Trial to Shorten Pharmacologic Treatment of Newborns With Neonatal Opioid Withdrawal Syndrome (NOWS)

Pragmatic, Randomized, Blinded Trial to Shorten Pharmacologic Treatment of Newborns With Neonatal Opioid Withdrawal Syndrome (NOWS)

The objective of this study is to evaluate the efficacy of a rapid wean intervention compared with a slow-wean intervention in reducing the number of days of opioid treatment from the first dose of weaning to cessation of opioid among infants receiving an opioid (defined as morphine or methadone) as the primary treatment for neonatal opioid withdrawal syndrome (NOWS).

This will be a pragmatic, randomized, blinded trial comparing a rapid-wean intervention (15% decrements from the stabilization dose) to a slow-wean intervention (10% decrements from the stabilization dose) to determine whether rapid weaning will reduce the number of treatment days among infants receiving morphine or methadone orally as the primary treatment for NOWS. Participating hospitals must provide pharmacologic treatment to at least an average of 12 opioid exposed infants each year, use a scoring system to assess for signs of NOWS (original or modified Finnegan Neonatal Abstinence Scoring system, Eat-Sleep or Console), and provide opioid replacement therapy with either morphine or methadone as the primary drug for treating NOWS. Hospitals may change use of these two opioids during the trial period. The investigators will stratify randomization by hospital. The study protocol will commence after NOWS signs have been controlled with an opioid (stabilization) and weaning of pharmacologic treatment is to be started. At or before each 24-hour interval, clinical team members will evaluate and score infants, per hospital practice, for signs of NOWS to determine if the infant will tolerate weaning of the study drug. After study drug cessation, the clinical team will observe infants in the hospital for at least 48 hours prior to discharge, which is similar to clinical practice. A trained examiner will administer the Neonatal Intensive Care Unit (NICU) Network Neurobehavioral Scale (NNNS) to assess neurobehavioral profiles after infants cease study drug and prior to discharge. At one month post discharge, primary caregivers will complete the Brief Symptom Inventory (BSI), the Maternal Postnatal Attachment Questionnaire (MPAQ) and a caregiver questionnaire. The site research team will contact the primary caregiver(s) to update contact information and/or complete questionnaires when the infant is 6 months, 12 months, 18 months, and 24 months of age. The questionnaires will assess infant wellness, neurobehavioral functioning and development, postnatal attachment and bonding, and caregiver well-being. At 24 months, the infants will be seen during which a, certified developmental specialists, blinded to the intervention, will administer the Bayley Scales of Infant and Toddler Development, Fourth Edition (Bayley-4) to assess infant neurodevelopment. The BSI and the Brief Infant Toddler Social Emotional Assessment (BITSEA) will also be administered during the 24 month visit along with measures of growth.

The study will have two intervention arms (rapid-wean and slow-wean) with 251 morphine/methadone treated infants per intervention arm, for a total of 502 morphine/methadone treated infants.

The pharmacy will track dose levels to know where an infant is within a rapid- or slow-wean intervention arm. The clinical team will be blinded to the dose level and will only be aware of the study steps. Both the rapid- and slow-wean intervention arms are depicted to indicate that if each intervention arm has the same number of escalations, the study steps will be identical. This is critical to maintaining the clinical team blinding.

The number of days of opioid treatment (used as primary treatment), including escalation, resumption, and spot treatment, from the first weaning dose to cessation of opioid.

The numbers of days of opioid treatment from the first weaning dose to cessation of opioid with a rapid and slow-wean interventions among infants treated with morphine.

The numbers of days of opioid treatment from the first weaning dose to cessation of opioid with a rapid and slow-wean interventions among infants treated with methadone.

The proportions of infants in the rapid and slow-wean intervention arms who have an escalation or resumption of opioid medication during weaning.

The total amounts of opioid from the first weaning dose to cessation of opioid among infants in the rapid and slow-wean intervention arms.

proportion of infants who experience initiation and/or escalation of second-line or third-line drugs to treat NOWS

The proportion of infants who experience initiation and/or escalation of second-line or third-line drugs to treat NOWS signs from the first weaning dose to cessation of opioid in the rapid-wean and slow-wean intervention arms.

proportion of infants in each intervention arm with safety outcomes of seizures (clinical or EEG), excessive stool output, respiratory disturbances, and feeding tolerance

The proportion of infants in each intervention arm with safety outcomes of seizures (clinical or EEG), excessive stool output, respiratory disturbances, and feeding tolerance.

The proportion of infants in each intervention arm with an atypical NNNS neurobehavioral profile prior to discharge.

The Brief Symptom Inventory (BSI) will be the measurement tool used to asses maternal well-being in each intervention arm. A regression model will be used to analyze BSI total scores. This model will include a fixed treatment effect (intervention arms), an adjustment for the stratifying variable, and fixed effects for the covariates of maternal treatment and stabilization dose. The BSI outcomes are measured at 1-month after discharge and 24-months of age. The models for BSI outcomes will include the 1-month after discharge BSI outcome as a covariate. The F-test of the intervention arm effect will be the primary test of interest. The intervention arm difference will be reported along with 95% CI. Analyses will be conducted among the entire group and among those where the biologic mother is the primary caretaker.

The Maternal Postnatal Attachment Questionnaire (MPAQ) will be the measurement tool used to asses maternal-infant attachment in each intervention arm. A regression model will be used to analyze MPAQ total scores. This model will include a fixed treatment effect (intervention arms), an adjustment for the stratifying variable, and fixed effects for the covariates of maternal treatment and stabilization dose. The intervention arm difference will be reported along with 95% CI. Analyses will be conducted among the entire group and among those where the biologic mother is the primary caretaker.

Anthropometric outcomes will be measured at birth, time of discharge, and 24 months. We will calculate anthropometric z-scores at each of the three assessment periods for the purpose of analysis based on age and gender specific WHO norms.We will provide the mean and SD of infants' weights (z-scores) separately for each treatment group. We will use a mixed linear model to evaluate the effect of treatment arm on weight (z-scores). We will examine the impact of the treatment arm on length, HC, and infant weight for length (z-scores). We will provide the mean and SD of infant BMI-z at 24-months for each treatment group. To compare Bayley-IV scores between intervention arms, we will perform a linear mixed-effects model with a fixed effect for the intervention group and a random effect for study site. We will report point estimates for the group mean difference along with a 95% CI, and the team will repeat this analytical approach for each of the Bayley-IV domains.

We will analyze the caregiver questionnaire outcomes and the death outcome using a longitudinal GLMM or GEE model appropriate for the outcome type since the data will be collected at multiple time points after discharge. Count data that tend to have more than 0 or 1 events counted will be analyzed using a Poisson model while binary or count data that rarely goes beyond 1 occurrence will be analyzed using a Logistic model. In case of count data that rarely goes beyond 1 occurrence, this data will be transformed to binary data (occurrence/no occurrence). We will present mean outcome ratios for count data (from Poisson models) and odds ratios for binary data (from logistic models) with respect to the intervention effect as well as 95% CI of the intervention effect. All analyses will be adjusted for repeated measures over time, so that patterns of change for these outcomes over time can be assessed by treatment group.

To compare Bayley-IV scores between intervention arms, we will perform a linear mixed-effects model with a fixed effect for the intervention group and a random effect for study site. We will report point estimates for the group mean difference along with a 95% CI, and the team will repeat this analytical approach for each of the Bayley-IV domains. Descriptive statistics (means, medians, SD, percentiles) for continuous secondary outcomes and frequency based statistics (N and percentages) for binary secondary outcomes will be generated and summarized in a tabular form by treatment group.

Inclusion Criteria: Hospital Level Hospital provides pharmacologic treatment to at least an average of 12 opioid exposed infants each year Hospital uses a scoring system to assess for signs of NOWS (original or modified Finnegan Neonatal Abstinence Scoring system, Eat-Sleep or Console) Hospital provides opioid replacement therapy with either morphine or methadone as part of pharmacologic treatment of NOWS Infant Level Gestational age ≥ 36 weeks Receiving scheduled pharmacological therapy with morphine or methadone as the primary drug treatment for NOWS secondary to maternal opioid use Tolerating enteral feeds and medications by mouth Exclusion Criteria: Hospital Level 1. Hospitals discharge > 10% of infants from the hospital on opioid replacement therapy on average per year Infant Level Major birth defect (e.g. gastroschisis) Any major surgery (minor surgery [e.g., circumcision, digit ligation, frenulectomy] is not an exclusion criterion) Hypoxic-ischemic encephalopathy Seizures from etiologies other than NOWS Treatment with opioids for reasons other than NOWS Respiratory support (nasal cannula or greater) for > 72 hours Planned discharge from the hospital on opioids Use of other opioids (e.g., buprenorphine) as primary drugs for treatment of NOWS Weaning of morphine or methadone as the primary treatment of NOWS has started