While some 1,142 started the survey, a total of 1,126 participants completed at least one of the survey questions: 1,078 questionnaires were from service users who had contacted the helpline or online media (social media or DIBM), with 48 web chat responses excluded. Five of the translated versions were started (two Polish and three Bengali), but as none of the survey questions were answered, they were removed from the final data set. The number of calls/contacts to the NBH services over the evaluation period and how many surveys were completed for each different type of support are presented in Table 1.

Table 1 Response rates by types of support

While overall the data suggests that the views of ~ 15% of service users had been captured (1,078/7,129), the NBH figures represent calls/contacts and not the number of service users. While ongoing online media conversations or helpline calls (e.g., if call is interrupted) are recorded as one contact, it was not possible to formally record a denominator of individual service users, as e.g., they may have re-contacted the NBH services about a different issue over the evaluation period. From our sample, 724 (67.16%) stated that this was their first contact (five could not remember (0.46%), two did not answer this question (0.19%)), and ~ 32% of participants had previously contacted NBH services.

Further analysis was undertaken to compare the demographics (age, ethnicity, education level) of those who took part in the evaluation (helpline only), with those who contacted the helpline service over the evaluation period. While it was intended that all callers to the helpline would be asked for demographic-related information (rather than the usual one in five callers), there was ~ 35% of data missing (related to callers wanting to remain anonymous, and/or how the contact was ended). Overall, however, there were similar percentages of callers who were White or non-White (White 84.2% vs. 77.1%), who were educated to GCSE or below or A level of above (GCSE or below 3.6% vs. 4.0%), and who were aged either below or above 35 years (below 35 years 61% vs. 68%) between the two groups (the percentages are the evaluation participants vs. general helpline service users where missing data were excluded), thereby indicating that the evaluation sample was generally representative of the wider population.

Overall, 942 (97.5%) (missing data removed) were extremely satisfied/satisfied with the support received, with similar levels of satisfaction reported across the three different types of support (helpline, social media, DIBM). Most service users contacted the NBH services about a ‘specific difficulty’ (n = 676; 62.71%), just under 9% made contact for ‘general support’ (n = 93; 8.63%), and 239 (22.17%) for both (70 (6.49%) of participants did not answer this question). When considering this question separately for those who accessed the helpline or online media, a slightly higher percentage contacted online media for a specific difficulty (73.84% (n = 367) vs. 60.47% (n = 309)). Most service users involved in the evaluation had contacted the service about an issue they were personally facing (n = 1042; 96.66%), with the remaining (n = 27, 2.50%) calling on behalf of their partner (n = 8), friend/family member (n = 5), client (n = 6), or other (n = 8) such as healthcare professionals.

Below we present the descriptive and inferential statistics. An interpretation of these findings contextualised by qualitative comments is then presented in key descriptive themes.

Comparing service user and call/contact characteristics

The descriptive statistics for service user characteristics and call/contact characteristics are presented in Table 2, and the Chi-squared tests, and Wilcoxon rank-sum (Mann–Whitney) tests that compare these variables across the two support options are detailed in Table 3.

Table 2 Descriptive statistics for servicer user characteristics and call/contact characteristics (frequency (%) unless otherwise stated) Table 3 Comparing service user characteristics and call/contact characteristics between helpline callers and online media usersFactors that influence overall satisfaction

Similar to the 2011 evaluation [11] and to make the results comparable, the impact of service user characteristics and call/contact characteristics on overall satisfaction were investigated first (Model 1), as these were not influenced by the experiences during the call. Service user characteristics included were age (re-categorised into three categories “under 24”, “25–34” and “over 35”), ethnicity (re-coded as “White” and “non-White”), whether the service user was a first-time mother, whether mother had breastfed before, whether English was a first language, and level of education (classified as GSCE or equivalent, A level or equivalent, degree, postgraduate degree). Call/contact characteristics included whether it was the first time they had used the NBH services and service user attitudes towards (a) how easy/difficult it was for their call to be answered (for helpline) or (b) how easy/difficult it was to use the NBH support (for online media). The participants’ responses of “extremely difficult” and “somewhat difficult” were combined to “difficult” due to “extremely difficult” being very lowly populated across all types of support.

After identifying significant factors in Model 1, the regression model was expanded to explore how service user views’ on service characteristics (attitudinal responses to the help and support they received) affected overall satisfaction (Model 2), to explore the influence of support on individual service users’ breastfeeding experiences on overall satisfaction (Model 3), and whether overall satisfaction was influenced by the service users’ wellbeing (feeling less worried, less stressed, more confident, reassured and more knowledgeable about breastfeeding) and whether follow-up support options were provided (Model 4).

Modelling results for all four models for the helpline and online media support are summarised in Table 4. The table also reports adjusted R-square statistics as an unbiased estimator of R-square in the population [19]. The adjusted R-square was consistently higher for the helpline than for online media across all the models which means that the variables in the helpline model explain more of the variability in overall satisfaction than the online media data. This finding may relate to telephone conversations being more focused and detailed than virtual contact. Alternatively, it may demonstrate that the online audience is more heterogeneous compared to helpline users. Also, Table 4 shows that Model 2 has the highest adjusted R-square compared to other models: this demonstrates that the service characteristics variables explain more of the variability in satisfaction than factors in other models.

Table 4 Results of regression modelling factors explaining overall satisfaction for helpline and online media responders

In the following sections, we synthesise the findings from the descriptive and inferential statistics and draw on insights from the qualitative findings to help explain the variations in findings across the two support models.

Demographic-related issues

Overall, there were significant differences in demographic profiles between the two groups. First in terms of age (p = 0.004), the proportion of younger people, under 24, in the online media group was higher, while the proportion of older mothers, 35 and over, was higher in the helpline group (Table 3). The proportion of service users who were first-time mothers and educated to degree level and above was significantly higher in the helpline when compared to the online media group (p < 0.005). The variable ‘Is English your first language’ was associated with overall satisfaction for the helpline (p = 0.001) rather than online media model within the regression models (Table 4). However, when comparing this variable across the two groups, the service users who had English as a first language (p = 0.008) and were White British (p < 0.005) were more likely to use online media, and non-White, non-native English speakers were more likely to use the helpline (Table 3). These insights thereby indicate that confidence to converse in English is perhaps more essential for helpline users when compared to online media.

Accessing the service

How ‘easy’ it was to access the service was significantly associated with overall satisfaction for the helpline and online media groups (see Table 4). Some service users referred to how the speed of response had been unexpected, ‘absolutely incredible! Could not believe how quickly I received support’ particularly when contacting the NBH services out of hours or during national holidays when other support is not available:

[there was] a quick response even during the holiday period (bank holiday – Boxing Day). [Online Media/DIBM] (Participant 751).

Others complained about the need to make multiple calls to the helpline, or that waiting for responses to voicemails and online media was ‘stressful’, with 169 (32.9%) (missing data removed) of helpline callers agreeing that the helpline opening hours should be extended. On occasion, those who called the helpline were critical about voicemails not being returned, or the support being too late to be of help:

We were not called back until the next day after which time we had more or less solved the original issue with the help of a health visitor and GP. [Helpline] (Participant 207).

When comparing the responses across the two groups, it appears that overall, significantly more online media users found accessing support easier when compared to those who called the helpline (p < 0.005) (Table 3). From the qualitative feedback, it is suggested that this may be due to the nature of the call/contact. Some online media users spoke of expecting a delay; with one referring to how they would have contacted ‘a different service‘ had the issue been urgent. As those contacting the helpline are potentially more likely to request immediate support, this is likely to have tempered their views on access when timely support was not provided:

I needed the help there and then. It felt like too long to wait when you are struggling to feed your baby. [Helpline] (Participant 227).

Online media users also referred to how these sources of support meant they did not have to ‘hold in a queue and often miss the opening hours due to working shifts‘, thereby enabling them to seek support around their work or home life patterns:

I’ve got a 16-month old to run after who is obviously a handful! I love that the social media messages can be done around work and I’m not pressed for an instant reply. [Online Media/DIBM] (Participant 679).

Overall, first-time users were significantly more likely to use the helpline when compared to online media (p < 0.005). This could reflect first-time users having more ‘urgent’ issues, whereas those returning to the service may require more general help that could be resolved using asynchonous support. It may also, in part, be associated with service users’ lack of awareness of the different service offers. For example, qualitative feedback highlighted that while some service users knew about the helpline, they were unaware of the online options offered:

I only knew there was a number to call, if knew could message would of contacted NBH earlier as was nervous phoning so put it off for days. [Online Media/DIBM] (Participant 947).

Other service users reflected on how they had not been aware of the DIBM and commented that ‘It’s fantastic, although I didn’t know about it for ages!’ and ‘I found out about the drug help on my own – wish my midwives had told me about it’.

Perceptions of support

Overall, there were similarities and differences in factors that influenced satisfaction between the helpline and online media models. Significant factors unique to helpline callers related to the volunteer understanding what the caller was talking about (p = 0.001), and the volunteer giving the support that was needed (p = 0.017) (see Table 4). These factors indicate a potential cause and effect of helpline callers’ feeling understood and the volunteer being able to offer needs-led care, as reflected in comments such as:

I found the person I spoke with on the hotline to be extremely helpful. It felt like I was talking to someone who just got it – who understood exactly what I was going through and gave me the most informed, supportive advice. [Helpline] (Participant 279).

Two variables that influenced overall satisfaction in both support models were the information being helpful (p = 0.03 helpline and p < 0.001 online media) and the support meeting their expectations (p < 0.001) (Table 4). These factors both point to how the utility of the support is crucial irrespective of how it is provided, with frequent comments praising the volunteers as ‘professional’, ‘very well trained’, ‘knowledgeable’ individuals who had a ‘breadth of knowledge’ and ‘knew exactly what they were talking about’. One online media user reported:

I love this service and find the information and support given surpasses that of other services. I used to spend considerable amounts of time in forums and search engines finding answers but they varied so much and we’re never straightforward. Here I get straightforward, honest responses which I know I can trust. [Online Media/Social Media] (Participant 720).

While service users referred to the helpline being the public face of the NBH service, some online media users felt that the contact still made them feel they had a personal connection with a volunteer. One commented how ‘I felt like I was talking to a person not just a chatbot’ and another stated:

I was very surprised how quickly I received a response, and it was a proper chat. Not just one response and then done, they asked me a few questions and we had a back-and-forth conversation. [Online Media/DIBM] (Participant 801).

Impact of support on breastfeeding

Regarding the influence of the support on service users’ breastfeeding experiences, for those who used online media, only the support helping them to resolve their breastfeeding issues was found to be a significant factor in overall satisfaction (p < 0.001). Several online media users referred to being given practical information such as being encouraged to ‘cluster feed’ or guidance on medication that enabled them to confidently continue to feed, which for one meant:

[I didn’t have to] choose between prioritising my own health and wellbeing through use of medications or stopping breastfeeding’ as ‘the advice meant I felt safe to continue with both’. [Online Media/DIBM] (Participant 1080).

Whereas overall satisfaction amongst helpline callers related to being able to put into practice the information provided (p = 0.016), and the support helping them to resolve their breastfeeding issues (p < 0.001) and/or encourage them to continue breastfeeding (p = 0.028) (see Table 4). These insights reflect that the support being able to resolve their breastfeeding challenges is crucial, and that being able to discuss the issues in real-time can be more encouraging as well as maximising the potential to translate the information into practice. One helpline caller explained the importance of:

Someone who offered real advice, and time to talk it through and help me. Issues can be difficult to resolve, and you often feel a bit isolated if it isn’t working out and having problems. They were able to really have a very open conversation and help you reduce worry and tackle the issue with their advice. [Helpline] (Participant 441).

While overall only 273 (24.24%) considered that they would not have been able to continue breastfeeding without the help of the NBH service (although not significantly associated with overall satisfaction for either group), some explicitly spoke of how contact with the NBH was crucial in providing the instrumental and/or emotional support to resolve their breastfeeding issues, and to continue breastfeeding:

The first phone call I made 10mths ago I was so done with feeding my newborn but the support the lady gave me was out of this world nearly three hours on the phone and that gave me the power to carry on. [Helpline] (Participant 231).

Impact of support on wellbeing

Emotional and cognitive-related factors significantly associated with overall satisfaction for both helpline callers and online media users related to feeling more confident (p = 0.011 helpline and p < 0.001 online media) and reassured (p < 0.001 and p = 0.034 respectively), whereas, feeling more knowledgeable about breastfeeding was a strong significant indicator for helpline callers only (p < 0.001 Table 4): a difference potentially related to online media users being more likely to contact the service about a specific issue, rather than for general support and advice. Service users frequently described how important and ‘invaluable’ the support was for their mental and emotional well-being at what can be a challenging time. One helpline caller described:

I felt incredibly supported and was given words to express some of how I’d been feeling. Although I didn’t come away with much practical advice, the emotional support and time I’d been given had been invaluable. [Helpline] (Participant 320).

Increased confidence could be related to the volunteer providing information about a particular situation that ‘put my mind at rest and gave me confidence’ to continue feeding. Or how increased knowledge enabled them to feel more empowered to talk to and challenge the advice provided by health professionals; ‘[I felt] more informed when talking to GP, armed with information’. Following contact with the NBH, service users spoke of feeling more ‘informed and confident in making choices that work for me and my family’ and more confident in their own decisions around breastfeeding. Some also reflected on how their increased confidence promoted positive help-seeking behaviours. One online media user referred to how:

[I would] seek out the help I need’ and how ‘[I would now] happily speak to someone if I am having any more issues. [Online Media/DIBM] (Participant 744).

Reassurance was evident in terms of how service users frequently commented that they felt ‘relieved’, ‘calmer’, and ‘much better after the call’. The NBH provided reassurance at critical points, for example, late at night when there were few other sources of support, when they had ‘a query which felt really urgent’, when having a ‘difficult time trying to establish breastfeeding’ or when they were feeling vulnerable in what some described as ‘the most vulnerable times of your life (4 days after birth)’.

Follow-up support

Volunteers suggesting that service users re-contact the NBH services as needed was a significant factor of overall satisfaction for the helpline callers (p = 0.002) only (Table 4). A few of the responders referred to how the support had continued to help them at various stages in their journey from the earliest of days to maintaining exclusive feeding and to maintaining breastfeeding into the postnatal period:

NBH Helpline has been a breastfeeding journey saver for me on several occasions where I have been struggling and have nearly given up altogether! I have not had any support regarding BF from the NHS etc. I simply internet searched for support and called, and received exceptional support and advice I will continue to use this service over and over in my BF journey and have and will continue to recommend to other breastfeeding mums. [Helpline] (Participant 216).

The NBH was described as ‘somewhere to turn to’ and how this was particularly beneficial when callers were ‘at breaking point’.