

Table 4. Summary Safety Outcomes for Infants in the Eat, Sleep, Console (ESC) and Finnegan Cohorts.

Discussion

Strengths and Limitations

Conclusions

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles