An Integrative Review of Exclusive Breast Milk Expression

Abstract

Introduction

A growing number of families are using exclusive breastmilk expression to feed their newborns. They need accurate information and support from their providers.

Methods

The purpose of this integrative review was to explore the prevalence, characteristics, practices, and outcomes of families who fed exclusively expressed breastmilk to their infants. The electronic databases of CINAHL, Scopus, PubMed, Web of Science, and article reference lists for articles on expressing human milk for infants were searched. Results were filtered to glean published studies between 2012 and 2022, academic journals, human studies, and English language journals while excluding secondary reviews and expert opinions. Both authors screened records within Covidence systematic review management software. Selected studies were evaluated for quality of evidence using the Johns Hopkins Research Evidence Appraisal Tool.

Results

Twenty‐seven studies were included in this review. Most families who exclusively expressed originally planned to directly breastfeed their newborns and only began pumping after encountering problems. The most frequently recurring theme was that advice from health care personnel was inconsistent. Women described a knowledge deficit and received more help with exclusive expression from informal social media groups than their providers. Prevalence of exclusive breastmilk expression is highest in Asian countries and in the neonatal intensive care unit environment. These women had a shorter duration of human milk feeding and higher likelihood of cessation compared with women who both fed expressed milk and directly breastfed.

Discussion

Only recently have breastfeeding researchers begun distinguishing the type of milk (human milk or artificial formula) from the feeding method in their publications. This approach allows clinicians to provide better guidance on the outcomes of each feeding method. As clinicians, we can help families by encouraging them and providing consistently accurate information. Solely feeding expressed human milk is a valid option when the lactating person is unable or unwilling to directly breastfeed.

INTRODUCTION

The benefits of breastfeeding are well established and have spurred many initiatives to increase breastfeeding exclusivity and duration. Yet, when families have difficulty breastfeeding, they may want to explore other options to continue to provide human milk to their newborns and infants. One option that may retain some breastfeeding benefits is feeding by exclusive human milk expression, commonly called exclusive expressing (or expression) (EE) or exclusive pumping (EP). ¹ , ² With EE, the newborn is not directly breastfed and is fed all human milk by bottle, nasogastric, orogastric, gastrostomy tube, or other alternative means. ³ These options allow for continued human milk feeding in situations in which the parent encounters barriers that prevent direct breastfeeding or chooses not to directly breastfeed. ⁴ Parents may be unaware of EE, and health care staff and providers should be able to support all infant feeding options.

QUICK POINTS

• ✦Individuals who are exclusively expressing breastmilk lack formal support and primarily use social media to guide their pumping practices.
• ✦Prevalence of exclusive expression varies widely by setting, region, and infant's age; feeding patterns can change frequently, and hence longitudinal studies are needed.
• ✦Exclusive expression is increasing among families encountering premature births and breastfeeding difficulties; although time‐consuming, it becomes easier over time.
• ✦Clinicians must be equipped with current, accurate information on exclusive expression to successfully guide individuals through this option for providing breastmilk to their infants.

Breastmilk Is the Biological Normative Substance for Human Infant Feeding

Both the American Academy of Pediatrics and the World Health Organization recommend exclusive breastfeeding for 6 months and continued breastfeeding with complementary foods for 2 years, and thereafter as long as desired by the couplet. ⁵ , ⁶ , ⁷ Children who were breastfed had lower rates of autism spectrum disorders, ⁸ fewer childhood infections such as diarrhea and respiratory illnesses, ⁹ fewer dental malocclusions, ⁹ less type 2 diabetes, ¹⁰ and improved cognitive development. ¹¹ Likewise, women who breastfed had lower rates of breast ⁹ , ¹² and ovarian cancer. ⁹ , ¹³ , ¹⁴

In 2019, more than 83% of US newborns received any human milk, but the prevalence rapidly decreased by 3 months of age to 45.3% exclusively breastfeeding and 24.9% receiving any breastmilk at 6 months. ¹⁵ With 3.6 million newborns born annually in the United States, this equates to 2 million infants who were breastfed initially but no longer exclusively breastfed at 6 months. ¹⁶ Globally, optimal breastfeeding could prevent an estimated 98,243 maternal deaths and 974,956 cases of childhood obesity annually. ¹⁷

The high human milk initiation rate indicates that most families desire to feed infants human milk but often experience barriers that interfere with effective breastfeeding. The most common reasons for early cessation of breastfeeding in term infants were perceived insufficient milk supply and breast or nipple pain. ¹⁸ EE allows the continuation of human milk feeding. It may be used as a short‐term solution until direct breastfeeding can be restarted, or it may be used as a long‐term feeding option. Rates of EE are difficult to determine because many studies do not separate the substance fed from the mode of delivery.

Individuals who chose to exclusively express breastmilk often did so because they valued the benefits of human milk feeding but likely encountered obstacles to direct breastfeeding, they elected EE from birth, or their infants could not directly breastfeed because of their current medical conditions. ⁴ , ¹⁹ , ²⁰ , ²¹ Currently, abundant information and support are available on breastfeeding, but community support and specific education on EE are lacking. ²² Furthermore, uninformed health care personnel provided conflicting or incomplete information and guidance on EE. ²³ Therefore, the purpose of this integrative review was to explore the prevalence, characteristics, practices, and outcomes of families who fed exclusively expressed human milk to their infants. These 4 categories were selected a priori to answer the research question, “What is the key information clinicians need to know to best care for exclusively expressing families?”

The authors acknowledge that not all lactating people identify as women, and some may choose to use the term chestfeeding. This article will use breastfeeding and gender‐neutral terms unless referring to previous research participant identifiers.

METHODS

Search Strategy

To strengthen rigor, we chose the Whittemore and Knafl ²⁴ integrative review method to guide the literature search, data evaluation, and article analysis. An integrative review is appropriate for this emerging topic due to the need to understand the subjective experience of using this feeding method, in addition to the overall prevalence and outcomes. In consultation with a medical librarian, the first author searched for peer‐reviewed primary research articles in PubMed, Scopus, Web of Science, and CINAHL. All articles about exclusively expressing human milk for infants were considered for inclusion, focusing on the prevalence, characteristics, practices, and outcomes. Search terms used included breast milk collection, OR breast pumping, OR breast milk expression, AND exclusive. Filters included English language, human studies, academic journals, and published within the previous 10 years. Article reference lists were also hand searched to identify additional publications. See Supporting Information: Table S1 for the full search strategy.

Secondary reviews, expert opinions, and articles in non‐English language were excluded. Articles focusing on hospital practice change or lacking explicit distinction between when the sample transitioned from EE to other feeding methods, those analyzing human milk components or fecal testing, or articles that described outcomes or interventions that did not address the research question were also excluded.

Study Selection

Record screenings were conducted by one master's‐prepared registered nurse who is also an international board‐certified lactation consultant (IBCLC) and one doctoral‐prepared certified nurse‐midwife. Records were retrieved, and both authors independently completed initial screenings based on the titles and abstracts within Covidence systematic review management software. ²⁵ Title and abstract screening yielded 100% agreement between authors. Both authors also conducted full‐text screening independently, and discrepancies were resolved by discussion.

The authors agreed before screening retrieved articles that for EE to be considered present in a lactating person and infant dyad, the following would be documented: (1) no direct breastfeeding, (2) consistent expression of breastmilk, and (3) an “enduring infant feeding pattern” of feeding all expressed human milk by bottle or other means such as gastrostomy tube.^{3(p 946)} For article inclusion, a minimum time of EE was not defined as this was not stated in the articles. If authors described a sample as EE and the description of the sample supported the definition above, they were considered EE for purposes of this integrative review. If there was not a clear sample of families that were EE for a period of time, the article was excluded. If the articles were infant‐focused, the articles were eligible regardless of whose human milk the infant consumed.

The database searches yielded 395 potential records and 23 additional records from reference list searches. After 137 duplicates were removed, the remaining 281 studies underwent the title and abstract screening. This screening excluded 201 studies based on the eligibility criteria. Full‐text versions were reviewed for the remaining 80 studies, and 53 were excluded for not meeting the eligibility criteria. Five studies were excluded for not being published within 10 years. Full‐text review yielded a final sample of 27 studies. The Preferred Reporting Items for Systematic Reviews and Meta‐Analyses Statement was followed in reporting this review (See Figure 1). ²⁶

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses Diagram

Data Evaluation

To evaluate quality, the authors used the Johns Hopkins Evidence‐Based Practice Model to rate the level of evidence. Then, they assessed the quality further using the Johns Hopkins Research Evidence Appraisal Tool. ²⁷ The authors started by identifying the type of evidence and then answered questions specific to the study methods, resulting in a final quality appraisal of A (high), B (good), or C (low) for the quantitative studies and A/B (high/good) or C (low) for the qualitative studies. ²⁷ The first author assigned each article a quality rating, and the second author reviewed and confirmed the results, which are displayed in Table 1.

Table 1.

Description of Included Studies

Authors, Year Design	Aim of Study	Participants	n	Quality a
Anders et al, 2022 4 Qualitative	To explore the experiences of women who fed their infants by EE to determine their feeding experience, their feeding goals, their feeding intentions, and the factors that influence their feeding experience.	Women who had given birth within the last year and who had exclusively expressed for at least 2 wk.	21	A/B
Bai et al, 2017 28 Quantitative	To (1) determine the prevalence of EE in 2006‐2007 and again in 2011‐2012; (2) identify associated factors with EE; and (3) explore the association between EE and duration of breastmilk feeding.	Women with healthy full‐term infants intending to breastfeed followed up for 12 mo.	2450	A
Bower et al, 2017 21 Qualitative	To describe women's experiences expressing milk for their very low birth weight NICU newborns.	Women with NICU infants >1500 g.	17	A/B
Clapton‐Caputo et al, 2021 40 Qualitative	To understand the expectations and experiences of women who used EE and also used social media for support.	Australian women who chose EE and accessed a social media support group for EE support within 5 y.	10	A/B
Demirci et al, 2018 41 Qualitative	To describe breastfeeding of infants with complex congenital surgical anomalies.	Women with infants <6 wk old who needed surgery for cardiac, neural tube, or gastrointestinal, defect.	15	C
Fan et al, 2020 32 Prospective cohort	To determine EE prevalence, explore the reasons of acquiring a breast pump, and identify predictive factors for EE at specific points within the first 6 mo of life.	Singleton term couplets intending to provide breastmilk or breastfeed.	761	A
Foong et al, 2021 29 Prospective cohort	To describe the breastfeeding experience and breastfeeding outcomes of women who stayed in confinement centers for 30 days compared with women who recovered from childbirth at home.	Women with partners who were intending to breastfeed their healthy term infants.	187	B
Froh et al, 2017 20 Qualitative	To describe the breastfeeding experience of women with infants who had CDH being cared for in the NICU.	Women who were feeding by EE for their infants with CDH in the NICU.	11	C
Jardine, 2019 42 Mixed methods	To determine when survey respondents first heard of EE, describe their information sources and the usefulness of these sources, and identify differences between those who learned about EE antepartum verses postpartum.	Respondents who currently or formerly fed their infants by EE.	1215	C
Jiang et al, 2015 33 Prospective cohort	To determine the impact of breastmilk expression in the early postpartum period on breastfeeding duration.	Women with healthy, full‐term infants following a singleton pregnancy and birth with >2 mo maternity leave.	340	A
Kaye et al, 2019 38 Retrospective cohort	To explore the experiences of mothers of infants with orofacial clefting and investigate factors associated with providing breastmilk to their infants.	Women with 1‐ to 2‐y‐old children who were previously diagnosed with orofacial clefting and referred for cleft team care.	50	A
Keim et al, 2017 30 Retrospective cohort	To compare women who fed by EE for their infants with those who directly breastfed with or without pumping.	Women with infants born at >24 weeks’ gestation and not intending to formula feed.	478	A
Larkin et al, 2013 46 Quantitative	To determine whether women who elected EE could achieve full milk production (700 mL/day) using only the Ameda Platinum pump within the first 2 wk after birth.	Women with NICU infants and planning to lactate.	26	C
Li et al, 2012 51 Prospective cohort	To investigate infant weight gain by both milk type (breastmilk vs formula) and mode of feeding (breast vs bottle).	Women with infants born after 35 weeks’ gestation who weighed at least 2.25 kg.	1899	A
Liu et al, 2018 47 Randomized control trial	To evaluate nursing education and support intervention on lactation among mothers separated from their infants.	Women who gave birth by cesarean.	260	B
Martino et al, 2015 37 Prospective cohort	To examine breastmilk feeding exclusivity and duration among infants with complex anomalies requiring surgery and NICU care.	Women with infants who had prenatally diagnosed anomalies requiring surgery and who received prenatal lactation consultations.	165	C
McGuire, 2021 43 Case report	To share a woman's experience of EE for her infants.		1	A
O'Sullivan et al, 2017 39 Qualitative	To identify behaviors for breastmilk production and breastmilk feeding.	Women who had ever pumped or expressed breastmilk and had a child between 1 and 3 y of age.	41	A/B
O'Sullivan et al, 2019 31 Quantitative	To compare different sets of infant feeding questions to determine how the prevalence of different feeding strategies differ based on the questions asked.	US women with children 19‐35 mo old.	456	A
Palmquist et al, 2020 44 Qualitative	“To understand the experiences of mothers who decided to breastfeed or express milk when their premature very low birth weight infant was admitted to a NICU” and how hypermedicalized NICU management affects postpartum health and well‐being.44(p2)	US women who had given birth within the past 3 y to a to a very low birthweight NICU newborn with a serious gastrointestinal disease.	17	C
Pang et al, 2017 34 Prospective cohort	To explore the predictors of type of infant feeding and its association with breastmilk feeding duration.	Women with singleton, term births whose infants still received breastmilk at 3 mo.	500	B
Pang et al, 2020 1 Prospective cohort	To explore the type of milk fed and the mode of infant feeding with child cognition between 6 and 54 mo of age.	Women with healthy, term, singleton infants from uncomplicated pregnancies.	491	B
Pinchevski‐Kadir et al, 2017 36 Retrospective cohort	To examine the effect of mode of breastmilk feeding, exclusivity, and maternal perceptions on breastmilk feeding duration among preterm infants.	Women with infants born <32 weeks’ gestation without major congenital malformations or genetic anomalies.	162	A
Ru et al, 2020 49 Prospective cohort	To investigate lactation initiation and milk production with hospital‐grade breast pumps in mothers separated from their preterm infants. The second aim was to “study the effect of breast pumping frequency on the volume and nutrient content of expressed milk.49(p2)”	Women with premature infants in the NICU willing to breastfeed.	30	B
Scholten et al, 2022 48 Retrospective cohort	To investigate lactation initiation and support in mothers of very low birth weight infants. The authors’ second objective was to relate these data to infants’ nutrition during hospitalization.	Women with infants between 6 and 18 months of age whose birthweight was <1500 g.	437 women	B
Schwarz et al, 2021 52 Case report	To share the experience of a healthy 28‐year‐old primigravid woman, who used breastmilk for her critically ill infant and developed complications.		1	A
Whipps et al, 2022 45 Qualitative	To analyze different breastfeeding pathways and differing descriptions of breastfeeding success.	Women who gave birth in the last 3 y with a variety of backgrounds.	38	A/B

Abbreviations: CDH, congenital diaphragmatic hernia; EE, exclusive expressing; NICU, neonatal intensive care unit.

^aQuality (A‐C) from the Johns Hopkins Evidence‐Based Practice Model rating scale.

Data Extraction and Synthesis

Within Covidence systematic review management software, the first author completed data extraction, and the second author reviewed and confirmed the accuracy of the data. There were no discrepancies. Key information extracted included country, aims, study design, dates of study, funding sources, conflicts of interests, setting, inclusion and exclusion criteria, recruitment methods, sample size, number that were exclusively expressing, prevalence, time of exclusivity and duration of EE, practices, themes, characteristics, outcomes, and limitations. Data were exported to Microsoft Excel for further analysis. To synthesize results, the authors grouped extracted data into the 4 categories of prevalence, characteristics, practices, and outcomes.

RESULTS

The final sample of 27 articles consisted of 16 quantitative studies, 1 mixed‐methods study, and 10 qualitative studies. The country most represented was the United States in 13 of the 27 studies, and 23 articles were published in 2017 or later.

The analysis of the level of evidence showed that 1 study was level I (true experimental), 24 studies were level III (nonexperimental), and 2 studies were level V (case studies). The authors carefully read each study to answer the questions on the appraisal tool. Of the 27 studies, 21 were high or good quality, with 6 being graded as low quality. Various weaknesses were identified among the low‐quality studies, such as unclear research questions, insufficient information on the data analysis process, and lack of justification for sample size.

Prevalence

Eight studies focused on samples of healthy infants outside the neonatal intensive care unit (NICU) and included discussion of EE prevalence. Six of these 8 were conducted in Asian countries and reported prevalence rates between 6.8% ²⁸ and 62%. ²⁹ The 2 remaining studies were conducted in the United States, where prevalence among breastfeeding women was lower at 6.9%, ³⁰ 7.5% for infants 3 months old,³¹ and 7.7% for infants at 6 months old. ³¹ Compared with US studies, the studies conducted in China showed a higher prevalence of EE among healthy breastfeeding infants, and most showed prevalence rates over 14% at various time points between 1 and 6 months of age. ²⁸ , ³² , ³³ Similarly, the studies in Singapore had EE rates of 16.5% among the breastfeeding sample ³⁴ and 19.8% among bottle‐fed infants. ¹ The highest EE rates were found in Malaysia, with EE rates between 39% and 62%. ²⁹ The women who gave birth most recently in 2017‐2018 had higher EE rates at 14.6% to 20.2% ³² and 39 to 62%. ²⁹

Breastmilk is even more important for infants who are critically ill, premature, or require surgery. ³⁵ Many are unable to directly breastfeed, and the EE rates were highest among this group, with a reported prevalence of 43.8%, ³⁶ 59.5%, ³⁷ and 79%. ³⁸ Infant feeding strategies may change multiple times throughout an infant's life, making it more challenging to identify trends and outcomes based on any particular infant feeding strategy. ³⁹

Characteristics

Participant characteristics, prior knowledge and sources of knowledge of EE, and how participants perceive the time and work of pumping will be discussed in the following section.

Participant Characteristics

Seven articles included information on the participant characteristics of women who exclusively expressed breastmilk. ⁴ , ²⁸ , ²⁹ , ³⁰ , ³³ , ³⁴ , ³⁶ When asked about their feeding intentions during pregnancy, most women who chose EE originally planned on direct breastfeeding or combination feeding. ⁴ , ³⁶ Women usually exclusively expressed because they encountered barriers to direct breastfeeding such as latch or sucking issues, ²⁹ , ³³ flat or inverted nipples, ³³ and a need to monitor the infant's intake. ²⁹ Women who exclusively expressed were also more likely to have given birth preterm (52% vs 8%, P < .0001), had a NICU newborn (58% vs 11%, P < .0001), had a longer newborn length of stay, and had a higher proportion of twin births (45.1% vs 29.3%, P = .049). ³⁰ , ³⁶ They were also more likely to have experienced a cesarean birth. ²⁸ , ³⁰

Maternal characteristics associated with EE include no prior breastmilk feeding experience ²⁸ and primiparity (adjusted odds ratio [aOR] 1.54; 95% CI, 1.04‐2.26). ³⁴ In a US study, the women who used EE were more likely to have less education, ³⁰ which differs from China, where high education was associated with EE (odds ratio [OR] 4.92; 95% CI, 1.66‐14.62; P < .01), ³³ and Singapore, where tertiary education was associated with EE (aOR 2.22; 95% CI, 1.22‐4.04). ³⁴ Maternal employment may also influence infant feeding decisions. Women who had exclusively expressed had higher rates of employment during pregnancy (aOR 2.53; 95% CI, 1.60‐4.02) ³⁴ and short maternity leave, ³³ and were more likely to return to work in the postpartum period. ²⁸

Women's Knowledge of EE

Six articles reported information on knowledge of EE. ²⁰ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ Studies of women who chose EE identified missed opportunities for nurses and providers to support them ⁴ , ⁴² yet studies focusing on women who were exclusively expressing breastmilk for their infants in the NICU described more positive experiences with lactation support. ²⁰ , ⁴⁴ Although 63% of individuals in one study attended prenatal breastfeeding classes, they did not find the information helpful for EE. ⁴² Most women did not learn of EE until after birth and needed additional information to pump adequately to establish and maintain a full milk supply. ⁴¹ , ⁴² Online support groups, websites, and social media groups were the primary sources of education and emotional support for families who chose EE. ⁴⁰ , ⁴² , ⁴³

Time and Work of Pumping

Eight articles provided information on the time and work of pumping. ⁴ , ²⁰ , ²¹ , ²⁹ , ³⁸ , ³⁹ , ⁴¹ , ⁴⁵ EE is a time‐intensive feeding method and was described as burdensome, ⁴¹ time‐consuming, ⁴ , ³⁸ similar to a full‐time job, ²¹ and impossible with working full‐time. ³⁹ With time and experience, lactating women said EE became more feasible, and they were able to develop a routine that worked. ⁴ , ²⁰ In contrast, other researchers reported that women chose EE because they believed direct breastfeeding was time‐consuming and interfered with sleeping. ²⁹

Practices

Within this review, 8 articles contained information on EE practices. ⁴ , ³⁷ , ⁴⁰ , ⁴² , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ Women are in the secretory initiation phase (lactogenesis I) at birth through the first 2 to 4 days. ⁵⁰ Then they transition to secretory activation (lactogenesis II) with the onset of copious milk production. ⁵⁰ Guidance on pumping initiation differed by study. For example, some women were told to pump within 2 hours of vaginal birth and 4 hours after cesarean birth, ³⁷ whereas others were advised to pump within 6 hours of birth. ⁴⁹ Directions on frequency were more consistent; women were told to pump every 2 to 3 hours and at least 8 times daily. ³⁷ , ⁴⁶ , ⁴⁹ Only one study gave specific instructions for nighttime pumping, allowing one break between pumping of no more than 5 hours at night. ⁴⁹ Larkin et al had women pump for at least 15 minutes, hand express after pumping during secretory initiation, and replace hand expression with hands‐on pumping after the onset of secretory activation. ⁴⁶

Women who pumped immediately after birth and 7 to 9 times daily entered secretory activation earlier, had higher milk volumes, and were most likely able to have an adequate supply to feed their infant an exclusive breastmilk diet. ⁴⁷ , ⁴⁸ In an observational prospective cohort study, women with NICU newborns who were exclusively expressing pumped between 6 to 8 times daily with an average of 6.7 pumps per day. ⁴⁹ Researchers identified that most women who used a hospital‐grade pump at least 6 times daily for EE could achieve a full milk supply. ⁴⁹

Infant feeding has a steep learning curve. In a study examining women's reflections on their EE journey, 71% of women said they did not know about EE until after birth. ⁴² In this retrospective study, women were asked to rate their initial EE knowledge on a 0 to 100 scale, with an average knowledge score of 32.98, indicating they lacked the necessary information for EE. ⁴² Informal online support networks provided many women with the tools necessary to succeed. ⁴ , ⁴⁰ , ⁴² Clapton‐Caputo et al identified several subthemes in their qualitative study exploring the experiences of women who access social media groups when pumping. ⁴⁰ Within the theme of receiving information to manage EE, subthemes included expression schedules and the “let‐down” reflex; learning EE; and resolving lactation and breast‐related issues.^{40(p 374)} As women gained more knowledge and experience, they integrated additional techniques that helped them learn the “tricks of the trade.”^{4(p 8)}

OUTCOMES

Infant Outcomes

Two studies reported information on infant outcomes with EE. ¹ , ⁵¹ Differences in infant weight gain by the type of milk given and mode of delivery were found. ⁵¹ Compared with infants directly breastfeeding, those fed by EE gained 89 g more monthly (95% CI, 13.19‐164.47; P = .02), and infants bottle‐fed formula gained an additional 71 g per month (95% CI, 56.03‐86.47; P < .001). ⁵¹ This was in a large study of 1899 couplets, but only 34 infants were fed by EE. ⁵¹

Only one study examined differences in cognitive outcomes when the type of milk given was compared with the mode of delivery. ¹ The 11 women who chose EE were incorporated within the larger group of women choosing mixed feeding (both direct breastfeeding and feeding expressed breastmilk). Infants fed some or only breastmilk in the first 3 months of life performed better on tests of cognition (P = .011), had better verbal intelligence scores (P = .046), and had higher overall intelligence scores (P = .02) at 54 months of age compared with infants fed only formula. In addition to differences in cognitive results, children fed by EE and mixed feeding also demonstrated better gross motor skills when compared with the children only fed formula (P = .038). ¹

Cessation of Human Milk Feedings

Eight studies provided data on cessation of breastmilk feedings for women who were exclusively expressing. ¹ , ⁴ , ²⁸ , ³⁰ , ³³ , ³⁴ , ³⁶ , ³⁸ Insufficient milk supply is a common problem among breastfeeding families and may be even more significant in women who are feeding their infants with EE. A retrospective cohort study found that women who were performing EE were more likely to have insufficient milk supply compared with those who combined direct breastfeeding and feeding expressed breastmilk (91% vs 75%; P = .04). ³⁰ Insufficient milk supply was the most frequently stated reason for discontinuing pumping among women who used EE for their infants in the NICU. ³⁶ , ³⁸ Women who used EE had a higher likelihood of formula feeding by 12 months (100% vs 86%; P = .01), and they initiated formula feeding earlier (median = 1 day of life vs 3 days; P < .01) compared to women using mixed feeding. ³⁰

Human milk feeding duration initially appears to be shorter in infants fed by EE compared to infants fed by mixed feedings. In a small study of women feeding their infants by EE, the average breastmilk feeding duration was 5.04 months. ⁴ In the studies conducted in Singapore and China, the EE duration ranged between 5.1 months and 7.3 months, much longer than the 56‐day duration seen in the only US study with duration information. ¹ , ³⁰ , ³³ , ³⁴

Women who use EE had a higher likelihood of earlier lactation cessation at each time point within the first 6 months of life compared with women directly breastfeeding. For women exclusively expressing for a one‐month‐old newborn, their risk of breastmilk feeding cessation was increased (adjusted hazard ratio [aHR], 1.25; 95% CI, 1.04‐1.51). ²⁸ Likewise, women exclusively expressing at 6 weeks postpartum had an increased risk of breastfeeding cessation (adjusted risk ratio, 1.77; 95% CI, 1.25‐2.48; P < .001). ³³ Among women providing any breastmilk for a 3‐month‐old infant, those exclusively expressing had a higher likelihood of stopping any breastmilk feeding (aHR, 1.43; 95% CI, 1.12‐1.84); ²⁸ (aHR, 2.20; 95% CI, 1.61‐3.02). ³⁴ In contrast, when analyzing duration among women feeding their infants only breastmilk by direct breastfeeding or mixed feeding, women exclusively pumping and providing 100% breastmilk did not have an increased risk of stopping providing only breastmilk (aHR, 0.93; 95% CI, 0.48‐1.79). ³⁴

EE Challenges

Five studies had information on EE challenges. ³⁷ , ³⁸ , ³⁹ , ⁴³ , ⁵² Among women using EE for their infants hospitalized in the NICU, the most cited reasons for discontinuing pumping were stress, low milk supply, returning to work, pain, and that pumping was time‐consuming. ³⁷ , ³⁸ Case studies highlight the challenges these women encountered. The first case described a woman who tried direct breastfeeding for 2 months unsuccessfully before deciding that EE would be less stressful. ⁴³ The couplet also struggled with the infant accepting human milk after it was previously frozen, which the parent perceived as caused by high‐lipase milk. The parent worked to counteract this by scalding all milk prior to freezing. ⁵³ This woman successfully pumped for 13 months and provided her infant with breastmilk until 20 months of age. ⁴³ The focus on increasing breastmilk feeding duration by creating a stockpile and differences in breastmilk production versus duration of breastmilk consumption were subthemes previously identified within qualitative studies. ³⁹

The second case reported a woman who used EE for her infant in the NICU when she contracted recurrent bilateral methicillin‐resistant Staphylococcus aureus mastitis. ⁵² It resulted in breast abscesses, 2 bilateral incision and drainage procedures, a 12‐day inpatient hospitalization, and home health care for wound care and monitoring the surgical drains. ⁵² Breastfeeding problems are a known risk factor for mastitis (OR, 5.0; 95% CI, 2.4‐10.5; P < .001); therefore, EE and not direct breastfeeding may have contributed to the severity of this case. ⁵⁴ Understanding EE challenges will help clinicians better educate and support these vulnerable families.

DISCUSSION

Families desire to provide human milk to their infants but often face challenges that cause many to prematurely change to artificial formula. Although direct breastfeeding is the biological normative option ⁶ and has unique benefits, EE is a valid second choice and can be sustained as a long‐term feeding option to extend the benefits of human milk to more infants. Many families are unaware of EE, and clinicians have an opportunity to impact families who want to use this option. Clinicians can share information on strategies, such as hands‐on pumping or manual expression of colostrum following pumping, encouragement to persevere, and advice on pumping frequency and duration to help families succeed.

EE may not be feasible for all families or compatible with full‐time employment. ³⁹ It is time‐consuming, and women may feel they have to advocate for themselves ⁴ and seek out solutions to the problems they encounter. However, some appreciate the sense of control and autonomy EE can provide them. ⁴ , ³⁹

Clinicians must carefully assess families pumping for a NICU infant to ensure they receive the needed resources and guidance to meet their infant feeding goals. Often, EE is more prevalent in this demographic; it is imperative that they be given accurate and consistent information to best establish and maintain a sufficient milk supply. Furthermore, families may be traumatized by the birth and still grieving for the loss of the healthy pregnancy, desired childbirth experience, and healthy infant they may have expected. When encountering lactation difficulties, they may perceive their birth and lactation complications as personal failures. ⁴⁴ After overcoming lactation and EE challenges, some women felt good about their accomplishments ⁴ and others still felt like their breastfeeding experience poorly aligned with their goals. ⁴⁵

Future research on infant feeding should include longitudinal prospective studies that track substance fed, mode of delivery (bottle or breast), pumping patterns, problems encountered, and health outcomes. Importantly, the act of providing breastmilk must be separated from the act of direct breastfeeding and the mode of delivery. ⁵⁵ , ⁵⁶ When assessing maternal outcomes, researchers should focus on the length of time the person has lactated. Likewise, infant outcomes such as respiratory tract infections, otitis media, diabetes, and obesity and the length of time the infant received breastmilk should be assessed. Breastfeeding is not always a simultaneous activity between a parent and infant; more than 10% of infants in a recent study received breastmilk for at least 4 weeks after the woman stopped expressing. ³¹ , ⁵⁷ , ⁵⁸

Strengths and Limitations

An integrative review approach was chosen because it would allow diverse perspectives about the wide range of variables that would answer the research question on EE. ²⁴ A medical librarian was consulted early in the planning and execution of the review. Only peer‐reviewed studies were included, with few limitations to capture pertinent literature comprehensively. The use of specific review software helped standardize and avoid errors. This comprehensive review includes articles from 8 countries and many study designs.

There are limitations to every review. Publication bias is a risk, as it is possible that there are research studies that did not get published due to nonsignificant findings. Secondly, there is a risk that some published literature was not found in our database searches. Another limitation is the small sample sizes in some studies, which could affect the significance. Lastly, there were significant differences in study aims and comparison groups within the studies, which limits the generalizability of the findings of this review.

Positionality

Both authors have extensive experience and hold certifications in maternity care. The first author is a registered nurse IBCLC, and the second author is a certified nurse‐midwife. The first author also fed their fourth child by EE for 13 months. Additionally, both are cisgender women who have personally breastfed their children.

Implications for Practice and Research

Improving data collection is essential to accurately determine the differences in outcomes and each practice's unique risks and benefits. Although human milk feeding duration may appear shorter in those who are fed by EE, one must consider that infants who are directly breastfeeding or mixed feeding may not be the best comparison group. Considering that EE often occurs after there have been barriers to direct breastfeeding, an appropriate comparison group may be infants having difficulties with direct breastfeeding in the first few weeks of infancy. Compared with parents wanting to breastfeed and not meeting their feeding goals, those who feed by EE may have better health outcomes than those who switched to formula feeding.

Clinicians must educate and support all families in reaching their feeding goals. For those separated from their infants or planning EE, pumping ideally should occur within one hour of birth. ⁵⁹ , ⁶⁰ Pumping should then occur every 2 to 3 hours, ⁶⁰ , ⁶¹ with a goal of at least 6, but preferably 8 to 12 pumping sessions in a 24‐hour period. ⁵⁹ , ⁶⁰ , ⁶¹ While expressing colostrum, pre‐expression breast massage and manual expression after pumping are beneficial. ⁶² Once in lactogenesis III, the desired milk volume is often reported as at least 500 ml in 24 hours, but more commonly, the aim is for 750 to 1000 mL in 24 hours. ⁶⁰ , ⁶¹ Additional guidance provided by the Association of Women's Health, Obstetric & Neonatal Nurses Practice Guideline on the Use of Human Milk During Parent‐Newborn Separation includes encouraging double pumping and pumping for 2 minutes following cessation of milk flow. ⁵⁹ The research is less clear about when families can start decreasing the frequency of pump sessions or the best strategies for maintaining long‐term pumping.

Information about the recommended pumping frequency is also helpful for those concerned about undersupply or maintaining adequate milk volume. Power pumping, a technique mimicking cluster feeding, is a common method to increase the human milk supply. ⁶³ A hands‐free pump may be helpful to allow for the continuation of work or interaction with other children in the family while continuing to pump. With the frequency of EE, the silicone pump parts (duckbill, membranes) need to be replaced more frequently to maintain pump efficacy. An extra set of pump accessories can help ensure families have a clean supply after each use. Pump supplies that touch the milk can be washed in a closed basket in the top rack of the dishwasher, hand washed in a dedicated wash basin, and optionally steam sterilized once daily. ⁶⁴

Clinicians should consider the key information they can provide for lactating individuals who are using EE versus what may necessitate a referral to an IBCLC. As a clinical expert, an IBCLC can provide the dyad with anticipatory guidance and problem‐solving to help overcome any unique challenges. ⁶⁵ This will be particularly helpful for accurate flange fitting, persistent breast or nipple issues, and managing an over‐ or undersupply of milk. ⁶⁶

CONCLUSION

This integrative review aimed to explore the prevalence, characteristics, practices, and outcomes of individuals who exclusively express breastmilk. Results of this review show an EE prevalence rate of at least 6.8% among breastfeeding families; most of these parents receive practical and social support online. Parents who exclusively expressed were more likely to lack prior breastmilk feeding experience, given birth preterm, have an infant with a NICU stay, and planned initially on direct breastfeeding or combination feeding. They received varying instructions on important issues such as how quickly to start pumping after birth and the frequency and duration of pumping throughout their wake and sleep cycles. Insufficient milk supply and a shorter breastmilk feeding duration were common themes for EE families. This review shows that EE is a feasible option for lactating people with consistent lactation education, social support, equipment, supplies, and time for pumping.

CONFLICT OF INTEREST

K.R. has received funding to support conference travel and attendance from Lansinoh Laboratories. B.M. has no conflicts of interest to disclose.

Supporting information

Table S1. Literature Search Strategies