In the past two years, there has been a growing commitment to the goal of universal health coverage (UHC) and wide reaching and high-level discussion about the centrality of UHC to the post-2015 Millennium Development Goal agenda and the development and formation of the Sustainable Development Goals. Taking forward UHC will be futile without specific effort and action in fragile states. While definitions and figures vary, some 1.2 billion people are estimated to live in fragile and post-conflict settings [1]. It estimated that ‘one third of the global poor lived in fragile states in 2010, and projections indicate that roughly half will do so by the year 2015’ [2]. In fragile settings, access to equitable and quality health services is not only vital, but of huge importance for rebuilding the social fabric of countries [1].

The health workforce is a key health systems building block that underpins the expansion of health services and UHC efforts. Most countries in the global South have a shortage of formal health workers and are increasingly looking to a range of close to community providers (e.g. community health workers (CHWs), village midwives and traditional birth attendants (TBAs)) to fill the gap and reach the poorest and most marginalised individuals, households and communities. Close to community providers are arguably critical players in fragile settings, settings where human resource shortages are particularly acute as health workers may have been killed during conflict, fled the country or simply unable to reach their work places as they are located in hard to reach areas. Maternal deaths and under-five mortality rates are much higher in fragile and post-conflict states and this is an urgent area for action [3–6]. Despite their potential, there is limited literature assessing the different actual and potential role of close to community providers (such as TBAs) in promoting maternal health in fragile settings.

The role of TBAs in maternal and newborn health has been widely debated. A TBA is defined by WHO “as a person who assists the mother at childbirth and who initially acquired her skills delivering babies by herself or by working with other birth attendants” [7]. TBAs have been in existence since at least the nineteenth century and were additionally trained to conduct safe deliveries in order to address the shortage of skilled birth attendants especially in developing countries where maternal mortality remained high [8, 9]. The evidence shows that TBAs have made no significant change in the reduction of maternal deaths [10–12], although some reductions in perinatal mortality attributable to TBA training has been documented [13]. Hence, as far back as 2002, the World Health Organization (WHO) suggested a rethink of the roles and responsibilities of TBAs within their communities and suggested they could work as health promoters to improve the utilization of skilled birth attendants, rather than directly as primary care providers who deliver babies themselves [14]. This has led to several new initiatives that partner with TBAs to improve maternal and newborn health [15, 16].

Health Poverty Action (HPA) is a non-governmental organization (NGO) which aims to strengthen “poor and marginalised people in their struggle for health, prioritising the communities almost everyone else in the world has forgotten” [18]. HPA works in 12 countries in Africa and Asia. Within Somaliland and Sierra Leone, HPA prioritized a focus on maternal health as this is a key priority area in both contexts. HPA has worked with partners to design and implement programmes that work in partnership with TBAs in ways that build on their trusting relationships with communities to try to improve maternal health experiences and indicators. In this new model, TBAs do not actually deliver babies, but rather give information and support and encourage women to give birth in health centres. The details of both programmes are as follows:

We present the four key themes which emerged from the joint analysis of participants’ perceptions and experiences of trained TBAs’ new role in promoting maternal and newborn health following the intervention supported by HPA in rural contexts in Sierra Leone and Somaliland, and these have been used to structure the results section Theme 1 presents participants’ perspectives on the value of the training received; theme 2 the impact of the trained TBAs new role; theme 3 opportunities to strengthen integration of trained TBAs into the health system and theme 4 the challenges to realizing and sustaining trained TBAs new role. In the results SL is used for shorthand for Sierra Leone and SMLD for Somaliland.

Theme 1: Perspectives on the value of the training received

Providing training to TBAs was considered an important step by all participants. An IDI with a Somaliland MOH staff acknowledged the shortcoming of TBAs in their usual role as providers of maternity care. These two quotes refer to TBAs before they were trained to become MHPs.

During an FGD session, women voiced their reservation about the knowledge and skills of the TBAs.

All trainees interviewed valued their training; which was seen as important in upgrading skills, level and in some cases as a boost to means of livelihood. For example:

Some Trained TBAs felt that the process of being engaged in training meant that they had earned more respect from the women and their husbands and were more recognized at the community level and beyond their immediate village.

A Somaliland TBA reflected on what she has gained from the training which in her view has changed the way she worked significantly.

Theme 2: Perceptions and experiences of the impact of the training and the new TBA role.

In both countries, the training of TBAs was perceived to have had a positive impact on the communities by increased community sensitization on the need for utilization of health care services, by women and other members of the community and reduction of maternal and child deaths.

In Sierra Leone, pregnant women participating in focus group discussions discussed the benefits of utilizing hospital services and discontinuing deliveries at home. Many of them spoke about the advice trained TBAs had given them about the dangers of delivering at home and the importance of health center utilization.

Some of the trained TBAs also said that the pregnant women were unaware of the consequences of giving birth at home but now they know because of their training.

Most pregnant women in Sierra Leone had very positive views about the roles of the trained TBAs and a few even pointed out some deficiencies of TBAs who have not been trained. Most of the women spoke of the roles of the TBAs assigned to their villages and many said the trained TBAs go to their houses regularly to check how they and their babies are doing and ensure they take their medicines.

In Somaliland, participants also provided similar perspectives. In their view, the training of TBAs has also influenced the health seeking behavior of pregnant women in their community.

Within Sierra Leone, and as part of the post conflict health reconstruction process the Free Health Care initiative was launched in 2010 with the promise of free health care for pregnant and lactating women and under-fives. Analysis of interviews with health workers and Ministry of Health staff in Sierra Leone highlighted that even with the free health care initiative there were still many pregnant women delivering at home but since the training was completed there has been an increase in the utilization of services in the PHUs.

Most pregnant and newly delivered mothers in Sierra Leone emphasized that trained TBAs had stopped taking deliveries at home and that pregnant woman were utilizing the health centers.

Different participants discussed similar positive increases in health facility delivery post training Somaliland.

One key informant from Somaliland gave an account of a referral for utilization she witnessed.

With regard to maternal and child deaths, almost every participant in Sierra Leone mentioned that there had been a noticeable reduction in the deaths of pregnant women and their babies compared to what they were experiencing prior to the training of the Trained TBAs.

The health center staff said since the training there has been no death recorded at all in the health center.

The findings from Somaliland are complemented by the output indicators (proportion of TBAs’ referral and facility delivery rates) from the project documents. During the first six months of the training, 56% of total deliveries were referred, while 72% were referred in second year and 67% in the third year of the project. On the other hand, the desk review of project documents highlighted that total numbers of women received maternity care at the five maternity centers increased from 779 in 2009 to 3296 in 2012. Routine project data elicited that there has been a steady rise in facility-based deliveries from baseline (2009) peaking by mid-2012.The SBAs working at those maternity centers reported that the changing role of TBAs and linking up with the TBA with the facility contributed partly to the increase in skilled deliveries at health facilities.

In Sierra Leone, the research was carried out within the first year of the training hence, and the number of referrals made by MHPs from June 2013–April 2014 which was 28,640 (HPA and Ministry of Health).

The analysis in both contexts highlight how all different participant groups from pregnant women, to trained TBAs to key informants perceive the training has had a positive impact on community sensitization, health center utilization and maternal and child health. Within the Sierra Leonean context, the training and new roles for TBAs coincided with new bylaws that were being enforced by local village chiefs: any woman found delivering outside the health center, and those supporting her (e.g. TBAs trained or untrained) would be made to pay a fine of 50,000 Leones, (approximately USD$10). It was unclear how often this was enforced but it emerged as an issue within the FGDs and is likely to have also had an impact on increased numbers of health center deliveries.

Theme 3: Opportunities to strengthen integration of trained TBAs into the health system.

Within the Sierra Leonean contexts, trained TBAs and key stakeholders mentioned that there were now in an era of a good working relationship between trained TBAs and the health center staff. Trained TBAs felt that there were increasingly cordial relationships between them and health workers. One trained TBA explained that whenever she brings women to the health center, she feels happy because they are given prompt care.

This improved relationship has also been echoed in Somaliland. TBAs developed relationships with skilled birth attendants, and could call and bring women along to health centres without fear of intimidation. In Sierra Leone, SBAs were sensitized to the new role and some of them were involved in the training of the TBAs to become MHPs. Some TBAs were invited to help in the facilities. Furthermore, TBAs received feedback from SBAs on some of the women they had referred or escorted. One TBA expressed her thoughts on this.

Health workers are beginning to increasingly recognize the role of TBAs in their communities and the value in working with them to reach the community. As one health worker in Somaliland surmised;

Theme 4: Challenges to realizing and sustaining trained TBAs new role.

Our qualitative analysis in two different fragile country contexts both confirmed that with appropriate training and support it is possible to change the behavior and practices of TBAs so that they become part of a partnership with formal health providers and have stronger links with health facilities and health systems. To our knowledge this is the first paper that has explored perspectives on new approaches to TBAs roles in two different contexts. Trained TBAs can interact effectively with their communities, and in two different contexts were perceived to help overcome barriers to acceptability, utilization and contribute to effective demand for maternal and newborn services and impact on utilization of skilled birth attendance. Trained TBAs were able to perform non-clinical roles, linking women to care, and this has the potential to transform health seeking behavior and health outcomes for mothers and newborns in fragile and conflict affected settings and beyond.

Training and reorientation was essential for TBAs to adopt their new roles. The training provided in Somaliland and Sierra Leone focused on early referral of women for delivery at a health facility and providing companionship during labour and delivery. This contrasts with previous training modules, which emphasized building TBAs’ skills to safely conduct deliveries, recognize and refer women with complications [24]. In the two different contexts, and at different stages in the implementation of the projects (later in the HPA project cycle in Somaliland and earlier in Sierra Leone) the training of TBAs was viewed positively by the trained TBAs themselves, the health workers and pregnant and newly delivered women at community level.

Improved utilization of SBAs and reduction of home delivery are key maternal behaviours that could arguably improve outcomes for mothers and their newborns [25, 26]. The TBA training and reorientation to perform new roles - including referral and companionship – were perceived by different stakeholders (including women, TBAs, health providers and participants from the Ministry of Health) to have impacted maternal behaviours. This is consistent with the findings of a recent systematic review which showed that women were more likely to be referred if they lived near a TBA trained to refer compared to women who lived elsewhere [12]. Byrne and Morgan’s systematic review demonstrated that when TBAs are trained and supervised, they are more likely to refer complicated cases [16]. With the approach taken in both Somaliland and Sierra Leone, TBAs were trained and required to refer all women (with or without complications) for a facility-based birth.

Qualitative perspectives from trained TBAs, communities and key stakeholders from Ministries of Health and health centres in both study countries confirmed that trained TBAs had a positive impact on community sensitization for uptake of health services by mothers. Most trained TBAs mentioned during interviews their ‘commitment’ to the community in which they lived and worked. TBAs, like other close to community providers, are embedded in communities, and often have trusting and respected relationships with women at the community level and shared language and cultural framing of health issues [21]. In addition to being accompanied by TBAs, it is also possible that pregnant women were motivated to use the health care facilities because of the importance they attach to traditional norms and values, of which listening to the advice of trained TBAs whom they have a lot of respect for and feel comfortable with, is a component. A study that assessed the contributions of TBAs in maternal health in Ghana confirmed that TBAs enjoyed a lot of trust and respect in their communities and were in a position to influence women’s health behaviour [8]. Hence training and partnerships with TBAs which builds on their embedded and trusted community positionality has great potential.

The recent outbreak of Ebola in Sierra Leone changed the landscape on multiple fronts and brought additional human resource constraints. At the peak of the outbreak, health workers were severely affected, and many have lost their lives and those who remained faced multiple challenges in responding to the epidemic, due to limited prior knowledge on infection prevention and control practices. Sierra Leone’s fragile health system has been further weakened and it is likely that recent gains made in maternal health are being eroded. The level of mistrust between health service users and health professionals has increased as the virus continued to spread [27]. In the post Ebola reconstruction phase, efforts must be placed on bridging this gap. At the peak of the outbreak, it was documented that women and children are staying away from hospitals and health centres due to the fear and stigma associated with Ebola [28]. The additional burden posed on the already fragile health care delivery infrastructure by the Ebola virus disease epidemic and the underutilization of health facilities by this group puts them at even greater risk for adverse outcomes. There are reports of pregnant women dying at home of preventable illness; an estimated 400 mothers died between 20th May 2014 and 15th July 2014 from preventable illness [29], with an estimated 31% decline in institutional deliveries from May to Sept 2014 resulting in a corresponding rise in maternal case fatality rate from 1.27 to 3.08 in facilities providing comprehensive emergency obstetric and newborn care, which peaked at 3.48 at the height of the outbreak in Nov 2014 [30]. The impact of Ebola on trained TBAs is largely unknown, but they are very likely to have been affected as their gendered caring roles within households and communities means they are especially vulnerable to infection. In Liberia, it is estimated that 75% of Ebola cases are female and in Sierra Leone women have comprised 55 to 60% of the dead [31]. Lessons learnt from containing the Ebola outbreak in Uganda in 2001 highlighted the importance of rebuilding trust and collaborative working relationships with different community groups and structures and the importance of community health workers in this respect [32]. Close to community providers, such as TBAs, embedded within and trusted by communities are arguably strategically placed to help rebuild trust, in the post Ebola reconstruction phase. Strategies to support and sustain them will be particularly critical.

So how can trained TBAs be best supported to realize their roles? Our analysis shows that during the fieldwork period a cordial working relationship between health centre staff and trained TBAs seemed to have evolved in both Sierra Leone and Somaliland. As stated earlier, most of the SBAs were involved in the training of the TBAs and they were trained on topics that SBAs already practiced. Most TBAs reported that they were well-received and well-treated by the healthcare workers (skilled birth attendants) each time they referred or escorted a woman and this was highly valued. These positive reports contradict the report of abusive relationships between health care workers and TBAs that have been previously documented [33]. Findings from a systematic review have also emphasized the need for a cordial relationship between Trained TBAs/TBAs and health workers [16].

In resource-constrained, largely donor dependent settings and especially fragile settings like Sierra Leone and Somaliland, remuneration for TBAs, and other cadres remains a contentious issue. This is partly attributable to the fragility of the economy and limited resources available for health care delivery. The gains and pitfalls of paying TBAs for their services has been persistently debated [34] and emerged as a key theme in both contexts in our study. Traditionally, TBAs have had no formal payment package for their services, rather depending on the goodwill accorded to their roles and payment in cash or kind by their clients [35, 36]. In our study, the incentives paid to TBAs as they performed their new roles in both countries were largely viewed to be inadequate; and in both contexts TBAs may receive less overall remuneration in their new roles (i.e. USD$3/month in Sierra Leone and USD$5/month in Somaliland may well be less than what they would have received in donations/in kind delivering babies at home in the past). In Sierra Leone participants (women from the community and trained TBAs) felt that the new role given to the trained TBAs disrupted their farming activities as they had less time to work on their farms. Most trained TBAs and pregnant women who were interviewed mentioned that incentives should be increased because in their new roles TBAs would have no other source of income. However, in Somaliland five years after the training, the incentives were withdrawn but reports confirm that some TBAs continued with their new roles and continue to refer and accompany women to health facilities for delivery.

So how should trained TBAs’ contributions be appropriately remunerated and sustained? Within both contexts MOH staff discussed strategies and policy initiatives to incorporate the new TBA role into the health system as Community Health Workers (Sierra Leone) or Lady Health Workers (Somaliland). This is a very positive development – formally integrating and supporting trained new-role TBAs within the health system is likely to promote positive relationships and outcomes. Consultation and fair and transparent approaches to remuneration will be critical in making this a success.

In summary, our findings from multiple perspectives across two different FCAS highlight the possible gains of new roles for TBAs through linking them with and integrating them into the formal health system. Ministry of Health staff, health workers, trained TBAs and communities all provided positive perspectives on the new role of TBAs. These findings may open further conversations and policy dialogue on how the formal health sector can leverage the non-clinical skills of TBAs to increase skilled birth attendance rates for marginalized rural women. Conversations on the integration of TBAs (in their traditional role) into the formal health sector has largely met with resistance by formal health sector actors [37]. Proponents of integration argue that TBAs are close to the community and provide vital links to reaching the community as demonstrated in our study while opponents fear the possibility of TBAs taking on more than they can handle, delaying referrals and causing further complications for women [35, 38–41]. Our qualitative analysis of different perspectives has highlighted that TBAs when engaged in well-defined and suitable largely non-clinical roles have major contributions to make to maternal and newborn health as close to community providers. Their potential is arguably critically important in promoting UHC in FCAS where human resources for health are additionally constrained and MCH needs particularly acute.

Strategies to increase skilled birth attendance are particularly critical in fragile contexts where maternal and newborn health needs are great and human resources constrained. Close to community providers, such as TBAs, have strategic potential in reaching women in the community and linking them to skilled providers of maternity care. Training TBAs to refer and support women to deliver in health centres was positively received by a range of stakeholders in both Sierra Leone and Somaliland. Key challenges include sustainability and appropriate and fair remuneration. Evaluating similar approaches in different contexts will be important to generate further insights and perspectives into approaches to working with TBAs as close to community providers to improve maternal and newborn health. Even if our findings are not enough at this time to drive a policy shift towards renewed recognition and integration of TBAs as maternal health care providers in a role as health promoters, they add new insights and enrich the conversation around suitable and strategic roles for TBAs, which build on their embedded positionality.