Supporting breastfeeding when clinical nutrition interventions are required in the paediatric healthcare setting: a systematic review | International Breastfeeding Journal

Selection of sources of evidence

The database search retrieved 14,064 citations; after duplicates were removed and additional 3 manuscripts from citation searching were retrieved a total of 13,177 studies were screened with 36 studies meeting inclusion criteria (Fig. 1.PRISMA Flow diagram). A systematic review by Hookway et al. was excluded to avoid duplicate reporting, as many of the evidence sources included in the Hookway et al. systematic review were identified from database searching as meeting inclusion criteria for this review [15]. One additional paper exploring the maternal experience was identified from reviewing the evidence sources included in the Hookway et al. systematic review [25]. A total of 36 manuscripts were included for the final data analysis.

Characteristics of evidence sources

Study designs included cohort studies (n = 6) [26,27,28,29,30,31]; quasi-interventional studies including quality improvement projects (n = 7) [32,33,34,35,36,37,38]; randomised control trial (n = 1) [39]; case series (n = 4) [40,41,42,43]; cross-sectional studies (n = 6) [44,45,46,47,48,49]; qualitative studies (n = 7) [18, 25, 50,51,52,53,54] and case reports (n = 5) [55,56,57,58,59]. All papers referred to infants admitted to a paediatric hospital ward or infants managed in outpatient and primary care settings across the USA (n = 12) [26, 27, 30, 32, 33, 35, 36, 40, 47, 53, 56, 57]; Canada (n = 3) [42, 55, 59]; USA and Canada (n = 3) [44, 50, 51]; UK (n = 4) [18, 25, 52, 54]; Europe (n = 8) [29, 31, 34, 38, 45, 46, 48, 49]; Asia (n = 3) [39, 41, 43]; New Zealand (n = 2) [28, 58] and Kenya (n = 1) [37].

Nineteen manuscripts were retrieved exploring how to support breastfeeding mothers in the context of their infants needing nutrition support, encompassing different clinical conditions: congenital heart disease (CHD n = 6) [26, 30, 32, 33, 39, 40]; foetal anomalies (n = 1) [27]; inborn errors of metabolism (IEM n = 5) [29, 31, 38, 41, 43]; bronchiolitis (n = 1) [45]; complex respiratory conditions (n = 3) [28, 36, 42]; faltering growth (n = 2) [34, 37] and hyperbilirubinaemia (n = 1) [35]. Twelve manuscripts exploring maternal experiences were retrieved considering mothers of infants with phenylketonuria (PKU n = 5) [44, 48,49,50,51]; CHD (n = 2) [53, 54]; Down Syndrome (n = 1) [52]; bronchiolitis (n = 2) [46, 47]; and mixed medical complexity (n = 2) [18, 25]. Another 5 case studies were included encompassing cow’s milk protein allergy (CMPA) [55]; CHD (n = 2) [56, 57]; ketogenic diet for refractory epilepsy [58] and tracheostomy-dependent infant [59].

There were a variety of clinical nutrition contexts across all manuscripts. Nutritional interventions included: nutrition support to meet increased requirements (n = 13) [18, 26, 27, 30, 32, 33, 36, 39, 52,53,54, 56, 57], breastmilk fortification or modification (n = 6) [26, 32, 33, 40, 42, 53], supplementary formula or calories (n = 12) [18, 26, 28, 34,35,36,37, 40, 46, 47, 52, 54], feeding difficulties (n = 18) [18, 25,26,27,28, 30, 33, 35, 37, 39, 40, 42, 52,53,54, 56,57,58], tube feeding (n = 12) [18, 25, 28, 30, 40, 42, 45,46,47, 53, 54, 57], parenteral nutrition (n = 2) [45, 46], therapeutic diet (n = 12) [29, 31, 38, 41, 43, 44, 48,49,50,51, 55, 58], specialised formula or modular supplements (n = 10) [29, 31, 38, 41, 43, 44, 48,49,50,51].

Critical appraisal

Many of the included intervention studies contributed low-level evidence as they were mostly not randomised and had no control group. Due to the ethical implications of studying breastfeeding when it is the ‘gold standard’ source of nutrition for infants, it is difficult to randomise participants into intervention and control groups when an intervention is likely to result in longer breastfeeding duration or improved breastfeeding outcomes. There was only one unblinded randomised control trial [39], and many of the intervention studies were quasi-experimental, which included quality improvement initiatives [32,33,34,35,36,37,38]. The quality improvement projects and case reports show what can be done in a practice-based setting, and it is important that these continue to be reported to add to existing literature [32, 33, 35, 36, 55,56,57,58,59]. There were a range of observational studies [26,27,28,29,30,31, 40,41,42,43,44,45,46,47,48,49], with most not adjusting for confounding factors and having a limited follow up period during the period of hospital admission, with a few studies noting it was difficult to establish breastfeeding before discharge due to short length of stay [30, 33].

In the qualitative studies [18, 25, 50,51,52,53,54], studies were strong in outlining their methodology and analysis of the data collected. All studies reported author positionality. In two studies, recruitment relied on convivence sampling, with participants being self-selected [18, 52], and in one of these studies many were unintentionally self-nominated healthcare professionals or breastfeeding peer supporters [18].

Synthesis of results

Measures to support breastfeeding

Studies exploring measures to support breastfeeding are summarised in Table 2. Support measures were grouped into three overarching categories: 1) Environmental modifications and hospital procedures that facilitate breastfeeding, 2) Access to breastfeeding support healthcare workers and breastfeeding education, and 3) Flexibility in the provision of nutrition to facilitate breastfeeding.

Category 1: environmental modifications and hospital procedures that facilitate breastfeeding

This category describes provisions in the paediatric hospital environment and adaptations to hospital procedures that can support breastfeeding. Common environmental modifications used across studies included: a special delivery unit integrated into a paediatric hospital designed to care for mothers of infants with congenital anomaly, which enabled breastfeeding support to start in the antenatal period and continuity of care [30]; improving access to breast pumps ± expressing kits provided on admission [26, 32, 45]; expanded eligibility for pasteurised donor human milk [32]; meals for lactating mothers [32]; information flyers and education material, specific to clinical presentation [45]. Hospital procedures that facilitated breastfeeding included: individualised feeding plans considering maternal breastfeeding goals [33, 36]; lactation care integrated into the daily multidisciplinary ward round and clinical handover [33, 37]; clinical pathway to support infants establishing feeding directly at the breast [42]; standardised Lactation Consultant (IBCLC) referral via electronic medical record order set on admission to facilitate early involvement of an IBCLC in patient care [35]; development of policies and procedures related to expressing and storing breastmilk [26, 45].

Category 2: access to breastfeeding support healthcare workers and breastfeeding education

This category describes support measures that facilitated access to breastfeeding support professionals such as Lactation Consultants (IBCLC) and access to breastfeeding education resources. Increasing breastfeeding supports encompassed: integrating lactation education and counselling into pregnancy care and support for antenatal expressing when there is an antenatal diagnosis of a congenital condition (e.g. congenital heart disease) that frequently necessitates nutrition support [27, 30, 32, 40]; antenatal group care model for breastfeeding education and support [27]; lactation education for hospital staff including addressing misconceptions about the ability of medically complex infants to breastfeed, alongside upskilling healthcare staff in breastfeeding support [26, 32, 33, 45]; use of social media as a communication forum to facilitate access to infant feeding and parental support post discharge from hospital [39]; peer support model to improve access to lactation counselling within the paediatric hospital environment in a low resource setting [37]; timely involvement of an Lactation Consultant in multi-disciplinary care and feeding assessments [29, 35, 36, 40, 42, 59].

Category 3: flexibility in the provision of nutrition to facilitate feeding at the breast

This category describes clinical practices that help to facilitate direct and responsive breastfeeding, which in turn supports maternal supply and enhances maternal-infant bonding. Strategies included: partial breastfeeding whilst supplementing with fortified EBM [42]; monitoring objective clinical parameters to gauge adequacy of the infant feeding at breast (e.g. weighted breastfeeds to assess milk transfer and fluid balance, monitoring biochemical markers, growth, developmental milestones) [26, 29, 31, 33, 41,42,43, 58]; use of breastfeeding friendly medical devices in respiratory patients that traditionally require tube feeding eg. FlexTend tracheostomy [42, 59]; optimised positioning to support airway patency and non-nutritive sucking at breast which facilitated transition to breastfeeding for complex airway patients, several who were receiving supportive naso-gastric feeds [28, 42].

Breastfeeding outcomes when supportive measures implemented

Measures such as improved pump access and staff/parent education saw an increase human milk feeds from 55 to 92% (76% mothers’ own milk) [32], and preadmission breastfeeding status sustained in 80% vs. 49% of patients admitted with acute respiratory conditions [45]. By incorporating test weights to monitor fluid balance and volumes of breastmilk consumed, a paediatric cardiac care unit saw a 45% increase in direct breastfeeding episodes after implementing weighted breastfeeds to objectively monitor milk transfer at the breast amongst babies with CHD [33]. Expressing initiation rates were higher amongst mothers receiving antenatal counselling and delivering in a special delivery unit within paediatric hospital vs. external hospital (96% vs. 67%) [30]; likewise, over 90% of mothers initiated lactation on the day of delivery after receiving antenatal group breastfeeding support [27]. There was a significant improvement in exclusive breastfeeding rates in an intervention vs. control group (73.3% vs. 46.7%) with the implementation of a social networking support group for follow up education post discharge [39]. When responsive and direct breastfeeding was possible for infants with inborn errors of metabolism, breastfeeding duration was observed to improve [29, 31, 41, 43]. Breastfeeding duration was longest amongst infants with PKU who were fed on demand with specialised formula given prior to feeding at the breast to satiety, with an average duration of 10.4 months vs. average of 7.2 months for all infants with PKU receiving breastmilk [31]. Early involvement of a Lactation Consultant (IBCLC) in patient care increased the percentage of infants from 57 to 94% who were receiving breastmilk past the second clinic visit for infants with cystic fibrosis [36], and infants readmitted with jaundice were almost 4 times more likely to exclusively breastfeed during hospitalisation when seen by an IBCLC on the first shift [35]. When mothers of infants with faltering growth were not actively supported to increase breastmilk production and were given advice to supplement their infants with unlimited volumes of formula, the contribution of breastmilk to an infant’s total milk intake significantly decreased from 100 to 41%, and formula intake nearly doubled [34].

Maternal experience

Studies describing the maternal experience are summarised in Table 3. There were 3 main themes that illustrated the maternal experience of breastfeeding a child requiring nutrition interventions in the paediatric healthcare context: 1) Breastfeeding was difficult with complex nutritional needs, 2) Desire to breastfeed informing maternal identity and emotions, and 3) Supportive paediatric healthcare team promoted breastfeeding.

Theme 1: breastfeeding was difficult with complex nutritional needs

For many mothers, providing optimal nutrition and immune support by breastfeeding their baby and enhancing the maternal-infant bond, motivated women to persist despite significant breastfeeding challenges including sustaining supply, routine expressing, need for formula supplementation and ensuring breastfeeding did not compromise medical stability [18, 25, 44, 50, 51, 53]. A mother’s breastmilk supply was impacted by not being able to feed her infant at the breast, especially for younger newborn infants when maternal breastmilk supply is in a critical phase of establishment [44, 46, 53]. Tube feeding affected a mother’s ability to directly breastfeed and bond with her baby [47]. Having more than one type of feeding method (e.g. Breast, bottle, tube), and the introduction of commercial milk formula (e.g. Specialised formula necessary in IEM) often compromised maternal supply and could influence a mother’s ability to continue breastfeeding [44, 46, 54].

A mother of an infant with PKU emphasised that ‘[Breast milk] supply issues were the biggest challenge’ ([50] p.6). Committing to the extra work of routine expressing to sustain supply was an additional demand on mother’s time and caregiving responsibilities, alongside the increased effort of breastfeeding given the complexity of dietary management necessitating frequent feed changes which also impacted supply [44, 50, 51, 53]. When a mother struggled to maintain her supply, more adaptations and changes to breastfeeding were required [18]. In some circumstances additional expressing was not sufficient on its own to sustain supply; some mothers of infants with PKU reported use of galactagogues such as domperidone [50], other mothers weaned earlier than expected in the context of their infant’s medical complexity and feeding difficulties [18, 52].

Whilst the benefits of breastfeeding were a strong motivator of breastfeeding their infants, universal breastfeeding problems such as latching issues, breast pain and nipple damage, were heightened by the additional challenges associated with breastfeeding an infant with PKU needing complex dietary management [44, 51]. One dietitian described modifying breastfeeding to accommodate clinical dietary needs as ‘an art not a science’ ([50] p.5), highlighting the fluidity of dietary management in the nutritionally complex breastfed infant. The constant changes to nutrition care plans made sustaining breastmilk supply challenging in the context of not being able to responsively feed: ‘Breastmilk works on supply and demand and is not a spigot. It was hard to suddenly reduce the number of breastfeeds as I would get engorged, and even harder to suddenly increase the number of breastfeeds because I wasn’t producing enough milk’ ([44] p.223).

Theme 2: desire to breastfeed informing maternal identity and emotions

For many mothers, a child with unconventional nutritional needs necessitated changes to maternal feeding intentions and expectations about what breastfeeding would be like [25, 50, 53]: ‘I remember explaining to one of the rounding teams: It’s really hard to produce milk without a child. Without actively touching, holding, connecting, bonding, feeling… The milk I’m producing is filled with tears as opposed to being filled with joy’ ([53] p2006). Some mothers adjusted by reframing their circumstances and aligned their desire to breastfeed with the realities of their situation, whereby an aspiration to exclusively breastfeed became a goal to exclusively provide breastmilk for her baby [25]. Breastfeeding also enabled mothers to stay grounded when the future seemed uncertain, ‘Breastfeeding is the one thing I could hold on to, of these dreams I had for this baby’ ([53] p.2005). Breastfeeding facilitated a biological and emotional connection to support mothers bonding with their infant at a time that was unsettling and fraught with the unknown [25, 47, 51].

Breastfeeding was an opportunity for mothers to contribute to their child’s care and provided a sense of normalcy amongst medical and nutritional complexity [18, 25, 51]. For some mothers, breastfeeding increased their confidence and self-efficacy, through mastering a complex dietary management routine and continuing to breastfeed in exceptionally challenging circumstances [50]: ‘I feel so proud of myself… I think that if I hadn’t [breastfed], I might have thought that well maybe I would have done something differently. So I think that [breastfeeding] really was a labour of love’ ([51] p.732). Opportunities to breastfeed were important for mothers to sustain hope for the future: ‘The fact that she could do it and we could do it was really powerful for me and it really meant a lot… because she’d done it once I felt fairly confident that we could go back and try to do it again’ ([25] p.796).

Mothers reported significant stress associated with the increased demands and complexity of dietary management, and infant feeding became an obsession: ‘Feeding was everything that was all I could think about, all I could focus on…’ ([52] p.136). Concerns about weight gain instilled fear and tube feeding contributed to feelings of disempowerment with the focus on quantifying intake, where volumes are prioritised over responsive feeding or feeding at the breast [18, 25, 50, 52, 53]. Many mothers described their lives evolving around feeding their infant: ‘I had to breastfeed him. Then I had to express… every 3 h my whole life was literally evolving around tube feeding, expressing and breastfeeding’ ([54] p.3). Breastfeeding was a demanding and time intensive aspect of mothering, with maintaining clinical stability and optimising nutritional status becoming the focus of breastfeeding [50, 52, 54]: ‘It was almost like PTSD [post-traumatic stress disorder]… I needed to know how many ounces went in her… I think having that pressure of wanting to [breastfeed] your child and being scared that your child’s not going to be gaining weight…’ ([53] p.2005). The obsession and significant demands associated with feeding their infants, in turn informed maternal identity: ‘I was so worried about every drop of milk… because that was how they made me feel… I was being more of a nurse than a mum’ ([54] p.6).

In some instances, mothers expressed mixed emotions continuing to breastfeed despite significant obstacles, and difficulties deciding to wean or cease breastfeeding earlier than expected, often in the context of low supply [18, 44, 48, 52]. There were feelings of maternal inadequacy and guilt strongly related to infant feeding decisions [52, 54]: ‘You’ve got this guilt constantly that you’re not doing the right thing, you should be breastfeeding, and obviously he’s getting formula, so there’s so much guilt around’ ([52] p. 136). Some mothers mentioned the impact of stress and trauma on the milk-ejection reflex, their ability to express breastmilk and ultimately compromising their capability to continue breastfeeding: ‘It was just so stressful… trying to express while she was so poorly, it just didn’t happen. And that was the end of our breastfeeding there’ ([18] p.10).

Theme 3: supportive paediatric healthcare team promoted breastfeeding

A supportive healthcare team facilitated maternal choice, reducing the maternal burden and need for self-advocacy [53]. The presence of healthcare staff that championed breastfeeding, took away the need for mothers to continually advocate to breastfeed their baby and preserved their breastfeeding journey [18]: ‘They came with breastfeeding first… They wanted to make sure that I had that connection and they were going to help me as much as they could’ ([53] p.2006). For some mothers a supportive healthcare professional was crucial in the establishment of breastfeeding: ‘[She] was the only reason that my baby was allowed to latch after her first surgery, because she pushed and pushed… thankfully, I had allies in the hospital to advocate for me and my baby’ ([53] p.2006).

In one study dietitians were identified as the primary support person for mothers within the multidisciplinary healthcare team [51]. However given a dietitian’s focus on the management of the infant’s nutritional status, they could also be unsupportive of breastfeeding when they lacked education and skills to support continued breastfeeding, underestimated the burden of complex dietary regimes, or gave mothers conflicting advice [50, 52, 54]: ‘The dietitian… got the wrong end of the stick and thought that I shouldn’t be breastfeeding him, because she said that my milk was dangerous to him… that’s not true… after the surgery he could breastfeed as much as he wanted’ ([54] p.6). Supportive breastfeeding policies were not the norm and there was a lack of consistent and reliable breastfeeding support for challenging clinical scenarios available in paediatric hospitals: ‘When people start saying, “Our policy is …” we have to have a much bigger discussion. Because my baby is not a protocol, and neither is our family’ ([53] p.2006). Some infant feeding specialists helped facilitate direct breastfeeding [53]; on the contrary there could also be difficulties accessing an IBCLC who had specialist knowledge in supporting medically complex infants: ‘The breast-feeding support worker that was attached to PICU bless her. She… didn’t know anything about it really, like she’s obviously read the facts, but it’s very different, isn’t it?’ ([18] p. 14).

Often there was a lack of consensus regarding clinical practices for supporting breastfeeding in medically complex situations and mothers were unable to access evidence-based breastfeeding resources specific to their child’s clinical condition [51,52,53,54]. Some healthcare professionals acknowledged the boundaries of their knowledge, and expertise to support breastfeeding in uncommon clinical scenarios was typically held by mothers with firsthand experience [52]. One mother felt validated as she navigated breastfeeding challenges through the humility of a physician: ‘Our complex care doctor is very much like, I know you’re on all those [social media] groups, why don’t you get on there and see what other people have done? I love that because he recognises that he has limitations in what he knows’ ([53] p.2006).

A mother’s interaction with her healthcare team could be more influential in shaping her feeding experiences, than the nutritional management principles related to the infant’s clinical presentation: ‘the assumptions of health professionals regarding the ability of the mother/baby to breastfeed were more of a barrier to successful breastfeeding than the infant’s diagnosed condition’ ([52] p.138). Many health professionals lacked knowledge and skills to support infants to breastfeed in nutritionally complex circumstances; their ambivalence towards encouraging breastfeeding often attributed to their focus on the medical and nutritional management of the patient [25, 46, 52, 54]. Some mothers sensed that breastfeeding was an inconvenience to healthcare staff: ‘A lot of them have sort of thought of breastfeeding as a barrier to her eating… they actually mentioned that a few times that it would be easier if she was bottle fed because we could see how much she was getting’ ([18] p.15). There was conflict between maternal feeding goals and medical/dietetic goals of care, with mothers opposing the healthcare team to protect feeding their infant at the breast or not attending appointments as they were not receiving the support they needed to continue breastfeeding [52, 53]. Hospital policies often left mothers needing to seek ‘permission’ to breastfeed in the context of clinical protocols that dictated infant feeding decisions and interfered with establishment or maintenance of breastfeeding [52, 53].

In some instances, feeding an infant at the breast shifted to the mother expressing so the healthcare team could measure volumes and maintain precision to quantify nutritional intake, often to address growth concerns and maintain strict fluid balance. A focus on quantifying intake resulted in imposed feeding schedules, pressure to introduce additional formula and bottle feed, and potential overuse of naso-gastric tubes [18, 25, 46, 52,53,54]. Many mothers expressed anxiety associated with infant weight gains and the need for the healthcare team to quantify the care received with mothers describing a ‘single-minded healthcare team’ whose focus was ‘all about the numbers’ ([53] p.2002). Breastfeeding is difficult for healthcare staff to quantify, and the need for exact measurements provoked anxiety in a lot of mothers: ‘They kept asking me how much has he had? Because I think it was very much tailored towards you know, formula fed babies… I couldn’t really give them kind of like the amounts that they were trying to document’ ([18] p.10). Healthcare teams made decisions based on their subjective assessment on the adequacy of maternal supply and mothers felt pressure to keep up expressing volumes of breastmilk that felt unachievable [18, 52].

Key learning from the case reports

Measures supporting breastfeeding in the case reports (Table 4) included antenatal education [56], expressing supplies [57], support and feeding assessments from lactation consultants [57, 59], optimising direct breastfeeding in response to growth outcomes [57], and adapting clinical practices through modifying maternal diet or clinical devices used to support sustained breastfeeding [55, 58, 59]. A positive maternal experience was reflected when healthcare professionals discussed infant feeding options to support informed infant feeding decisions [56], and supported continued breastfeeding with novel measures [58]. Responsive healthcare teams supported direct and exclusive breastfeeding for motivated mothers [57, 59], and individualised care specific to the infant feeding goals of the mother [57]. Misconceptions and inconsistent healthcare practices with regards to breastfeeding infants with complex nutritional needs, were barriers to a mother breastfeeding [57].