Weaning buprenorphine in pregnant patients^#

Abstract

Objective

Evaluate maternal and neonatal outcomes after buprenorphine wean compared to patients maintained on buprenorphine throughout pregnancy.

Methods

Prospective cohort study of pregnant patients with opioid use disorder enrolled in a multidisciplinary treatment program between 2015 and 2022. All patients were offered Medications to treat Opioid Use Disorder (MOUD) primarily with buprenorphine. Patients had at least 2 prenatal visits and negative urine drug tests (UDT) prior to weaning. The experimental group underwent a buprenorphine wean greater than 20% of their baseline dose. The control group was maintained on buprenorphine throughout the pregnancy. Relapse was defined as patient reported use or positive UDT during weekly assessments. Mass spectrophotometer was used for detection of drugs in samples. Fisher’s exact tests were used to compare outcomes in weaned and control groups.

Results

334 of 456 (73%) patients were treated with buprenorphine during pregnancy, with 39 in the experimental group and 295 in the control group. The mean dose for buprenorphine was similar between the groups (wean: 10.6 mg ± 5.6 vs. control: 10.3 mg ± 4.6, p = 0.76) but was significantly lower at delivery (wean: 4.4 ± 4.6 mg vs. control: 13.0 ± 4.7, p < 0.0001). Mean gestational age at initiation of the buprenorphine wean was 22.7 weeks. 10 of 39 (26%) who weaned were able to completely discontinue buprenorphine prior to delivery. Demographic data was similar between the groups, including overdose history. Overdose history at time of enrollment had a higher trend in the non-weaning group. neonatal opioid withdrawal syndrome (NOWS) treatment was significantly lower in the wean group (23 vs. 47%, p = 0.006), as was highest Finnegan score (9.6 ± 4.5 vs. 12.3 ± 4.0, p = 0.0003). Birthweight percentile was significantly higher in the wean group (44.3 ± 29.9 vs. 34.8 ± 24.4, p = 0.03). Gestational age at delivery, mode of delivery, and complications (HTN, DM, preterm labor, or short cervix) at delivery did not significantly differ between the groups.

Conclusion

Despite counseling to stay on buprenorphine, there are patients who desire to wean. The NOWS rate in the weaned cohort was significantly lower than the controls with no observed increase in maternal or neonatal morbidity. There were no maternal overdoses or deaths during the pregnancy. Larger studies are needed to evaluate this approach.

Keywords:

• Buprenorphine wean
• opioid use disorder
• pregnancy
• MAT
• MOUD

Introduction

Opioid Use Disorder (OUD) in pregnancy is a multifaceted clinical dilemma, with substantial implications for both maternal and neonatal outcomes [1]. In the context of OUD, Neonatal Opioid Withdrawal Syndrome (NOWS), formerly known as neonatal abstinence syndrome, represents a significant clinical challenge. NOWS manifests through a range of behavioral and physiological signs and symptoms, which can necessitate a prolonged hospitalization and intensive treatment [2].

A cornerstone in the management of OUD in pregnant patients has been Medications to treat Opioid Use Disorder (MOUD). Both buprenorphine and methadone were recently compared in a large randomized controlled trial by Suarez et al. Buprenorphine was associated with a lower risk of adverse neonatal outcomes when compared to methadone, with comparable maternal outcomes [3]. Nevertheless, the question of optimal buprenorphine management throughout gestation remains a point of significant debate and ongoing research.

Existing guidelines from bodies such as the American College of Obstetricians and Gynecologists (ACOG) and Substance Abuse and Mental Health Services Administration (SAMHSA) discourage weaning from buprenorphine during pregnancy due to the high risk of relapse and subsequent adverse outcomes [2,4]. However, some pregnant patients express a strong desire to wean from buprenorphine, putting healthcare providers in a challenging position.

Our study navigates this complex issue by evaluating the maternal and neonatal outcomes of pregnant patients who voluntarily underwent a buprenorphine wean compared to those maintained on buprenorphine throughout their pregnancy. This research holds important implications for individualizing treatment for pregnant patients with OUD and expanding our understanding of buprenorphine dose management during pregnancy.

Materials and methods

This was a prospective observational cohort design, conducted within a safety net hospital system in Kentucky from 2015 through 2023. Institutional Review Board approval was received from the University of Kentucky (Protocol #43698). The study participants were composed of pregnant and postpartum individuals aged 18 and above, with Opioid Use Disorder (OUD) and seeking obstetric healthcare or medication for OUD during the perinatal period.

Implemented in a comprehensive, multidisciplinary perinatal substance treatment program, the study was supported by grants from the Kentucky Opioid Response Effort (KORE). This program is situated within the Department of Obstetrics and Gynecology, Division of Maternal Fetal Medicine at the University of Kentucky Healthcare (UKHC). Notably, the program offers a spectrum of services from high-risk obstetric healthcare to behavioral health support and peer counseling. Due to limitations in resources, this study only included English-speaking patients.

Inclusion criteria consisted of individuals aged 18 and above with a confirmed diagnosis of opioid use disorder, consented to observational research, and attended at least two prenatal care visits. Exclusion criteria excluded those who withdrew their participation at any stage of the program. They were offered MOUD with buprenorphine, contingent on their willingness to participate and consent to undergo urine drug testing.

The study design differentiated the participants into two groups: the experimental group, which voluntarily underwent a buprenorphine reduction exceeding 20%, and the control group, which continued the recommended buprenorphine treatment throughout the pregnancy. Primary outcome aimed to determine the incidence of NOWS. Secondary outcomes included maternal and neonatal morbidity and mortality, birth weight, gestational age at delivery, Neonatal Intensive Care Unit (NICU) admission, and NOWS treatment. All neonates from patients on MOUD or opioid use were observed by newborn nursery providers for signs of withdrawal for the first five days. Only infants meeting treatment criteria (consecutive Finnegan scores >8 or one score >12) were admitted to the NICU. NOWS treatment utilized morphine as first line pharmacological therapy. After pharmacologic treatment initiation and neonatal stabilization based on Finnegan scores of less than 6 for at least 24–48 h, a slow medication wean in decrements of 10% took place.

Measures

The study commenced with an exhaustive patient evaluation process to ensure the provision of personalized, comprehensive care. Registered nurses assessed each patient’s cognitive status, ensuring they were alert and oriented to their person, place, and time. Following the verification of immediate healthcare needs and the patient’s suitability for the program, informed consent was obtained for both program participation and involvement in the observational research, followed by a meticulous walkthrough and signing of the program contracts.

The patient-centric approach extended to a range of screenings and evaluations that delved into the patient’s pregnancy and social determinants of health (SDOH). This entailed exploring various factors such as Abuse Assessment Screen (AAS), Adverse Childhood Experience (ACE), housing security, depression levels, relapse rates and more. Relapse was measured by patient reported use or positive urine drug tests during weekly assessments. Drug analysis testing was done via mass spectrophotometry. All relevant data collected during these processes was securely captured and managed via the Research Electronic Data Capture (REDCap) tool, a web-based platform known for its security and efficiency in data management.

Analysis

For each demographic and clinical outcome variable, differences between the wean and control groups were assessed using a two-sample t-test or Fisher’s exact test, as appropriate. Across all analyses, a p-value less than 0.05 was considered significant. All analyses were completed in R 4.2.1 (R Foundation for Statistical Computing; Vienna, Austria).

Results

A total of 334 out of 456 (73.2%) pregnant patients with OUD were treated with buprenorphine as part of a multidisciplinary treatment program between 2015 and 2022. 20% of both wean and control groups were stabilized on buprenorphine prior to pregnancy, 7/39 (20.6%) vs 55 of 295 (20.6%). Both groups shared comparable average starting doses of buprenorphine: 10.6 mg ± 5.6 for the wean group and 10.3 mg ± 4.6 for the control group (p = 0.76). However, a significant difference was observed in the mean dose at delivery: 4.4 mg ± 4.6 for those who weaned compared to 13.0 mg ± 4.7 in the control group (p < 0.001).

Weaning from buprenorphine was initiated at a mean of 22.7 weeks of gestation. Interestingly, 10 out of 39 patients (25.6%) in the wean group were successful in completely discontinuing the medication prior to delivery. Overdose history and other demographic data were similar between the two groups, as shown in Table 1. Overdose history during enrollment had a higher trend in the non-weaning group. This may have been due to a selection bias as patients in the wean group were highly motivated patients.

Table 1. Comparative maternal characteristics of pregnant patients with OUD: buprenorphine weaned vs. maintained.

	Buprenorphine wean N = 39	Did not wean buprenorphine N = 295	p Value
Maternal characteristics
Age	28.9 ± 5.4	29.2 ± 4.9	0.73
Race
Asian	0	1 (0.3%)	1.00
More than 1 race	0	3 (1.0%)
Black or African American	0	4 (1.4%)
White	39 (100%)	287 (97.3%)
Insurance type
Insured*	39 (100%)	286 (97.3%)	0.60
Uninsured	0	8 (2.7%)
Tobacco use in pregnancy	33 (84.6%)	250 (85.3%)	0.31
Electronic cigarette use	11 (28.9%)	29 (10.4%)	0.02
Fagerstrom Test	5.3 ± 1.4	4.7 ± 2.1	0.35
MAOD started prior to the pregnancy	7 (20.6%)	55 (20.8%)	1.00
Total buprenorphine dose (mg/day)	10.6 ± 5.6	10.3 ± 4.6	0.76
Total naloxone dose (mg/day)	2.3 ± 1.3	2.6 ± 1.1	0.29
History of overdose	5 (33.3%)	53 (47.3%)	0.41
Concurrent psychiatric diagnoses at enrollment
Bipolar disorder	6 (17.6%)	35 (12.5%)	0.41
Anxiety	15 (44.1%)	156 (55.5%)	0.27
Panic disorder	0 (0%)	7 (2.5%)	1.00
ADHD	2 (5.9%)	12 (4.3%)	0.65
PTSD	1 (2.9%)	20 (7.1%)	0.71
AAS Score = 1**	19 (48.7%)	148 (50.2%)	1.00
Enrollment GAD-7 Score ≥10***	5 (41.7%)	59 (50.4%)	0.76
ACE Score****	3.9 ± 3.1	3.9 ± 2.5	0.88
Enrollment Edinburgh score ≥12	7 (53.8%)	56 (47.5%)	0.77
Average Enrollment Edinburgh Score	12.4 ± 6.0	11.5 ± 6.3	0.64

*Most insurance was government/public assisted. **AAS, Abuse Assessment Score; ***GAD, Generalized anxiety disorder; ****ACE, adverse childhood experiences.

Maternal and neonatal outcome summaries are presented in Table 2. A significant decrease in the occurrence of NOWS was noted in the group that weaned; only 23.1% of those neonates required treatment for NOWS, compared with 46.6% of the control group (p=0.006). Correspondingly, the average highest Finnegan score, an indicator of NOWS severity, was significantly reduced in the wean group (9.6 ± 4.5 vs. 12.3 ± 4.0, p<0.001).

Table 2. Comparative maternal and neonatal outcomes of pregnant patients with OUD: buprenorphine weaned vs. maintained.

	Buprenorphine wean N = 39	Did not wean buprenorphine N = 295	p Value
Maternal outcomes
Mode of delivery			0.17
Vaginal delivery	20 (51.3%)	157 (53.8%)
Operative vaginal delivery	5 (12.8%)	15 (5.1%)
Cesarean section	14 (35.9%)	120 (41.1%)
MAOD dose at delivery	4.4 ± 4.6	13.0 ± 4.7	<0.001
Postpartum Edinburgh score ≥12	1 (9.1%)	15 (22.7%)	0.44
Average postpartum Edinburgh Score	4.7 ± 4.1	8.2 ± 5.7	0.05
Maternal deaths	0 (0%)	4 (1.4%)	1.00
Prenatal adverse event
(Any of the following): arrested, birth <28 gestation, miscarriage (<20 weeks), fetal demise >20 weeks, illness or injury associated with substance abuse, relapse, victimization	16 (41%)	148 (50.2%)	0.31
Total number of prenatal visits	14.1+/−7.2	13.6+/−6.7	0.69
Number of inappropriate prenatal visits
Self-reported illicit drug use, inappropriate UDT, no show appointment	4.5+/−5.9	4.7+/−6.1	0.82
PP severe adverse events
Prolongation of hospitalization (including medical or psychiatric hospitalization)	1 (2.6%)	14 (4.7%)	1.00
PP adverse events
Relapse illness	10 (25.6%) 8	114 (38.6%)	0.16
Injury associated with SUD	(20.5%)	51 (17.3%)	0.66
Total number of PP Visits	3.9+/−3.3	4.5+/−2.7	0.29
Breastfeeding at discharge	10 (26.3%)	63 (20.9%)	0.53
Rooming in with infant	18 (50%)	152 (53%)	0.86
Neonatal outcomes
Birth weight	3047 ± 632	2932 ± 517	0.96
Birth weight percentile	44.3 ± 29.9	34.8 ± 24.4	0.03
Small for gestational age births	3 (7.9%)	43 (14.6%)	0.03
APGAR 1 min	7.7 ± 1.7	7.9 ± 1.6	0.33
APGAR 5 min	8.4 ± 1.2	8.6 ± 1.0	0.86
Infant treated for NOWS	9 (23.1%)	137 (46.6%)	0.01
Highest Finnegan Score	9.6 ± 4.5	12.3 ± 4.0	<0.001
Length of NOWS treatment (days)	13.0 ± 5.6	16.1 ± 9.3	0.33
Infant admitted to NICU	26 (66.7%)	227 (77.2%)	0.16
Total length of hospital stay (in days)	11.1 ± 15.2	15.6 ± 18.1	0.15
MDT with buprenorphine metabolite*	24 (72.7%)	258 (94.9%)	<0.001
UDT with buprenorphine metabolite**	23 (59%)	264 (90.7%)	<0.001

*MDT, Meconium drug test; **UDT, urine drug test.

Furthermore, the wean group exhibited significantly higher birth weight percentiles for their infants compared to the control group (44.3 ± 29.9 vs. 34.8 ± 24.4, p = 0.03). However, analysis of other variables such as gestational age at delivery, mode of delivery, complications during delivery, total number of prenatal and postpartum visits, and total number of inappropriate visits revealed no significant differences between the two groups. There were no maternal overdoses or deaths during the pregnancy. However, there were 4 maternal deaths in the postpartum period, none were in the buprenorphine wean group. All four deaths were due to pregnancy associated causes.

Discussion

Opioid use in pregnancy has escalated dramatically in recent years, paralleling the epidemic observed in the general population. Correspondingly there has been a significant increase in the incidence of NOWS. Pregnancy provides an important window of opportunity to identify and treat people with substance use disorders. For pregnant people with an opioid use disorder, MOUD with either buprenorphine or methadone is the recommended therapy and has been shown to improve outcomes [4].

Neonatal Opioid Withdrawal Syndrome is an expected and treatable condition seen in 30–80% of infants born to women taking opioid agonist therapies [4]. Nevertheless, there are some patients that desire weaning of buprenorphine to decrease the risk of NOWS. These patients are counseled regarding the concern of destabilization of their recovery with weaning and the increased risk of returning to illicit drug use [5–7]. Of particular concern is that these patients have a lower tolerance to opiates, and illicit opiates are increasingly laced with potent synthetic opiates such as fentanyl, increasing the risk of overdose. Despite these conversations, some patients will still request weaning, which requires shared decision making. As healthcare providers our primary goal is to continue to have these patients engaged in their recovery including comprehensive prenatal care, social services, pain management, mental health care, and health education to improve both maternal & neonatal outcomes. Patients who underwent buprenorphine wean were informed this was contrary to the standard of care and they were encouraged to restart MOUD immediately postpartum, particularly as most maternal overdose deaths occur in the postpartum period [8]. For those patients that decline restarting buprenorphine postpartum, long acting Naltexone is offered which is also a medication used to treat opioid use disorder.

Our study provides evidence of the potential benefits of buprenorphine weaning during pregnancy in stable and motivated patients, particularly in reducing the incidence and severity of NOWS. This contrasts with the control group, who maintained their buprenorphine treatment throughout pregnancy and exhibited higher rates of NOWS and lower birth weight percentiles. Edinburgh Postnatal Depression Scale (EPDS) was attempted in at least two postpartum visits (1–2 and 6 week visits). Interestingly, EPDS scores were significantly lower in the buprenorphine weaning group. Although multifactorial, less NOWS may have contributed to that finding. Anxiety medications and tobacco use were similar between both groups as these drug classes can impact NOWS. A strength of our study includes our diverse recovery support services that may have helped the success of the wean group. Patients were strongly encouraged to continue their recovery care in our program for a full year postpartum. It would be interesting to look at sexually transmitted disease conversion in the late antepartum and postpartum period as another way to assess the feasibility of this approach. Patients that experience destabilization from being weaned from buprenorphine may engage in risky sexual behaviors to obtain illicit drugs.

Limitations

There are several limitations to consider in this study. First, the sample size of the wean group was relatively small, which may limit the statistical power to detect smaller but potentially clinically significant differences. Second, our results may not be generalizable to broader populations as this was a single center study with wrap around services in a predominately white population. Patients that requested weaning were stable and highly motivated. Not all patients were followed up in our program for a full year postpartum. Furthermore, potential confounding factors such as socioeconomic status, support system, and mental health conditions were not completely accounted for in the control measures.

Conclusion

In conclusion, our results suggest that buprenorphine weaning during pregnancy may offer potential benefits for neonatal outcomes. Benefits particularly include reducing the incidence and severity of NOWS with no apparent increased maternal or neonatal morbidity in patients who have self-determination to wean and are adherent to the prescribed comprehensive treatment regimen. However, given the limitations of this study, further large-scale, multicenter studies are needed to confirm these findings and explore additional outcomes.

Acknowledgments

Asmita Shrestha, MPH; Arnold Stromberg, PhD

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Due to the nature of the research, due to ethical and privacy concerns, supporting data is not available.

Additional information

Funding

This research was supported by the Kentucky Opioid Response Effort (KORE) via Substance Abuse and Mental Health Services Administration (SAMHSA) [Grants, H79TI081704, H79TI083283], as well as the data management system that is hosted by the University of Kentucky with grant support from [NIH CTSA UL1TR001998]. This work was supported by Commonwealth of Kentucky.