Barriers and Facilitators of Breastfeeding Reported by Postpartum Women in Methadone Maintenance Therapy

Abstract

Introduction:

This study utilized a cross-sectional qualitative and quantitative interview-based survey to capture the infant feeding practices and barriers to exclusive breastfeeding for women in methadone maintenance therapy. Participants were recruited from an opioid dependence treatment center in an urban setting in the Southeastern United States.

Materials and Methods:

A convenience sample of women in treatment (n = 30) were interviewed using an adapted instrument designed to capture decisions and intentions to formula feed or breastfeed; support from friends and family; hospital experience; support from healthcare personnel; and maternal knowledge of breastfeeding while taking methadone.

Results:

The majority of women in the sample initiated breastfeeding, but only 10% continued for >1 month. Challenges related to infant hospital stay posed a significant barrier. Two-thirds of infants remained hospitalized after the mother was discharged. Out of the 24 women who initiated breastfeeding, 11 reported that they discontinued because of issues related to infant's neonatal intensive care unit (NICU) stay. Eleven women reported that their healthcare providers did not discuss breastfeeding with them. Women who were encouraged to breastfeed by healthcare staff were more likely to breastfeed for longer durations.

Conclusions:

Women in treatment for opioid dependence both desire and attempt to establish breastfeeding, but encounter significant challenges, including long NICU stays and lack of support and education, that compromise their success. These findings should inform the development of future programs or interventions geared toward increasing breastfeeding initiation, support, and duration among women who give birth to babies while in treatment for opioid addiction.

Introduction

Between 2000 and 2009, there was a fivefold increase in opiate abuse among U.S. women aged 15–44 years, and a corresponding increase in the number of newborns with neonatal abstinence syndrome (NAS).^1,2 NAS is a set of symptoms caused by acute opioid withdrawal and affects 60–90% of babies born to women who used opioids during pregnancy.^3,4 Although not universally recommended, the American College of Obstetrics and Gynecology (ACOG), The American Academy of Pediatrics, and the Academy of Breastfeeding Medicine endorse breastfeeding as safe and encourage women who are HIV negative and stable on opioid agonist treatment postpartum.^3,5,6 The very low transfer of both methadone and buprenorphine into human breast milk carry no proven adverse effect and has been shown to decrease the risk and severity of NAS for infants exposed to opioids in utero.^3,4

Despite these recommendations, women in methadone maintenance therapy have breastfeeding rates of less than half the national rate, and of those who initiate breastfeeding, more than half will stop within 1 week.⁷ Although women in treatment experience many of the same struggles with breastfeeding that other women do, there are complex physical, emotional, and social challenges that are unique to women with opioid dependence.⁸ These include, but are not limited to a history of sexual abuse or trauma, fear of hepatitis C transmission, although not a contraindication, and the high risk of relapse to illicit drugs, which are contraindicated for breastfeeding.⁹ Babies with NAS have increased irritability, poorer sucking reflex, and greater likelihood of neonatal intensive care unit (NICU) admission, each of which complicates breastfeeding.⁸ Likewise, support structures that facilitate successful initiation of breastfeeding are often weak or absent in this population.¹⁰ Furthermore, if providers and other members of the treatment team lack awareness of the current recommendations regarding breastfeeding and methadone, women in treatment receive less support and anticipatory guidance during pre- and postnatal care.¹⁰

Methadone therapy involves frequent contacts with the healthcare system, and a unique opportunity for health promotion and counseling. This study aims to inform strategies for breastfeeding promotion in this context, by illuminating some of the barriers to successful breastfeeding in this vulnerable population. An adapted survey instrument was used to examine intention to breastfeed among postpartum women in methadone therapy, services and supports they report receiving, self-reported breastfeeding rates, and reasons for attrition.

Materials and Methods

This study utilized a cross-sectional qualitative and quantitative interview-based survey to capture the infant feeding practices, barriers, and beliefs of women in methadone maintenance therapy. Participants were recruited from an opioid dependence treatment center in an urban setting in the southeastern United States. All study procedures were approved by the Belmont University Institutional Review Board.

Setting

Behavioral Health Group (BHG) is an outpatient opioid dependence treatment program that provides counseling and treatment with methadone or buprenorphine. Patients receive annual physical examinations, diagnostics and referrals, and education on safe sex and HIV. Between January 2013 and May 2015, BHG served ∼1956 patients, 39 of whom were pregnant. All pregnant patients were being treated with methadone according to provider and patient preference. At the time of the study, all the patients were self-pay and BHG did not bill insurance for Medication-Assisted Treatment (MAT). Methadone is more affordable than Buprenorphine and was therefore preferred by patients (A. Darlington, personal communication, November 2015).

Participants

Women were eligible to participate if they were at least 18 years of age, delivered a baby while in treatment between January 2013 and May 2015, and lived with their babies while continuing treatment. A convenience sample of eligible women was recruited between October 2015 and November 2015.

Procedures

Study flyers were posted in the clinic's lobby and dosing booths and instructed women interested in participating to notify their dosing nurse, who referred them to the interviewer. The interviewer met with potential participants in a private room at BHG to explain the study and offer informed consent. Participants were given the option to complete the interview that day or make an appointment to meet with the interviewer at a more convenient time. Questions were read aloud to participants, and answers were input into a computer using SurveyMonkey, Inc.^® by the interviewer in real time. At the end of the 20- to 30-minute session, participants were given a 5-dollar gift card to a local supermarket as compensation for their time.

Survey instrument

A 47-item instrument was developed, including both original questions and questions adapted, with permission, from the 2006 Infant Feeding Survey.¹¹ The survey instrument was designed to capture decisions and intentions to formula feed or breastfeed (11 items); exposure to formula marketing (6 items); social support or influence from healthcare providers (12 items); and hospital experience of woman and infant (8 items). Adaptations and additional questions were anchored by Bandura's 2003 Triadic Reciprocal Causation model and included open and closed-ended questions designed to capture the experience of women breastfeeding while taking methadone. These items included perception of support received from treatment center healthcare personnel (3 items) and personal knowledge of the risks and benefits of breastfeeding while taking methadone (2 items).

Analysis plan

For responses to each of the three open-ended questions, the project leader developed category schemes by identifying underlying concepts and then reported frequencies of responses by category. Coding of data was supervised and reviewed by the project leader's DNP faculty; there was no second coder.

Breastfeeding duration was the dependent variable in quantitative analyses. If a woman exclusively formula fed, breastfeeding duration was recorded as 0. A preliminary variable plot showed that the correlation between breastfeeding duration in days and infant hospital stay in days was not linear and had outliers. However, there was a clear curvilinear relationship. The combination of a relatively small sample and the presence of outliers made any estimate that compared mean values (such as t tests) problematic. Thus, winsorizing was used to reduce the disproportionate impact of outliers.¹² Through this process, outliers are replaced with less extreme values.¹² Figure 1 shows the winsorized data wherein values >62 days were winsorized to 62 days (using a common standard of 10% of observations trimmed, in this case three observations). Correlation coefficients were then used to measure associations between breastfeeding duration in days and other variables of interest. Pearson correlations were calculated when the requirement of two continuous variables was met. However, when one variable was dichotomous and the other variable was continuous, point-biserial correlations were calculated. In an alternative measure, breastfeeding duration was dichotomized as “seven days or less or not at all” versus “more than seven days.” For this measure, chi-square tests examined the relationship between breastfeeding duration and other variables; Fisher's exact tests were used when cell sizes were <5. The p-value for all tests was 0.10, and all tests were two-tailed. Data were analyzed using IBM^® SPSS^® Statistics version 23.0 software.

FIG. 1.

Scatterplot of winsorized data demonstrating relationship between infant hospital stay in days and breastfeeding duration in days among women in methadone maintenance therapy. r = 0.124, p < 0.1.

Results

Breastfeeding duration

Eighty percent (n = 24) of women initiated breastfeeding within 2 days postpartum and reported breastfeeding duration from 1 day to 10 months. The average breastfeeding duration was 41.9 days. None breastfed exclusively until starting the babies on solid food: 37.5% (n = 9) of women stopped breastfeeding after 1 week, 37.5% (n = 9) of women breastfed for <1 month but >1 week, 16.7% (n = 4) breastfed for 2 to 3 months, and 8.3% (n = 2) for 6 months or longer. See Figure 2 for proportion of women breastfeeding over time. Responses to the open-ended question asking women to give reasons for why they stopped breastfeeding were organized into five major categories and are reported in Table 1. The most common theme for breastfeeding cessation was “infant hospital stay,” the category which included all references (n = 11) to the inconvenience and/or expense of traveling back and forth to the hospital (n = 6), lack of transportation (n = 3), and living far from the hospital (n = 2). These responses are unique to women who are not discharged with their babies and for whom the challenge of establishing breastfeeding is compounded by the logistics of traveling to and from the NICU from home for pumping, breast milk delivery, and feedings. In addition, four women reported that they stopped breastfeeding because of “fear or uncertainty about the effects of methadone on their newborn.”

FIG. 2.

Proportion of women breastfeeding from birth to 7 months.

Table 1.

Category Schemes for Discontinuation of Breastfeeding

Reason	% endorsed^*
Baby in NICU, % (n)	45.8 (11)
Inconvenience (not related to NICU), % (n)	37.5 (9)
Baby medical condition/issues, % (n)	25.0 (6)
Insufficient supply, % (n)	20.8 (5)
Fear/uncertainty about effects of exposure to methadone on baby, % (n)	16.7 (4)

Some women gave more than one reason for breastfeeding cessation.

NICU, neonatal intensive care unit.

Sociodemographics and smoking tobacco during pregnancy

A total of 30 women participated in the study, with an average age of 28.9 years. Breastfeeding duration was positively associated with education (p < 0.1) and nonsmoking status (p < 0.1). All other sociodemographic variables were not significantly correlated with breastfeeding duration and are displayed in Table 2.

Table 2.

Participant Sociodemographics and Smoking Tobacco During Pregnancy

Age (years)
Range	22–41
Mean	28.9
Marital status, % (n)
Single, never married	53.3 (16)
Married	33.3 (10)
Widowed	3.3 (1)
Divorced	3.3 (1)
Separated	6.7 (2)
Level of education, % (n)
Elementary or some high school	23.3 (7)
High school diploma or GED	36.7 (11)
Some college	26.7 (8)
Bachelor's degree	10.0 (3)
Master's degree or higher	3.3 (1)
Parity, % (n)
Primipara	26.7 (8)
Multipara	73.3 (22)
Breastfeeding duration
Range	1–300 days
Mean	41.9 days
Tobacco use during pregnancy, % (n)
Yes	83.3 (25)
No	16.7 (5)
Ethnicity, % (n)
Caucasian	96.7 (29)
African American	3.3 (1)
Living arrangement, % (n)
Partner/partner and other children	73.3 (22)
Alone with child/other children	13.3 (4)
Other	13.3 (4)
Annual income, % (n)
<$24,999	13.3 (4)
$25,000–$49,999	46.7 (14)
$50,000–$74,999	36.7 (11)
$75,000–$99,999	3.3 (1)

GED, general educational development.

Feeding decision-making

Sixty percent (n = 18) of women reported intending to breastfeed exclusively, 20.0% (n = 6) reported intending to feed their infants a combination of breast milk and formula, and 20.0% (n = 6) reported intending to formula feed exclusively. A chi-square test showed that women who intended to breastfeed exclusively, and made the decision before or during pregnancy, were more likely to breastfeed for a longer duration than women who intended to feed with formula or with a combination breast milk and formula (p < 0.05).

Among women who reported intent to exclusively breastfeed or feed with a combination of breast milk and formula (n = 24), reasons for planning to breastfeed were categorized into five themes as follows: (1) “Healthier for baby” (79.2%; n = 19), (2) “Helps infant withdrawal/NAS” (41.7%; n = 10), (3) “Financial reasons” (33.3%; n = 8), (4) “Colostrum” (20.8%; n = 5), and (5) “Bonding” (16.7%; n = 4). When asked how long they intended to breastfeed their babies, 50.0% (n = 12) reported 6 months or less, 20.8% (n = 5) reported 1 year, and 29.2% (n = 7) stated that they did not have a specific timeframe. Reasons women reported intention to exclusively formula feed (n = 6) were categorized into two themes. Half (n = 3) reported “fear of methadone being transferred to baby through breast milk” and the other half (n = 3) reported “convenience” as the main reason for exclusively formula feeding.

When asked who or what helped them with their infant feeding decision 50% (n = 15) reported help from a significant other, healthcare professional, or family members. A chi-square test suggests that women who reported receiving help making the infant feeding decision were more likely to plan to breastfeed exclusively than women who reported not receiving help, although it missed significance (p = 0.14).

Services and support

When asked about breastfeeding support women received from their treatment center, 23.3% (n = 7) reported that healthcare staff discussed breastfeeding with them. Of the women, 63.3% (n = 19) indicated that they were exclusively formula fed as infants, 26.7% (n = 8) reported that they were exclusively breastfed, and 6.7% (n = 2) women reported that they were fed a combination of formula and breast milk. One woman (n = 1) did not know how she was fed. A point-biserial correlation coefficient showed longer breastfeeding duration was associated with having been breastfed as an infant (p < 0.05).

When asked about assistance with feeding such as latching, positioning, or preparing formula, 83.3% (n = 25) reported receiving help while in the hospital. When asked about support in the hospital, 66.7% (n = 20) felt that the hospital staff encouraged breastfeeding, while 33.3% (n = 10) felt that hospital staff did not show preference. A chi-square analysis showed that women who breastfed for shorter amounts of time or not at all were more likely to report that the hospital encouraged breastfeeding (p < 0.05). Length of infant hospital stay was inversely correlated with breastfeeding duration (p < 0.1), and the winsorized data are demonstrated in Figure 1.

When asked about what they heard or knew about breastfeeding while taking methadone, 46.7% (n = 14) reported that it can lessen the impact of NAS, only 16.7% (n = 5) stated it was safe, 16.7% (n = 5) stated it was unsafe, and 20.0% (n = 6) were unsure or had not heard anything.

Infant feeding experience

Five women (20.8%) reported some level of satisfaction with their breastfeeding experience. A point-biserial correlation coefficient showed significant correlation between satisfaction and breastfeeding duration (p < 0.001). When asked how they would feed their infant if they had another child, 33.0% (n = 10) reported intent to exclusively breastfeed, 30.0% (n = 9) reported intent to exclusively formula feed, 30.0% (n = 9) reported intent to feed a combination of breast milk and formula, and 7.0% (n = 2) reported they were not planning on having another child. A point-biserial correlation indicated that longer breastfeeding duration was associated with intent to feed their future child breast milk exclusively or with a combination of breast milk and formula (p < 0.05). Additional analyses were performed on predictors of breastfeeding duration, although none were significant (Table 3).

Table 3.

Correlations Between Demographic Variables and Breastfeeding Duration

Variable	Pearson's correlation	Point biserial	P	Codes
Sociodemographics and smoking tobacco during pregnancy
Age	−0.082		0.67
Marital status		0.24	0.19	0 = Single
Marital status		0.24	0.19	1 = Married
Living arrangement		0.24	0.23	0 = Partner/partner and other children
Living arrangement		0.24	0.23	1 = Alone with child/other children
Level of education		0.35	0.06	0 = Elementary or some high school
				1 = High school diploma
				2 = Some college
				3 = Bachelor's degree
				4 = Master's degree or higher
Parity status		−0.20	0.27	0 = Primiparous
Parity status		−0.20	0.27	1 = Multiparous
Annual income		0.15	0.42	0 = <$50,000
Annual income		0.15	0.42	1 = ≥$50,000
Tobacco use during pregnancy		0.33	0.07	0 = Yes
Tobacco use during pregnancy		0.33	0.07	1 = No
Decision/intent to feed
When did you first decide how you would feed your feed your baby?		0.19	0.92	0 = Before pregnancy
		0.19	0.92	1 = During pregnancy
When you were pregnant, what did you plan to feed your baby when your baby was born?				Breast milk exclusively
				Formula/combo of formula and breast milk
When you were pregnant, how long did you plan to breastfeed? (n = 24)		0.17	0.42	0 = Had no timeframe
		0.17	0.42	1 = Had a timeframe
Services and support
When you were pregnant, did you get any information about feeding your baby?		0.06	0.75	0 = yes
		0.06	0.75	1 = no
Did your doctor or nurse talk to you about breastfeeding? (prenatal care not at opioid dependence treatment center)		0.27	0.15	0 = Yes
		0.27	0.15	1 = No
Did your doctor or nurse encourage you to breastfeed while taking methadone? (prenatal care not at opioid dependence treatment center)		0.01	0.95	0 = Yes
		0.01	0.95	1 = No
Did the healthcare staff at your methadone maintenance therapy clinic discuss breastfeeding with you before you had your baby?		−0.29	0.12	0 = Yes
		−0.29	0.12	1 = No
How did your mother feed you as a newborn?		−0.41	0.03	0 = Breast milk
How did your mother feed you as a newborn?		−0.41	0.03	1 = Formula exclusively
Did you have a close friend or relative who breastfed her baby?		−0.04	0.85	0 = Yes
		−0.04	0.85	1 = No
Days baby stayed in hospital after mother was discharged	−0.352		0.06
While in the hospital, did you receive any help feeding your baby?		−0.13	0.49	0 = Yes
		−0.13	0.49	1 = No
Did anyone ever suggest that you stop breastfeeding?		0.023	0.91	0 = Yes
Did anyone ever suggest that you stop breastfeeding?		0.023	0.91	1 = No
Infant feeding experience
How satisfied were you with your breastfeeding experience?		−0.83	<0.01	0 = Satisfied
How satisfied were you with your breastfeeding experience?		−0.83	<0.01	1 = Not satisfied
If you had another baby, how would you feed your baby?		−0.43	0.02	0 = Breast milk
If you had another baby, how would you feed your baby?		−0.43	0.02	1 = Formula exclusive

Discussion

Breastfeeding among women taking methadone is encouraged by ACOG and provides numerous benefits to both child and woman alike; still, breastfeeding rates remain low in this population. Twenty-four women (80.0%) in the sample initiated breastfeeding, but only 6 (20.0% of the sample) continued for >1 month. The high rate of attrition and the descriptive data presented confirm that women in methadone therapy face a set of unique barriers to breastfeeding, including lack of support and education from healthcare providers, challenges related to infants' hospitalization, and an unclear understanding of the advantages, safety, and basic mechanics of breastfeeding while taking methadone.

Role of healthcare providers

Eleven women (36.7%) reported that their prenatal care providers did not discuss breastfeeding with them. Only seven women (23.3%) reported that health staff at their treatment center discussed breastfeeding with them during their pregnancy. While discouragement of breastfeeding was not explicitly captured by the survey, it can be extrapolated that if 11 women reported receiving no information from prenatal care providers and 14 reported they were encouraged to breastfeed, the remaining five women may have been actively discouraged or advised against breastfeeding. Providers' lack of awareness of current guidelines may influence their active promotion of breastfeeding in this population.¹⁰ The data suggest that women who were encouraged to breastfeed by their provider or healthcare staff were more likely to breastfeed for longer durations, although the statistical test was not significant (p = 0.11). Larger studies have shown that prenatal and postnatal support programs and promotion among the general population have increased breastfeeding rates and duration.¹³ These are substantial and modifiable clinical procedures that can increase breastfeeding initiation and duration in this population.

Infant hospitalization

Logistical challenges related to infant hospital stay posed a significant barrier to breastfeeding for women taking methadone. In the study sample, 73.0% (n = 22) of infants remained hospitalized after the mother was discharged. Out of the 24 women who attempted to breastfeed, 11 reported that they discontinued breastfeeding because of issues related to infant's NICU stay. There was a significant negative correlation between days a woman breastfed her infant and the number of days the infant stayed in the hospital after birth. Other studies show babies who are born preterm and admitted to the NICU have reduced breastfeeding rates.^14,15

Along with the logistical challenge of mother–infant separation while in the NICU, babies with NAS often experience poor sucking reflex or weak latch.³ Early intervention with lactation support is warranted in these cases to provide monitoring and hands-on assistance with technique. Women who formula fed or who breastfed for 1 week or less were more likely to report that the hospital encouraged breastfeeding. Although this seems counterintuitive, it is possible that staff noted difficulty with breastfeeding or formula feeding in the hospital and intervened. Targeting health promotion during pregnancy about the importance of breastfeeding and offering breastfeeding support are essential to the successful establishment of healthy and safe infant feeding practices for all women, but may have particular impact with this population. In efforts to prepare mothers for the possibility of a long NICU stay, sensitive and individualized counseling, education on NAS and potential breastfeeding experience and benefits, breastfeeding training and techniques, and breastfeeding supplies will assist these women.¹⁶

Many women stated that they faced challenges with transportation to and from the NICU, creating an additional logistical barrier to breastfeeding. This finding was consistent with the literature among women in the general population with preterm babies in the NICU.¹⁶ Living situation may have contributed to this barrier; 26.7% (n = 8) reported living without a partner and 66.7% (n = 20) reported living with other children at home. The authors recommend further research or development of a program for transportation of breast milk or mother to the NICU. Studies in preterm infants admitted to the NICU have shown that separation of infant and mother causes a loss in connection resulting in difficulties in breastfeeding.¹⁶ All measures should be taken to avoid mother and infant separation while in the NICU.¹⁶

Knowledge of breastfeeding

Support from family members, friends, and significant others were reported as influencing factors in infant feeding decisions. Women who were breastfed as infants were more likely to breastfeed their babies, supporting existing evidence that family members and environment can strongly influence infant feeding practices. Although not significant, the results suggested that women who received some form of external help with their feeding decision were more likely to plan to feed their infants breast milk, compared with those who reported receiving no form of help with the decision. This finding accentuates the opportunity healthcare providers have on influencing women to make safe infant feeding decisions. However, instruction on safe infant feeding practices should extend to family members and significant others.

Maternal level of education was positively correlated with breastfeeding duration, a finding that is consistent with literature looking at the general population.^17,18 More than half of the women had a high school education or less, which may have significantly influenced breastfeeding rates in this sample, before being compounded by illicit substance abuse history. Less than half of the women (46.7%; n = 14) reported awareness of the potential that breastfeeding can help the infant with NAS, representing a lack of knowledge on the importance of breastfeeding as a part of illicit substance abuse treatment. Two-thirds of the sample (36.6%; n = 11) reported their misunderstanding or lack of clear understanding about the safety of methadone exposure through breast milk and almost a quarter of the sample discontinued breastfeeding (n = 4) or never attempted breastfeeding (n = 3) because they were uncertain or fearful about the safety of methadone exposure to their newborns. These data illuminate the opportunity to target maternal health education on the safety of breastfeeding while taking methadone and benefits of breastfeeding overall. Doing so may increase rates of breastfeeding attempts and continuation.

Limitations

This study looked at infant feeding beliefs and practices in a small sample of postpartum women in methadone maintenance therapy at a single treatment center. The generalizability of results may be limited. The majority of participants were Caucasian, and it is known that breastfeeding rates vary across cultures, independent of illicit substance use.¹⁹ With voluntary participation, women who felt self-conscious or ambivalent about breastfeeding may be underrepresented. In addition, participants were still enrolled in therapy and may not represent views of women who discontinued treatment due to relapse or successful completion. The decision to collect retrospective data from women in treatment increases the risk of recall bias; a longitudinal design may yield more sensitive data but a lower response rate with high risk of attrition. The authors acknowledge the value of ecological data related to the hospital context, in which breastfeeding is being attempted, such as the presence or absence of clear hospital breastfeeding and lactation support policies, and parental presence at the bedside for babies with NAS in the NICU. Although not measured in this study, the authors acknowledge evidence describing additional barriers or contraindications to breastfeeding, including HIV status, gestational age, use of psychiatric medications with variable safety data, and maternal custody of the infant on both breastfeeding rates and NAS outcomes.^9,20

Conclusions

This is one of the first studies to assess infant feeding practices, barriers, and beliefs among women in methadone maintenance therapy. Women in treatment both desire and attempt to establish breastfeeding, but encounter challenges, including long NICU stays and lack of support and education, that compromise their success. These findings should inform future programs or interventions geared toward increasing breastfeeding initiation and duration among pregnant women in methadone therapy. With daily dosing visits to their treatment center, the patient–provider encounters during the prenatal and postpartum periods are prime opportunities for healthcare providers to assess, educate, and encourage this group of vulnerable women to safely breastfeed.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Substance Abuse and Mental Health Services Administration. Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings. NSDUH Series H-48, HHS Publication No. (SMA) 14-4863. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2014.

Patrick

, Davis

, Lehman

, et al. Increasing incidence and geographic distribution of neonatal abstinence syndrome: United States 2009 to 2012. J Perinatol, 2015; 35:650–655.

Hudak

, Tan

. Neonatal drug withdrawal. Pediatrics, 2012; 129:e540–e560.

Patrick

, Schumacher

, Benneyworth

, et al. Neonatal abstinence syndrome and associated health care expenditures: United States, 2000–2009. J Am Med Assoc, 2012; 307:1934–1940.

Reece-Stremtan

, Marinelli

. ABM clinical protocol #21: Guidelines for breastfeeding and substance use or substance use disorder, revised 2015. Breastfeed Med, 2015; 10:135–141.

Mascola

, Borders

, Terplan

. Committee opinion no. 711: Opioid use and opioid use disorder in pregnancy. Obstet Gynecol, 2017; 130:81–94.

Wachman

, Byun

, Philipp

. Breastfeeding rates among mothers of infants with neonatal abstinence syndrome. Breastfeed Med, 2010; 5:159–164.

Demirci

, Bogen

, Klionsky

. Breastfeeding and methadone therapy: The maternal experience. Subst Abus, 2015; 36:203–208.

Jansson

, Velez

, Butz

. The effect of sexual abuse and prenatal substance use on successful breastfeeding. J Obstetr Gynecol Neonatal Nurs, 2017; 46:480–484.

10.

Jansson

, Velez

, Harrow

. Methadone maintenance and lactation: A review of the literature and current management guidelines. J Hum Lact, 2004; 20:62–71.

11.

Cunningham

, Jackson

, Oickle

. 2006. Infant Feeding Survey: Factor Influencing Breastfeeding Initiation, Duration, and the Introduction of Solids. Leeds Grenville and Lanark District Health Unit, 2009.

12.

Tuğran

, Kocak

, Mirtagioğlu

, et al. A simulation based comparison of correlation coefficients with regard to type I error rate and power. J Data Anal Inform Proces, 2015; 3:87–101.

13.

Schanler

, Potak

. Breastfeeding: Parental education and support. 2015. www.uptodate.com (accessed August 6, 2017).

14.

Mulready-Ward

, Sackoff

. Outcomes and factors associated with breastfeeding for <8 weeks among preterm infants: Findings from 6 states and NYC, 2004–2007. Matern Child Health J, 2013; 17:1648–1657.

15.

Scott

, Binns

, Graham

, et al. Temporal changes in the determinants of breastfeeding initiation. Birth, 2006; 33:37–45.

16.

Ikonen

, Paavilainen

, Kaunonen

. Preterm infants' mothers' experiences with milk expression and breastfeeding: An integrative review. Adv Neonatal Care, 2015; 15:394–406.

17.

Celi

, Rich-Edwards

, Richardson

, et al. Immigration, race/ethnicity, and social and economic factors as predictors of breastfeeding initiation. Arch Pediatr Adolesc Med, 2005; 159:255–260.

18.

Heck

, Braveman

, Cubbin

, et al. Socioeconomic status and breastfeeding initiation among California mothers. Public Health Rep, 2006; 121:51–59.

19.

Perrin

, Rothman

, Sanders

, et al. Racial and ethnic differences associated with feeding- and activity-related behaviors in infants. Pediatrics, 2014; 133:e857–e867.

20.

Howard

, Schiff

, Penwill

, et al. Impact of parental presence at infants' bedside on neonatal abstinence syndrome. Hosp Pediatr, 2017; 7:63–69.