Evaluation of the use of health care assistants to support disadvantaged women breastfeeding in the community

Abstract

There is little experience of the use of health care assistants in the community to support breastfeeding in the UK. The aim of this project was to evaluate the implementation of a small‐scale pilot project using health care assistants in the community to support disadvantaged women breastfeeding. The evaluation was funded as part of the Department of Health's Infant Feeding Initiative. A longitudinal observational and quasi‐experimental design was used. The project involved women, who had recently given birth, living in an area of London identified by the government's Sure Start scheme as socio‐economically disadvantaged. This paper focuses mainly on the findings drawn from the qualitative data focusing on the process of implementation, the role of the Support Worker and women's perceptions of the support. The findings suggest that the use of health care assistants in the community may offer a practical and encouraging approach in supporting breastfeeding which is acceptable to both breastfeeding women and health care professionals. More research is needed to establish whether the intervention significantly increases breastfeeding rates.

Introduction

This paper describes the initial evaluation of a small‐scale project that piloted and evaluated the appointment of a health care assistant in the community as an Infant Feeding Support Worker. The project was set up in 2001 as a collaboration between Sure Start and the maternity services in a socially disadvantaged area of London in order to meet national health targets regarding breastfeeding. The evaluation was funded as part of the Department of Health's Infant Feeding Initiative (Dykes 2003).

The project had a number of aims: in particular to increase rates of women making an informed choice to breastfeed; secondly to enhance general levels of support to new mothers in an area of social deprivation. The agencies also hoped to develop more communication and interdisciplinary work and to explore how far the role of a health care assistant in maternity could be developed with a community base. The evaluation strategy was intended to explore and assess how far these project aims could be met in practice. In addition, the research team aimed to explore and develop further the evidence on the forms of support likely to be effective in helping women to breastfeed.

The evaluation gathered qualitative and quantitative data on the planning and implementation of the project, on the views and experiences of the different stakeholders involved and on the rates of breast, mixed and formula feeding in the project area, before and after implementation. Both primary and routinely collected hospital data were included.

As this was an innovative scheme with little prior experience of the workings of such roles in the UK, the evaluation needed to monitor and document the nature of the intervention, and to develop a good, analytical description of the implementation of the project. This paper outlines the background to the project, key issues arising in the process, the role of the health care assistant in practice, the experiences of local mothers and service providers. The initial outcome data and the potential impact of the project are only briefly and tentatively discussed here.

Background to the project

This project was developed in an area of London that had recently been identified as part of the government's Sure Start scheme. Although very diverse, it included a high level of temporary (B&B type) accommodation, a high number of refugees and high levels of relevant indicators including teenage pregnancy, low birthweight, childhood accidents and health problems, and low levels of literacy and numeracy. Key Sure Start targets include reducing infant emergency hospital admissions, reducing smoking and giving guidance on breastfeeding. Data for breastfeeding rates were not available locally. However, an earlier maternity services research project in the local area that included this ward found that in 1994/95, at 2 weeks, only 39% of mothers were fully breastfeeding and 26% partly. By 3 months this had fallen to 19% and 24%, respectively. In 1997/98, at 4 weeks, 41% were breastfeeding fully and 21% partly (McCourt and Page 1996; Beake et al. 2001), suggesting a slight improvement in the period leading up to this study. At the time of the study the local Trust was working towards achieving Baby Friendly Initiative status and had received the certificate of achievement the previous year; however, audit figures for the unit as a whole did not show any improvement in breastfeeding rates.

Initial Sure Start documentation indicated around 174 births per year in the area. Local women generally booked for maternity care with the nearby National Health Service (NHS) Trust that was the main partner in this project. This was a large, obstetrically led teaching hospital but midwifery care locally was provided by caseload midwives working in group practices, with a high level of continuity of carer and community‐based and midwifery‐led care. Most women received postnatal care from midwives who had provided care antenatally and for labour/birth. The midwives were able to operate a selective visiting policy, enabling them to vary the pattern of visits and time given, to focus care where needed. Earlier research had shown that the visit time was longer and more varied than in the conventional community midwifery service (Piercy et al. 1996). The Sure Start project was newly established, providing a range of drop‐in facilities for local families, a psychology service and a health visiting service, with two health visitors able to offer home visits to families needing additional support. Nonetheless, a need for more feeding support to new mothers was perceived. The project was intended to supplement rather than substitute for existing levels of midwife and health visitor support and to provide a different form of support.

Literature on breastfeeding support

Existing research and service audit has identified gaps and problems in the support women receive for breastfeeding, as well as for other aspects of care postnatally (Ball 1994; Audit Commission 1997; Garcia et al. 1998; McCourt et al. 1998; Sikorski et al. 2004). The value of practical and social support for adjustment to parenthood and for breastfeeding is well established in the research literature. However, there is little evidence that additional professional midwife support is necessary or superior to other possible arrangements for additional support, for example, from peers or community members (Schafer et al. 1998; Sikorski et al. 2004). Much of the literature on lay support is based in the USA, however, where there is no community‐based midwifery and postnatal home visiting is not the norm.

A literature search found little research into the effects of nonprofessional support for breastfeeding in the community in the UK. Additionally, a variety of different peer support initiatives have been set up, often as small‐scale local schemes and it is not clear from the literature how far they share similar principles and features, or differ. A recent in‐depth study found that practical, role‐modelling forms of support were most likely to have a positive impact on socially disadvantaged mothers’ intentions and success in breastfeeding (Hoddinott and Pill 1999). A recent trial of peer counsellor support also found that women valued nonjudgmental listening, reassurance and encouragement to keep going. It did not show evidence of effect on breastfeeding rates. However, the support available was only by telephone, relied on busy volunteers and on the women needing support to initiate the contact (Graffy et al. 2004). In a socially disadvantaged area of Glasgow, a study of a community‐based peer breastfeeding support programme suggested an increase in initial breastfeeding rates with peer support being acceptable to both mothers and professionals (McInnes et al. 2000; McInnes and Stone 2001). A further UK evaluation of breastfeeding support workers, although only a small project, suggested that the use of support workers in an area of social deprivation had been of benefit in promoting and encouraging mothers to breastfeed (Battersby and Sabin 2002). As lengths of hospital stay decrease in the UK, it might be anticipated that many women will need additional support at home. The research available suggests that early discharge home does not affect breastfeeding rates (Brown et al. 2002), perhaps because of low levels of support available in postnatal wards (Garcia et al. 1998). However, there may be particular implications for disadvantaged women who may benefit more from extra support with breastfeeding (Jones and West 1985).

A prospective, randomized controlled trial of the effectiveness of community postnatal support workers in the UK, although much valued by the women, concluded that there were no health improvements in the study group (Morrell and Stapleton 2000; Morrell et al. 2000). However, this study offered women general practical and emotional support and did not focus on infant feeding, nor did it target women living in an area of deprivation or lacking social support.

Although there has been very little experience of use of health care assistants in the community to support breastfeeding in the UK, there are some other models of community‐ or home‐based practice that might be drawn on. These include the maternity aides established as a service in the Netherlands – where home birth has remained relatively high – and use of trained nursery nurses as assistants to health visitors. However, the support provided by Dutch maternity aides is more general and takes place within a very different system of maternity care. Similarly, the work of health visitors is somewhat different in that the assistants have greater formal training and they work in a more tightly planned way, with the health visitor. There was little evidence, therefore, apart from the Glasgow study, of how such a role, with a high level of independence and a relatively open remit, might work in practice (McInnes and Stone 2001) or of what its effects might be.

The intervention

Because of its innovative nature, it was not possible to clearly describe the intervention at the outset of the project – this was a key aim of the evaluation and so this is presented as part of our findings and discussion. The project's aim, however, was to establish a post of Infant Feeding Support Worker, which would be created and managed within the midwifery service, although funded by Sure Start, on the health care assistant scale. Health care assistants were established within the local maternity unit, working closely with midwives, but this post would function with a community base. This involved visiting women in their own homes as well as hospital and working closely with Sure Start workers, across professional and agency boundaries. Independent home visits would be conducted.

Estimates of births in the Sure Start area were available, but levels of support and number of women requiring additional support had not been precisely estimated, so that it was difficult to plan for work patterns, boundaries and inclusion or exclusion criteria. A provisional framework was drawn up by an interagency working group with very broad criteria: any woman who felt she needed additional support, covering the period from 32 weeks of pregnancy to 4 months postnatally. Similarly, it was not clear at the outset what type of support needs would be identified or prioritized, but the aim of working with women antenatally recognized the potential need for information and encouragement for some women in making decisions around how to feed their baby.

The working group included midwives, health visitors and managers, a consumer representative (J.T.) who had previous experience of breastfeeding research and eventually the Infant Feeding Support Worker and Researchers. Although the primary aim of the researchers was to evaluate the project, the nature and stage of the work meant that this was approached very much in the manner of action research (Elliot 1991) with researchers contributing to the project initially by raising questions and then by providing feedback and the project group contributed considerably to the research process.

Design, materials and methods

The study used a longitudinal observational and quasi‐experimental design, with historical controls, describing and analysing the implementation of the scheme and comparing levels of care and breastfeeding rates before and during the first year of the project. The research methods included audit of activities and outcome data and qualitative analysis of the implementation of and responses to the scheme. The data were gathered in line with three key phases: baseline (data on practices and women's experiences in the 6 months before the project commenced), implementation and outcome. The entire study period was fixed to 1 year as it was part of the Department of Health's Infant Feeding Initiative, which limited the postimplementation data gathered. However, following the fixed period the Support Worker continued to gather data concerning feeding rates for a further 4 months. All the structured data collection tools were piloted in the preimplementation period and minor amendments made.

The methods of data collection were as follows:

• 1 Women's questionnaires. All women receiving care in the area before and during the implementation (n = 84) were sent a brief structured questionnaire, 6 weeks postnatally that included closed and open questions to record their infant feeding patterns and explore their perceptions of the postnatal care received. These were designed to be attractive and straightforward to complete without needing a high level of literacy and fluency in English. Reminders were given to nonresponders, by telephone when possible.
• 2 Women's interviews. Semi‐structured interviews were conducted with women postnatally. These were planned to avoid loss of data resulting from potential skewed or poor questionnaire response rates in areas of social deprivation and to obtain greater depth of understanding of what women find helpful in supporting feeding. The women included for interview also received questionnaires. Because of delays in the project implementation giving very small numbers, all women who had received support in the initial period from mid‐May to June (i.e. the first to receive support) were invited to participate and six interviews were conducted. Interviews were conducted in women's homes, from about 6 weeks postnatally, and with permission; they were all audio‐taped. All the interviews were transcribed; however, due to the tight time frame there was no time for the women to review the transcripts.
• 3 Focus group discussion with midwives. A group discussion was held with midwives working in the area to explore their perception of the project and their early experience of working with the Support Worker. To ensure ease of attendance, we arranged to hold this within the regular meeting of all community and group practice midwives for this area, and so a large number of midwives were present, not all of whom had direct experience of working with the new postholder. Initially this was planned to take place early in and again towards end of pilot. Because of project implementation delays only the initial focus group was conducted in this phase of the research.
• 4 Interviews with working group members. Semi‐structured interviews were conducted towards the end of the project with all working group members as follows: Sure Start and community midwifery managers, health visitor, link midwife and Infant Feeding Support Worker. These used a brief topic guide to elicit and explore experiences of the implementation and workings of the project, the nature of the Support Worker role and their reflections on its initial impact.
• 5 Support Worker case record forms. A report form was designed and piloted for use both for this research project and for routine project records. The Support Worker completed a simple data sheet for each woman she visited, recording the type, duration and frequency of support given, the woman's feeding problems and patterns.
• 6 Midwife record forms. A similar simple form was designed and piloted for project and routine use. Midwives providing care to women in the project area were asked to complete a form for all women they cared for giving birth in the relevant period, to be kept in the women's hand‐held medical notes.

Ethics

Ethical permission was obtained from the local Trust's research ethics committee. Letters and information sheets for women and practitioners made clear that there was no obligation to take part in the research, that participation would not affect their care in any way and that they could withdraw at any point. All data sheets and transcripts were coded for confidentiality. Working group members were aware from their participation and discussions that in a small project, confidentiality would not guarantee full anonymity for their comments, and were reminded of this before interviews. All had opportunities to read and comment on a draft of the report before wider publication. All midwives participating in the focus group discussion, whether involved with the project or not, were advised regarding confidentiality in our handling of the data and the need for confidentiality within the group.

Response

Support Worker data sheets were completed for all women receiving care as part of the project in the initial study period (total n = 25 postimplementation) with 55 in total in the first 10 months of implementation. From all cases included in the study, only 23/84 midwife record sheets were returned, and only 2/25 of these were postimplementation. During the study period, the community services were completely overhauled with caseload midwifery extended to the whole community service, approximately doubling the number of caseload midwives. With such a major reorganization of the community midwifery services, the priorities of the midwives were with settling into a new way of working and level of awareness of the project was extremely low, despite frequent reminders by a research midwife. The midwives’ focus group was attended by 14 midwives.

Women (n = 9) who had received care from the Support Worker and were approximately 6 weeks postnatal were offered an interview and of these six interviews were conducted, three could not be contacted or did not wish to be interviewed. One woman interviewed did not speak English fluently, and the need for interpreters or bilingual researchers would need to be addressed in any larger‐scale study as would the need to provide appropriate access to the service for women who do not speak English. Only women recruited within a tight time frame were eligible for interview given the completion date of the project and the need to interview women who were 6 weeks postnatal.

Because of bureaucratic delays, an existing member of staff in a health care assistant post was initially seconded to the project for approximately 2 months. This allowed a person with experience of working with midwives locally to establish the position, and also provided researchers with two individual perspectives on the role (interviews were conducted with both).

Of the 59 preimplementation and 25 postimplementation women's questionnaires sent out, 33 and 11, respectively, were completed: response rates of 56% preimplementation and 44% postimplementation. Telephone reminders proved relatively effective, and a number of women who had forgotten to complete the questionnaire preferred to complete it by telephone with the researcher. As the end of the project period fell during the summer holiday, and with a tight reporting deadline for the initial phase, not all postimplementation women received a reminder.

Data analysis

All structured data sheets and interview questions were entered onto Excel spreadsheets and were analysed using descriptive statistics. Open questions and qualitative data from interviews were analysed thematically. In the case of interviews, the researcher who conducted the interview read and re‐read the transcript for overall meaning and then annotated each with potential codes and theme areas. This was then repeated independently by another member of the research team. The team then met to discuss the themes emerging in the interviews and to agree a set of key codes and categories.

Findings and discussion

As indicated, the findings discussed here are mainly drawn from qualitative data. The focus was on the process of implementation, understanding the nature of the Support Worker role and exploring women's perceptions of support, in relation to both their needs and experiences, from both before and after the Support Worker began her work. Accordingly, the themes are discussed under these headers.

Planning and management

The working group met at regular intervals first to develop and then to co‐ordinate and monitor the project. Day‐to‐day management was provided by a midwife co‐ordinator with a specific interest and expertise in breastfeeding and line management by the Trust's Community Midwifery Manager. Although clinical supervision was by midwives, the Support Worker was expected to liaise closely with the Sure Start health visitors, as part of the Sure Start team and using the Sure Start programme as a work base.

These arrangements were formally planned but also required refinement in practice once the project started. This project was also seen as a chance to re‐establish communication and collaboration between hospital‐ and community‐based services and between midwives and health visitors that had been undermined by the way services were organized from the 1970s to 1990s. The Support Worker needed to work across two organizations with different structures and ways of working, where traditionally women had been passed on from midwifery to health visiting, with little contact or overlap. Once a clear job description, specification and criteria had been developed, the next important step was to refine and review supervision and management arrangements, role definition and boundaries.

A 2‐week induction programme was planned shared between the two organizations, with 1 week spent at Sure Start and 1 week in the hospital. The only formal training given was a 2‐day workshop that is offered to midwives as part of the hospitals ongoing Baby Friendly Initiative training programme.

Defining the role

Despite the lack of prior collaboration, the philosophy and understanding of the qualities sought for the post were shared and there was a strong commitment to working together around the needs of women and families. The Support Worker role was seen as additional to and different from those of professionals so specific ‘expertise’ on infant feeding was not required. However, a candidate with personal experience of breastfeeding  and  some  relevant  experience  – such as working with community groups, working with mothers/babies – was sought by the service employing the candidate. Qualities specified were:

• 1ability to listen;
• 2ability to understand and work with women's and families’ needs;
• 3interpersonal and communication skills;
• 4maturity and life experience;
• 5ability to manage autonomy and boundaries; and
• 6ability to ‘engage’ rather than ‘teach’.

The title of ‘Infant Feeding Support Worker’ was chosen primarily not to alienate women who might initially consider bottle feeding and it was accepted that she would support women however they chose to feed their baby, even though her primary aim was to support breastfeeding.

The role carried a high level of autonomy and responsibility as, although supervised, the Support Worker would visit women independently at home and plan support with them. Additionally, although clearly focused on infant feeding, as the support was intended to be different from that offered traditionally by professionals, it was thought likely to be somewhat broader. It was important, but difficult, for all those involved to define what forms of support would be included and what the limits to this were or should be – when should the Support Worker refer on to others.

Initially, it was planned that midwives would refer women for support, using a simple pro‐forma, either antenatally or postnatally. This did not prove effective in practice, as we discuss below, and a form of Support Worker/maternal self‐referral was developed.

Access to the service

In the early weeks of the project, with busy midwives unfamiliar with this way of working, there were few referrals. Therefore, the Support Worker made an introductory visit to all new mothers in the area where the Sure Start facilities were introduced and the mother's needs around feeding assessed informally. If the woman wanted additional support, further visits would be arranged, taking the woman's desire  for  support  as  the  cue.  In  effect,  this  meant the project was highly centred on the women's own definition of need, but did not depend on women having the confidence or knowledge to seek this out independently.

The contact started antenatally if a woman was referred by a midwife; for example, if the woman expressed concerns about feeding difficulties with a previous baby or uncertainty about whether to breastfeed. These referrals were only beginning to be established at the end of the study period, however, and it was not possible to form any view on the potential benefits of antenatal contact. The possibility of the Support Worker participating in Sure Start antenatal parent groups was also being discussed. However, she felt it was beneficial to have made contact with women before birth. Women also commented on the value of having met someone previously who they could call on if problems arose.

To contact women postnatally, the Support Worker checked the birth register regularly, although it was hoped that this time‐consuming approach would be replaced by regular listings from the hospital's computer records. She then made a brief visit to women in hospital, or at home, in the early postnatal period. If women wanted more support, further visits would then be arranged. In the initial visit she introduced the Sure Start services and generally enquired about how the woman was ‘getting on’ before discussing feeding – partly to avoid feelings of defensiveness in women who might otherwise feel pressured about breastfeeding.

The Support Worker role in practice

In general the Support Worker saw the need to listen to women, sit with them and encourage them as central to the role. She also noted that many women had broader problems that related to or impacted on their ability to breastfeed: if the woman was stressed or anxious for other reasons, feeding would become more difficult. This was taken into account in her approach, but where more complex general needs for support arose, these were referred to the Sure Start health visitor.

We identified a number of key themes from the Support Worker interviews relating to her perception of women's support needs, and how to respond. These are summarized below and then compared with women's perceptions, as reported in questionnaires and interviews.

Making yourself comfortable

The Support Worker observed that many women were attempting to feed in an unrelaxed position and with poor posture that could cause pain and fatigue. She used a practical, trial and error approach to comfort and positioning. In additional to seating and posture, this approach would include measures such as ‘making sure you have a drink by your side’ or sitting a potentially fretful toddler beside you with a book to share and a drink.

Confidence about sufficiency of milk

She reported that many mothers, especially with first babies, expressed anxieties about whether the baby was getting enough milk – as they cannot measure or see breast milk as with bottled milk. She discussed other ways that women could ‘see’ or ‘know’ the baby was getting enough milk that would increase the mother's confidence, including her own observation and knowledge of her baby's patterns, development and contentment and her own ‘embodied’ knowledge such as feeling the let‐down reflex and changes in her breasts.

Not expecting life to go ‘by the book’

She observed that many women were anxious and disappointed because their experiences seemed to differ from what was presented in books and magazines. She encouraged women to feel reassured that all experiences were different, not conforming to an ideal and that they would gain confidence with experience and practice.

Reinforcing knowledge

Although most women were aware of key benefits of breastfeeding, she felt able to keep them informed about less well‐known benefits. She kept a file of magazine and paper cuttings to share with women. She also informed women about the underlying workings of aspects of feeding they were less familiar with. This included the importance of latching on effectively, different qualities of breast milk during the feed, the relationship between suckling and supply and how ‘top‐ups’ with formula could interfere with this, reinforcing problems with sufficiency of milk.

Establishing feeding

She noted that as many women leave hospital very early, the initial few days of establishing are usually undertaken at home, at a time when women may previously have been in hospital with staff constantly present, even if very busy. She was also aware from women's reports that many found care in hospital in the early period inadequate, confusing or unhelpful.

More general help for women lacking social support

She noted that many women did not have family around to help and that many had no, or very limited, experience of young babies. Consequently they often lacked confidence and basic practical knowledge such as how to change a nappy or bath a baby. This was increasingly important with very early hospital discharge. However, rather than trying to provide all support herself, after the very early days she encouraged women to attend community groups and took opportunities to put women in touch with others, for mutual support.

Referral to others

Where she felt women might be depressed or have more long‐term or complex needs, she put them in touch with health visitors or other appropriate local services (e.g. Babytalk, Weaning Group, Parents To Be Group). This might include referral back to midwives, a breastfeeding specialist or general practitioner where the breastfeeding problems might require this, for example, mastitis or suspected infections.

Women's perceptions of feeding support

The themes identified in women's questionnaire and interview responses related closely to those highlighted by the Support Worker, suggesting that she had been able to form a responsive view of women's self‐perceived needs, what they found helpful, or not. From the women's accounts, the Support Worker role can be divided into three main areas: practical/technical support, information and general or social support. All were considered important but general or social support was more highly valued and emphasized by the women, particularly during interviews. In contrast, women's accounts of midwives’ roles tended to describe mainly technical/practical support and information that tended to be didactic. While some women received good midwifery support, others were highly critical of the nature of the support offered.

Practical/technical support involved activities such as help with ‘positioning’ and ‘latching on’. The fact that the Support Worker had time to sit with women and observe them was seen as particularly valuable in this respect. For example:

She offered me very practical advice, she was watching me do the breastfeeding and trying to give me pointers on how to improve. She was encouraging, positive and supportive. She said it was fine to do what you are doing but try it this way and try that. (Miranda)

A number of women identified this, before implementation, as a gap in the provision of care. For example:

home visit from person specialising in breastfeeding and with time just for that would have been really welcome, and beyond the initial two weeks. (Preimplementation questionnaire – open question)

Breastfeeding isn’t easy and I would have loved it to be and still be able to breastfeed my second child. I hope if I have another child more help will be available. (Preimplementation questionnaire)

Information fell into two main types: ‘tips’ (such as suggestions on how to prevent and deal with problems such as soreness) and underpinning information, for example, on the mechanisms of breastfeeding, sufficiency of milk and so on. It was apparent from the women's accounts that the type of information needed was quite different from that sometimes found in health promotional literature and antenatal visits or classes. All were aware of the main health benefits of breastfeeding but their knowledge of the practical aspects and their underlying physiology – such as the relationship between frequency of suckling and supply of milk – was sometimes less full and women appreciated more detailed information about this. For example:

she told me one thing about breastfeeding, when you breastfeeding it's alright to continue, you know, continue, no stop the baby and change another breast because first milk, second milk, third milk, you know what I mean? (Fatima)

more explanation about breast feeding beforehand. Ex – I did not knew that the actual milk comes after couple of days. So simple but I did not knew and no‐one told me. (Postimplementation questionnaire)

Echoing the Support Workers’ comments, a number of women expressed concerns about sufficiency of milk. Several women expressed concerns about or a focus on measurement, and these tended to be women who had introduced feeding by bottle. It appeared that for some women, external reassurance such as the ability to visualize and formally measure the amount of milk taken was important.

General or social support was discussed in great depth in the interviews and was clearly highly valued by the women. They emphasized the importance of general encouragement, gaining confidence and knowing there was someone available to help and to talk to.

but there is a thing in your mind thinking OK there is support already there and I’m not on my own. (Miranda – talking about the value of meeting the Support Worker antenatally)

she's like colic baby, she has always got wind and all that and she would always come and find out how she's doing and how she is feeding and how she is getting on, you know, it's encouraging. (Femi)

Some specifically felt this had made a difference to ability to continue breastfeeding.

it just encouraged me, because I was planning to mixed feed as well … so it just encouraged me really to just keep it on the breast and it was just nice for me to see that, you know, you have people that comes round to talk to about things like that because that has never been. (Ola)

The importance of a friendly, encouraging, nondogmatic and nondidactic approach was evident. For example:

because she's a friendly person I found it useful. Let me say that because I enjoy her coming round because she's nice, you know, when she comes round she feels at home and you’re comfortable around her kind of thing, so I love her coming round. (Ola)

Similarly, the continuity of antenatal and postnatal visits by one person and the relationship that could be formed was valued. This woman continued:

that made a big difference because you don’t often see, when people come round like that they just do what they need to do and go. There's no relationship or anything, but her coming round is also relationship‐based, She's not coming round just to do her duty, she comes to build a relationship and that actually makes you feel comfortable around her, to actually talk to her and open up to her. (Ola)

While some women received this kind of support from a caseload midwife, where this was not available the approach was sometimes contrasted with that of professionals, who were seen by some women as too dogmatic, or unrealistic. The following quotes illustrate the strength of feeling among such women about the negative potential of a didactic, impersonal approach:

it's all very well saying you must breastfeed, yes, you must do this, but they don’t know, they haven’t done it. (Miranda – twins)

my gut feeling is that sadly the vast majority of professionals offering advice to new mothers on breastfeeding have no experience of breastfeeding themselves and this creates a confusing discrepancy between advice offered and the realities of the experience. I put the reason why so many people stop breastfeeding relatively early on down to this fact. (Preimplementation questionnaire – open question)

feel that pressure to breastfeed exclusively of ‘NCT style’ breastfeeding Nazis approach actually puts lots of women off – surely some feeding is better than none. (Preimplementation questionnaire – open question, referring to midwives)

Clearly, a proportion of women felt pressurized by the approaches to support taken by some professionals, and this appear to have an alienating rather than supportive effect, where women would simply tend to dismiss their advice as unrealistic, lacking a basis in personal experience and not really tuned in to how women feel postnatally when faced with feeding problems.

Initial indictors of outcomes

As our initial evaluation took place very early in the scheme, and owing to delays in implementation, it was only possible to obtain very limited figures on outcomes during the study period. The figures given here should therefore be treated with great caution.

Figures from the routine hospital maternity data system, which records feeding pattern at birth, can be seen in Table 1. This compares with women's self‐reported feeding patterns as seen in Table 2. These initial figures, as discussed earlier, were extremely small, so must be viewed particularly cautiously. However, the Support Worker records give slightly larger numbers for feeding patterns postimplementation (see Table 3).

Table 1.

Routine maternity data – feeding at birth

Feeding at birth	Before IFSW input 1 September−2 April (%)	With IFSW 2 May−2 August (%)
Exclusive breastfeeding	60 (59)	43 (67)
Mixed – breast/formula	0	0
Standard formula	4 (4)	3 (5)
Special formula	9 (9)	3 (5)
Missing	29 (28)	15 (23)

IFSW, Infant Feeding Support Worker.

Table 2.

Women's questionnaire reports of feeding, 6 weeks postnatally

	Before IFSW (%)	With IFSW (%)
Exclusive breastfeeding	18 (55)	7 (64)
Mixed – breast/formula	8 (24)	4 (36)
Formula	7 (21)	0

IFSW, Infant Feeding Support Worker.

Table 3.

Infant Feeding Support Worker/Sure Start feeding records

	At birth (%)	At 6 weeks (%)	At 4 months
Exclusive breastfeeding	45 (81)	41 (74)	32 (58)
Mixed – breast/formula	4 (7)	8 (15)	12 (22)
Formula	4 (7)	4 (7)	5 (9)
Missing	2 (4)	2 (4)	6 (11)

This table only includes those women for whom records could be completed through to 4 months postnatally during the first 10 months of the Infant Feeding Support Worker.

The outcome data suggest that rates of initiation and continuation of breastfeeding, particularly at the later stage between 6 weeks and 4 months when many women introduce formula or mixed feeding, may be increasing in association with the implementation of the project. This was very early in the life of the project and the figures at this stage should not be considered reliable, nonetheless the initial findings are encouraging and suggest that further research would be worthwhile.

Concluding points

The experience of the implementation of this scheme was encouraging. Two quite different organizations concerned with maternal and infant health were able to work together effectively to establish a Support Worker role. At the end of the pilot period, the post was continued and the closer communication between the agencies and professions continued to develop.

As an innovative role, working across organizational and professional boundaries, and providing support that could potentially be quite diffuse, the need to develop appropriate boundaries was seen as important. The professionals involved expressed initial concerns about the need to define the boundaries of the role, in terms of type of support to be provided, when and how to refer to them, and avoiding attempts to provide a ‘professional’ type role. The intention of the scheme was, in any case, to provide a different, complementary form of support to that provided by midwives or health visitors, as well as additional time and this appears to have been achieved.

The Support Worker's understanding of her role reflected this and women's comments suggest that they saw this support as helpful rather than undermining. Women valued knowing she was readily available to them. They liked the way she was knowledgeable, reassuring, encouraging and that she had time for them whether this be in their home, hospital or a community setting. Some contrasted it with the approach of midwives, who they felt were trying to tell them what to do. The value of facilitating the women's own sources of support, encouraging participation in community activities and making links with other mothers was also recognized and was reflected in the Support Worker's approach and activities.

Traditionally, health education has tended to use didactic approaches, based on the assumption of a knowledge deficit or gap that needs to be filled. Information may be provided in a theoretical, rather than person‐centred or experiential form and professionals may assume that their clients lack information about the benefits of certain health behaviours (Jones et al. 2002). The responses of both the Support Worker and the women in this study suggest that women are generally knowledgeable about the benefits of breastfeeding, but may lack some practical and theoretical information that will help them to cope with breastfeeding in practice and have confidence in it. A practical approach to offering this is appreciated as well as adding to their theoretical knowledge, particularly where this is based on observing and responding to the woman's own situation.

The degree of concern expressed by the women and reported by the Support Worker about sufficiency of milk was an important issue, and again, an experiential approach appeared to be more effective than a didactic one where professionals’ information, although useful, simply told women they will have enough milk. This experiential approach was then reinforced by more ‘theoretical’ information, offered in the form of tips and ideas, in a way that appeared to be more empowering for women than the more partial information that women often report being offered.

It was too early to say, in the life of this project, whether the project made a measurable impact on rates of breastfeeding. Apart from limited figures available, comparison would be extremely difficult without the option of conducting a randomized controlled trial. However, we suggest that the findings were sufficiently encouraging to warrant the conduct of further research.

Acknowledgements

With thanks to the women, the midwives, health visitors, health care assistants and service managers who took part in the project and its evaluation. The Department of Health's Infant Feeding Initiative funded the evaluation and Sure Start Coningham funded the support worker post.