In this study, we explored HCP decision-making around use of BMS at TOH and identified factors contributing to their use with and without medical indications. Our findings suggest that several multilevel barriers influenced the provision of BMS at the patient, HCP, and institution-level.

Patient-level factors such as medical indications, physical factors, and parental requests contributed to the use of BMS. These findings are similar to other studies investigating in-hospital formula supplementation including a study conducted across hospitals in the United States that reported medical indications, maternal request/preference/feelings, and lactation management-related issues as the top three most common reasons for formula supplementation reported by hospital staff [24]. Furthermore, our findings were consistent with several of the top reasons for maternal request of BMS reported by a public health unit in Ottawa, Canada including: medical conditions of mother or infant, inconvenience/fatigue/lack of time/finding breastfeeding too demanding, and milk supply concerns [25]. Maternal complaint of not producing enough milk, often referred to in the literature as “self-reported insufficient milk” (SRIM), has been frequently cited as a common reason for introduction of BMS [26, 27] and overall breastfeeding cessation [28]. Avoiding in-hospital formula supplementation, improving breastfeeding counselling, and increasing maternal breastfeeding self-efficacy are integral in protecting against SRIM and unnecessary BMS use [26, 27].

Additionally, parental request for BMS may be attributed to poor prenatal breastfeeding education and preparation for newborn care, which was a recurring theme in our interviews. A qualitative study conducted in Northern California investigating decision-making around maternal request for formula reported similar results with inadequate preparation for breastfeeding and using formula as a solution to breastfeeding problems as key reasons for in-hospital supplementation [29]. Furthermore, a systematic review assessing the effectiveness of prenatal education on breastfeeding outcomes found that caregivers participating in prenatal programs had higher breastfeeding uptake and knowledge, increased positive attitudes towards breastfeeding, as well as better self-efficacy [30].

Additionally, HCP-related factors associated with BMS use was another main theme that emerged from our interviews. Clinicians’ beliefs and attitudes on breastfeeding and recommendations for use of BMS are significantly associated with in-hospital breastfeeding initiation and EBF [31, 32]. In our findings, while most HCPs were highly supportive of EBF and indicated they would review the benefits of breastfeeding with parents who requested BMS, others stated they would provide a BMS without question. Most HCPs didn’t want mothers to feel pressured to breastfeed or evoke feelings of guilt for wanting to supplement with formula. “Formula-shaming” is a controversial topic in the literature and consequent negative effects on the mother have been explored. A recent systematic review revealed that formula feeders experienced guilt more often than breastfeeding individuals and that external guilt was most associated with HCP influence [33].

Additionally, inconsistency in training amongst staff may further exacerbate this issue as it facilitates confusion and miscommunication among HCPs, and thereby to the patient as well. Key points HCPs highlighted was that some of the training may be outdated or inconsistent, or that they were unfamiliar with hospital protocols and policies. The development of skills and knowledge needed to support a breastfeeding individual is essential for ensuring breastfeeding in-hospital and continuity after discharge. One study showed that resident physicians who had higher participation in a breastfeeding education program were more likely to have better breastfeeding rates amongst their patients [31]. Similarly, another study found that the odds of breastfeeding were greater in hospitals where there was breastfeeding education for new employees, nurses received breastfeeding education in the past year, and there was a written breastfeeding policy [34]. The same study found that the time invested by hospitals in staff training is proportionally related to improved breastfeeding outcomes [34]. This can be compared to the BFI’s Breastfeeding Success Steps, which highlights the importance of having a written breastfeeding policy routinely communicated to HCPs and to ensure all HCPs have the knowledge and skills to implement the policy [8]. Thus, along with updated training, it is key that institutions facilitate straightforward communication of policies to staff to bridge any gaps between current knowledge and clinical practice.

Lastly, we identified institutional-level factors that challenged HCPs’ ability to provide adequate breastfeeding support to their patients and promote potential use of BMS. Resource limitations, including budget and staffing shortages, was a dominant subtheme identified in this study. Most HCPs highlighted how low staff numbers hindered the amount of staff time spent available to provide breastfeeding support to patients, coupled with the need to monitor multiple patients for the duration of their shift. This is similar to studies conducted in other institutions which found that nurse staffing shortages were major barriers to implementing the BFI’s Ten Steps to Successful Breastfeeding, particularly due to insufficient time in providing adequate breastfeeding support [35, 36]. HCPs also highlighted the need for additional LC coverage, especially after regular working hours such as evenings, weekends, and holidays. A recent systematic review demonstrated that lactation consultants play a key role in improving breastfeeding initiation and EBF rates [37]. LCs are often referred to cases with medical indications and may not have time to speak with other patients, which may also contribute to the use of BMS if effective support isn’t provided by other staff members to patients in need. HCPs also noted there may be incomplete documentation of BMS use on the hospital system that may contribute to vague breastfeeding data.

Beyond factors that we identified at the patient, HCP, and institution levels specific to our study site, there may be additional external factors influencing overall infant feeding decisions. Many conceptual models have been developed to describe the multilevel interactions between the determinants of breastfeeding, such as the 2023 Lancet breastfeeding series framework [38]. Along with mother-infant factors, the authors highlight the impact of different structural factors (i.e., marketing and political economy of commercial formula) and settings (i.e., workplace and employment) on infant feeding practices. For example, although the topic of BMS marketing was not a major theme identified in our study, it is important to note how the commercial milk formula industry has long lasting impacts on breastfeeding practices. Currently, our institutional policies prohibit marketing of BMS within the hospital and to parents. This may contribute to why this issue was not raised during the interviews. However, Canada remains one of several countries that have no legal measures to regulate marking of BMS as suggested by the International Code of Marketing of BMS [17]. Thus, future studies investigating the effect of BMS marketing on breastfeeding practices at a local, provincial, and national scale are warranted. Additionally, maternity employment legislations are also key determinants of parents’ infant feeding decisions [39]. Studies have shown that extended paid maternity leave are associated with positive breastfeeding outcomes including higher breastfeeding initiation and duration rates [40,41,42]. Currently, the province of Ontario supports up to 63 weeks of parental leave which promotes retention of breastfeeding beyond 6 months. However, this leave is unpaid, and parents may only receive benefits through the federal government depending on specific eligibility criteria.

Our findings highlight several modifiable barriers to EBF at TOH. Although we identified three separate categories contributing to EBF rates and use of BMS, understanding the interconnectedness between all categories is critical in improving breastfeeding outcomes in this institution. For example, institutional policies and training may undermine an HCPs knowledge and skills on providing breastfeeding support which may further influence the mother’s confidence and ability to breastfeed their infant. This may ultimately contribute to BMS use. Likewise, shortage of funding and staff may increase the stress and workload of HCPs which may lead to inadequate in-hospital breastfeeding support for the mother and subsequent use of BMS that may continue post-hospital discharge. Thus, developing multilevel interventions is critical to maintaining long-term change. Figure 1 portrays our initial codes, subthemes, and main themes and the interactions between these themes.

Factors influencing institutional exclusive breastfeeding (EBF) rates and use of breastmilk substitutes (BMS)

Many HCPs proposed strategies that could help overcome current barriers to increase EBF rates and reduce use of BMS. Developing and distributing resources that promote prenatal breastfeeding education among patients may address issues related to misinformation and negative beliefs and attitudes about breastfeeding. HCPs also suggested that more frequent and updated training could help limit differences in clinical practice between staff members and improve overall skills and knowledge in providing breastfeeding support. This can be addressed through developing annual electronic learning modules for this population or obligatory workshops delivered throughout the year. Budgeting and staffing shortages may not be easily addressed. However, identifying breastfeeding champions from existing staff may improve patient access to lactation support. Additional suggestions included developing resources such as a virtual breastfeeding support platform, walk-in clinic hours with LCs, and establishing a connected hospital-to-community perinatal care system to support patients’ post-hospital discharge. Lastly, another suggestion to reduce BMS use included implementing a mandatory note in patients’ medical chart. This would help capture in-depth information and support regular audits on the frequency and circumstances for supplementation at TOH. Although our findings were similar to those of many other international studies, we advise further research to be conducted on this topic in Canada to better understand the targeted interventions required to improve national breastfeeding rates and decrease use of BMS. Further research is also required to understand how lack of legal measures on BMS marketing may impact infant feeding practices throughout Canada. To achieve the WHO’s goal of 70% EBF in the first 6 months of life by 2030, implementing provisions from the International Code of Marketing of BMS is a pivotal first step.

This study offers valuable insight into HCP decision-making around use of BMS; however, limitations in our study exist. First, there were a few cases where it was difficult to differentiate between situations that were medically indicated versus NMI as such situations were not further elaborated on during interviews. For example, low milk supply and premature birth may be medically indicated in circumstances defined by the BCC such as if due to primary glandular insufficiency or if the infant was born at less than 32 weeks of gestation [8]. Likewise, separation of the mother-infant dyad was listed as a physical factor but may be due to medically indicated reasons. In addition, the voluntary nature of the study may contribute to selection bias as participants may be more interested in issues surrounding breastfeeding and the use of BMS. As the sample was only a subset of all eligible HCPs, our findings are not representative of the beliefs and attitudes of all eligible clinical staff at TOH. Self-reporting bias is also a potential limitation as HCPs may report information that may be inaccurate and reflects the HCP’s recollection of the situation or event. Social desirability bias must also be considered as HCPs may answer in a manner that is positively regarded by the researchers and the public but is not reflective of their own views.

Strengths of this study include the diversity of HCPs interviewed for the study across their clinical roles and extent of experience in maternity and infant care. In addition, a variety of methods were used to mitigate potential biases. The study’s semi-structured design enabled flexibility when conducting interviews where emerging themes may be probed with new questions. Likewise, the interviewer is a research trainee at the institution but is not directly affiliated with the department and does not hold a position of authority relative to the participants to ensure an impartial interview process. Lastly, the results of this study will be used to inform a larger department-wide survey to investigate the determinants of BMS use and low EBF rates in this institution. The results of our research can be used to drive similar studies both regionally and nationally.