To state that breastfeeding is best for infants is an understatement as its benefits are countless [1]. However, for many years, the declining rate of breastfeeding is attributable in part to a lack of appropriate support by many medical providers. Consequently, mothers did not receive consistent, timely, and adequate advice and assistance regarding breastfeeding [2, 3]. The Baby Friendly Hospital Initiative (BFHI) was a response to the Innocenti Declaration on the Protection, Promotion and Support of Breastfeeding call for action proposal, which was adopted by the forty-fifth World Health Assembly in May 1992 [4, 5]. Since then, hospitals in many countries implemented the initiative, which gained an increased support in the US [6,7,8,9]. In the recent years, however, controversy emerged regarding its efficacy and safety [10,11,12].

Table 1 presents the Ten Steps for Successful Breastfeeding, the main foundation of the BFHI. [13]. We will critically evaluate each of the Ten Steps to understand not only the clinical evidence for inclusion and effectiveness of each step, but also the possible associated risks.

Table 1 Ten steps for successful breastfeeding
Full size table

Step 1: Have a written breastfeeding policy that is routinely communicated to all health care staff

The BFHI and the Ten Steps resulted from a lack of a much needed national breastfeeding policies and guidelines [4, 13]. Implementing public practices at a national level was demonstrated to be an effective way in increasing breastfeeding initiation rates and duration [3, 14,15,16]. Therefore, in 1989, the US Surgeon General gave the first public endorsement of the benefits of breastfeeding and then reaffirmed in 2011 through the Surgeon General’s Call to support breastfeeding [3, 17, 18]. Outside the US, another initiative was the Promotion of Breastfeeding Intervention Trial (PROBIT) which compared the implementation of the BFHI practices against the common practices in hospitals in the Republic of Belarus [19]. The PROBIT trial showed that implementation of the BFHI led to increased rate and the duration of breastfeeding [19].

Australia and Switzerland also implemented the BFHI [2, 20, 21]. Australia began implementing the BFHI in 1992, shortly after the guidelines were released [2]. Although having a universal health care system, Australia faced unique difficulties in adopting the BFHI universally, because the health system for each region functioned independent of each other [2, 20, 22, 23]. However, some progress became evident with having the BFHI initiative based on the 2010 National survey that showed a 96% breastfeeding initiation rate. However, only 39 and 15% of mothers breastfed their infants exclusively until 4 and 6 months of age respectively [20].

Switzerland introduced the BFHI in 1993 and the success of the initiative was evaluated by a national survey in 1994 and in 2003 [21]. Results showed that the median duration of breastfeeding increased along with the duration of exclusive breastfeeding. Infants born in hospitals where the Ten Steps were implemented were more likely to be breastfeed for a longer time compared to those born in non-BF facilities [21]. However, a critical review of the results questioned to what extent BFHI influenced the rates and duration of breastfeeding amidst the country’s generalized change in culture or attitude toward breastfeeding [21].

Working mothers who choose to breastfeed are faced with many challenges once returning to the workplace, the most common reason for discontinuing breastfeeding [16, 24,25,26]. Heymann et al. in 2013 reviewed the national policies which intended to guarantee breastfeeding breaks in the workplace in an attempt to increase duration and frequency of breastfeeding [24]. Breastfeeding breaks with pay were guaranteed in 71% of countries compared to 25% that did not have a policy in place. After controlling for national gross domestic product, the female literacy rate, and the percentage of the population living in urban areas, exclusive breastfeeding rates were significantly higher in countries that guaranteed breastfeeding breaks [24].

There is no doubt that a national and local engagement in promoting breastfeeding is necessary [1, 15]. However, breastfeeding policies need to take in consideration different factors within a state or locality that can influence the success of breastfeeding initiatives [27]. Breastfeeding rates vary by state or by city. Further, rates differ because of factors such as income, educational level, prevalence of obesity, and smoking rate; these are risks that may be accountable for low breastfeeding rates [27]. As stated in the Ten Steps: “Ideally policies should also come as a commitment from parents, health professionals, the mass media, and other community groups.” [13]

Step 2: Train all health care staff in the skills necessary to implement this policy

Health care providers should not only know the policy but understand the reasoning for such policy [28]. However, there is controversy with regard to what education or training would be appropriate, who would provide the education or training, and the outlays or fees associated with such training as a requisite from the certifying agencies. Health providers have the knowledge and experience in various medical areas and should have a significant input in evaluating the personnel and the course content for training of personnel. Certified lactation nurses are an essential part of any program that aims to improve breastfeeding in a community; these lactation specialists are proficient in both theory and techniques related to breastfeeding. Therefore, education and training do not have to be necessarily obtained outside one’s institution.

It is undeniable that improvement in education content and training methods are needed. Pediatric residents usually receive little breastfeeding education during their training resulting in being non-active breastfeeding promoters [1, 16, 29,30,31,32]. On the other hand, there is little evidence that shows an improvement in the breastfeeding rates and duration after completion of training of health care providers [33]. But if the goal is to develop a national breastfeeding culture or environment, education of pediatric residents, nurses and practioners should be enhanced as part of their curriculum toward attaining their respective degree [1].

Step 3, 5, and 10. Inform all pregnant women about the benefits and management of breastfeeding; show mothers how to breastfeed and how to maintain lactation, even if they are separated from their infants and; foster the establishment of breastfeeding support groups and refer mothers to them on discharge from the hospital or birth center

There is substantial evidence that clearly shows the importance of prenatal and postnatal education and support in the improvement of breastfeeding initiation and duration [16, 34,35,36,37,38,39,40]. Primiparous mothers perceive breastfeeding as a natural phenomenon that will happen without problems. Therefore, when difficulties arise, the easier solution is to switch to artificial milk [41]. Mothers who delivered by C-section are astounded when they are expected to immediately start breastfeeding without any regard to postsurgical recovery and giving rise to complaints that breastfeeding is tiring and painful [42]. Roll et al. did a systematic review analyzing the factors that are related to a mother’s choice on how she will feed her infant [43]. Researchers have proposed a model representing how maternal and external factors influence the decision on breastfeeding (Fig. 1) [43].

Fig. 1
figure 1

Social Ecological Model of the influences that affects the maternal decision of breastfeeding. Reproduce with permission from ref. [43]

Full size image

Robert et al. in 2014 evaluated more than 1000 mothers in Belgium to assess their reasons for stopping breastfeeding [26]. At birth, a third of the mothers already decided not to breastfeed for personal reasons; however, introduction of breastfeeding substitutes was most likely due to intrinsic problems related to breastfeeding. One primary reason for breastfeeding cessation was the mothers’ perception of insufficient milk [26], in spite of the low incidence of this problem [25]. These findings further support the necessity of efficient and effective education for mothers.

A review in 2000 by Fairbank et al. in the United Kingdom found that breastfeeding information through printed material alone or delivered via a non-interactive or impersonal method of education has limited impact on breastfeeding initiation rates [33]. However, if the information is provided in small, informal groups, this format of information dissemination influenced the breastfeeding rates and duration; therefore, supporting the use of peer support programs [33].

Labarere et al. evaluated 231 mother-infant pairs in France and applied an intervention that included one extra visit postnatally to provide breastfeeding support [44]. The study reported an increase in the rate of breastfeeding at 4 weeks from 81.6% to 89.3%; the median duration of breastfeeding also increased from 13 weeks to 18 weeks [44]. There was also a decrease in the reporting of breastfeeding difficulties and an increase in breastfeeding satisfaction [44].

Enrollment in the Women, Infants and Children program (WIC) has been historically associated with lower rates of breastfeeding initiation compared to mothers who were not enrolled [45, 46]. Factors identified to influence these low rates were: lack of support inside/outside the hospital, returning to work, practical issues, WIC-related issues, and social/cultural barriers [47, 48]. However, Ahluwalia et al. reported in 2000 that, implementing 5 strategies to the WIC program in Georgia increased the breastfeeding initiation rate by almost 10% [49]. These strategies included: enhanced breastfeeding education, breast pump loans, hospital-based programs, peer counseling, and community coalitions [49]. This was later supported by a review by Hedberg et al. who found that interventions resulting in positive outcomes included peer counseling, better communication between hospital lactation consultants and WIC staff, breast-pump programs, and discouraging routine formula provision in the hospital and by WIC [47].

In 2009, the WIC program made an extensive revision of the package policies to comply with the recommendations of the Institute of Medicine [50]. After the 2009 WIC food package revisions, mixed results were published regarding the effect of program revision on breastfeeding. Overall the breastfeeding rate and duration increased among WIC participants. However, these rates are still lower than those who are not in need of accessing WIC services. Efforts must be directed to enhance access to peer-counseling programs that support breastfeeding made possible through increase in funding for the program [50].

A 2016 Cochrane review reported that education by health care, non-health care professionals and peer support interventions can result in improvement in the number of women initiating breastfeeding [51]. A 2017 systematic review of the step three by Wouk et al. found that prenatal and postnatal education are associated with better rates of initiation, duration and exclusivity of breastfeeding especially when given in conjunction with interpersonal support [35].

Steps 4, 7 and 8. Help mothers initiate breastfeeding within one hour of birth; practice rooming-in—allow mothers and infants to remain together 24 h a day, and encourage breastfeeding on demand

Skin to skin care has been reported in the literature since 1976 and was introduced after the positive results from the Kangaroo Care program. Women who experience early skin to skin are more likely to breastfeed longer and effectively, have decreased maternal stress and postpartum hemorrhage [52,53,54]. Infants experience better stability of their cardiorespiratory system and have higher glucose and better temperature control [52,53,54,55]. Early breastfeeding is related to a better breastfeeding method with a more organized breastfeeding pattern and improved overall success [53].

Likewise, rooming-in is suggested to benefit the initiation and length of breastfeeding [54]. Infants who room-in cry less, get soothed quicker and take more breastmilk [54]. However, a 2016 Cochrane review did not encounter conclusive evidence of better breastfeeding initiation and longer duration, or increase in the frequency of breastfeeding among infants rooming-in with their parents [56].

Conversely, the practices of early skin to skin care and rooming-in have associated risks. There are multiple reports on cases of sudden unexpected postnatal collapse in the neonate (SUPC) since 1994 [55, 57,58,59,60,61,62,63,64,65,66,67,68,69]. Sudden Infant Death Syndrome (SIDS) [70] differs from SUPC as the latter tends to occur in term or near-term infants who were well at birth and collapses unexpectedly in a state of cardiorespiratory compromise within the first 7 days of life [53]. As the BFHI expands in the US, there is also a growing concern of the associated risks because of the increase in the reported cases of SUPC [10, 11].

Risk factors identified to be associated with SUPC include primiparous delivery, parents being alone, maternal fatigue, and the infant being in a potentially asphyxiating position (prone position) [60, 62,63,64]. The high risk situations that may predispose infants to SUPC include extensive resuscitation (and the use of positive pressure ventilation), low Apgar scores, late preterm and early term infants (37 to 39 weeks), difficult delivery, mother receiving codeine or other medications (general anesthesia and magnesium sulfate) that may affect the neonate, and sedated mothers or sleepy mothers and newborns [53]. Therefore, recommendations were made for a safe skin to skin care [53, 55]. In Table 2 we present a suggested order set to help the personnel identify those neonates at risk of SUPC and provide a safe skin to skin care. This order set is based on the recommendations provided by the AAP [53] and Davanzo et al. [55].

Table 2 Proposed guidelines for safe implementation of early skin to skin and rooming-in
Full size table

Step 6. Give infants no food or drink other than breast-milk, unless medically indicated

Exclusive breastfeeding for the first six months of life represents the gold standard in infant nutrition [1, 71]. However, at the time of the Ten Steps implementation, health or medical centers had the liberal practice to give formula, glucose, or plain water which affected breastfeeding rates [13]. As the BFHI was implemented, health care facilities started to restrict the access to feeding supplements. Conversely, the lack of adequate feeding supplementation may result in excessive weight loss and hyperbilirubinemia among other medical conditions. Therefore, this generated a controversy regarding the safe use of supplemental formula while at the same time focusing on increasing breastfeeding rates.

Flaherman et al. performed a randomized control trial (RCT) in 2013 which gave limited formula feedings to 40 term infants with ≥ 5% of weight loss at 24 to 48 h of age by using a syringe [72]. They found that these infants had decreased formula intake at 1 week of life and continued breastfeeding for longer duration to 3 months of age. A similar trial in 2016 by Stranak et al. with 100 infants, found no differences in the rate of breast feeding initiation and its duration. [73]. Schbiger et al. randomized 602 infants to either restrictive supplement or pacifiers vs. conventional feeding practices during the first 5 days of life (supplementation after breastfeeding and pacifiers were offered without restriction) [74]. When comparing the groups, the study did not find a difference in breastfeeding rates at six months of life [74]. However the RCT that evaluated the use of cup vs bottle supplemental feeding showed a negative impact in the breastfeeding rates at 6 months of age [75]. From the 2016 Cochrane review, formula supplementation during the first few days did not affect breastfeeding rates at discharge and may have increase the rate of breastfeeding that continued to 3 months of age (low-quality evidence) [76].

Step 9. Give no pacifiers or artificial nipples to breastfeeding infants

How breastfeeding is influenced by the use of pacifiers or artificial nipples remains a topic of controversy [74, 77]. Medical benefits associated with the use of pacifiers include providing comfort, contributing towards neurobehavioral organization, and reducing the risk of SIDS [78, 79]. In the context of BFHI, the use of pacifiers is justified in low-birth weight, prematurity, and infants at risk for hypoglycemia [79]. Pacifiers are recommended to be offered to term infants when breastfeeding has been established [1, 78]. However, there is not an adequate, evidence-based definition of when breastfeeding is established. Therefore, controversy has evolved regarding the current recommendation that a pacifier can be introduced after 3 to 4 weeks of life [1].

Howard et al. reported that early use of pacifiers had a negative effect on any breastfeeding in a randomized study of 700 infants [75]. Jenik et al. evaluated the use of pacifiers on 1021 infants of mothers who were highly motivated to breastfeed and noted that offering pacifiers to 15 day-old infants (who regained their birth weight) was not detrimental [80]. The study of Schbiger et al. did not find differences in breastfeeding rates when comparing infants who used pacifiers vs. those that did not during the first 5 days of life [74]. Pincombe et al. analyzed 317 women who gave birth at a hospital that complied with the BFHI and explored the adherence of mothers to Steps 4 to 981. They initially found detrimental effects in the use of a bottle, pacifier, dummy, or a nipple shield during their postnatal stay; however, this became insignificant after adjusting for socio-demographic factors, intended duration of breastfeeding and method of delivery of the infant [81].

Systematic reviews by Karabulut et al. [82] and Nelson et al. [77] found detrimental effects in the use of pacifiers for the duration of any breastfeeding. However, a review by O’Connor et al. in 2009 indicated that the use of pacifiers did not influence breastfeeding duration or exclusivity [83]. A 2016 Cochrane analysis of two trials involving 1302 infants found no significant effect of pacifier use in healthy term breastfeeding infants either started from birth or after lactation was established [84].

Effectiveness of the BFHI

A landmark study being referred to as the example of a successful BFHI program is the Belarus intervention which was a cluster randomized study with a long-term follow-up [19, 85]. The study enrolled more than 8000 mother-infant pairs with the objective of assessing the effects of breastfeeding promotion on duration and exclusivity of breastfeeding. The study demonstrated a positive influence of the BFHI in rates of breastfeeding initiation and duration [19]. However, a criticism to the study is that only breastfeeding pairs were enrolled, making it difficult to evaluate the effect on breastfeeding initiation rates [20, 86]. Mothers came from a high educational background with low cesarean section and smoking rates [19]. Also mothers were discharged after vaginal delivery on day 6 to 7 of the infant’s life allowing for an adequate establishment of breastfeeding. This discharge practice is not consistent with Western standards of early discharge [19, 20]. Moreover, the Bellarus intervention had a rapid implementation in an underdeveloped health care system which encountered little resistance to the policy change. This scenario is quite different from settings and population in developed countries [20].

The systematic review by Fairbank et al. showed that institutional changes in hospitals using initiatives like the BFHI can be effective in increasing breastfeeding initiation and duration rates [33]. Merewood et al. surveyed 29 hospitals that earned a BFHI certification and reported increased rates of breastfeeding initiation and exclusivity [87]. These results were supported by a review in 2016 by Munn et al. [37] However, these studies failed to report length or duration of exclusive breastfeeding [37, 87, 88]. Yotebieng et al. reported the results of the implementation of the BFHI in Congo and found an increase in long-term exclusive breastfeeding rates after the implementation [89]. A retrospective analysis of the implementation of the BFHI in a hospital in Boston showed that the breastfeeding initiation rates remarkably improved after the implementation [88]. A major pitfall in getting BF certified is that the hospital must pay for all infant formula and ancillary items resulting in extra expenses that could be as high as $70,000 [88] and as low as $20,000. A cost analysis study in 2011 found that hospitals that adopted BF status had 1.6 to 5% higher expenses per delivery, compared to those that did not [90].

Contrary to the reported advantages of BFHI, Robert et al. in Belgium reported their experience with over a thousand infants and observed that being born in a BFHI facility did not influence the breastfeeding rates and duration [26]. A survey from 6752 women in Australia showed that infants born at a BFHI hospital had lower odds of breastfeeding at 1 and 4 months of age [91]. The conclusion conveyed was that in places where breastfeeding rates are high and evidence-based practices that support breastfeeding are in place, the BFHI accreditation does not have an influence on breastfeeding rates [91]. This finding was also supported by the study of Yotebieng et al [89].

Hawkins et al. analyzed the data from the Pregnancy Risk Assessment Monitoring System of 11723 mothers in five US states and compared the breastfeeding initiation rates of BF hospitals vs. non-BF hospital [92]. Breastfeeding initiation and duration rates were not different; however, among women with lower education there was a positive influence on rate of breastfeeding initiation and duration [92]. Another analysis from the same group compared the compliance with the Ten Steps among the BF accredited and non-accredited institutions [93]. Contrary to expectation, half of the mothers reported compliance with 6–7 steps regardless of the BF accreditation status. Findings of the study suggest that compliance rather than accreditation would increase breastfeeding rates [93].

Howe-Heyman et al. did an extensive review of the literature in 2016 and reported that the majority of studies reach a conclusion that BFHI is an intervention to increase initiation, long-term duration, and exclusivity rates of breastfeeding [94]. However, study design, setting and disparate methods were limitations that challenge the conclusion from the many studies. The review concluded that there is no clear evidence to support the positive influence of the BFHI to improve breastfeeding [94]. Clinicians should focus on evidence-based practices that have shown a significant influence on initiation, duration and exclusivity of breastfeeding [94].

Discussion

Breastfeeding should be the default choice for infants’ nutrition. However, the decision to breastfeed is individualized and it is made by the parents usually in the prenatal period [39]. Maternal factors predisposing to lower breastfeeding rates include low-income, non-Hispanic ethnicity, obesity, depression, younger maternal age, and or less than high school education [47, 95]. These factors need consideration in prenatal education if programs were to have a positive influence in improving initiation and duration of breastfeeding. In addition to accounting for the socio-demographic population characteristics, each institution must adapt its program according to the geographic area being served, as well as the current breastfeeding rates, practices, and perceptions [30, 32, 95]. Health care facilities need to identify champions and educate the health and non-health care personnel in the implementation of its policy.

The WIC program has demonstrated to be an effective way of improving overall nutrition and increasing breastfeeding initiation in the participating women whose socio-economic characteristics would otherwise predispose them not to breastfeed. For continued and further success, there is a need for improvement in peer-counseling to support breastfeeding in the WIC program.

Health care facilities need to evaluate current prenatal and postnatal educational programs to improve and expand BFHI. As part of the prenatal education, determination is made of the mother’s choices for skin to skin practices, rooming-in, and type of feeding. Maternal confidence in achieving exclusive breastfeeding, established as early as 32 weeks of gestation, is predictive of breastfeeding at 6 months of age [39]. Mothers need to understand that breastfeeding is not an easy task and that they may encounter problems especially during the early days postpartum when the goal is to establish successful breastfeeding [39, 42, 43]. In addition, during the postnatal period, providing peer support and education from health care providers comprise a proven and effective intervention [34, 40].

There is little evidence to support the practice of rooming-in [56]. However, we believe that rooming-in should be encouraged in a safe environment [10, 53]. A common objection from mothers to the BFHI is that they are made to feel guilty when not choosing to rooming-in. Also, this requirement has encouraged certain health care facilities to close their nurseries. However, we question that this practice may lead to increased NICU admissions for conditions that may appear to be concerning but may merely be a part of an infant’s postnatal transition period and be safely monitored in the newborn nursery.

Strong evidence exists in support of the benefits of the skin to skin care practice [52, 53]. The desire for skin to skin care as well as for rooming-in should be decided in the prenatal period and parental decisions must be respected. However, mothers need to be informed of what to expect and the pros and cons of skin to skin care. Skin to skin care should be safely promoted with mother and infant who are constantly supervised by a trained nurse. The risks associated with skin to skin care and rooming-in must be recognized and preventive measures have to be in place [10, 11, 53, 55]. Formulating policies will help to assure safe practices.

Exclusive breastfeeding is difficult to establish especially for primiparous women and evidence suggests that supplemental feedings are not related to decrease exclusive breastfeeding rates [18, 72, 73]. Term infants with weight loss ≥ 5% from birth weight, are at special risk and supplemental feedings should be considered [72, 73]. Currently, supplementation of feeding is still controversial. If a mother chooses to formula feed, health care personnel should explore the reasons for this decision. However, this should be done preferably during prenatal education; otherwise there is the risk of discomfort or guilt that as mothers they are providing suboptimal care to their infants.

The restrictive use of pacifiers is being challenged with the recent evidence indicating that its use is not related to decreased breastfeeding rates [18, 74, 83, 84]. Presently, there is no evidence as to when it is appropriate to offer pacifiers; therefore, each case should be evaluated individually. However, early pacifier use may actually benefit the infants if mothers are experienced and committed to breastfeeding.

The Ten Steps for Successful Breastfeeding and the BFHI were developed at a time where breastfeeding rates were historically low. Since then, there has been an enormous amount of research published of factors to be taken into consideration for a successful breastfeeding program. We do agree with the statement that the Ten Steps are in urgent need of an update [12]. Moreover, evidence is non-conclusive and not in full support of the BFHI as a program that can successfully increase initiation and long-term breastfeeding rates. Therefore, using the increase of breastfeeding initiation rates does not serve as a suitable or appropriate outcome to reflect the success of the BFHI [94]. Consequently, it would be problematic to regard the BFHI as best practice for the improvement of breastfeeding initiation rates and duration.

Certification of a hospital or health center as BF is by no means the only option for a successful breastfeeding program or to be designated as a BF institution [12, 18]. However, its structure can be carefully considered by health care facilities as a means to improve their policies on breastfeeding practices. A suitable approach to improve breastfeeding rates addresses geographic and population factors, early prenatal education, and postnatal support as the main components of an ideal program. A breastfeeding culture requires local, state and national interventions for optimal success.