Women's and peer supporters' experiences of an assets‐based peer support intervention for increasing breastfeeding initiation and continuation: A qualitative study

Abstract Background and context Breastfeeding peer support is valued by women, but UK trials have not demonstrated efficacy. The ABA feasibility trial offered proactive peer support underpinned by behaviour change theory and an assets‐based approach to women having their first baby, regardless of feeding intention. This paper explores women's and infant feeding helpers' (IFHs) views of the different components of the ABA intervention. Setting and participants Trained IFHs offered 50 women an antenatal meeting to discuss infant feeding and identify community assets in two English sites—one with a paid peer support service and the other volunteer‐led. Postnatally, daily contact was offered for the first 2 weeks, followed by less frequent contact until 5 months. Methods Interviews with 21 women and focus groups/interviews with 13 IFHs were analysed using thematic and framework methods. Results Five themes are reported highlighting that women talked positively about the antenatal meeting, mapping their network of support, receiving proactive contact from their IFH, keeping in touch using text messaging and access to local groups. The face‐to‐face antenatal visit facilitated regular text‐based communication both in pregnancy and in the early weeks after birth. Volunteer IFHs were supportive of and enthusiastic about the intervention, whereas some of the paid IFHs disliked some intervention components and struggled with the distances to travel to participants. Conclusions This proactive community assets‐based approach with a woman‐centred focus was acceptable to women and IFHs and is a promising intervention warranting further research as to its effect on infant feeding outcomes.


| BACKG ROU N D
Peer support is a method of delivering social support to others who share common experiences. Internationally, breastfeeding peer support interventions have been shown to have a significantly greater effect on any and exclusive breastfeeding in low-or middle-income countries compared to high-income countries. 1 While UK randomized controlled trials of breastfeeding peer support have not demonstrated efficacy, policy recommends peer support for socially disadvantaged women. [1][2][3][4] Qualitative studies report that women value peer support and disparities in outcomes may be due to implementation and context. 5,6 Currently, there are a range of breastfeeding peer support programmes (both paid and volunteer) available in the UK. To increase acceptability, effectiveness and inclusiveness, programmes are recommended to be woman-centred (including help with formula and mixed feeding), be offered proactively and focus on the early weeks. [6][7][8][9][10][11][12] The ABA (Assets-based feeding help Before and After birth) intervention was developed and offered within a feasibility randomized controlled trial. It combined proactive peer support underpinned by behaviour change theory, particularly providing social support and restructuring the environment, (COM-B model) 13 with an assets-based approach to women, regardless of their feeding intention. 14 Assets-based approaches focus on positive capabilities of individuals and communities, rather than their needs, deficits and problems. 15 The ABA intervention was delivered by trained infant feeding helpers (IFHs) who offered women an antenatal meeting to discuss infant feeding and help to identify their community assets (including local groups) and used a conversational approach to develop a friends and family tree diagram of infant feeding experiences and potential support (Infant Feeding Genogram 16 (see Figure 1)).
Postnatally, daily contact was offered for the first two weeks after birth, followed by less frequent contact until five months as women wanted, through face-to-face contacts, phone calls and text messages.
The feasibility trial was successful in recruiting primiparous women, including those from areas of socioeconomic disadvantage, with adequate follow-up rates; recruiting and training existing peer supporters to the new ABA role; and delivering the intervention with satisfactory fidelity; and it was acceptable to women, IFHs and maternity services. 17 The proportion of ABA intervention women reporting breastfeeding initiation and any breastfeeding at 8 weeks and 6 months was consistently higher than in the usual care group. 17 The aim of this paper was to understand the views of women and IFHs of the ABA intervention components when delivered by two different peer support services.

| Setting
The ABA feasibility trial was undertaken in two geographical sites in England. The sites were selected because they had contrasting volunteer and paid peer support services operating, in areas with high levels of socioeconomic disadvantage and low rates of breastfeeding initiation and continuation. Existing breastfeeding peer supporters (n = 13) at the two sites received six hours of ABA IFH training. 14 At Site A, the ABA intervention was delivered by paid IFHs (n = 6) in an inner-city setting; at Site B, the IFHs (n = 7) were volunteers in a more rural setting. As part of their existing job, IFHs in Site A generally worked in a more ethnically diverse area of the city some distance from our study site. In Site B, IFHs were volunteers at local neighbourhood breastfeeding groups. To deliver the antenatal session, Site B IFHs met women at local children's centres and cafes as they were not insured to visit women in their homes, whereas the paid workers in Site A were able to provide home visits.

| Participants
Women, regardless of feeding intention, were recruited to the ABA trial between February and August 2017 through community midwifery clinics. Midwives provided women with study information, and then, a researcher approached women at antenatal clinics to gain informed consent. Overall, 103 primiparous women were recruited, 50 of whom received the ABA intervention. 14 Semi-structured qualitative interviews were undertaken with a sample of 21 women who received the intervention, who returned their 8-week outcome questionnaire and who had agreed to be interviewed. Women with different ages, feeding experiences and levels of engagement with ABA were purposively selected and interviewed at home when their babies were aged 4-21 weeks. These interviews explored their views and experiences of the ABA intervention and ranged from 45 to 90 minutes in duration. All 13 IFHs took part in one of two focus groups (n = 9-four at Site A and five at Site B) or a telephone interview (n = 4; two from each site). IFH focus groups/interviews were led by GT, who had no prior interactions with the IFHs. Other researchers (JI, JC, DJ) attended as note takers. The peer supporter co-ordinator at Site B also attended the focus group to offer insights from her perspective. The focus groups/interviews explored experiences of the ABA intervention and its delivery; focus groups were ~100 minutes and interviews were ~30 minutes long.

| Analysis
Data analysis was carried out by trained qualitative researchers (JC, DJ, GT, JI) who have extensive experience of qualitative research and evaluation of breastfeeding peer support services, from psychology, health services research and midwifery backgrounds. Interviews used a topic guide (see Appenidx S1), were audio-recorded, transcribed verbatim, anonymized and imported into NVivo 11 (QSR International Pty Ltd) for coding. Transcripts were analysed using thematic methods by developing a coding framework and a series of themes to describe women's and IFHs' experiences. 18 Subsequently, views of the women and IFHs were compared using framework analysis. 19 A subset of four transcripts was independently coded by GT, JC and DJ, followed by discussions to agree the coding framework. This framework was used by two researchers (JC, DJ) to code the remaining transcripts, with ongoing discussions to consider and agree any changes as needed.
All analytical decisions were shared with the wider research team using a consensus process to agree the final coding and thematic framework.
Patient and public involvement (PPI) was essential in shaping the development of the ABA study and intervention. Several different groups of new mothers and fathers, serving deprived populations, were involved in PPI group discussions as described in the main study report. 20 They discussed the interpretation and dissemination of the results and agreed that participants should be sent an easyto-read study summary leaflet by post or email which they approved and has been done.  Table 1 shows the characteristics of the women interviewed compared to all who received the ABA intervention. The women had an average age of 28 years, and their baby's ages at interview ranged from four to 21 weeks (mean 12 weeks). Participants' quotes are attributed to Site A or B, with their baby's age at interview (in weeks) and whether they were breastfed (bf) (including any breastfeeding) or formula-fed (ff) at 8 weeks. Similarly, the IFHs (1-13) were attributed to Site A (n = 6) or B (n = 7). Each theme reports the perspectives of women and the IFHs.

| RE SULTS
Overall, women valued the opportunity of receiving support from someone with similar experiences and learning about what community assets were available. The volunteer supporters were excited by new opportunities to meet different women and provide support for several months, and the paid supporters appreciated the content but found that arranging visits to the women was difficult due to their workloads and distance to participants.
Five themes ( and family tree', 'keeping in touch using proactive text messages', 'knowing about local groups and assets' and 'a woman-centred approach'.

| Early opportunities for infant feeding conversations
Women recalled antenatal meetings with their IFHs as being a relaxed discussion and welcomed the opportunity to have a 'chat' about infant feeding whatever their preference. The 'face-to-face' element of the antenatal meeting was considered an important part of being able to develop a relationship with the IFH and encourage contact after their babies were born. While for some, the meeting with their IFH resulted in them 'feeling a lot more positive' about feeding, one woman who had been intending to formula feed described how it helped her to reconsider her feeding decision.
It made me rethink about breastfeeding… but having that chat with her it did re-jog my memory there is another option sort of thing, yeah it did, definitely.

(P6 Site A, 11w, started bf, ff by 8 weeks)
The antenatal meeting was less interesting to women when it seemed to be a fact-giving exercise or when they had decided how they wanted to feed their baby and already felt well-informed.

| 'Mapping the friends and family tree'
Women provided mostly positive views about the mapping exercise to create their Infant Feeding Genogram (see Figure 1). Many found genogram completion to be useful as it helped them recognize how much support was available. Some women described the process as 'reassuring' as it reminded them how fortunate they were to have support.
She did a really useful thing actually, which was we did

| 'Knowing about local groups and assets'
Women provided positive comments about the assets leaflet which contained information about local groups, websites and phone lines for support. They mentioned being aware of some, but not all, of the resources listed, and that there was more support available than they had expected. One woman reflected that while she had already been thinking about going to local groups, the assets leaflet helped to remind her where and when these activities were, and she particularly valued the offer from the IFH to accompany her.

| 'A woman-centred approach'
A key feature of the ABA intervention was offering support using a 'woman-centred approach' rather than having a breastfeeding-centred focus in all discussions.
Women mostly felt that this had been achieved when they described being reassured that they knew where to go for appropriate advice and support, not feeling that they were being pressured to breastfeed and receiving positive feedback and encouragement from their IFH. IFHs also reported that they understood and tried to use a woman-centred approach when reflecting on the training and in describing some of their early contacts with women.

| Early and proactive support
Proactive support has been reported by others to be effective in increasing breastfeeding rates. [8][9][10]21 Continuity of targeted peer support with an antenatal visit and postnatal support from the same local supporter has been shown to be associated with psychosocial benefits for mothers, health professionals and peer supporters. 11,22 Proactive women-centred contact providing continuity of care from pregnancy to several weeks after birth was also valued by women in a small study, 9 and very early postnatal support has been reported as an important factor for effective breastfeeding support. 12

| Assets-based approaches incorporating the infant feeding genogram
The assets-based approach via use of the genogram and the assets leaflet were highly valued features of the ABA intervention. Such approaches are in line with sustainable models of community development via extending networks and building social capital. 23 Assets-based approaches have been used in a range of public health studies. 15,24 For breastfeeding, these could focus on both intrinsic personal resources such as self-efficacy in relation to infant feeding 25 and motivation and drive to maintain feeding, 26 and external resources such as family and friends; wider social networks of women who have breastfed; and community assets such as children's centres, mother and baby or breasteeding groups, and local breasteeding peer supporters. 27 The theory of change approach for assets-based working proceeds through four stages: (a) reframing thinking, goals and outcomes; (b) recognizing the assets available to achieve the change; (c) mobilizing assets for a purpose; and (d) coproducing outcomes. 28 The discussion with the IFH, with prompting via the infant feeding genogram and assets leaflet, facilitated movement through these stages towards a co-produced map of their existing assets landscape, which helped women restructure their social environment and increase their personal and external resources to support feeding their baby.
The Infant Feeding Genogram was developed in 2014 as part of a study exploring how women who were the first to breasteed in a family made sense of their decisions. 16 Our study is the first to explore its wider acceptability, and this is further analysed by Thomson et al. 29 The genogram gives detailed information about the family structure and the interactions between generations, but it does not show relationships with a wider social group. The way that the IFHs used the genogram with women might be better described as a sociogram, another family therapist tool or a mixture of the two, giving a picture of the many supportive relationships available to women. 30 Strengthening the use of the modified genogram in a refined ABA intervention would help IFHs understand the processes involved.

| Peer supporters using behaviour change techniques
Using peer supporters to provide social support and restructuring the social environment with a woman-centred approach through encouragement and advice has been recommended by others. A meta-synthesis of women's experiences and perceptions of breasteeding support found that a person-centred approach was more acceptable than breasteeding-focussed discussions. 6 Women in other studies have welcomed a peer supporter approach that helped them mobilize external and personal resources to achieve their breastfeeding goals through words of praise and reassurance. 5 A recent feasibility study using motivational interviewing techniques as their peer supporter intervention (Mam-Kind) reported that supporters found it quite challenging to move from an 'expert-by-experience' role to a more collaborative approach when giving information. 31 A similar challenge was also implied in our study by some IFHs (Site A) who felt that many ABA intervention components, such as being women-centred, were already part of their role, and some failed to perceive the value of co-creating the genogram.
Other studies have examined the influences of significant others on women's feeding behaviour and emphasized the importance of holistic family-centred approaches to supporting women. 32 Similarly, helping women to become familiar antenatally with the venues where postnatal groups are held to facilitate return after birth, with someone who can introduce them to a group on the first occasion (such as an IFH), has been shown to influence why interventions work in some places and not others. 4,27,33 We will use the findings from this study to modify the design of the information materials for women and training given to IFHs in our future trial. We will provide more explanation of how to incorporate and deliver the behaviour change techniques of restructuring the social environment and providing social support using an assets-based approach and more practical discussion about how to deliver the assets materials.

| Strengths and limitations
The study strengths include exploring a novel assets-based approach to delivering infant feeding support and including all women regardless of feeding intention. We compared the perspectives of IFHs and women who received the ABA intervention and included two different sites with different delivery models (paid workers and volunteers).
We achieved rigour in this study by use of detailed data analysis, undertaken by multiple researchers and analytical decisions being shared with all team members to achieve credibility. The researchers have a range of health-related backgrounds with prior experience of evaluating peer support. None of them were involved in direct delivery of the intervention, and all were involved in the data analysis.
We have included a wide range of quotes, from different individuals across the two sites to illustrate the final interpretations. All quotes are supported by demographic details to enhance transferability of the findings.
Although the use of PPI within the ABA study provided us with a vital user perspective, it was challenging to sustain relationships with some. Pregnancy and caring for young children take up a relatively short period of women's lives, and inevitably, they move on by returning to full-time work or being involved with school activities, which can make it difficult to have continuity with PPI contributors.
Limitations include our sample of ABA intervention women interviewed; all returned the 8-week questionnaire, and so the views of the nine questionnaire non-responders in the trial are unknown. A slightly higher proportion of women interviewed were breastfeeding at 8 weeks than for the whole intervention group, but otherwise those interviewed were similar to the women in the trial.

| CON CLUS IONS
Women who received the ABA intervention and paid and volunteer IFHs who delivered it welcomed this approach, despite some challenges in its delivery. The components of the intervention, including the infant feeding genogram and local assets information, were perceived to be useful in exploring and highlighting available sources of help that women could draw on for advice and support.
This proactive community assets-based approach with a woman-centred focus is a promising intervention that warrants further research to explore its effect on infant feeding outcomes.

ACK N OWLED G EM ENTS
We would like to thank all the women, infant feeding helpers and health-care professionals who took part in the study; all the ABA study team for contributions during team discussions; and PPI con- HT is a Trustee for NCT Charity, which provides infant feeding support, and a breastfeeding counsellor for the same organization.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.