The search identified 877 titles. After removal of duplicates, 452 articles underwent title/abstract screening, and 312 articles were excluded as they did not address exclusive breastfeeding promotion. Thus, 140 full-text articles were assessed for eligibility, and 44 articles were eligible for inclusion. The matching full-text articles were acquired for review. Four articles could not be accessed and were not included in the review. Therefore, 40 articles were selected and included for analysis in the scoping review. Figure 1 (PRISMA flowchart) showed the process of article selection.
Characteristics of the studies: The articles selected for this review varied in the study design and the setting in which the studies were conducted (Table 7). Most of the studies were conducted in United States (n=10), followed by China (n=8), Indonesia (n=4), Iran (n=3), Vietnam (n=3), Australia (n=2), Netherlands (n=2), Egypt (n=1), New Zealand (n=1), Norway (n=1), Turkey (n=1), Malaysia (n=1), Niger (n=1), Thailand (n=1) and Taiwan (n=1). Six studies were published after 2020, 29 studies were published from 2010 – 2019, and five from 2002 - 2009. Study designs included, randomized control trials (RCT; n = 20), quasi-experimental (n = 5), correlational (n = 2), descriptive (n = 5), and mixed methods (n=3).
Almost 9000 mother-child pairs and family pairs participated in the 39 studies. The sociodemographic characteristics were reported in 35 studies. Participants ranged from only mothers (n = 34), mother-infant pairs (n = 3), family (n = 1), and hospitals (n = 2). No study included fathers only or extended family. The setting of the articles ranged widely from the hospital (2), prenatal/maternity clinics (n =29) primary health clinics (6), Local Implementing Agencies (LIAs) (ID-05) (n = 1) and communities (2). Thirty-one studies assessed the prevalence of postpartum exclusive breastfeeding at different time intervals while nine studies suggested measures to promote exclusive breastfeeding. Most studies reported exclusive breastfeeding at the individual level, only three studies reported at the family and hospital levels.
Overview of Theories, Frameworks, and Philosophies
Eighteen theories, six frameworks and two philosophies were extracted. Ajzen’s theory of planned behaviour, Bandura’s theory of self-efficacy and Dennis’ breastfeeding self-efficacy theory were the most used theories in the studies. These theories were also the most effective at increasing exclusive breastfeeding rates. Bandura’s self-efficacy theory and Dennis’ breastfeeding self-efficacy theory increased EBF rates in these studies- 11% vs. 6% (intervention vs. control) (45), 25% vs. 2% (intervention vs. control) (53), 80% vs. 75% (intervention vs. control) (55), 67% vs. 59% (intervention vs. control) (71), 56% vs. 37% (intervention vs. control) (72). Ajzen’s theory of planned behaviour increased EBF rates in these studies- 42% vs. 10% (intervention vs. control) (50), 88% vs. 77% (intervention vs. control) (76). Exclusive breastfeeding rates also increased in other studies that used Titler’s Iowa’s model of evidence-based practice-36% (43), Kotter’s change theory-65% (51) and intervention mapping-48% vs. 27% (intervention vs. control) (57). Most studies used self-efficacy theory for data collection whereas, studies using theory of planned behaviour was used for program content development and implementation.
Self-efficacy and Planned Behaviour Theories
Bandura’s theory of self-efficacy (10,15,26,28) and Dennis’ breastfeeding self-efficacy theory (10,15,26–28) were the most common theories. Self-efficacy is defined as the belief in a person’s ability to organize and accomplish actions required to manage prospective situations (77). Self-efficacy influences thinking and decision-making, effort and persistence, and choice. Thus, breastfeeding self-efficacy is the mother’s perceived ability to breastfeed her infant (42). One of the sources of self-efficacy is information received through verbal persuasion (78). Dennis’ breastfeeding self-efficacy questionnaire in its short form (79) measured breastfeeding self-efficacy in women during pre-partum and/or postpartum and/or assessed the effect of an intervention (usually breastfeeding education) on breastfeeding self-efficacy. Social cognitive theory was used to select suitable educational strategies in breastfeeding program (80). Hence, utilizing the breastfeeding self-efficacy theory, health professionals may be able to influence the practice of breastfeeding by modifying this information (81). In her model of infant feeding behaviours, Lutter recognized the importance of self-efficacy in the achievement of a behaviour. The model posited that infant feeding depends on two factors - the interaction between a woman’s choice to breastfeed and her ability to act upon the choice (self-efficacy). Lutter further described that these factors are influenced by three determinants including proximate, intermediate, and underlying determinants (82). Moussa Abba et al (2010) used the model of infant feeding behaviours to identify determinants of breastfeeding (56).
Theory of planned behaviour (TPB) was used in previous studies to successfully predict dishonest actions (83), leisure behaviours (84), and implement interventions that will be effective to change behaviours (85). The goal of TPB is to predict and explain breastfeeding behaviour. TPB guided the design of study interventions and to predict breastfeeding outcomes (70,74). Whereas other studies used the theory to design questionnaires and explain exclusive breastfeeding behaviour (65). TPB and Reasoned Action Approach (RAA) developed by Fishbein and Ajzen (2010) originated from the theory of reasoned action (TRA). RAA posited that attitude towards behaviour, perceived norm, and perceived behavioural control, determine intention, which predicts behaviour (86). This approach was used to design impact pathway of interpersonal counselling and mass media on breastfeeding practices in Vietnam (59). The Attitude-social influence-self-efficacy model, influenced by TPB, TRA and Bandura’s theory of self-efficacy was used to develop an educational programme (intervention) which increased EBF at 6 months (48). The model, developed by de Vries et al (1988) suggests that attitude, social influence, and self-efficacy determine behavioural intention which in turn predicts behaviour (87,88). Similarly, Information-Motivation-Behavioural-Skills (IMB) model suggested that health-related information, motivation, and behavioural skills are primary determinants of performance of health behaviours (89). It was used to design counselling sessions that focused on enhancing IMB breastfeeding determinants among HIV-infected women to exclusive breastfeeding utilizing motivational interviewing techniques (68).
Mann’s Adolescent decision-making competence theory was used to design developmentally sensitive, education and counselling intervention (69). ADM competence theory suggests that competent decision-making has nine elements including choice, comprehension, creativity, compromise, consequentiality, correctness, credibility, consistency, and commitment (90). The four stages of cognitive development are: sensorimotor, pre-operational, sensory operational, and formal operational.
Breastfeeding self-efficacy theory is limited by the interaction between self-efficacy and previous breastfeeding experience, which may have biased the actual effectiveness of the theory on EBF. McCarter-Spaulding and Gore (2009) reported that breastfeeding self-efficacy scores were higher among mothers who had previous breastfeeding experience (54). Further, in cultures where breast pumps are not accepted or settings where breast pumps are not easily accessed, use of breastfeeding self-efficacy questionnaire may not be appropriate, as it contains an item about using breast pumps. Theory of Planned Behaviour has no standard questionnaire (91), thus there were no unified variables to test the theories in the included studies.
Alianmoghaddam and colleagues used Granovetter’s strength of weak ties theory and Milligan and Wiles’ theory of landscapes of care to explain importance of social relationships, social interactions and social support within virtual communities that are associated with breastfeeding (92). Granovetter posited that individuals’ personal experiences is embedded within the larger social structure beyond the control of some individuals (93). Whereas Milligan and Wiles described landscapes of care as the result of interaction between socio-structural processes and structures that shape experiences and practices of care (94). In addition, Mercer affirmed the significance of social support in her Theory of Maternal Role Attainment. The theory suggested that maternal role attainment is influenced by maternal age, socioeconomic status, perception of birth experience, early mother-infant separation, social stress, social support, personality traits, self-concept, child-rearing attitudes, perception of infant, role strain, and health status (95). Rahayu (2017) used this theory to identify support systems that account for exclusive breastfeeding success (62).
Social Franchise Model was used to design breastfeeding intervention - infant and young child feeding (IYCF) counselling services (58). Social franchise model for IYCF suggested that a franchise facility must provide these services - exclusive breastfeeding promotion, support and management, and complementary feeding education and management (96).
The social theories constructs of strength of weak ties, landscapes of care, and social franchise model cannot be tested using empirical indicators, this restricts their use in quantitative studies with correlational and experimental designs.
Theory for Systems Interventions
Breakthrough Series (BTS) collaborative model was developed by Institute of Healthcare Improvement with the goal to bring large number of hospital teams together to seek improvement in a specific topic or field (97). Arbour et al (2019) used this model to plan a programme - Home Visiting Collaborative Improvement and Innovation Network (HV CoIIN) - which increased breastfeeding initiation and exclusive breastfeeding duration (37). Moreover, Titler’s Iowa’s model of evidence-based practice was developed to empower health professionals to translate research findings into practice to provide quality care (98). Brockman used this model to guide the implementation of a new in-patient model of nursing care - mother-child couplet care - which effectively increased exclusive breastfeeding rates (43). Brockman also used Lewin’s change management model which posits that change occurs in three stages including unfreeze, move/transition, and unfreeze (99). Brockman applied Lewin’s model to manage the complex change processes in the transition from the traditional model of obstetric nursing to mother-child couplet care. Further, the primary role of health professionals is to promote health. Thus, the Health Promotion Model, developed by Pender (1982) will assist professionals to understand health behaviour determinants and empower them to provide quality behavioural counselling (100). Health promotion model was used in included studies to design an intervention - breastfeeding motivation program (44) and explain research findings (73).
Kotter’s change theory was used to initiate culture change for a successful adoption of Baby Friendly Hospital Initiative. This in turn led to an increase in exclusive breastfeeding rates (51). Kotter posited that there are eight steps in the process of change including creating a sense of urgency, forming guiding coalitions, vision development, communicating the vision, removing obstacles and employee empowerment, creating short-term wins, consolidating gains, and strengthening change by anchoring change in the culture (101).
A common weakness of the system intervention theories is their unsuitability to design or evaluate individual-focused interventions.
Green’s Proceed-Precede model was first published as an evaluation framework in 1974 (102), as Precede in 1980 (103) and as a full framework in 1991 (104). Precede-Proceed framework comprises eight phases to guide professionals to develop, implement and evaluate health promotion programmes (105), using socio-ecological model to assess individual characteristics and socio-political conditions (106). Ahmed (2014) used the Precede model to design a five-session breastfeeding educational program which increased exclusive breastfeeding rates (80). The Proceed-Precede model was also used to explain family support factors that promoted exclusive breastfeeding rates (64). The Belief, Attitudes, Subjective Norms and Enabling factors (BASNEF) model, developed by Hubley (1988) originated from Precede model and TRA. It posited that belief, attitude and subjective norms determine behavioural intention, which supports enabling factors for a behaviour (107). The BASNEF model has been used to positively influence nutritional behaviours to reduce risk factors for cardiovascular diseases (108). Ahmadi et al (2016) used BASNEF and GATHER (Greet, ask, tell, help, explain and return) models to guide breastfeeding consultation sessions for the intervention group. The intervention group reported higher rates of exclusive breastfeeding (36).
Nicholson (1990) developed an analytical framework to facilitate adaptation - transition cycle. The cycle consisted of four stages including preparation, encounter, adjustment, and stabilization (109). The stages are useful to enhance readiness, reduce negative emotions, support personal change and role development, and maintain successful adaptation outcomes (110). Transition cycle was used to illustrate and explain mothers’ transition to breastfeeding (46).
Concept maps are tools for organizing and representing knowledge. They illustrate specific label for a concept in a box with lines showing linking words that create a meaningful statement (111). Thepha et al (2019) used concept mapping during all three intervention meetings to provide information and findings regarding identifying and prioritising facilitators and barriers to 6-month exclusive breastfeeding (66). Bartholomew et al (1998) designed intervention mapping as a framework for the development of health education intervention. The framework has five steps: matrix creation, intervention methods selection, program design, identifying adoption and implementation plans, and program evaluation plan generation (112). Intervention mapping was used as a logic model to guide development of educational program which increased exclusive breastfeeding rates (57).
The use of a framework to guide a study is limited as the included frameworks have several stages, but most studies need to implement only a few stages to meet their goals. Thus, limiting the generalizability of the frameworks across studies.
Pragmatism and Phenomenology were the two philosophies used in included studies. Pragmatism is an American philosophy first developed by Charles Pierce. It is a way of doing philosophy, it is concerned with actions (113). Pragmatism evaluates the truth of the meaning of theories in terms of the successful application of those theories. That is, theories are meaningful only if they are practically applicable. Pragmatists subscribe to the notion of instrumentalism because they view theories as instruments for problem solving. Baerug et al (2016) used pragmatism as the basis for their quasi-randomized control trial study (38). In pragmatism, the whole of a concept or phenomenon is found in the consequences of the concept or phenomenon (114). Therefore, Baerug and colleagues applied this in their study as they were interested in the effect (consequence) of baby-friendly community health services on exclusive breastfeeding. The intervention increased exclusive breastfeeding at 6 months. Further, phenomenology is a philosophy developed by Husserl which involves description of lived experience, free from preconceived ideas about the phenomenon. Phenomenology attempts to describe experience from the perspective of the person who had the experience first-hand. Lestari et al (2019) used this philosophy to describe participants’ involvement in exclusive breastfeeding promotion activities (52). In addition, Froehlich and colleagues also used it to qualitatively analyse data collected from participants and to formulate essence descriptors of their breastfeeding experiences and daily routine (46).