Abstract
Considerable effort has been made in recent years to gain a better understanding of the effectiveness of different interventions for supporting breastfeeding. However, research has tended to focus primarily on measuring outcomes and has paid comparatively little attention to the relational, organizational, and wider contextual processes that may impact delivery of an intervention. Supporting a woman with breastfeeding is an interpersonal encounter that may play out differently in different contexts, despite the apparently consistent aims and structure of an intervention. We consider the limitations of randomized controlled trials for building understanding of the ways in which different components of an intervention may impact breastfeeding women and how the messages conveyed through interactions with breastfeeding supporters might be received. We argue that qualitative methods are ideally suited to understanding psychosocial processes within breastfeeding interventions and have been underused. After briefly reviewing qualitative research to date into experiences of receiving and delivering breastfeeding support, we discuss the potential of theoretically informed qualitative methodologies to provide fuller understanding of intervention processes by focusing on three examples: phenomenology, ethnography, and discourse analysis. The paper concludes by noting some of the epistemological differences between the broadly positivist approach of trials and qualitative methodologies, and we suggest there is a need for further dialog as to how researchers might bridge these differences in order to develop a fuller and more holistic understanding of how best to support breastfeeding women.
Keywords: breastfeeding support, discourse analysis, ethnography, phenomenology, process evaluation, qualitative research
1. INTRODUCTION
Interventions to support breastfeeding are many and varied, with evidence that some forms of support can increase the duration of breastfeeding (Renfrew, McCormick, Wade, Quinn, & Dowswell, 2012; Sinha et al., 2015), though not always (Hoddinott, Seyara, & Marais, 2011). The focus of intervention studies has usually been on measuring outcomes, using randomized controlled trials (RCTs). However, RCTs usually focus only on establishing causal relationships between interventions and outcomes and rarely illuminate the nature of the relationship and how interventions might contribute to outcome (Elliot, 2010; Pawson & Tilley, 1997; Thomson & Trickey, 2013). Therefore, less is known about the interpersonal and organizational processes that may be key to successful breastfeeding interventions, and how interventions to support breastfeeding are experienced by women. In this paper, we discuss the limitations of evidence gained from measuring large‐scale outcomes in RCTs and argue that greater use of qualitative research can enhance understanding of key processes in supporting breastfeeding. Our focus is primarily on high‐income countries as this is where the majority of the English language research on support for breastfeeding has been conducted. However, several of our arguments, particularly about the context‐dependent nature of most interventions, have broader relevance.
Key messages.
Supporting women with breastfeeding is a complex process and requires sensitive interaction and attention to contextual issues.
Women are likely to make sense of and experience breastfeeding support interventions differently, depending on their cultural context and prior experiences and preferences.
Research into the effectiveness and helpfulness of support interventions for breastfeeding has tended to focus on measuring outcomes, and there is a need for more research into intrapersonal, relational, and organizational processes involved in delivering interventions.
Qualitative research methods are ideally suited to examining intervention processes but have been underused in evaluations of breastfeeding support interventions.
2. THE COMPLEXITIES OF SUPPORTING BREASTFEEDING
Due to continued concern that rates of breastfeeding, particularly beyond the early weeks in developed countries, do not meet the World Health Organization and UNICEF's (2003) recommendations (WHO, 2015), interventions to support breastfeeding have proliferated. They may be delivered antenatally, postnatally, or both, and may focus on organizational change and staff training or on direct support to mothers. Some, such as the Baby Friendly Initiative, may involve all or several of these (UNICEF Baby Friendly Initiative UK, 2012). Direct support, the focus of this paper, can vary considerably, hence the concern with researching best practice. Support comes in a variety of guises and may be offered by health professionals or volunteers (e.g., peer supporters); via group, individual, or family consultations; by phone, face to face, or via written information; over shorter or longer time periods; in community or healthcare settings; proactively or reactively; and may or may not be “hands‐on.” Training given to supporters also varies and whether the emphasis is on information provision, technical support, or emotional support. The breastfeeding challenges addressed by support interventions can include expectations, positioning at the breast, pain, concerns about milk supply, and strategies for managing others' responses to breastfeeding (see Feldman‐Winter, 2013 for overview of support interventions for breastfeeding).
This range of interventions is not surprising given the complexities of breastfeeding (Hoddinott et al., 2011). Infant feeding practices have different meanings in different contexts (Burns, Schmied, Sheehan, & Fenwick, 2010), and some mothers may struggle to manage perceived tensions between breastfeeding and the needs, expectations, and comfort of others (Leeming, Williamson, Lyttle, & Johnson, 2013). Moreover, an extensive literature demonstrates how frequent framing of infant feeding decisions in moral terms can leave mothers, certainly in high‐income countries, feeling potentially judged and obliged to account for themselves to others, including infant‐feeding supporters (e.g., Miller, Bonas, & Dixon‐Woods, 2007; Ryan, Bissell, & Alexander, 2010; Thomson, Ebisch‐Burton, & Flacking, 2015). This may be particularly difficult where women interpret unanticipated difficulties with breastfeeding as failure at something that should be “natural” (Larsen, Hall, & Aargaard, 2008; Williamson, Leeming, Lyttle, & Johnson, 2012). Therefore, those who support mothers face dilemmas such as how to promote the health benefits of breastfeeding whilst not undermining the maternal identity of women who formula‐feed (Trickey & Newburn, 2014). As such, breastfeeding support needs to be considered a complex intervention (Thomson & Trickey, 2013), involving relational aspects that warrant close attention by researchers, including the language used and messages conveyed. These issues are not usually directly examined by outcome studies and, as Thomson and Trickey (2013) suggest, might be addressed more fully by wider use of qualitative methodologies.
3. THE OUTCOME RESEARCH PARADIGM AND ITS LIMITATIONS
RCTs are generally considered the gold standard for assessing the effectiveness of healthcare interventions (Moore et al., 2015), based on the assumption that the most useful data about interventions are outcomes (such as rates of initiation, exclusivity, and maintenance of breastfeeding) for those who receive a particular intervention, compared to outcomes for those who do not. A further assumption is that where participants are randomly allocated to intervention or control from a sufficiently large and representative sample confounding factors will be eliminated and findings will be generalizable to a wider population (Walach & Loef, 2015). Therefore, the results of RCTs are attractive to policy‐makers and budget holders (Shaw, Larkin, & Flowers, 2014), where the aim is not to explain the processes through which interventions may or may not facilitate change, but instead to predict the likely effectiveness of an intervention across a population.
Systematic reviews of findings from RCTs have drawn a number of conclusions about the effectiveness of support for breastfeeding: for example, that interventions within primary‐care settings that combine antenatal and postnatal elements are more likely to lead to higher rates of breastfeeding (Chung, Raman, Trikalinos, Lau, & Ip, 2008; Patnode, Henninger, Senger, Perdue, & Whitlock, 2016); that peer support may be of more value in low‐income and middle‐income countries than in higher income countries with routine postnatal care (Jolly et al., 2012); and that face‐to‐face support may be of more value than telephone support (Renfrew et al., 2012). However, there are limits to the applicability of these findings. RCTs generally proceed as if the particular intervention being trialed is a homogenous entity, which is delivered similarly across participants, and may ignore variation in implementation (Thomson & Trickey, 2013). Assumptions of homogeneity also gloss over the way in which breastfeeding support is an interaction between individuals that is not as easily defined as, say, a drug intervention. Despite increased emphasis on assuring and assessing implementation fidelity in trials of public health interventions (e.g., Moore et al., 2015), support and guidance are an intersubjective phenomenon that is likely to be experienced differently by participants, particularly in different cultural and subcultural contexts (Shaw et al., 2014). Those using healthcare services are not passive recipients but are active in negotiating and influencing the nature of the intervention (Pawson, Greenhalgh, Harvey, & Walshe, 2004) and may take up or resist aspects of an intervention in varied ways. Skilled breastfeeding supporters are therefore likely to adapt their support to the needs of different clients in ways that may be difficult to specify at the outset of a planned intervention. This suggests the value of researching how standardized interventions may be delivered and experienced differently, especially where the intervention is delivered in cultural settings that differ considerably from those in which the intervention was developed.
Assumptions about homogeneity of interventions become even more questionable where systematic reviews combine findings from several RCTs into categories of intervention and compare these. For example, although face‐to‐face support was found by Renfrew et al. (2012) to be more effective than telephone support, the authors advocate caution interpreting this finding as the two forms of communication can vary considerably. The appropriateness of each may depend on factors such as prior relationships, support needs, practicalities such as transport, and whether phone contact increases accessibility of support out‐of‐hours. Therefore, although aggregated outcome data may be of value for making decisions at a population level, further research may be needed to help practitioners understand how interventions might be received by different mothers and when and where different aspects of support might be more or less helpful.
Outcome data tell us only about the end point and not how it was reached. The story of how a mother sustained breastfeeding (or did not), of her relationships with those supporting her and others, and how she responded to the different components of support, may tell us more about how to support women in similar situations than measuring the beginning and end points. Furthermore, a focus on outcome measurement requires researchers to define desired outcomes, when these may be valued differently by different stakeholders (Sanders, Egger, Donovan, Tallon, & Frankel, 1998). In the case of breastfeeding support, whilst there may be broad agreement among service providers about the desirability of outcomes such as increased duration or exclusivity of breastfeeding, interventions may achieve more than this. Women may value approval and validation of their mothering skills (Dykes, Hall, Burt, & Edwards, 2003; Sheehan, Schmied, & Barclay, 2009). Becoming a “good mother” with a contented, thriving baby and family seem more important outcomes for many mothers than breastfeeding per se (Hoddinott, Britten, & McInnes, 2012; Marshall, Godfrey, & Renfrew, 2007). And, within a Western cultural context where breastfeeding might sometimes be seen as embarrassing or even disgusting (Dowling, Naidoo, & Pontin, 2012), feeling good about breastfeeding may also be an important outcome of a support intervention. Therefore, it is useful for evaluation studies to incorporate open‐ended exploration of participants' views on the outcomes that were valuable for them and how these were achieved, in addition to measuring achievement of outcomes specified by researchers. It might also be useful to explore mothers' reasons for not engaging with aspects of an intervention, and whether this was because these aspects worked against outcomes that were important to them. For example, Hunt and Thomson (2016) found that expectations of judgment and anticipation of a rules‐based approach discouraged mothers from engaging with peer support.
4. INVESTIGATING PROCESS IN BREASTFEEDING SUPPORT
Within healthcare research, there is now greater recognition of the need to understand intervention processes as well as outcomes. For example, the UK Medical Research Council recommends collecting data as part of RCTs that illuminate how interventions achieve their effects, why they fail, and how interventions can be optimized (Craig et al., 2006) and has recently developed guidance for process evaluation (Moore et al., 2015). This follows the development of “realist evaluation” (Pawson & Tilley, 1997), which aims to explain mechanisms by which healthcare interventions achieve their effects in particular contexts and places particular value on theoretical understanding of how complex social processes influence outcome. Although process evaluation can include objective measurement of predicted components of change mechanisms, a crucial element is exploring the subjective experience and perspectives of stakeholders. However, as Renfrew et al. (2012) note, with a few exceptions, the perspectives of stakeholders (e.g., mothers) on intervention processes have so far received limited attention within trials of breastfeeding support interventions. These perspectives may be extremely useful in explaining varying trial outcomes (Hoddinott et al., 2011).
Despite the relative absence of stakeholder perspectives within trials, there is growing emphasis elsewhere on understanding women's breastfeeding experiences from their own perspective (see Afoakwah, Smyth, & Lavender, 2013; Burns et al., 2010 for reviews). This includes understanding experiences of standard postnatal support from healthcare professionals and peer supporters. Two reviews of English language research prior to 2007, focusing on mothers' experiences of breastfeeding support in mostly high‐income countries (McInnes & Chambers, 2008; Schmied, Beake, Sheehan, McCourt, & Dykes, 2011) and several studies since, have identified factors related to whether or not support offered is considered helpful. For example, the use of inflexible, directive, and standardized advice, particularly if there are any inconsistencies in this, seems particularly problematic (McInnes & Chambers, 2008; Schmied et al., 2011), whereas support that is viewed by mothers as mother‐centered and responsive to their needs appears to be strongly valued, especially if it facilitates mothers' own decision‐making (Hoddinott et al., 2012; McInnes & Chambers, 2008; Schmied et al., 2011; Sheehan et al., 2009). The perceived quality and authenticity of the relationship with supporters seem important (McInnes & Chambers, 2008; Schmied et al., 2011), and for some research participants, any sense of exposure, critical judgment, or surveillance by health professionals seems particularly difficult (Leeming, Williamson, Johnson, & Lyttle, 2015; McInnes & Chambers, 2008; Palmér, Carlsson, Mollberg, & Nystrom, 2012; Sheehan et al., 2009). Mothers view useful interventions as those that build confidence in breastfeeding and self‐esteem as a mother, as well as helping with technical aspects of feeding and providing realistic discussion of potential breastfeeding difficulties (Bäckström, Hertfelt Wahn, & Ekström, 2010; Hoddinott et al., 2012; Schmied et al., 2011; Sheehan et al., 2009). Investigation of the perspectives of those supporting breastfeeding has highlighted some of the challenges of offering effective support including: maintaining consistency whilst responding to women's individual needs, choices, vulnerabilities, and concerns; time restrictions; poor continuity of care; inappropriate professional norms; and inaccessibility of training (e.g., Bäckström et al., 2010; McInnes & Chambers, 2008; Nelson, 2007; Tennant, Wallace, & Law, 2006). However, this literature on mothers' and supporters' perspectives is not well integrated with outcome studies. Much of the research above has explored experiences of varied standard care, rather than trials of particular interventions, and it is not always clear how breastfeeding support was delivered and structured. Therefore, it is difficult to know how factors outlined above might have impacted women's responses to breastfeeding interventions evaluated in trials.
Some trials of breastfeeding interventions have sought participants' perspectives, often using postintervention questionnaires which include either likert‐scale or open‐ended questions (e.g., Ekström, Widström & Nissen, 2006; Graffy & Taylor, 2005). However, although able to provide information about the acceptability of different aspects of an intervention from a sizeable sample, questionnaires provide only a brief retrospective snapshot of process, shaped by the use of a limited number of predetermined questions. There may be important aspects of a woman's experience of an intervention, which she may not frame in terms of the researcher's query. Nor does a questionnaire enable close examination of interactions between those delivering and receiving interventions. A limited number of studies have demonstrated the value of more extensive exploration of how women and supporters engage with a particular intervention. For example, through interviews with mothers, midwives, and peer supporters, Ingram (2013) explored in detail, from varied viewpoints, aspects of breastfeeding peer support that were perceived as helpful and/or challenging to deliver. Using both interview, survey, and observational data, Hoddinott, Chalmers, and Pill (2006) showed how women not only considered potential gains when deciding whether or not to engage with postnatal breastfeeding groups or one‐to‐one peer support but also weighed up perceived risks such as being undermined, pressured, or having their privacy invaded. However, studies such as these that give a central role to participants' experiences when evaluating specific breastfeeding support interventions (e.g., Dennis, 2002; Fox, McMullen, & Newburn, 2015; Memmott & Bonuck, 2006; Raine, 2003; Thomson, Crossland, & Dykes, 2012b) are still the exception rather than the rule.
In order to understand further why some breastfeeding interventions seem more successful than others, some researchers have also begun to explore implementation processes from the perspective of service providers. For example, in the UK, Hoddinott, Britten, and Pill (2010) used interview, observational, and survey data from varied stakeholders to identify aspects of organizations (e.g., leadership and communication) and community context (e.g., social class) that impacted implementation of breastfeeding support groups positively and negatively in different settings. Similarly in the US, Nickel, Taylor, Labbok, Weiner, and Williamson (2013) used interviews and the concept of “organizational readiness to change” to examine staff perceptions of barriers and facilitators when implementing the World Health Organization and UNICEF's (1989) Ten Steps for Successful Breastfeeding. Thomson, Bilson, and Dykes (2012a) collected data via interviews with a range of professionals that demonstrated the importance of credible leadership for implementing a community Baby Friendly Initiative in the UK and showed the ways in which credibility was achieved.
5. THE VALUE OF QUALITATIVE METHODS FOR AN INVESTIGATING PROCESS
Much of the research discussed in the section above is qualitative. Although quantitative data may provide useful information about interpersonal and organizational processes in interventions (Moore et al., 2015), qualitative methodologies are particularly suited to this because of their concern with understanding, describing, and interpreting the nature of psychosocial phenomena. Qualitative research aims to understand “how” rather than test “whether”. It also tends to assume that knowledge is context specific and therefore incorporates methods for attending to the way in which particular psychosocial processes arise in particular contexts. However, collecting brief qualitative data about satisfaction or acceptability within a clinical trial (e.g., via questionnaires), though useful, does not make maximum use of the potential of qualitative research to explore intervention processes. Fuller use of specific qualitative methodologies can provide researchers with alternative theoretical concepts for thinking about the fluid and complex nature of psychosocial aspects of interventions (e.g., “discourse,” “narrative,” “lifeworld,” “intersubjectivity,” and “embodiment”) so that understanding of intervention processes is not restricted by reliance on mechanistic assumptions about linear cause and effect. There are now several qualitative methodologies in use in healthcare research, each with particular aims and theoretical frameworks. Below, we briefly discuss three of these (phenomenology, ethnography, and discourse analysis), and how they could guide the researcher towards more detailed exploration of what goes on between people, and how this might be experienced, when practitioners implement breastfeeding interventions.
6. PHENOMENOLOGY
6.1. Overview
Phenomenological research explores detailed first‐person accounts in order to understand lived experience holistically and in some depth (Langdridge, 2007). Phenomenology rejects the idea of examining a phenomenon such as an infant feeding intervention objectively, as if it is a fixed entity. Instead, the aim is to understand how this phenomenon appears in the consciousness of those engaging with it (Brooks, 2015). Therefore, there is a commitment to participant‐led methods of collecting detailed data (e.g., unstructured or semistructured interviews and open‐ended diaries) and an inductive rather than theory‐led approach to analysis (Howitt, 2016). Some phenomenological methodologies (e.g., Giorgi, 2009) aim for rich description of participants' experiences as lived prereflectively, whereas other phenomenological approaches (e.g., Smith, Flowers, & Larkin, 2009) argue for the importance of interpretation and are concerned with how participants make sense of meaningful personal experiences, sometimes relating this to broader contextual issues (e.g., Langdridge, 2007), or to phenomenological ideas about the nature of the lifeworld (lived experience) such as temporality or embodiment (Davidsen, 2013).
6.2. Use of phenomenology in breastfeeding research
Phenomenological methods are becoming more established within maternal and child health research (e.g., Thomson, Dykes, & Downe, 2011) and have highlighted important features of the lived experience of breastfeeding; for example, that breastfeeding may be experienced as confirming or challenging maternal identity (Palmér, Carlsson, Mollberg, & Nyström, 2010, Palmér et al., 2012, Palmér, Carlsson, Brunt, & Nyström, 2014; Williamson et al., 2012), as requiring determination (e.g., Bottorff, 1990; Hauck, Langton, & Coyle, 2002; Palmér et al., 2010), and as an activity scrutinized by others (McBride‐Henry, 2010; Palmér et al., 2012, Palmér, Carlsson, Brunt, & Nyström, 2015). Phenomenological research has also drawn attention to the shifting ways in which the breastfeeding body can be experienced (Dykes & Williams, 1999; McBride‐Henry, White, & Benn, 2009; Palmér et al., 2010, 2015; Ryan, Todres, & Alexander, 2011) and the complexity of emotional responses to breastfeeding (e.g., Guyer, Millward, & Berger, 2012; Mozingo, Davis, Droppleman, & Merideth, 2000; Palmér et al., 2012, Palmér et al., 2015).
6.3. Value of phenomenology for investigating breastfeeding support interventions
Phenomenological studies have explored relations with breastfeeding supporters as part of the intersubjective experience of breastfeeding (e.g., Hauck et al., 2002; Guyer et al., 2012; Mozingo et al., 2000; Palmér et al., 2012) and have explored the meaning and experience for practitioners of challenges in providing support to breastfeeding women (e.g., Nelson, 2007). However, the phenomenological approach has been less visible in studies explicitly investigating novel interventions to support breastfeeding. This is not surprising given the tension between the emphasis on researcher control within trials and phenomenology's aim to explore subjective experience from the participant's perspective in an open‐ended manner. Using a phenomenological lens to research a breastfeeding intervention means shifting the research question from the researcher's concern with “does it work?” or even “how does it work?” to “what was it like to take part and what did the experience mean?”
Exploring retrospective accounts of any one group (e.g., intervention recipients or providers) is best considered as only one line of enquiry (Elliot, 2010). Individuals may not be able to articulate some aspects of their experience or may have an emotional investment in particular interpretations. Moreover, a fuller understanding of interpersonal and social processes in supporting breastfeeding women might be gained by the addition of ethnographic and discursive methods. However, we would argue that ignoring the lived experience of participating in an intervention places significant limitations on understanding what has taken place.
7. ETHNOGRAPHY
7.1. Overview
Ethnography focuses on cultural interpretation of a setting or interaction. Detailed descriptions of situations are used to understand implicit and explicit meanings and relevant cultural constructions within the area of study (Dykes & Flacking, 2016). Although there is no clear accepted definition of ethnography, it usually involves the researcher participating in people's daily lives to conduct an in‐depth study of accounts and actions within their everyday context (Hammersley & Atkinson, 2007). Methods of data generation are particularly sensitive to social context, and researchers attempt to view the world through the eyes of the participants to understand how their social worlds are interpreted and experienced (Mason, 2002; Miles, Hubermann, & Saldana, 2014). Data are usually unstructured rather than predetermined and can be generated from a range of sources, but these often involve observation and informal conversations (Hammersley & Atkinson, 2007). Participant observation involves the researcher immersing himself or herself in the research setting and systematically observing dimensions of that setting, such as interactions, relationships, actions, and events (Lofland, Snow, Anderson, & Lofland, 2006; Mason, 2002).
7.2. Use of ethnography in breastfeeding research
This emphasis on collecting data in “natural” everyday settings and the open‐ended exploratory nature of ethnography means it has (like phenomenology and discourse analysis) mostly been used for examining existing care practices and “treatment‐as‐usual” relations between women and healthcare professionals. Examples include studies observing community midwives and health visitors (Marshall & Godfrey, 2011; Marshall et al., 2007) and practices on postnatal wards (Dykes, 2005a, 2005b) in Northern England, and nursing practices in a neonatal unit in North America (Cricco‐Lizza, 2011, 2014, 2016). The detailed understanding that comes from such studies can provide useful insights that inform breastfeeding support such as the way in which breastfeeding involves labor‐intensive care‐work (Dykes, 2005a) and the complexity of the different roles women attempt to fulfill whilst breastfeeding—striving to maintain their identity as good mothers, within differing perceptions of this, whilst also maintaining identities as wives and workers (Marshall et al., 2007). Cricco‐Lizza's (2014) ethnography demonstrated the complexity of nurses' emotions and the largely unrecognized emotional labor required to support women in a busy, stressful neonatal intensive care unit, highlighting the range of coping strategies used.
7.3. Value of ethnography for investigating breastfeeding support interventions
Ethnography is less commonly used for exploring novel, more structured interventions that are being studied within RCTs. However, Hoddinott et al. (2010) used a realist version of ethnography within a cluster RCT to produce an explanatory model for variability in the implementation and effectiveness of breastfeeding support groups. This demonstrates ethnography's potential for understanding the varying and complex systems and cultures within which interventions are implemented. Additionally, Young and Pelto (2016) suggest that formative ethnographic research can enhance understanding of key cultural features of a local context prior to intervention.
There are a number of challenges to using more traditional versions of ethnography as part of clinical trials because ethnography has its roots in naturalistic enquiry that involves the researcher studying actions and accounts within an existing context (Savage, 2006). The very nature of an intervention means that this has been disrupted or changed. However, overcoming these challenges and finding ways to use ethnographic methods alongside trials would enable better understanding of the meaning and impact of interventions, of what goes on between those delivering and receiving the intervention, and of the social context within which the intervention is being implemented (Tumilowicz, Neufeld, & Pelto, 2015).
8. DISCOURSE ANALYSIS
8.1. Overview
Discourse analytic approaches are generally informed by a social constructionist perspective that demands an alternative stance on language to much of the realist perspective pervasive through health promotion research. For discourse analysts, participants' language is more than just a way for individuals to describe and communicate internal thoughts or feelings. Instead, language is used in varied ways to construct different versions of events and social realities (Burr, 2015). Access by researchers to the way in which language constructs versions of social “reality” is invariably gained through a discourse analysis of talk and text. Although there are varying strands of discourse analysis (e.g., discursive psychology and Foucauldian discourse analysis), the overall concern is with the ways in which language and discourse constitute versions of our social world. From this perspective, a discourse may be thought of as a set of linguistic resources that constructs a particular version of something in the social world, such as an object, event, or category of person. For example, depending on the language used, breastfeeding can be made to appear as a lifestyle choice or an uncontested, normative aspect of mothering. A central consideration of the discursive approach is that language is performative in that it holds a function for the speaker or author of the discourse. Discourse analysis is applicable to different methods of data collection from open‐ended interviews to naturalistic data and analysis of written text (policy documents, etc.). As such, it can be combined with ethnographic methods.
8.2. Use of discourse analysis in breastfeeding research
Discourse analysis is relevant for breastfeeding research in its focus on the discourses and constructions that are inherent in health promotion around breastfeeding and mothers' subjectivities in relation to infant feeding. It is therefore of no surprise that many of the discourse analysis papers in breastfeeding research look at the ways in which breastfeeding is constructed through health promotion materials and interactions. Many of these have noted the moral discourses that work through much breastfeeding promotion and discussion of breastfeeding practice (Ryan et al., 2010; Wall, 2001). Discourse analysis has been used as a methodology to explore discourses around managing breastfeeding (Payne & Nicholls, 2010), constructions of the breastfeeding infant (Burns, Fenwick, Sheehan, & Schmied, 2016), and managing decisions around infant feeding in particular populations (Hoddinott & Pill, 2000).
A stream of research has focused on the discourses inherent in the teaching and promoting of breastfeeding in prenatal classes. For example, Locke (2009) examined the two competing discourses of prenatal breastfeeding classes—the seemingly oxymoronic construction of breastfeeding as the “natural” skill that also needed to be “taught.” Similarly, research by Fenwick, Burns, Sheehan, and Schmied (2013) revealed that the language and practices of midwives were often limited to convincing women to breastfeed rather than engaging with them in conversations that facilitated exploration and discovery of how breastfeeding might be experienced within the mother–infant relationship and broader social or cultural context. Burns, Schmied, Fenwick, and Sheehan (2012) have also examined how midwives construct breastfeeding in interactions with breastfeeding mothers, noting the discourse of breast milk as “liquid gold.” They argue that this discourse privileges the nutritional aspects of breast milk over both the practice of breastfeeding and the support needs of the breastfeeding mother.
8.3. Value of discourse analysis for investigating breastfeeding support interventions
Although not commonly used for exploring the structured interventions evaluated by RCTs, discourse analysis provides tools for reaching a deeper understanding of what varying support interventions communicate about infant feeding and how they do this. Fine‐grained analysis of the interactions within an intervention can show how the promotion and support of breastfeeding is being negotiated and performed in practice. Discourse analysis is also readily applicable to both deconstruct and assist with the design of information sheets and health promotion materials. By mapping the key discourses around breastfeeding for a variety of demographics, discourse analysis offers the opportunity for tailored interventions that pay close attention to the language used by practitioners.
9. CONCLUDING COMMENTS
Understanding how women can be best supported to breastfeed is a priority for researchers concerned with maternal and child nutrition, and the past couple of decades have seen an explosion of research in this area. However, we suggest that research is still constrained by a limited range of research questions, methodologies, and theoretical standpoints. Researchers could more frequently be asking questions about the lived experience of participating in and implementing different forms of breastfeeding support, the meanings of interactions intended to be supportive from the perspectives of different stakeholders, and the different versions of infant feeding that are talked into being by the language used in particular promotional and supportive interventions. Research in the neighboring field of psychotherapy effectiveness has long benefitted from process research, using a range of qualitative and quantitative methods to examine, for example, significant events in therapy, aspects of therapeutic relationships, varying therapist practices, and interactional sequences (see Hardy & Llewelyn, 2015 for an overview). It is our hope that similar advances can be made in understanding interpersonal processes in breastfeeding support interventions. We would endorse the recent guidelines within the UK for examining process within public health interventions (e.g., Moore et al., 2015) but argue that this should not be limited to measuring theoretical change mechanisms hypothesized by researchers. Use of qualitative methodologies and engagement with participants' meanings enable unanticipated insights into the impact of different aspects of an intervention that may challenge the original theoretical concepts underlying an intervention.
However, it is important to recognize some of the challenges of expanding process research by greater use of qualitative research. In particular, qualitative methodologies are unlikely to share the positivist assumptions of outcome research. Their strengths arise from being more in tune with interpretivist epistemologies, and there is a danger that if qualitative research is “shoehorned” into an RCT, then the full potential of an alternative epistemological position is lost (Morse, 2005; Savage, 2006). For example, if the aim of a trial is to establish decontextualized and generalizable “truths” regarding the effectiveness of a breastfeeding intervention, then it is unlikely that any qualitative data collected will be used to illuminate ways in which the intervention may have very different meanings for participants in different contexts. Epistemological differences have perhaps contributed to the tendency for separate streams of research focusing, on the one hand, on women and practitioners' experiences and practices in relation to nonspecified breastfeeding support using qualitative methods and, on the other hand, trials of specific support interventions using largely quantitative measurement of outcomes. However, this risks the formation of research “silos” where a more holistic understanding of the usefulness of support interventions and variations in how they are received and enacted is foregone in favor of retaining methodological consistency.
Integrating process and outcome research and quantitative and qualitative findings, either from one study or several, remains a tricky issue requiring further discussion (Moore et al., 2015; Shaw et al., 2014). Some breastfeeding support researchers have drawn on a realist approach to evaluation in order to either embed qualitative process enquiry alongside an RCT in a single project (Hoddinott et al., 2010) or as a suggested framework for synthesizing quantitative and qualitative findings across several studies (Thomson & Trickey, 2013). Such approaches are particularly useful for contextualizing the findings of RCTs in a manner that develops theoretical understanding, and the use of an overarching theoretical framework such as realist synthesis can enable more coherent integration of research findings. However, we also need to be careful that the unique value of different methodological approaches is not lost if subsumed within a theoretical framework that may not quite fit. For example, the inductive approach of methodologies such as phenomenology can be valuable as an alternative to more theory‐driven realist synthesis, and much of the value of discourse analysis lies in its recognition that participants' talk does not simply represent and illuminate mechanisms of change (as might be the case in a realist synthesis) but is in itself constitutive of change. Therefore, researchers might alternatively consider how to draw together disparate findings in ways that embrace differences in methodological, epistemological, and ontological assumptions. For example, they might treat these differences as a set of interchangeable lenses through which the world can be viewed tentatively. As such, it becomes possible, even within one research program, to see what might be gained by tentatively viewing a breastfeeding intervention through a “cause‐and‐effect” lens, trying to draw general conclusions that might be applicable in other settings too, before then using a more interpretivist lens to explore the nuances of meaning, language and interaction, and the way in which the intervention may in fact be a slightly different phenomenon to different participants in different contexts. This, though, requires reviewers and funders, as well as researchers, to be open not just to a variety of methodological approaches but also to broader definitions of “good” research and to recognize that criteria for quantitative research may not be appropriate for qualitative research. There is continued debate about quality criteria for qualitative healthcare research (e.g., Dixon‐Woods et al., 2007; Walsh & Downe, 2006), which is not surprising given the different aims and assumptions of different qualitative approaches. However, one criterion in deciding the value of research into breastfeeding support ought to be whether or not the findings will help practitioners to recognize the varied ways in which women might respond to their attempts to support breastfeeding and to adjust their interventions accordingly.
SOURCE OF FUNDING
There was no external funding for the paper.
CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.
CONTRIBUTIONS
DL developed the initial ideas for the paper and led the development of the manuscript. All authors contributed to writing and to further development of ideas and read and approved the final submission.
Leeming D, Marshall J, Locke A. Understanding process and context in breastfeeding support interventions: The potential of qualitative research. Matern Child Nutr. 2017;13:e12407 10.1111/mcn.12407
REFERENCES
- Afoakwah, G. , Smyth, R. , & Lavender, T. (2013). Women's experiences of breastfeeding: A narrative review of qualitative studies. African Journal of Midwifery and Women's Health, 7, 71–77. [Google Scholar]
- Bäckström, C. A. , Hertfelt Wahn, E. I. , & Ekström, A. C. (2010). Two sides of breastfeeding support: Experiences of women and midwives. International Breastfeeding Journal, 5, 20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bottorff, J. L. (1990). Persistence in breastfeeding: A phenomenological investigation. Journal of Advanced Nursing, 15, 201–209. [DOI] [PubMed] [Google Scholar]
- Brooks, J. (2015). Learning from the lifeworld: Introducing alternative approaches to phenomenology in psychology. The Psychologist, 28, 642–643. [Google Scholar]
- Burns, E. , Fenwick, J. , Sheehan, A. , & Schmied, V. (2016). ‘This little piranha’: A qualitative analysis of the language used by health professionals and mothers to describe infant behaviour during breastfeeding. Maternal & Child Nutrition, 12, 111–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burns, E. , Schmied, V. , Fenwick, J. , & Sheehan, A. (2012). Liquid gold from the milkbar: Constructions of breastmilk and breastfeeding women in the language and practices of midwives. Social Science and Medicine, 75, 1737–1745. [DOI] [PubMed] [Google Scholar]
- Burns, E. , Schmied, V. , Sheehan, A. , & Fenwick, J. (2010). A meta‐ethnographic synthesis of women's experience of breastfeeding. Maternal & Child Nutrition, 6, 201–219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burr, V. (2015). Social constructionism (Third ed.). London: Routledge. [Google Scholar]
- Chung, M. , Raman, G. , Trikalinos, T. , Lau, J. , & Ip, S. (2008). Interventions in primary care to promote breastfeeding: An evidence review for the U.S. Preventive Services Task Force. Annals of Internal Medicine, 149, 565–582. [DOI] [PubMed] [Google Scholar]
- Craig, P. , Dieppe, P. , Macintyre, S. , Michie, S. , Nazareth, I. , & Petticrew, M. (2006). Developing and evaluating complex interventions: new guidance. Medical Research Council.Retrieved from http://www.mrc.ac.uk/documents/pdf/complex-interventions-guidance/.
- Cricco‐Lizza, R. (2011). Everyday nursing practice values in the NICU and their reflection on breastfeeding promotion. Qualitative Health Research, 21, 399–409. [DOI] [PubMed] [Google Scholar]
- Cricco‐Lizza, R. (2014). The need to nurse the nurse: Emotional labor in neonatal intensive care. Qualitative Health Research, 24, 615–628. [DOI] [PubMed] [Google Scholar]
- Cricco‐Lizza, R. (2016). Infant feeding beliefs and day‐to‐day feeding practices of NICU nurses. Journal of Pediatric Nursing, 31, e91–e98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Davidsen, A. S. (2013). Phenomenological approaches in psychology and health sciences. Qualitative Research in Psychology, 10, 318–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dennis, C. L. (2002). Breastfeeding peer support: Maternal and volunteer perceptions from a randomized controlled trial. Birth, 29, 169–176. [DOI] [PubMed] [Google Scholar]
- Dixon‐Woods, M. , Sutton, A. , Shaw, R. , Miller, T. , Smith, J. , Young, B. , … Jones, D. (2007). Appraising qualitative research for inclusion in systematic reviews: A quantitative and qualitative comparison of three methods. Journal of Health Services Research & Policy, 12, 42–47. [DOI] [PubMed] [Google Scholar]
- Dowling, S. , Naidoo, J. , & Pontin, D. (2012). Breastfeeding in public: Women's bodies, women's milk In Hall Smith P., Hausman B. L., & Labbok M. (Eds.), Beyond health, beyond choice: Breastfeeding constraints and realities). Chapel Hill: Rutgers Press. [Google Scholar]
- Dykes, F. (2005a). ‘Supply’ and ‘demand’: Breastfeeding as labour. Social Science and Medicine, 60, 2283–2293. [DOI] [PubMed] [Google Scholar]
- Dykes, F. (2005b). A critical ethnographic study of encounters between midwives and breast‐feeding women in postnatal wards in England. Midwifery, 21, 241–252. [DOI] [PubMed] [Google Scholar]
- Dykes, F. , & Flacking, R. (2016). Introducing the theory and practice of ethnography In Dykes F., & Flacking R. (Eds.), Ethnographic research in maternal and child health). Abingdon: Routledge. [Google Scholar]
- Dykes, F. , Hall, M. V. , Burt, S. , & Edwards, J. (2003). Adolescent mothers and breastfeeding: Experiences and support needs—an exploratory study. Journal of Human Lactation, 19, 391–400. [DOI] [PubMed] [Google Scholar]
- Dykes, F. , & Williams, C. (1999). Falling by the wayside: A phenomenological exploration of perceived breast‐milk inadequacy in lactating women. Midwifery, 15, 232–246. [DOI] [PubMed] [Google Scholar]
- Ekström, A. , Widström, A. , & Nissen, E. (2006). Does continuity of care by well-trained breastfeeding counselors improve a mother's perception of support? BIRTH, 33(2), 123–130. [DOI] [PubMed] [Google Scholar]
- Elliot, R. (2010). Psychotherapy change process research: Realizing the promise. Psychotherapy Research, 20, 123–135. [DOI] [PubMed] [Google Scholar]
- Feldman‐Winter, L. (2013). Evidence‐based interventions to support breastfeeding. Pediatric Clinics of North America, 60, 169–187. [DOI] [PubMed] [Google Scholar]
- Fenwick, J. , Burns, E. , Sheehan, A. , & Schmied, V. (2013). We only talk about breast feeding’: A discourse analysis of infant feeding messages in antenatal group‐based education. Midwifery, 29, 425–433. [DOI] [PubMed] [Google Scholar]
- Fox, R. , McMullen, S. , & Newburn, M. (2015). UK women's experiences of breastfeeding and additional breastfeeding support: A qualitative study of Baby Café services. BMC Pregnancy and Childbirth, 15, 147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Giorgi, A. (2009). The descriptive phenomenological method in psychology. Pittsburgh: Duquesne University Press. [Google Scholar]
- Graffy, J. , & Taylor, J. (2005). What information, advice, and support do women want with breastfeeding? BIRTH, 32(3), 179–186. [DOI] [PubMed] [Google Scholar]
- Guyer, J. , Millward, L. J. , & Berger, I. (2012). Mothers' breastfeeding experiences and implications for professionals. British Journal of Midwifery, 20, 724–733. [Google Scholar]
- Hammersley, M. , & Atkinson, P. (2007). Ethnography: Principles in practice (3rd ed.). Abingdon: Routledge. [Google Scholar]
- Hardy, G. E. , & Llewelyn, S. (2015). Introduction to psychotherapy process research In Gelo O. C. G., Pritz A., & Rieken B. (Eds.), Psychotherapy research: Foundations, process and outcome. (pp. 183–194). New York: Springer. [Google Scholar]
- Hauck, Y. , Langton, D. , & Coyle, K. (2002). The path of determination: Exploring the lived experience of breastfeeding difficulties. Breastfeeding Review, 10, 5–12. [PubMed] [Google Scholar]
- Hoddinott, P. , Britten, J. , & Pill, R. (2010). Why do interventions work in some places and not others: A breastfeeding support group trial. Social Science and Medicine, 70, 769–778. [DOI] [PubMed] [Google Scholar]
- Hoddinott, P. , Chalmers, M. , & Pill, R. (2006). One‐to‐one or group‐based peer support for breastfeeding? Women's perceptions of a breastfeeding peer coaching intervention. Birth, 33, 139–146. [DOI] [PubMed] [Google Scholar]
- Hoddinott, P. , Craig, L. C. A. , Britten, J. , & McInnes, R. M. (2012). A serial qualitative interview study of infant feeding experiences: Idealism meets realism. BMJ Open, 2, e000504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hoddinott, P. , & Pill, R. (2000). A qualitative study of women's views about how health professionals communicate about infant feeding. Health Expectations, 3(4), 224–233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hoddinott, P. , Seyara, R. , & Marais, D. (2011). Global evidence synthesis and UK idiosyncrasy: Why have recent UK trials had no significant effects on breastfeeding rates? Maternal & Child Nutrition, 7, 221–227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Howitt, D. (2016). Introduction to qualitative research methods in psychology. Harlow: Pearson Education Ltd. [Google Scholar]
- Hunt, L. , & Thomson, G. (2016). Pressure and judgement within a dichotomous landscape of infant feeding: A grounded theory study to explore why breastfeeding women do not access peer support provision. Maternal & Child Nutrition. doi: 10.1111/mcn.12279 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ingram, J. (2013). A mixed methods evaluation of peer support in Bristol, UK: Mothers', midwives' and peer supporters' views and the effects on breastfeeding. BMC Pregnancy and Childbirth, 13, 192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jolly, K. , Ingram, L. , Khan, K. S. , Deeks, J. J. , Freemantle, N. , & MacArthur, C. (2012). Systematic review of peer support for breastfeeding continuation: Metaregression analysis of the effect of setting, intensity and timing. BMJ (Clinical research ed.), 344, d8287. [DOI] [PubMed] [Google Scholar]
- Langdridge, D. (2007). Phenomenological psychology: Theory, research and method. Harlow: Prentice Hall. [Google Scholar]
- Larsen, J. S. , Hall, E. O. C. , & Aargaard, H. (2008). Shattered expectations: When mothers' confidence in breastfeeding is undermined—a metasynthesis. Scandinavian Journal of Caring Sciences, 22, 653–661. [DOI] [PubMed] [Google Scholar]
- Leeming, D. , Williamson, I. , Johnson, S. , & Lyttle, S. (2015). Making use of expertise: A qualitative analysis of the experience of breastfeeding support for first‐time mothers. Maternal & Child Nutrition, 11, 687–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leeming, D. , Williamson, I. , Lyttle, S. , & Johnson, S. (2013). Socially sensitive lactation: Exploring the social context of breastfeeding. Psychology and Health, 28, 450–468. [DOI] [PubMed] [Google Scholar]
- Locke, A. (2009). ‘Natural’ versus ‘taught’: Competing discourses in antenatal breastfeeding workshops. Journal of Health Psychology, 14, 435–446. [DOI] [PubMed] [Google Scholar]
- Lofland, J. , Snow, D. , Anderson, L. , & Lofland, L. (2006). Analyzing social settings: A guide to qualitative observation and analysis (4th ed., Vol. 4th). London: Wadsworth. [Google Scholar]
- Marshall, J. L. , & Godfrey, M. (2011). Shifting identities: Social and cultural factors that shape decision making around sustaining breastfeeding In Liamputtong P. (Ed.), Infant feeding practices: A cross‐cultural perspective). New York: Springer. [Google Scholar]
- Marshall, J. L. , Godfrey, M. , & Renfrew, M. J. (2007). Being a ‘good mother’: Managing breastfeeding and merging identities. Social Science and Medicine, 65, 2147–2159. [DOI] [PubMed] [Google Scholar]
- Mason, J. (2002). Qualitative researching (2nd ed.). Sage: London. [Google Scholar]
- McBride‐Henry, K. , White, G. , & Benn, C. (2009). Inherited understandings: The breast as object. Nursing Inquiry, 16, 33–42. [DOI] [PubMed] [Google Scholar]
- McBride‐Henry, K. (2010). The influence of the ‘They’: an interpretation of breastfeeding culture in New Zealand. Qualitative Health Research, 20, 768–777. [DOI] [PubMed] [Google Scholar]
- McInnes, R. J. , & Chambers, J. A. (2008). Supporting breastfeeding mothers: Qualitative synthesis. Journal of Advanced Nursing, 62, 407–427. [DOI] [PubMed] [Google Scholar]
- Memmott, M. M. , & Bonuck, K. A. (2006). Mothers' reactions to a skills‐based breastfeeding promotion intervention. Maternal & Child Nutrition, 2, 40–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miles, M. , Hubermann, M. , & Saldana, J. (2014). Qualitative data analysis: A methods sourcebook (3rd ed.). Sage: London. [Google Scholar]
- Miller, T. , Bonas, S. , & Dixon‐Woods, M. (2007). Qualitative research on breastfeeding in the UK: A narrative review and methodological reflection. Evidence and Policy, 3, 197–230. [Google Scholar]
- Moore, G. F. , Audrey, S. , Barker, M. , Bond, L. , Bonell, C. , Hardeman, W. , … Baird, J. (2015). Process evaluation of complex interventions: Medical Research Council Guidance. BMJ, 350, h1258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morse, J. M. (2005). Beyond the clinical trial: Expanding criteria for evidence. Qualitative Health Research, 15, 3–4. [DOI] [PubMed] [Google Scholar]
- Mozingo, J. N. , Davis, M. W. , Droppleman, P. G. , & Merideth, A. (2000). “It wasn't working”: Women's experiences with short‐term breastfeeding. The American Journal of Maternal Child Nursing, 25, 120–126. [DOI] [PubMed] [Google Scholar]
- Nelson, A. M. (2007). Maternal‐newborn nurses' experiences of inconsistent professional breastfeeding support. Journal of Advanced Nursing, 60, 29–38. [DOI] [PubMed] [Google Scholar]
- Nickel, N. C. , Taylor, E. C. , Labbok, M. H. , Weiner, B. J. , & Williamson, N. E. (2013). Applying organisation theory to understand barriers and facilitators to the implementation of baby‐friendly: A multisite qualitative study. Midwifery, 29, 956–964. [DOI] [PubMed] [Google Scholar]
- Palmér, L. , Carlsson, G. , Brunt, D. , & Nyström, M. (2014). Existential vulnerability can be evoked by severe difficulties with initial breastfeeding: A lifeworld hermeneutical single case study for research on complex breastfeeding phenomena. Breastfeeding Review, 22, 21–32. [PubMed] [Google Scholar]
- Palmér, L. , Carlsson, G. , Brunt, D. , & Nyström, M. (2015). Existential security is a necessary condition for continued breastfeeding despite severe initial difficulties: A lifeworld hermeneutical study. International Breastfeeding Journal, 10, 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Palmér, L. , Carlsson, G. , Mollberg, M. , & Nyström, M. (2010). Breastfeeding: An existential challenge: Women's lived experiences of initiating breastfeeding within the context of early home discharge. International journal of qualitative studies on health and well‐being, 5, 5397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Palmér, L. , Carlsson, G. , Mollberg, M. , & Nystrom, M. (2012). Severe breastfeeding difficulties: Existential lostness as a mother—women's lived experiences of initiating breastfeeding under severe difficulties. International journal of qualitative studies on health and well‐being, 7, 10846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Patnode, C. D. , Henninger, M. L. , Senger, C. A. , Perdue, L. A. , & Whitlock, E. P. (2016). Primary care interventions to support breastfeeding: Updated evidence report and systematic review for the US Preventive Services Task Force. JAMA, 316, 1694–1705. [DOI] [PubMed] [Google Scholar]
- Pawson, R. , Greenhalgh, T. , Harvey, G. , & Walshe, K. (2004). Realist synthesis: An introduction. ESRC Research Methods Programme University of Manchester. [Google Scholar]
- Pawson, T. , & Tilley, N. (1997). Realistic evaluation. London: Sage. [Google Scholar]
- Payne, D. , & Nicholls, D. A. (2010). Managing breastfeeding and work: A Foucauldian secondary analysis. Journal of Advanced Nursing, 66(8), 1810–1818. [DOI] [PubMed] [Google Scholar]
- Raine, P. (2003). Promoting breast‐feeding in a deprived area: The influence of a peer support initiative. Health & Social Care in the Community, 11, 463–469. [DOI] [PubMed] [Google Scholar]
- Renfrew, M. J. , McCormick, F. M. , Wade, A. , Quinn, B. , & Dowswell, T. (2012). Support for health breastfeeding mothers with healthy term babies. Cochrane Database of Systematic Reviews, 16(5) CD001141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ryan, K. , Todres, L. , & Alexander, J. (2011). Calling, permission and fulfilment: The interembodied experience of breastfeeding. Qualitative Health Research, 21, 731–742. [DOI] [PubMed] [Google Scholar]
- Ryan, K. , Bissell, P. , & Alexander, J. (2010). Moral work in women's narratives of breastfeeding. Social Science and Medicine, 70, 951–958. [DOI] [PubMed] [Google Scholar]
- Sanders, C. , Egger, M. , Donovan, J. , Tallon, D. , & Frankel, S. (1998). Reporting on quality of life in randomised controlled trials: Bibliographic study. British Medical Journal, 317, 1191–1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Savage, J. (2006). Ethnographic evidence: The value of applied ethnography in healthcare. Journal of Research in Nursing, 11, 383–393. [Google Scholar]
- Schmied, V. , Beake, S. , Sheehan, A. , McCourt, C. , & Dykes, F. (2011). Women's perceptions and experiences of breastfeeding support: A metasynthesis. Birth, 38, 49–60. [DOI] [PubMed] [Google Scholar]
- Shaw, R. , Larkin, M. , & Flowers, P. (2014). Expanding the evidence within evidence‐based healthcare: Thinking about the context, acceptability and feasibility of interventions. Evidence‐Based Medicine, 19, 201–203. [DOI] [PubMed] [Google Scholar]
- Sheehan, A. , Schmied, V. , & Barclay, L. (2009). Women's experiences of infant feeding support in the first 6 weeks post‐birth. Maternal & Child Nutrition, 5, 138–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sinha, B. , Chowdhury, R. , Sankar, M. J. , Martines, J. , Taneja, S. , Mazumder, S. , … Bhandari, N. (2015). Interventions to improve breastfeeding outcomes: A systematic review and meta‐analysis. Acta Paediatrica, 104, 114–135. [DOI] [PubMed] [Google Scholar]
- Smith, J. , Flowers, P. , & Larkin, M. (2009). Interpretative phenomenological analysis: Theory, method and research. London: Sage. [Google Scholar]
- Tennant, R. , Wallace, L. M. , & Law, S. (2006). Barriers to breastfeeding: A qualitative study of the views of health professionals and lay counsellors. Community Practitioner, 79, 152–156. [PubMed] [Google Scholar]
- Thomson, G. , Bilson, A. , & Dykes, F. (2012a). Implementing the WHO/UNICEF Baby Friendly Initiative in the community: A ‘hearts and minds’ approach. Midwifery, 28, 258–264. [DOI] [PubMed] [Google Scholar]
- Thomson, G. , Crossland, N. , & Dykes, F. (2012b). Giving me hope: Women's reflections on a breastfeeding peer support service. Maternal & Child Nutrition, 8, 340–353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thomson, G. , Dykes, F. , & Downe, S. (2011). Qualitative research in midwifery and childbirth: Phenomenological approaches. London: Routledge. [Google Scholar]
- Thomson, G. , Ebisch‐Burton, K. , & Flacking, R. (2015). Shame if you do – shame if you don't: Women's experiences of infant feeding. Maternal & Child Nutrition, 11, 33–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thomson, G. , & Trickey, H. (2013). What works for breastfeeding peer support: Time to get real. EMJ Gyn Obs, 1, 15–22. [Google Scholar]
- Trickey, H. , & Newburn, M. (2014). Goals, dilemmas and assumptions in infant feeding education and support. Applying theory of constraints thinking tools to develop new priorities for action. Maternal & Child Nutrition, 10, 72–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tumilowicz, A. , Neufeld, L. M. , & Pelto, G. H. (2015). Using ethnography in implementation research to improve nutrition interventions in populations. Maternal & Child Nutrition, 11, 55–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- UNICEF Baby Friendly Initiative UK . (2012). Guide to the Baby Friendly Initiative standards. UNICEF Baby Friendly Initiative UK, London.
- Walach, H. , & Loef, M. (2015). Using a matrix‐analytical approach to synthesizing evidence solved incompatibility problem in the hierarchy of evidence. Journal of Clinical Epidemiology, 68, 1251–1260. [DOI] [PubMed] [Google Scholar]
- Wall, G. (2001). Moral constructions of motherhood in breastfeeding discourse. Gender and Society, 15, 592–610. [Google Scholar]
- Walsh, D. , & Downe, S. (2006). Appraising the quality of qualitative research. Midwifery, 22, 108–119. [DOI] [PubMed] [Google Scholar]
- Williamson, I. , Leeming, D. , Lyttle, S. , & Johnson, S. (2012). ‘It should be the most natural thing in the world’: Exploring first‐time mothers' breastfeeding difficulties in the UK using audio‐diaries and interviews. Maternal & Child Nutrition, 8, 434–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Health Organization and UNICEF . (1989). Protecting, promoting and supporting breast‐feeding: The special role of maternity services. Geneva: WHO Press. [Google Scholar]
- World Health Organization and UNICEF. (2003). Global strategy for infant and young child feeding. Geneva: WHO Press. [Google Scholar]
- World Health Organisation . (2015). Infant and young child feeding by country. http://www.who.int/nutrition/databases/infantfeeding/countries/en/index.html (accessed June 2016).
- Young, S. , & Pelto, G. H. (2016). Evaluative ethnography for maternal and child nutrition interventions In Dykes F., & Flacking R. (Eds.), Ethnographic research in maternal and child health). Abingdon: Routledge. [Google Scholar]