In this study we evaluated women's experiences of peer counselling for exclusive breastfeeding. The vast majority of women were very positive about peer counselling. The few who expressed negative feelings did not discredit peer counselling as such, but rather were dissatisfied with specific aspects of the counselling they received, for example having received too few visits or inadequate information.
The women who had received at least the stipulated five visits were more satisfied with the peer counselling. Most of the women expressed general satisfaction with the time the peer counsellor spent with them and considered the visits useful. They felt they were able to interact freely with the peer counsellors. The women identified needs that were not met by the peer counsellors as mainly knowledge about complementary feeding and family planning, which had not been specifically targeted by the intervention.
Exploring satisfaction with care often presents methodological challenges. When and where the evaluation is undertaken may influence client response and some may give socially desirable answers. Using more than one interviewer may introduce some inter-observer errors. In this round of exit interviews we tried to minimise this type of bias by training the research assistant who carried out the interviews and stressing the importance of recording the women's responses to questions verbatim. The research assistant explained to the mothers the importance of ascertaining their true feelings about the peer counselling as it would help to improve it in the future. JN was assisted during the interviews with the women by an interpreter who was not part of the intervention team. The responses to the open-ended questions were coded by a team that included members not directly involved in the intervention.
The time between the peer counsellor visits and the interviews was different for the different mothers, and long periods between peer counselling and interviews may have introduced recall bias. In an effort to minimise this bias, the interviews were conducted soon after the women completed the peer counsellor schedule. Where the woman was not found at home, the research assistant kept returning to the home for the interview until she was found or confirmed as lost to follow-up. In this study the mothers received multiple visits from the peer counsellors and this could stimulate them to remember more than if they were visited only once. Using both open-ended and closed-ended questions could have helped stimulate the women to remember more about their experiences with the peer counsellors. A large number of women participated in these interviews and this increased the chances of establishing a more realistic picture of the participating women's experiences.
A number of factors may have influenced the women's positive experiences with the peer counselling intervention. The time the women spent with the peer counsellors could be viewed in terms of the number of visits each one received as well as the quality of the visits. The women who received more visits expressed greater satisfaction with the different aspects of the peer counselling. It is possible that with more visits the rapport between the peer counsellor and the woman improved, giving the woman the opportunity to learn more and ask for clarification of any unclear messages. A similar finding was previously reported in Toronto, Canada, where the total number of contacts between the peer volunteer and the women, as well as the length of the peer volunteer relationship, correlated with the mothers' evaluation of their peer support experience .
Furthermore, the peer counsellor's attitude during each visit could have influenced the woman's experience. This may be especially true if the peer counsellor showed the woman that she had time to discuss with her and address all her concerns. A peer counsellor who seems to be in a hurry may make a woman feel less free to discuss her issues and ask questions. The woman might feel she was burdening the counsellor by asking questions. This may lead to the woman feeling she was not given enough time by the peer counsellor. In this study, many women reported that the peer counsellor took time with them and did not appear to be in a hurry and this could have been responsible for the women's positive attitudes.
It is noteworthy, however, that the issue of "enough time" may be relative as it may be influenced by a woman's understanding of the issue under discussion. One who understands well an issue she wants to discuss might consider the time spent as "enough" simply because her concerns have been addressed to her satisfaction. Furthermore, the cultural context of the study needs to be understood. Since visitors to a home are treated well, the women could have felt obliged to say positive things about the peer counsellors whom they treated as visitors to their homes.
The women generally felt that the peer counsellor visits were useful to them. Some even felt that peer counselling had empowered them to make decisions about how to feed their babies since they now had the knowledge. This is important, as it implies that peer counsellors have to be well prepared through training in order to help the women feel empowered to make decisions about feeding their infants. The benefit to the babies from the visits was echoed by many women and this could have influenced how the women felt towards the intervention. Similar findings have been reported in high-income countries [22, 23, 35] and in Asia  where women expressed a feeling of being empowered to breastfeed their babies.
There were a number of components to the interaction between the peer counsellors and women, such as time spent and social aspects of the relationship. Women highlighted issues of trust, identifying with the peer counsellors, as factors that may have facilitated free discussion and better understanding. The approach of the peer counsellors was highlighted as important as they were able to sit with the women and discuss breastfeeding issues as equals. This reiterates what was reported in the United Kingdom, where support workers had time to sit with the women and observe them, an action that was valued by the women . This further highlights the concept of "peer" to these women and how peer counselling is appreciated by them depending on how it is packaged. It may be important to consider who appropriate peers in different communities are. The importance of the social aspects of peer counselling that are highlighted in the current study was noted in earlier studies . In Bangladesh, certain attributes of the peer counsellors, such as their occupation or trade, made them more acceptable to the mothers .
The issue of the women feeling respected by the peer counsellors is important as it may affect the success of the intervention. Showing respect is very much about general politeness as defined in a certain cultural setting. In this study setting, going to someone's home may reflect acceptance on the part of the visitor, so the women could have considered the peer counsellor visits as a sign of being accepted by them. This may have influenced how the women received the messages brought by the peer counsellors. However, as much as visitors are valued and respected, they are also expected to show respect to their hosts. The issue of age seems to be related to respect as the woman who felt she was not respected by the peer counsellor attributed it to the age difference between them. Older women may not respect the views of younger ones, or younger women might just feel that older ones do not respect their opinions. This is rooted in most cultures where increasing age is believed to be associated with more wisdom and older persons expect to be respected because of their age. This may be an important aspect to consider when identifying appropriate peers to allow free interaction with the women.
More than half the women felt there was nothing they missed learning from the peer counsellors, a finding similar to that reported in Toronto, Canada, where most mothers felt there was nothing they would have wanted the peer counsellor to do differently . However, the women who had needs that the peer counselling did not meet identified complementary feeding and family planning as areas about which they wished to learn more. This is not surprising as the current intervention focused mainly on exclusive breastfeeding for the first six months of life, but the women were concerned about what to feed their babies after the age of six months and how to gain control over their fertility. This highlights the importance of planning a complete infant feeding package that covers exclusive breastfeeding, complementary feeding and possibly mothers' health. If the scope of their counselling activities is widened, the challenge will be how much information the peer counsellors can handle in the short training period in order to be able to provide adequate explanations to the women. This finding also supports the argument for integrating breastfeeding promotion into wider child and maternal health interventions, an approach that has been well discussed in an earlier publication .
The women generally had positive feelings about the peer counselling process. Acquisition of knowledge and benefits to the baby were highlighted by the women as important explanations for their positive responses. This is similar to previous reports in which women appreciated peer counsellors' support towards successful exclusive breastfeeding [16, 23, 28]. The women who expressed negative feelings about the intervention voiced concerns about having received insufficient information about breastfeeding or having received few visits, rather than being generally negative. Their concern seemed to be related to how the peer counselling was presented to them rather than not wanting peer counselling at all, and their comments could be used to improve the intervention. Similar sentiments were expressed by some Bangladeshi women, who complained that though some peer counsellors could deliver the messages about breastfeeding, they could not give them adequate explanations, hence leaving them unconvinced .
One of the less frequently given reasons for having positive feelings towards the peer counselling process was women having received some form of financial assistance from the peer counsellor. This was a rarely practised gesture by the peer counsellors and was not encouraged by the study team. That notwithstanding, there were reported instances where a peer counsellor encountered a desperate situation during a visit where a young mother had been abandoned by her husband with no support and she felt compelled to give some money to buy soap for washing the baby's clothes. Such instances highlight some of the dilemmas faced by the peer counsellors as they visited the women to help them with breastfeeding.
Almost all the women expressed a desire to have a peer counsellor for their next pregnancy and felt they would recommend a peer counsellor for a friend. The positive effect they felt the peer counselling for breastfeeding had on their babies' health might have influenced this. Similar sentiments were expressed by mothers in Canada, where 85% of the supported women stated they would have a peer volunteer again and suggested that every new mother should be offered peer counselling .
Most of the supported women preferred a peer counsellor to a health worker. The reasons they gave for this preference were related to the peer counsellor's good approach to them during visits and the fact that nobody had helped them with breastfeeding before. Similar findings were reported in earlier studies, where supported women favoured a peer counsellor, who they considered to have helped them more than any other people they knew  and to have been their most important source of infant feeding advice . The issues of familiarity, availability and living in the same community and hence easy accessibility are important points raised by the women, and could be considered in future planning of similar interventions. However, for those who preferred a health worker, their concern was mainly the conviction that the health workers were well trained for the job.