History of Neonatal Abstinence Syndrome
The first surviving records of opium addiction date from the end of the eighteenth century although there are accounts of opium use from many centuries before that [1]. Morphine was isolated in 1804 and heroin was synthesized in 1874 [2]. Addiction to heroin and morphine became more common after their commercial production. As early as the nineteenth century [3], an increase in the incidence of morphine and heroin addiction among women was reported. Common thinking was that infants were not affected because morphine use among women was associated with sterility and a loss of sexual desire. In 1875, this was proven to be false after a case in a neonate who manifested signs of opioid withdrawal at birth was reported [4]. The diagnosis made was Congenital Morphinism and this case was followed by a number of similar reports [5], but sadly, most of the in-utero-exposed infants died after no specific treatment was offered [6]. In 1894, a series of 12 infants with Congenital Morphinism were treated with paregoric and 9 died. It is suspected that most of the infants died from prematurity and its complications as well as the lack of treatment for opioid withdrawal. In 1903, a report described an infant who survived after being treated with morphine [7]. When the successful treatment of seizures in an infant exposed to morphine in utero was reported in 1947 [8], Congenital Morphinism no longer was considered a medical curiosity. With the appearance of more reports [9, 10] significant attention to the condition occurred from obstetricians and pediatricians. Congenital Morphinism was subsequently renamed Neonatal Abstinence Syndrome (NAS) because of the constellation of signs culminating in one disease process.
Recently, the FDA has used the term neonatal opioid withdrawal syndrome (NOWS) when referring to maternal use of opioids during pregnancy with an outcome of withdrawal in the infant. However, there is some concern that this term is misplaced for use in a clinical setting. When NAS occurs as a result of prenatal exposure to an opioid, it occurs in a number of different contexts and the presentation and severity is related to a number of factors in addition to the opioid(s). Accordingly, this chapter will use the NAS terminology.
Between 1953 and 1955, 18 infants were reported with NAS. There was a 25 percent mortality rate due to prematurity and respiratory failure. The NAS was treated with methadone, phenobarbital, or paregoric. In 1956, 32 babies with NAS were reported with a 50 percent prematurity rate and a 34–93 percent mortality. When methadone was introduced for treating opioid addiction, it was thought that NAS would not occur [11]; however, it did and generally was more severe than that associated with heroin [12]. This could have been anticipated since patients received a steady dose of the medication whereas heroin doses on the street were variable.
Burgeoning numbers of babies with NAS were presenting to large city hospitals and reports by Blinick [13] and Zelson [14] from New York described the observed signs and the treatment that they were using, i.e., chlorpromazine [14]. Case series continued to be reported and then in 1975, Desmond and Wilson published a comprehensive paper on recognition and diagnosing NAS.
In 1974, realizing the need for research in the area of drug abuse in general and in perinatal drug dependence specifically, the Federal Government established the National Institute on Drug Abuse (NIDA) within the Alcohol Drug Abuse Mental Health Administration (ADAMHA). Areas researched included maternal medical effects on pregnancy and the newborn, psychological, and sociological aspects of maternal addiction, and comprehensive services for mother and the newborn. Many improvements in diagnosing and treating the maternal/infant dyad were established as a result of NIDA (now at NIH) funding. Recognizing that an assessment tool was important, in 1975, Finnegan et al. [15] (The Neonatal Abstinence Syndrome Score) and Lipsitz [16] (The Neonatal Withdrawal Scoring System) published their respective scoring tools.
Buprenorphine was approved in Europe in 1996 and in the United States in 2002 and NAS was found to be associated with this medication as well [17–20]. Neonatal abstinence associated with maternal use of prescription pain medications is included in the cases currently reported [21, 22]. Some of these opioid prescriptions and their date of reaching the marketplace include Percocet® (1976), Vicodin® (1984), and Oxycontin® (1996). The numbers of cases continued to escalate and have reached epidemic proportions in the past 10 years.
Incidence of NAS
Historically, the incidence of NAS has been described in terms of the percentage of infants prenatally exposed to opioids who required treatment for NAS. These data are all based on cohort studies with a wide range of percentages reported, i.e. from 29 percent to as high as 94 percent [22, 23]. Such a wide range is often due to a lack of consistent operational definitions, either in clinical settings or research findings, when NAS is reported. Some restrict the diagnosis of NAS to when treatment is required; others base the diagnosis on any reported signs. Moreover, there are numerous factors, in addition to prenatal opioid exposure, that contribute to severity of withdrawal and are described in other sections within this chapter.
Within the past 5 years, numerous researchers have utilized national and state data sets based on IDC-9 hospital codes of neonatal drug withdrawal [24]. The initial study published in 2012 found that the incidence of NAS had increased almost fivefold from 2000 to 2009 [25]. Subsequent studies have found that the incidence varies significantly between states [26], that the geographic variations in NAS are consistent with the variations in opioid pain prescriptions [27], and that the incidence of NAS and maternal opioid use increased disproportionately in rural counties relative to urban counties [28]. While this research has helped to emphasize consequences of the current prescription opioid epidemic, there are many limitations to these data, specifically in relation to whether or not the infant required treatment. Additionally, these data do not differentiate NAS that occurs from illicit opioid use, e.g. drugs such as heroin and misuse of prescription opioids, and NAS subsequent to appropriate maternal use of opioids under a physician’s care, e.g. medication for treatment for opioid use disorder or pain management. While the risk of NAS may be comparable, the overall risk to the fetus and neonate may differ significantly.
Morbidity in Infants Exposed to Opioids In Utero
Reports during the 1970s and 1980s showed that newborns of women using heroin had low birth weight and prematurity as the two most prominent medical issues. Conditions such as asphyxia neonatorum (respiratory failure due to inadequate intake of oxygen before, during, or just after birth), intracranial hemorrhage, nutritional deprivation, hypoglycemia (low blood sugar), hypocalcemia (low blood calcium), septicemia (blood-borne infection), and hyperbilirubinemia (jaundice) should also be anticipated in heroin-exposed babies. Infants exposed to methadone in utero are more likely to have higher birth weights and a decreased incidence of premature birth. Medical issues in infants born to women using heroin are generally influenced by the following: amount of prenatal care received by the mother; whether the mother suffered obstetrical or medical complications during the pregnancy; and whether the fetus was exposed to multiple drugs which can produce an unstable intrauterine milieu that can be further complicated by symptoms of withdrawal and overdose in both the woman and the fetus [29]. A number of reports [30, 31] demonstrate differences in heroin versus methadone exposed babies with regard to birth weight, neonatal problems, length of hospital stays, premature birth, and mortality. Rorke performed careful neuropathological examinations on preterm heroin-exposed babies whose mothers did not attend prenatal care or addictions services. She found lesions in the brains of the babies which demonstrated the result of hypoxia which probably occurred during maternal episodes of withdrawal or overdose, causing a lack of oxygen to the fetal brains [32].
It must be noted that data are lacking regarding morbidity in infants exposed to opioid pain medications. It is also unclear how morbidities differ in the contemporary population of infants born to women using illicit drugs. Some evidence is provided by Holbrook et al. [33] in which obstetric and neonatal outcomes in pregnant women with opioid use disorder (OUD) maintained on methadone or buprenorphine were examined. In a sample of 131 pregnant women with a history of heroin use or prescription misuse, the incidence of infectious medical complications and obstetric complications were much lower than historically reported rates among heroin users. Of import is that morbidity and mortality rates in pregnancy and in newborns can be decreased if women enroll in medication assisted treatment (MAT), receive prenatal care, and receive comprehensive services for the OUD [34].
NAS and Concomitant Drug Use
Signs similar to opioid withdrawal in neonates have been described for alcohol, benzodiazepines, and Selective Serotonin Reuptake Inhibitors (SSRIs) [35]. While not all require treatment, the interaction of nonopioids and opioids can affect the presentation of NAS. SSRIs and benzodiazepines have consistently been found to contribute to NAS severity.
Studies have found SSRIs to be related to both the presentation and treatment of NAS, with higher peak scores of NAS [36] and higher doses of medication required for treatment [36, 37]. However, the concomitant use of SSRIs with opioids has not been found to predict whether infants would require treatment for NAS or length of treatment [36–39]. It should be emphasized that in the case of a pregnant woman requiring medication treatments for OUD and depression, these data should not be used to support treating one disorder over the other.
A number of studies have found benzodiazepines to be associated with prolonged length of stay in infants being treated for prenatal opioid exposure [38, 40, 41]. Benzodiazepine use and misuse is prevalent among pregnant women with opioid use disorders and presents both a risk to the fetus from maternal overdose sedation and possible death, and to the neonate from NAS that requires lengthy pharmacotherapy with multiple drugs.
Cigarette smoking in women with opioid use disorders has also been found to affect the timing and length of treatment of NAS [42, 43] as well as the total amount of medication required to treat NAS [43]. With appropriate medical management and/or behavioral intervention, the impact of benzodiazepines and cigarettes on NAS can be avoided.
A caveat must be included for this section in that all of the data that support the information provided here are derived from infants prenatally exposed to methadone or buprenorphine. This information may or may not be generalizable to other opioids.
Variability in the Expression of NAS
In addition to concomitant drug use, there are several other factors related to the presentation of NAS. The gestational age of the infant has been found to be related to NAS with preterm infants requiring less treatment. One study found no difference in the need for treatment but it was the only study that limited the preterm group to late preterm, i.e. 34–36 weeks [44]. Studies have consistently found that preterm infants require less medication for treatment of NAS and have shorter duration of treatment [44–47]. Reasons for differences in opioid withdrawal for preterm infants are unknown with several possible explanations: (a) an immature hepatic metabolic system resulting in a gradual decline in serum concentration, (b) the developmental immaturity of the brain and associated opiate receptors, (c) less exposure over time to opioids, and (d) increased placental transfer of opioids in the third trimester [47]. It should be noted that all but one of these studies used the Finnegan Scoring Tool to assess NAS. The Finnegan tool was validated only for term infants and the use of this tool for preterm infants has not been validated. A validated NAS assessment for preterm infants may yield different results.
For decades, NAS due to prenatal opioid exposure was primarily the result of maternal heroin use and/or treatment with methadone for maternal OUD. With the addition of buprenorphine for the treatment of OUD, questions arose as to whether the frequencies of signs of NAS differ between the two medications. Data from the MOTHER study [48] were used to compare the NAS profile based on the MOTHER NAS (MNS) scale before treatment or in the absence of treatment in infants prenatally exposed to methadone or buprenorphine. The incidence of nasal stuffiness, sneezing, and loose stools was greater in the buprenorphine-exposed infants than the methadone-exposed infants. Additionally, methadone-exposed infants had higher mean scores for hyperactive Moro, disturbed and undisturbed tremors, failure to thrive, and excessive irritability [49]. Differences may exist between the two medications because of their pharmacological characteristics (full agonist versus partial agonist). These findings help explain differences in NAS severity between methadone and buprenorphine. There are no studies that have examined the NAS profile of infants prenatally exposed to opioid pain medications such as oxycodone, hydrocodone.
Several investigators have posited that the sex of the infant might play a role in predicting NAS severity since past studies have shown an increased vulnerability to adverse outcomes among males. Unger and colleagues [50], looking at sex-based differences in birth weight and length, head circumference, and NAS duration and severity, found that while males had a significantly higher birth weight and head circumference, there were no significant sex-related differences for NAS development, severity, duration, or medication administered. They reported that a similar postnatal vulnerability exists for both males and females, suggesting that factors other than sex are the major determinants of clinically significant NAS. No differences were found by Holbrook and Kaltenbach [51] in male versus female newborns with regard to the need for NAS treatment, length of treatment and peak dose of medication required.
Within the past 4 years, there has been significant, albeit preliminary, research to identify genetic and epigenetic factors that contribute to NAS variability. A series of studies from the same lab have progressively identified variants in the OPRMI and COMT genes associated with shorter length of hospital stay and less need for treatment in infants with NAS [52]; increased methylation within the OPRMI promoter associated with length of stay and the need for more than 2 medications to treat NAS [53]; and PNOC variants associated with NAS severity [54]. This work has exciting potential to provide individualized treatment regimens tailored specifically for infants with high-risk genetic profiles.
NAS and Agonist Treatment for Opioid Use Disorders
Questions often arise regarding the use of MAT for opioid use disorders during pregnancy due to the expected occurrence of NAS. However, the risk to the infant for withdrawal is considered much less harmful than the benefits of MAT for the mother and fetus. The 1993 and 2004 SAMHSA Treatment Improvement Protocols for Opioid Use Disorders, the 1997 NIH Consensus Panel on Effective Medical Treatment of Opioid Addiction, the 2012 American College of Obstetricians and Gynecologists and American Society of Addiction Medicine Joint Opinion, the WHO 2014 Guidelines for the Identification and Management of Substance Use and Substance Use Disorders in Pregnancy and the 2016 SAMHSA Collaborative Approach to the Treatment of Pregnant Women with Opioid Use Disorders all recommend MAT for pregnant women as the standard of care. Medications used to treat opioid use disorders in women who are pregnant are methadone, buprenorphine, and buprenorphine with naloxone (see Chapter 10 for a discussion on the use of these medications).
The relationship between maternal medication dose and NAS is often a question of clinicians, with the expectation that lower maternal medication dose will result in less severe NAS. This is especially applicable to methadone because of studies in the 1970s that recommended pregnant women be maintained on very low doses. An extensive literature from 1966 to 2009 was the subject of a systematic review and meta-analysis that found the severity of NAS does not differ between mothers on high or low doses of methadone [55]. A more recent study [56] examined the relationship between maternal methadone dose at delivery and neonatal outcomes with data from the MOTHER study and also found no relationship between maternal methadone dose and severity of NAS. These data are important to note because other drugs of abuse did not confound the sample and those who conducted the NAS assessment were trained to a gold standard of inter-rater reliability.
Due to buprenorphine’s more recent approval, there is less literature on the relationship between maternal buprenorphine dose and severity of NAS. However, the available literature consistently reports no relationship. This includes a study with data from the MOTHER study that has the unique strengths mentioned above [57].
Differences in NAS as a function of whether the infant has been exposed to methadone or buprenorphine have been reported in a number of studies. The most widely referenced study is the MOTHER multisite randomized controlled trial [48]. This study (n = 131) found no difference in the number of infants in each medication group that required treatment for NAS. However, infants prenatally exposed to buprenorphine (n = 58) required 89 percent less morphine to treat NA, spent 58 percent less time in the hospital being medicated for NAS, and spent 43 percent less time in the hospital overall than infants prenatally exposed to methadone (n = 73). A systematic review and meta-analysis of 12 studies, including the MOTHER study, found similar results in that NAS treatment duration was shorter and morphine dose lower for buprenorphine-exposed infants than methadone-exposed infants [58].
While NAS outcomes are superior with buprenorphine than methadone, practitioners should not use that as the sole criteria for which medication to use during pregnancy. As a partial agonist, buprenorphine may not be effective for certain women and methadone may be the medication of choice for those women (refer to Chapter 10). Current recommendations suggest that women maintained on methadone or buprenorphine who become pregnant should remain on their respective medications. A pregnant woman who becomes pregnant who is naive to agonist treatment may be a good candidate for buprenorphine. If she does not respond well to buprenorphine she can easily be transitioned to methadone. Overall, the woman’s medical, psychological, and substance use history should be taken into consideration in any treatment decision [59].
NAS Onset and Signs
Infants with in utero exposure to opioids have a chance to be born with NAS. Reported percentages differ with the population studied. Psychoactive drugs, such as opioids, have low molecular weights and lipid solubility which permits them to pass easily from the mother to the fetus through the placenta. Once the drugs accumulate in the fetus, equilibrium is established between the maternal and fetal blood. When the umbilical cord is cut at birth, there is a disruption of the trans-placental passage of drugs potentially permitting the development of signs of abstinence. Affecting the central nervous, autonomic nervous, gastrointestinal, and respiratory systems, this constellation of signs constitutes the multisystem disorder called Neonatal Abstinence Syndrome. NAS can be a serious medical condition because it affects the vital functions that permit growth and normalcy in the newborn, i.e., feeding, elimination, and sleep. If untreated, NAS can cause death as was noted in the historical accounts. Death can occur because of excess fluid loss, high temperatures, seizures, respiratory instability, and aspiration of fluid into the lungs with cessation of breathing. Current medical knowledge about drug use during pregnancy and newborn care eliminates the chance of death from NAS [60].
Neonatal abstinence resulting from opioids is characterized by signs of CNS hyperirritability, gastrointestinal dysfunction, and respiratory and autonomic nervous system signs.
Central nervous system – Tremors, high pitched crying, hypertonia, irritability, increased deep tendon reflexes and an exaggerated startle reflex are all characteristics of neonatal abstinence. An exaggerated rooting reflex and a voracious appetite manifested as sucking of fists or thumbs are also common, yet when feedings are administered the infant may have extreme difficulty because of an uncoordinated sucking and swallowing mechanism. Kron et al. [61] found that both heroin and methadone-exposed infants showed reductions in sucking rates and pressures, disordered sucking organization and a reduction in the amounts of nutrient consumed. As a result of these issues, an infant affected by NAS may have difficulty during feeding. Breastfeeding can be challenging because of their disorganized sucking reflex but with support it can be accomplished. Seizures, a serious sign, fortunately occur infrequently. However, because they may be subtle or confused with exaggerated tremors, the reported incidence of seizures varies, ranging from 1 percent [60] to 5.9 percent [62] and 7.8 percent [63] of newborns exposed to heroin or methadone during pregnancy. Certain practices used in the treatment of NAS, such as tightly swaddling the infant in a darkened room, can also make it difficult to observe seizure movements.
Gastrointestinal system – Regurgitation, projectile vomiting, and loose stools are gastrointestinal manifestations of NAS. Dehydration because of poor intake, coupled with increased losses from the gastrointestinal tract, can cause excessive weight loss, electrolyte imbalance, shock, coma, and even death, however with current neonatal practices, this should not occur. Babies with mild NAS not requiring treatment can lose about 4 percent of their birth weight, regaining it by the 7th day of life. Newborns displaying more severe abstinence lose more weight and do not regain their weight until an average of 2 weeks after birth, suggesting that timely and appropriate pharmacological control of abstinence, combined with the provision of extra fluids and calories to offset the weight loss, are important in the management of NAS [64]. Although babies that exhibit mainly gastrointestinal symptoms might not meet the criteria for pharmacological treatment of their NAS, it is still extremely important for these babies to be monitored following discharge from the hospital because of the potential for water losses and poor intake, both of which can lead to dehydration.
Respiratory system – Increased respiratory rate can be accompanied by retractions between the ribs of the chest wall, intermittent cyanosis, and apnea [65, 66]. Severe respiratory distress occurs most often when the infant regurgitates, aspirates, and develops aspiration pneumonia. Infants with acute heroin withdrawal were found to have increased respiratory rates, leading to hypocapnia and an increase in blood alkalinity during the first week of life [67]. Glass found that the incidence of respiratory distress syndrome was actually decreased in NAS-affected infants, possibly because of chronic intrauterine stress, accelerated heroin-mediated maturation of lung function or perhaps both [67]. This decrease, however, is just one benefit among a long list of adverse effects resulting from intrauterine heroin exposure.
Autonomic nervous system – The signs seen during abstinence include sneezing, yawning, skin color changes (mottling), and water loss caused by increases in temperature and the shedding of tears. Behrendt and Green [68] found that approximately 40 percent of low-birth-weight infants born to heroin-dependent mothers had spontaneous generalized sweating. In comparison, this condition appeared in less than 1 percent of healthy low-birth-weight babies.
In an assessment of 135 newborns exposed in utero to heroin and/or methadone undergoing NAS, the following signs were seen in 50 percent or more of newborns: tremors (mild or marked and disturbed and undisturbed), high pitched cry, continuous high pitched cry, sneezing, increased muscle tone, frantic sucking of the fists, regurgitation, sleeps < 3 hours after feeding, respiratory rate > 60/minute, poor feeding, hyperactive Moro reflex, and loose stools [69].
Assessment and Diagnosis
The level of fetal drug exposure can be affected by: the amount and purity of the drugs taken by the mother; the length of drug use and the mother’s drug metabolism; and the individual kinetics of placental drug transfer. Because of these variables, it is difficult to clinically predict at birth whether a baby will develop NAS or, if developed, whether that NAS will be of a mild, moderate, or severe degree. As a result, careful assessment and diagnosis of newborns exposed to drugs in utero is the only way to determine if they will develop symptoms and warrant subsequent treatment. The use of a toxicology screening protocol at birth is helpful in monitoring for the onset of NAS. Urine or meconium toxicology screens are used. Meconium is the first intestinal discharge of the newborn infant. Consisting of epithelial cells, mucus, and bile, it also includes drugs the fetus was exposed to from about the 20th week of gestation [29].
Assessment of NAS using a Scoring Tool
Not all drug-exposed newborns experience abstinence symptoms, therefore, routine prophylactic treatment is not recommended for NAS. Close observation of clinical signs is very important. The use of a severity score allows for the accurate evaluation of signs, avoids unnecessary treatment of mildly affected infants, and provides a methodology for the effective tapering of medications.
In the 1970s when NAS was being recognized and clinicians were considering treatment, no systematic, feasible, assessment, and treatment guideline was available for neonates. Scores to assess and treat abstinence in adults existed, i.e., the Himmelsbach Score [70], but those were not feasible for babies. With the undulating pattern of symptoms in the neonate vacillating from exacerbations of vigorous movements to sleep from exhaustion, a 24/7 assessment is appropriate. In the development of a score for newborns, the signs needed to be observed and categorized.
Signs of NAS are generally listed on score sheets according to the system that they affect: central nervous system, gastrointestinal system, respiratory system, and autonomic nervous system. It is important to confirm the diagnosis by maternal history, clinical findings of opioid use (e.g. track marks), urine toxicology screens, or meconium assessments. The laboratory assessments are confirmatory of exposure, but clinical examinations will determine if NAS appears. Scores permit administration of care systematically and selectively with avoidance of prolonged need for medication and hospital stay. A score allows for a “common language,” decreased variability, and improves parent communication and involvement.
Assessment for other neonatal conditions with similar signs is important since the neonates are born to high risk women who generally have various morbidities. The following are conditions that can present with some of the signs similar to NAS which may have to be ruled out: septicemia, encephalitis, meningitis, postanoxic CNS irritation, hypoglycemia, hypocalcemia, and cerebral hemorrhage.
A number of scoring tools were developed during the 1970s–1990s and reported in the pediatric literature. The following scores have been published from 1975 to 2010: Neonatal Abstinence Score [15]; The Neonatal Drug Withdrawal Scoring System [16]; Ostrea Tool [71]; The Neonatal Narcotic Withdrawal Index [72]; Neonatal Withdrawal Inventory (NWI) [73]; MOTHER NAS Scale (modified Finnegan) [48]. All the scores except one use a dose of medication related to the weight (mg per kg per day divided into doses for the day). The MOTHER NAS Scale bases dose according to the score irrespective of weight.
The Finnegan Neonatal Abstinence Score is mentioned by the American Academy of Pediatrics in their guideline article for measuring the onset, progression and diminution of symptoms of abstinence [35] and is the most widely used tool in the United States today. The Lipsitz Neonatal Drug Withdrawal Scoring System [16] is used in some institutions but is not represented in the current literature. Use of the MOTHER NAS Scale is the standard instrument used in a number of clinical trials [74]. Figure 14.1 is a Finnegan scoring sheet.
Figure 14.1 The Finnegan Neonatal Abstinence Score
The Neonatal Abstinence Score by Finnegan et al. [15] has an instructional manual and a DVD to train examiners to correctly administer the score. The manual assists with the following: accurately assesses opioid-exposed infants for the presence of abstinence signs; describes appropriate examination techniques required to evaluate NAS; reliability testing to achieve 90 percent reliability with other examiners in using the score. After orientation with this manual and DVD, an examiner can perform the score in 4–5 minutes. Since 1983, the score sheet has been translated into other languages [75].
Nonpharmacological Interventions
Traditionally, comforting techniques such as swaddling, the use of a pacifier, and caregiving within a quiet, dimly lit environment have been recommended for reducing the signs of NAS. There is some clinical evidence to suggest that much NAS is really maternal abstinence syndrome, meaning separating babies from their primary caregivers causes distress. Treating babies in bright, noisy environments such as the NICU without maternal soothing has been anecdotally shown to worsen infant outcomes and increase the incidence of NAS. Such interventions are based on clinical experience rather than systematic evaluation and are primarily related to nursing care. This is not to say these techniques are not effective but to highlight a shift to a broader approach in which nonpharmacological measures focus on the importance of the mother and infant as an interactional dyad [76]. This has led to a number of studies that have identified breastfeeding as an important factor in decreasing NAS scores, the need for treatment, length of pharmacotherapy, and length of hospital stay in infants prenatally exposed to methadone or buprenorphine as important in decreasing [40, 77–79]. The act of breastfeeding is imbedded with behaviors that are soothing and comforting to the baby and promotes mother/infant attachment. This has led to a number of medical academies/societies, e.g. the Academy of Breastfeeding Medicine, 2015; the American Society of Addiction Medicine, 2012; and the World Health Organization 2014, to include in their Guideline recommendations that pregnant women maintained on methadone or buprenorphine who are not using illicit drugs, are engaged in treatment and are not HIV positive, be encouraged and supported in breastfeeding.
This same focus on the mother–infant dyad has led to integrating mothers as partners in care. A number of studies have examined the use of providing care within a rooming-in setting rather than in the usual NICU setting. Such an approach was first developed in Canada and received support through the Vermont Oxford Network (VON) NAS, a 3-year quality improvement project of 2013–15 [80]. Prior to the VON project, the majority of studies were conducted in Canada [81, 82], with 2 from England and Germany, respectively [83, 84]. A recent publication from the Dartmouth University hospital system that implemented a Plan-Do-Study-Act (PDSA) program in response to the VON project found that by providing extensive training on the NAS scoring tool, scoring the infant after on-demand feeding, using a standardized physician score interpretation, and keeping the infant in a rooming-in setting reduced the length of stay from 16.9 to 12.3 days and cumulative morphine dose from 13.7 to 6.6 mg [85].
The VON project also has led to studies examining the importance of standardizing hospital practices such as treatment protocols. Studies have found that utilizing a standard NAS treatment and weaning protocol for either morphine or methadone reduced duration of treatment and length of hospital stay [86], that staff education and the use of a standardized morphine protocol reduced length of stay [87], and that the use of explicit weaning guidelines resulted in shorter duration of treatment, length of stay, and lower rate of adjunctive drug therapy [88].
Pharmacological Treatment
The need for and appropriate pharmacotherapy of neonatal opioid abstinence is based on the principles of accurate diagnosis and assessment. Since the signs of abstinence are nonspecific, the clinician must consider other serious neonatal conditions. In the neonatal period, pharmacotherapy is aimed at rapid clinical stabilization of the opioid-exposed infant followed by gradual reduction of the medication under careful medical supervision. Close assessment of the infant will ensure that mildly affected infants will not be treated unnecessarily and that infants who do require pharmacotherapy will not be allowed to develop severe abstinence with its attendant risk of significant morbidity and even mortality. For moderate to severely affected infants, medication dosage, in mg per kg, is related to the score and titrated up or down according to changes in scores. Medications are recommended when the score reaches a critical level and the neonate cannot be managed by supportive nursing measures. Dependent on a variety of factors, the average newborn will recover from NAS in 5–30 days and no untoward effects should occur as a result of this condition if appropriate assessment and treatment principles are provided.
The clinician should not procrastinate in providing medication for an infant whose NAS scores are at the treatment level. Waiting for severe symptoms to appear such as a seizure or dehydration is not warranted and precious time is lost in controlling the NAS. In the past, paregoric and tincture of opium were used to treat NAS. For a variety of reasons, not the least of which was the other substances within the medication (alcohol, anise oil, camphor, etc.) they are no longer prescribed. Diazepam and chlorpromazine are also not used because of their prolonged half-lives and associated complications [89].
Medications currently used by clinicians to treat NAS include: opioids (e.g. morphine, methadone, buprenorphine), anti-convulsants (e.g. phenobarbital), and centrally acting α2 agonists (e.g. clonidine). All of these medications have side effects and many include alcohol in the preparations. Morphine reduces bowel motility and loose stools in addition to facilitating feeding and interpersonal interaction. Morphine is also advantageous in that it has no alcohol and it has a short half-life. However, as with all opioids, it may cause apnea, sedation, hypotension, delayed gastric emptying, loss of bowel motility, and urinary retention.
It is generally assumed that opioid exposure precipitating NAS is best treated with another opioid; therefore, medications such as morphine and methadone have become the treatment of choice for NAS [35]. The recent recommendations by the American Academy of Pediatrics suggest carefully outlined regimens for orally administered methadone or morphine [35, 90]. Once the infant is clinically stabilized on a medication (based on decreasing abstinence scores), the total daily dosage can be lowered by 10 percent each day. When the medication is discontinued, the infant should be monitored in the hospital for rebound abstinence signs for at least 2 days [91, 92].
Kraft and colleagues conducted a series of open label randomized trials to examine the feasibility of using buprenorphine to treat NAS. The initial study compared sublingual buprenorphine versus a neonatal opium solution and found a nonsignificant reduction in length of treatment and duration of hospitalization in the buprenorphine group [93]. As this was the first study using buprenorphine to treat infants, a conservative dose was used. A subsequent study using an adjusted higher dose of buprenorphine found that the length of treatment was reduced by 40 percent (p = 0.01) and length of stay decreased by 24 percent (p = 0.05) for the buprenorphine treated infants compared to infants treated with morphine [94]. This group recently completed a double blind double dummy RCT comparing buprenorphine to morphine and found the duration of treatment to be significantly shorter with buprenorphine than methadone (15 days vs. 28 days, p < 0.001) [95].
Phenobarbital is frequently used as a second drug if morphine treatment does not adequately control the signs of abstinence, especially if exposure has been complicated by polydrug use [96, 97]. Phenobarbital is a nonspecific CNS depressant that offers the advantage of a broad spectrum of sedation controlling symptoms of irritability and insomnia in 50 percent of infants regardless of the mother’s choice of drug. However, it does not control non-CNS signs such as loose stools; it depresses sucking; and, in larger doses, it can depress respirations and may mask the severity of NAS symptoms, thus infants need to be closely monitored for the possibility of over-sedation. But if physicians properly monitor the blood level of the phenobarbital, there should be no problem when using this medication for the treatment of NAS [98, 99]. Generally, phenobarbital is not recommended as the primary agent for treating NAS but as adjunct therapy in cases of polydrug exposure.
The Cochrane Controlled Trials Register [100, 101] examined the effectiveness and safety of opioid treatment compared to the use of sedatives and supportive care. In 285 infants meeting inclusion criteria in five studies, they concluded there is no evidence that phenobarbital (compared to supportive care alone) reduces treatment failure. In seven studies with 585 infants meeting inclusion criteria, they concluded that opioids (compared to supportive care alone) appeared to reduce both the time needed to regain birth weight and the duration of supportive care, but increased the duration of hospital stay. When compared to phenobarbital, opioids might reduce the incidence of seizures but there is no evidence of effect on treatment failure, further illustrating the superiority of opioids for the treatment of NAS. An excellent review is found in an article by Kocherlakoda [102]. Figure 14.2 is a sample management plan based on this work.
