Community-based doulas for migrant and refugee women: a mixed-method systematic review and narrative synthesis.

BACKGROUND: Community-based doulas share the same cultural, linguistic, ethnic backgrounds or social experiences as the women they support. Community-based doulas may be able to bridge gaps for migrant and refugee women in maternity settings in high-income countries (HICs). The aim of this review was to explore key stakeholders' perceptions and experiences of community-based doula programmes for migrant and refugee women during labour and birth in HICs, and identify factors affecting implementation and sustainability of such programmes.
METHODS: We conducted a mixed-method systematic review, searching MEDLINE, CINAHL, Web of Science, Embase and grey literature databases from inception to 20th January 2022. Primary qualitative, quantitative and mixed-methods studies focusing on stakeholders' perspectives and experiences of community-based doula support during labour and birth in any HIC and any type of health facility were eligible for inclusion. We used a narrative synthesis approach to analysis and GRADE-CERQual approach to assess confidence in qualitative findings.
RESULTS: Twelve included studies were from four countries (USA, Sweden, England and Australia). There were 26 findings categorised under three domains: (1) community-based doulas' role in increasing capacity of existing maternity services; (2) impact on migrant and refugee women's experiences and health; and (3) factors associated with implementing and sustaining a community-based doula programme.
CONCLUSION: Community-based doula programmes can provide culturally-responsive care to migrant and refugee women in HICs. These findings can inform community-based doula organisations, maternity healthcare services and policymakers. Further exploration of the factors that impact programme implementation, sustainability, strategic partnership potential and possible wider-reaching benefits is needed. © Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

Migrant and refugee women from low-income and middle-income countries (LMICs) are at increased risk of poor health outcomes and maternity care experiences in high-income countries compared with other women.

There is limited synthesis of research on community-based doula support for these communities.

This paper demonstrates that community-based doulas can improve migrant and refugee women’s maternity care experiences.

Community-based doula programmes increase the accessibility of doula care for women from migrant and refugee backgrounds.

Community-based doula programmes may enhance the cultural responsiveness of maternity care for these women.

Community-based doulas best complemented the maternity care team when doula roles were clearly defined and boundaries understood by both doulas and providers.

Community-based doula programmes should be promoted to ensure migrant and refugee women’s access to culturally-responsive maternity care.

New models and ways of financing need to be developed to support and sustain programmes.

Further research is needed to understand the possible social capital impact of community-based doula programmes for doulas, migrant women, their partners and families.

Introduction

Migration is a significant social determinant of health.1–3 Migrant populations in high-income countries (HICs) contribute considerably to the birth rate: in some HICs, migrants can represent up to 20% of all women giving birth.1 Migrant women from low-income and middle-income countries (LMICs), who have resettled in HICs often have poorer health outcomes and experiences of maternity care compared with non-migrant women in the same settings,4 and can face substantial barriers to accessing and using maternal health services.5 For example, migrant women in HIC have higher rates of maternal mortality,5 stillbirth, perinatal, neonatal and infant mortality, compared with non-migrant women.6 7

Inequalities in maternal health outcomes and experiences of care between migrant and non-migrant women may be due to suboptimal quality care.8 Migrant women may experience communication barriers and mistreatment, such as discrimination, racism, physical and verbal abuse.4 9 10 They may feel rushed through appointments, ignored in decision-making or disregarded during maternity care11 12 and experience frustration with a lack of continuity of care and trust.11 13–15 Although these experiences may not be unique to women from migrant backgrounds, intersecting social identities and experiences (such as race, ethnicity, religion, economic status, employment status) add to layers of stigma discrimination and mistreatment.15–17

Migrant women in high-income settings value empathetic and respectful healthcare providers, who listen to and address their concerns and cultural differences.11 14 18–20 These positive healthcare interactions increase rapport, and increase confidence and sense of identity in maternity settings.11 21 22 A Cochrane qualitative evidence synthesis found that community-based doula support may be a way to enhance respectful interactions and culturally-responsive care.23 Culturally-responsive care refers to care that respects a person’s cultural needs, values and traditions.24 The absence of this type of care can compromise the health outcomes and experiences for migrant women.25–28

Community-based doulas

Doulas are people trained in providing continuous emotional, social, physical and advocacy support throughout a woman’s labour and birth.23 29–32 A Cochrane review found that women who received continuous support throughout labour and childbirth (eg, from a doula) had better outcomes, including reduced risk of caesarean section, shorter duration of labour and more positive childbirth experiences.32 Doulas establish rapport and develop therapeutic relationships with women; provide knowledge to make informed decisions; and facilitate open communication and respectful relationships with providers.23 30 31 33–36 Doulas may provide continuity of care during pregnancy, labour and birth, postnatally or throughout a combination of these periods.37

Community-based doulas share the same cultural, linguistic, ethnic backgrounds and/or social experiences (ie, sociodemographic status) as the women or birthing people they support and may be beneficial in bridging gaps in the provision of culturally-responsive maternity care.38 39 Community-based doulas may also not share these characteristics and be matched with women based on a needs-based approach (ie, socioeconomic hardship).40 Community-based doula programmes have emerged in response to the cultural needs of women, particularly from underserved communities (eg, lower-income, adolescents, black or minority ethnic women), not being met by maternity systems.32 36 These programmes may offer accessible, free doula support and are often operated by not-for-profit or community-based organisations with doulas working as volunteers.39–41

Community-based doula programmes may be a strategy to address the negative experiences and poorer health outcomes faced by migrant women in high-income maternity settings. However, questions remain about the structure and impact of these programmes, how they may provide culturally-responsive healthcare, including the perceptions and experiences of migrant women, their families and providers of community-based doula support. Existing reviews on similar topics23 32 42 have explored providing continuous support or labour companionship to all women and from different types of companions. No reviews specifically explored community-based doulas supporting migrant women. The aim of this paper was to explore key stakeholders’ perceptions and experiences of community-based doula programmes for migrant women in HICs, and to identify factors affecting successful implementation and sustainability of community-based doula programmes.

Methods

We conducted a mixed-method systematic review and report according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA43) (online supplemental appendix 1), Enhancing transparency in reporting the synthesis of qualitative research: (ENTREQ) statement44 (online supplemental appendix 2) and based on guidance from the Cochrane Effective Practice and Organisation of Care group.45 The review protocol has been published (PROSPERO: CRD42020193216).

Types of studies

We included primary studies that used qualitative, quantitative or mixed-methods designs. We excluded studies that were secondary analyses, reviews, news articles, commentaries, opinions, editorials, case studies, protocols or conference abstracts.

Topic of interest

We included studies that primarily focused on perceptions and experiences of community-based doulas supporting migrant or refugee women during labour and birth in any HICs (per World Bank category46), and in any type of health facility (eg, hospitals, birth centres). We defined key stakeholders as women from migrant or refugee backgrounds who were from LMICs and resettled in HICs, their partners, community-based doulas, healthcare providers and others such as programme managers or policymakers. We included studies with both community-based doulas and doulas who were experienced in supporting migrant women and their communities, but did not have ethnic, cultural or linguistic commonalities.

We excluded private-practising doulas and hospital-based doula programmes39–41 47; studies that did not explore community-based doula care during labour and birth and did not specify their clients included migrant women from LMICs; and doula care that took place during home birth, due to the inherently different processes and nature of care from facility-based settings.

Search methods for identification of studies

We searched MEDLINE; CINAHL (EBSCO); Web of Science; Embase (Ovid) databases from inception to 20th January 2022, as well as grey literature databases (online supplemental appendix 3). Search strategies were developed based on two existing Cochrane reviews on similar topics23 32 and consultations with a research librarian (online supplemental appendix 3). There were no limits on language or date of publication. Reference lists of included studies were searched, and forward citation of included studies was also conducted using Google Scholar.

Selection of studies

We imported the search results into Covidence48 and removed duplicates. Two authors (SMK and RIZ) independently reviewed titles and abstracts. Full-text articles were uploaded into Covidence and independently reviewed by two authors (SMK and RIZ). Discrepancies were managed through consensus with a third author (MAB) as needed. Where multiple papers from the same study were identified, these articles were grouped together as one study.

Data extraction and assessing methodological limitations

Two authors (SMK and RIZ) extracted data using a template including: study setting, research questions, research design, participants, programme characteristics, ethical considerations, data collection and analysis, findings (including themes, quotations and interpretations), limitations, conclusions and relevant tables or figures.

The same two authors independently critically appraised all included studies. Qualitative studies were appraised using the Critical Appraisal Skills Programme tool (online supplemental appendix 4) assessing research aims and design, recruitment, reflexivity, ethics, data collection and analysis and contribution to research.49 The quantitative and mixed-methods studies were assessed using the Mixed-Methods Appraisal Tool (online supplemental appendix 5).50 These tools enabled us to assess the methodological limitations of included studies.

Data management, analysis and synthesis

We used Popay’s (2006) narrative synthesis approach across four stages: (1) developing a preliminary synthesis; (2) exploring the relationships between studies; (3) assessing the robustness of the synthesis; and (4) developing a theoretical model.51 These steps were conducted iteratively and concurrently and are described in the following sections.

Developing a preliminary synthesis

We conducted an inductive qualitative thematic synthesis (based on Thomas and Harden’s approach52 and quantitative data synthesis through textual descriptions synthesis of the study designs and findings (SMK, RIZ and MAB).51 One highly relevant article was analysed by SMK using a line-by-line free coding approach for the foundations of a code book.23 32 These initial codes were checked on three other articles to ensure that these concepts were relevant and meaningful to other studies.23 32 We conducted line-by-line free coding on the articles’ results section data (eg, themes, participant quotes) and where authors summarised their findings. Text supporting each code was organised and colour coded to differentiate perspectives. We used NVivo software for analysis.53

Exploring relationships within and between studies

Descriptive themes were developed, reflecting those in the included studies’ findings. Then hierarchical analytical theme analysis was conducted to investigate key themes from the preliminary synthesis and to understand the relationships within and between included studies.51 The codebook was iteratively developed and refined, and higher order analytical themes were represented through summary of qualitative finding statements.52

Due to limited quantitative data, meta-analysis was not possible and we used textual descriptions to synthesise information on study design, health outcome measures and other main results.51 Articles containing relevant quantitative data was coded in NVivo. Textual summaries of quantitative data were created in an Excel spreadsheet and then analysed using textual descriptive analysis (SMK and RIZ).51

Assessing the robustness of the synthesis

The GRADE-CERQual approach (online supplemental appendix 9) was used to assess the confidence in each qualitative findings,54 55 using the following domains: methodological limitations,56 relevance,57 adequacy58 and coherence59 to assess the confidence. After reviewing each domain, we assessed the overall confidence55 as very low confidence, low confidence, moderate confidence or high confidence (SMK, RIZ and MAB).

Developing a theoretical model

Findings contributed to developing a ‘theory of change’ to understand how and why community-based doula support worked as an intervention, who benefited from it and how it may provide culturally-responsive care.51

Findings

Eighteen papers from 12 studies were included (figure 1; online supplemental appendies 6 and 7).38–41 60–73 Ten included studies were published in peer-review journals,38–40 60 61 63–68 70–74 one was a published book chapter41 and one a dissertation.62 The included studies were diverse in terms of study setting, methods, target population, community-based doula training and community-based doula characteristics. Studies were from four countries: Sweden (n=4),39 41 60 68–70 the USA (n=6),38 61 62 71–73 England (n=1)40 63–66 and Australia (n=1).67 Most studies were conducted in maternity clinics and hospitals in urban settings,38 39 60 61 67–69 71–73 75 with some focusing on low-income communities in urban settings40 63–65 and others a mix of urban, rural and small towns.41 70 Eleven studies used qualitative methods only (semi-structured interviews, focus groups and/or participant observation).38 39 41 60–62 67–73 The one remaining study from England used mixed-methods, consisting of five papers.40 63–66

Figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram. HIC, high-income country.

Different participant perspectives were included: one study with migrant women,39 three studies with community-based doulas,61 62 72 one study with healthcare providers,60 one study with key informants of a community-based doula programme67 and six studies with mixed-perspectives.38 41 68–71 73

There were variations in the community-based doula programmes related to terminology, clientele and doula remuneration (detailed in online supplemental appendix 8). First, support was described as provided by community-based doulas6 38 39 60–62 73; volunteer doulas2 40 63–67; community-based bilingual doulas1 68 69; and cultural interpreter doulas.1 41 70 The remaining two studies used the name of the programme to describe their doulas.71 72 Most programmes focused on: newly arrived migrant women39 41 60 68–71; refugee women38 72; women of colour including migrant and refugee women61 62 73; and women from underserved communities including women from migrant and/or refugee backgrounds.40 63–67 In terms of doula remuneration, four studies reported on programmes with unpaid volunteer doulas,39 40 60 63–67 four had salaried doulas,41 62 68–70 73 one had limited reimbursement paid per client (US$100)71 and three did not report reimbursement.38 61 72

Detailed critical appraisals of included studies are available in online supplemental appendies 4 and 5. Many qualitative studies provided limited justifications about research design and recruitment strategies, data collection and analysis methods and ethical considerations, and most qualitative studies did not discuss reflexivity. For the included mixed-method and quantitative studies, the primary concern identified was limited reporting of non-response bias.

Twenty-six qualitative review findings were developed (table 1); using the GRADE-CERQual approach, 10 were high confidence, 8 were moderate confidence, 6 were low confidence and 2 were very low confidence (online supplemental appendix 9).54 55

Table 1

Summary of qualitative findings

#	Summary of review finding	Studies contributing to the review finding	Overall CERQual assessment	Explanation of overall assessment
F1.0	Continuous woman-centred supportCommunity-based doulas provided individualised, woman-centred, continuous emotional, social and physical support for migrant women throughout their labour and birth. Community-based doulas may fill the gap in trained continuous emotional, social and physical support in labour and birth for migrant women in maternity settings.	39–41 60–62 65 66 68 69 71 72	High confidence	No or very minor concerns on coherence and adequacy, minor concerns on relevance (3 out of 9 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F2.0	Knowledgeable in childbirth and navigating maternity systemsCommunity-based doulas were perceived as knowledgeable sources of informational support, providing migrant women with childbirth education and guidance in accessing and navigating their new country’s maternity systems.	39–41 60 65 66 68–71	High confidence	No concerns on coherence, minor concerns on relevance (1 out of 6 studies may include non-migrant women’s perspectives), minor concerns on adequacy (6 out of 12 studies supported this finding) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F3.0	Engaging partnersCommunity-based doulas provided support to the partners of migrant women by enhancing their connection and involvement during labour and birth, particularly when there were differing cultural expectations of partners in providing birth support in high-income countries.	39 41 60 62 65 66 69	High confidence	Minor concerns on adequacy (6 out of 12 studies supported this finding), no or very minor concerns on coherence, minor concerns on relevance (2 out of 6 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F4.0	Shared language and culture benefitsCommunity-based doulas created a culturally safe space in labour and birth for migrant women, particularly if they were newly arrived or experiencing social isolation in their new countries. This was enhanced when community-based doulas shared the same language and/or culture as the woman, as this helped to reduce the barriers (eg, cultural or language) migrant women may experience when accessing and using maternity care. These community-based doulas were also perceived as cultural facilitators for both migrant women and healthcare providers.	38–41 60–62 66–72	High confidence	No or very minor concerns on coherence and adequacy, minor concerns on relevance (4 out of 12 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F5.0	Respectful treatment and advocacyCommunity-based doulas’ presence and support enhanced respectful treatment of migrant women from healthcare providers. In circumstances where community-based doulas perceived mistreatment, they actively advocated for migrant women by communicating to healthcare providers (direct advocacy).	38 40 41 61 62 66 67 69	Moderate confidence	No concerns on coherence, minor concerns on relevance (4 out of 7 studies may include non-migrant women’s perspectives), moderate concerns on adequacy (7 out of 12 studies supported this finding, 3 moderate to thick data and 4 thin data) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F6.0	Going above and beyond due to social justice work and volunteeringCommunity-based doulas demonstrated high motivation and commitment to continuous support for migrant women in labour and birth where in some circumstances doulas went above and beyond their expected roles taking on added responsibilities, despite being unpaid volunteers. This was often driven by their motivation to advocate and support migrant women from their own communities due to a strong sense of social justice.	40 41 60–62 65 66 68 69 71 72	High confidence	No or very minor concerns on coherence, minor concerns on adequacy (6 moderate to thick data, 2 thin data), minor concerns on relevance (3 out of 8 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F7.0	Considerations with non-matched community-based doulasCommunity-based doulas who were carefully matched with migrant women by a needs-based approach (eg, socio-economic hardship) rather than culturally or linguistically, were well received however, training into cultural backgrounds and having access to interpreters was recommended to ensure quality of care, positive relationships and rapport was established.	40 66 67 72	Moderate confidence	No or very minor concerns on coherence, minor concerns on relevance (2 out of 3 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, recruitment, reflexivity, data collection) and moderate concerns on adequacy (3 out of the 12 studies supported this finding).
F8.0	Role of community-based doula and position in maternity healthcare teamCommunity-based doulas were perceived as different from interpreters and trusted members of the maternity healthcare team, particularly when healthcare providers understood and valued their roles—which was enhanced when they gained experience working together. The roles and expectations of community-based doulas in providing non-clinical support to migrant women (eg, emotional and social) in contrast to healthcare providers’ clinical care should be clearly defined to strengthen collaborative care opportunities and relationships.	39–41 60 62 64 66–72	High confidence	No or very minor concerns on coherence and adequacy, minor concerns on relevance (3 out of 9 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data analysis, support for findings, data collection and analysis).
F9.0	Communication, interpretation and informed decision making/consentCommunity-based doulas enhanced communication between both migrant women and their healthcare providers during labour and birth, through language interpretation or translating to plain language. They also validated and encouraged migrant women to communicate with healthcare providers (indirect advocacy) and enhanced migrant women’s understanding of informed consent in labour and birth.	39–41 60 61 65 66 68 69 71 72	High confidence	No or very minor concerns on coherence and adequacy, minor concerns on relevance (2 out of 8 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F10.0	Difference from private practising doulasSome healthcare providers were often cautious when working with community-based doulas due to previous negative experiences with private-practicing doulas. However, community-based doulas were perceived as favourable and different by healthcare providers once they had more experience working with doulas and understood programmes were often not-for-profit volunteer organisations, supporting underserved communities such as migrant and refugee populations.	40 64 69	Low confidence	No concerns on coherence, minor concerns on relevance (1 out of 2 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, reflexivity, data collection and analysis) and serious concerns on adequacy (2 out of the 12 studies supported this finding).
F11.0	Social connectivity and emotional relationshipsCommunity-based doulas created emotional connections with migrant women similar to the intimate relationships shared with a family member or friend, which established a sense of trust. This relationship helped migrant women to feel a sense of community and social connectedness in their new country of residence.	39–41 60–63 65 66 68 69 71 72	High confidence	No or very minor concerns on coherence and adequacy, minor concerns on relevance (3 out of 9 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F12.0	EmpowermentMigrant women, community-based doulas and healthcare providers felt women were empowered during their labour and birth through the encouragement and reassurance that community-based doulas offered. Migrant women felt that the community-based doula’s belief in their abilities to labour and birth transformed their own beliefs of feeling fear, incompetence or self-doubt to confidence in their own capabilities.	39 60–62 65 66 71	High confidence	No or very minor concerns on coherence, minor concerns on relevance (3 out of 6 studies may include non-migrant women’s perspectives), minor concerns on adequacy (6 out of 12 studies supported this finding) and moderate concerns on methodological limitations (research design, recruitment, reflexivity, ethics, data collection, support for findings).
F13.0	Longer-term benefits for community-based doula supportMigrant women and their families may experience long-term benefits to their health and well-being due to community-based doulas signposting migrant women to relevant support services or through health promotion of specific public health priorities. Community-based doula programmes may provide doulas and their local communities with educational or employment opportunities that may have been limited to them prior to their involvement with the programme. A database is recommended to record and demonstrate the possible long-term family-centred impacts of community-based doula programmes on underserved populations and to gain potential investment opportunities.	40 41 62 64 68 73	Moderate confidence	No concerns on coherence, minor concerns on relevance (3 out of 5 studies may include non-migrant women’s perspectives) moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings), moderate concerns on adequacy (5 out of 12 studies supported this finding).
F14.0	Impact of limited clarity of community-based doula role in the maternity care teamWhen the purpose, roles and boundaries of community-based doulas within labour and birth were unclear among community-based doulas and healthcare providers, tension and conflict may result. Community-based doulas who engaged in clinical decision-making or roles outside of their scope of practice were negatively received by healthcare providers. Whereas healthcare providers either misperceived them as interpreters often rejecting their roles (ie, advocacy, birth support), delegated them clinical tasks, or perceived them as replacing providers' or women’s partners' roles.	40 41 60 62 64 66 68–70	High confidence	No or very minor concerns on coherence, minor concerns on adequacy (5 out of 12 studies supported this finding), minor concerns on relevance (2 out of 5 studies may include non-migrant women’s perspectives) and moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings).
F15.0	Impact of limited clarity of community-based doula role with being on-callSome community-based doulas were unable to provide on-call support for migrant women in labour and birth due to their own individual or cultural limitations working during night-time hours. Migrant women expressed disappointment in their doulas’ absence at their labour and birth as they had desired continuity of doula support.	39 40 64 65	Low confidence	No or very minor concerns on coherence and relevancy (1 out of 2 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, recruitment, reflexivity, ethics, data collection) and serious concerns on adequacy (2 out of 12 studies supported this data).
F16.0	Limited continuity due to meeting too lateSome migrant women who first met their community-based doulas in labour and birth were disappointed in the limited continuity of support offered by their doulas and would have preferred to establish relationships with their doulas earlier in their pregnancy.	39 40 65	Low confidence	No or very minor concerns with coherence, minor concerns with relevance (1 out of 2 studies may include non-migrant women’s perspectives), moderate concerns for methodological limitations (research design, recruitment, reflexivity, ethics, data collection) and serious concerns with adequacy (2 out of 12 studies supported this finding).
F17.0	Limited childbirth knowledgeSome migrant women expressed dissatisfaction and feeling unsupported during labour and birth when they perceived a gap in their community-based doulas’ training, specifically when doulas lacked specialised childbirth knowledge, birth support training and culturally competent care. Similarly, some doulas expressed inadequate knowledge and desired more education.	39 40 68 71	Moderate confidence	No or very minor concerns on coherence, minor concerns on relevance (1 out of 4 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, recruitment, reflexivity, ethics, data collection and analysis, support for findings) and moderate concerns for adequacy (4 out of 12 studies supported this finding).
F18.0	Establishing credibility and marketing of community-based doula programmesCommunity-based doula programmes implemented by a non-profit organisation external to maternity settings experienced concerns with networking opportunities and accountability that led to challenges in establishing credibility and receiving referrals to the programme. Similarly, volunteer programmes with doula-like support may limit recruitment and funding opportunities available for community-based doula programmes. Community-based doula programmes which engaged in community networking opportunities and advertised doula services to prospective clients and organisations outside of maternity settings may improve the credibility, marketing and uptake of programme services.	40 41 64	Low confidence	No concerns on coherence, minor concerns on relevance (1 out of 2 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings) and serious concerns on adequacy (2 out of 12 studies supported this finding).
F19.0	Funding and sustainabilityThe impact of partially or unfunded community-based doula programmes meant that programme staff and local champions were tasked with sourcing alternative funding streams (eg, paid doula work) in the short-term. This included some non-profit organisations using fundraising or paid community-based doula models to sustain their programmes. Networking and establishing strategic partnerships may appear to be valuable in strengthening community-based doula programme funding applications and pursuing alternative collaborative funding streams. Community-based doula programmes may have to consider aligning their priorities to local public health priorities areas to be considered for funding opportunities and as a possible cost-effective measure.	40 62 64 73	Moderate confidence	No or very minor concerns on coherence, minor concerns on relevance (3 out of 3 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, reflexivity, ethics, data collection) and moderate concerns on adequacy (3 out of 12 studies supported this finding).
F20.0	Professionalisation and organisation of community-based doula programmesCommunity-based doula programmes may need to change current organisational processes to attract prospective clients and to motivate and retain volunteers. This may include flexibility among recruitment of clients (eg, broadening eligibility criteria) and volunteers (eg, alternative referee options, contracts) and including volunteers in strategic programme meetings and mandatory training opportunities.	40 62 64	Moderate confidence	No concerns on coherence, minor concerns on relevance (2 out of 2 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, reflexivity, ethics, data collection) and serious concerns with adequacy (2 out of 12 studies supported this finding).
F21.0	Challenges with the resource-intensiveness of community-based doula programmesCommunity-based doula programmes are resource intensive and are often faced with logistical and financial challenges, particularly if there are additional expenses associated with maintaining the programme. Programmes often required a dedicated team to run them, in which staff appeared to go beyond their role by volunteering their time to cover shortfalls in programme activities.	40 64	Low confidence	No concerns on coherence, minor concerns with relevance (the only included study may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research aims and design, recruitment, reflexivity, ethics, data collection and analysis, support for findings) and serious concerns on adequacy (1 out of 12 studies supported this finding).
F22.0	Motivation of community-based doulas and their engagement with the programmeCommunity-based doula’s own motivation and time available to be a doula may impact how motivated and engaged they were in their role. Similarly, community-based doula programmes which had limited clients to support found doulas became unmotivated if they remained unmatched.	40 64 67	Low confidence	No concerns on coherence, minor concerns on relevance (2 out of 2 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, recruitment, reflexivity, data collection) and serious concerns on adequacy (2 out of 12 studies supported this finding).
F23.0	Challenges with receiving limited or no numeration and demanding workSome community-based doulas perceived being unpaid volunteers or receiving minimal reimbursement for their services as challenging especially if they were from the same communities as the migrant women they were supporting. Similarly, some women expressed feeling challenged knowing that their community-based doulas were unpaid volunteers.	62 65 68	Very low confidence	No concerns on coherence, minor concerns on relevance (2 out of 3 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, reflexivity, ethics, data collection) and serious concerns with adequacy (3 out of 12 studies supported this finding).
F24.0	Demanding nature of community-based doula workCommunity-based doulas expressed the demanding nature of being on-call through the physical and emotional toll on themselves, their partners and families. For some employed doulas, they found timekeeping their birth work hours challenging given the on-call nature of their work, and possibly are restricted in how much support they can provide, particularly when balancing other commitments (eg, paid work).	62 68 73	Moderate confidence	No or very minor concerns with coherence, minor concerns with relevance (2 of the 3 studies may include non-migrant women’s perspectives), moderate concerns with methodological limitations (reflexivity, ethics, data analysis) and moderate concerns with adequacy (3 out of 12 studies supported this finding).
F25.0	Mentorship and support opportunities may facilitate community-based doulas’ motivation and engagement with programmeWhen programmes ensured that community-based doulas felt supported and valued within their organisation, a sense of community among doulas and programme staff ensued. Mentorship opportunities may be valuable in motivating and creating the sense of community among community-based doulas and programme staff, especially for newly recruited volunteer doulas.	40 62 64 68	Moderate confidence	No concerns on coherence, minor concerns on relevance (2 out of 3 studies may include non-migrant women’s perspectives), moderate concerns on methodological limitations (research design, reflexivity, ethics, data collection) and moderate concerns with adequacy (3 out of 12 studies supported this finding).
F26.0	Moving towards paid community-based doula models may sustain the workforce and provide financial and health benefit for doulas and their familiesSome community-based doulas working as employees of a community-based doula programme received hourly remuneration; a set number of hours per week; paid attendance to meetings and training, leave entitlements; and health insurance for both doulas and their families. Doulas felt this paid-model symbolically represented value and recognition of the importance of their birth work, particularly as they were supporting women from their communities.	73	Very low confidence	No concerns on coherence, minor concerns on relevance (the only included study may include non-migrant doulas' perspectives), minor concerns on methodological limitations (ethics) and serious concerns with adequacy (1 out of 12 studies supported this finding).

Narrative synthesis

Three domains were developed from qualitative findings: (1) community-based doulas’ role in increasing capacity in maternity services; (2) impact on migrant and refugee women’s maternity experience and health; and (3) factors associated with implementing and sustaining a community-based doula programme (table 2). As there was limited quantitative evidence (one mixed-method study and three papers) a summary of which quantitative results supported qualitative review findings will be discussed after qualitative findings.

Table 2

Domains and themes

Domain	Themes
1. Community-based doulas’ role in increasing capacity in maternity services.	Trained labour and birth support. Culturally-responsive and respectful care. Complementary support to the maternity care team.
2. Impact on migrant and refugee women’s maternity experiences and health.	Immediate and short-term benefits. Longer-term benefits beyond maternity care.
3. Factors associated with implementing and sustaining a community-based doula programme.	Limited community-based doula role clarity. Sustaining the community-based doula programme as an external organisation. Sustaining the community-based doula workforce.

Community-based doulas’ role in increasing capacity in maternity services

Trained labour and birth support

Community-based doulas were trained in providing individualised, woman-centred, continuous, emotional, social and physical birth support which included non-pharmacological pain-relief measures.39–41 60–62 65 66 68 69 71 72 Doulas often established relationships with migrant women during pregnancy and were on-call for labour and birth support.39 60–63 65 66 69 71 72 Doulas were perceived by both migrant women and professional maternity care providers as approachable and having expertise in supporting labour and birth, and navigating the maternity system, particularly if they were former service users or experienced in birth support.39–41 60 65 66 68–71 Migrant women, doulas and providers recognised that doulas who had shared culture or language potentially bridged cultural understandings and differences, which allowed partners to actively participate in supporting providing birth support, especially if they had limited birth support knowledge.39 41 60 62 65 66 69

Culturally-responsive and respectful care

Community-based doulas who had shared culture or language helped create a culturally safe space in labour and birth for migrant women.38–41 60–62 66–72 Bicultural doulas felt strongly about protecting their shared culture and values and appeared to be valued by migrant women for providing a sense of familiarity through cultural connection, security and rapport in the birth space.40 60–62 69 71 72 They were also seen as more approachable when addressing sensitive issues (eg, family violence).40 71 However, being cultural facilitators places doulas in challenging situations such as where conflicting expectations arise between women and providers.40 41 69 Two studies matching migrant women through a needs-based approach rather than by culture or language, were well received; however, in some cases doulas faced communication challenges and difficulties establishing rapport with migrant women, were dependent on interpreters who were often difficult to access, and experienced criticism from providers.40 66 67 72

Migrant women, doulas and programme managers shared the perspective that a community-based doula’s presence held providers accountable and perhaps changed their practices.41 61 62 66 67 This became evident in circumstances where informed consent was not explicitly provided61 66; in preserving birth preferences40 62; or questioning mistreatment witnessed.61 71

Doulas appeared to be motivated by their sense of connection, advocacy and satisfaction in supporting migrant women in labour and birth as racism and discrimination can be systemic and structurally ingrained in maternity settings.40 41 60–62 65 66 68 69 71 72

Complementary support to the maternity care team

Community-based doulas were valued as important members of maternity teams when their non-clinical support roles were understood.39–41 60 62 64 66–72 This was demonstrated by midwives and obstetricians expressing relief in sharing labour support responsibilities with doulas, especially when busy and overstretched, which allowed them to focus on providing essential clinical care.40 41 60 62 66 69 71 72

Migrant women, doulas and healthcare providers valued how doulas provided non-judgemental knowledge on childbirth information and presented options available for medical interventions enhancing informed decision-making.40 41 60 61 65 66 69 71 72 Doulas engaged in non-verbal communication or translated medical terminology into plain language to support women’s comprehension.40 66 69 Similarly, if doulas perceived mistreatment (ie, providers making decisions without informed consent) they would encourage migrant women to communicate with providers to ask questions as a form of indirect advocacy.66 Some healthcare providers perceived community-based doulas more receptively than private practising doulas as they gained more experience with the programme and working with the doulas.40 64 69

Impact on migrant and refugee women’s maternity experiences and health

Immediate and short-term benefits

Community-based doulas established trust and social connectedness for migrant women in their new countries.39–41 60–63 65 66 68 69 71 72 Doulas were commonly perceived as family members or friends when they shared culture or language,39 41 60–62 66 69–72 whereas doulas with no commonalities were perceived by their clients as friends.40 Migrant women, community-based doulas and providers felt women were empowered during their labour and birth by the doula’s encouragement and reassurance.39 60–62 65 66 71 Professional healthcare providers also valued doulas who were confident and competent in their birth support abilities, particularly when they empowered their client to make decisions as this appeared to enhance their client’s confidence.60

Longer-term benefits beyond maternity care

Longer-term benefits associated with community-based doula support were also important.40 41 62 64 68 73 Intersecting barriers possibly impacting migrant women may be addressed through doulas’ signposting and health promotion of specific antenatal priorities.40 41 64 Similarly, education and employment opportunities (eg, private doula practice, midwifery qualifications) were afforded to doulas because of the skills gained from doula accreditation and volunteer work experience.40 64 73

Factors associated with implementing and sustaining a community-based doula programme

Limited community-based doula role clarity

Limited clarity on the role of community-based doulas in the maternity care team influenced their level of acceptance.40 41 60 62 64 66 68–70 Doulas who only provided language support, acted unprofessionally, provided clinical support beyond their scope or participated in decision-making against providers’ advice, were negatively received and created tension among providers.41 60 69 70 Providers’ perception of doulas duplicating or taking away their emotional and social support roles, leaving them with clinical roles often led to providers feeling threatened, which made providers either physically ignore or be dismissive towards doulas.40 41 60 66 69 70

Some studies discussed how doulas may not understand aspects of their role, specifically the expectations in being on-call 24/7 around their client’s due date.39 40 64 65 Migrant women in the same studies39 40 65 expressed disappointment in not meeting their doulas during their pregnancy or too late in pregnancy.39 40 65 Meeting and developing rapport with their backup doula if their primary doula was unable to be present at birth was also valued by women.65

Another issue identified was community-based doulas requiring further childbirth education.39 40 68 71 This highlighted the need for some programmes to educate and train doulas in emotional and physical support and recruit doulas who are compassionate and supportive.39 71 Likewise, migrant women expressed that culturally competent care is essential especially when supporting women from refugee or asylum-seeking backgrounds, to ensure that the care provided is responsive to their needs during labour and birth.40

Sustaining the community-based doula programme as an external organisation

Community-based doula programme staff which included managers and workers responsible in supporting doulas and matching processes suggested establishing the credibility of community-based doula programmes by clearly differentiating the programme’s aims and client eligibility criteria when marketing services and accepting referrals.40 41 64 Programme staff also valued incorporating an interdisciplinary approach or collaborative partnerships with relevant community organisations or hospitals.64

Community-based doulas and leaders, programme staff and commissioners suggested strategies in sourcing funding and sustaining the programme.40 62 64 73 Commissioners, responsible for the planning, funding and monitoring of healthcare services in England, suggested promoting the cost-effectiveness of volunteer programmes to align with broader public health agendas.40 Other strategies included relevant organisations or hospitals providing additional services (eg, training, interpreters) to support programmes at reduced or no expense which may be symbolic of reciprocal collaborative relationships.40 64

Improving the professionalisation and organisation of community-based doula programmes could potentially lead to recruiting and retaining motivated and committed volunteers.40 62 64 Programme staff perceived including non-negotiable terms within volunteers’ contracts before attaining accreditation or leaving the programme as essential to achieve meaningful financial investment in training and participation of doulas.40 64

Programme staff highlighted additional resource-intensive processes which included: the recruitment, external accreditation and training of doulas; security and interpreter services; retraining of essential programme staff due to turnover; and supporting women with complex social support needs which required dedicated debriefing, supervision and ongoing professional training.40 64 Programme staff expressed that the short-term strategies they engaged with (eg, doula support, administration or security) and the impact of reduced staffing, diverting funding and postponing volunteer training was often to their own expense and programme offerings.40 64

Sustaining the community-based doula workforce

Community-based doulas’ motivation and engagement with the programme was identified as influential in sustaining the programme’s workforce.40 64 67 The overarching issue is the apparent misalignment of both programmes’ and doulas’ own motivation and expectations in their role.40 64 67 Programme staff expressed that too few or too many client referrals also impacted allocation and matching processes.64

There were challenges with doulas receiving limited or no renumeration and the demanding nature of the work.62 65 68 In one study, some women felt they were unable to ask their volunteer doulas for additional support.65 In other circumstances, doulas themselves could be experiencing financial hardships.62 Programmes offering reimbursement typically did not reimburse at rates equivalent to hours served.62 68 Participants in one study suggested that sustainability could be improved by having salaried community-based doula programmes rather than working as paid independent contractors.73 These doulas expressed the potential benefits they received with secure employment for both themselves and their families.73

The unpredictability of being on-call and supporting clients for extensive hours was perceived by doulas as being both emotionally and physically taxing.62 68 73 The difficulties in disconnecting from work when home and the reliance on support from their own partners and families may demonstrate the need for more support being available for doulas.62

Both doulas and programme staff proposed mentorship and support opportunities may facilitate community-based doulas’ motivation and engagement with programme.40 62 64 68 Strategies in strengthening supportive relationships included: availability of programme staff to support doulas; accessible debriefing opportunities and counselling services; and supervised training opportunities.40 62 64 Mentorship opportunities would involve experienced doulas supporting new doulas to orientate them to the programme.40 64

Quantitative findings

Sixteen of the 26 qualitative themes (table 2) were also reflected in the included quantitative evidence (online supplemental appendix 10 presents all quantitative findings). All quantitative data was from three papers in the Spiby et al (2015) volunteer doula study for women from underserved communities including minority ethnic backgrounds.40 63 65 The two qualitative themes regarding the short-term and long-term benefits of community-based doulas were supported by quantitative evidence such as the positive relationship between community-based doula support and increase in knowledge about childbirth and skills (eg, caring for child) among recently arrived migrant women.40 The longer-term benefits also extended to the community-based doulas themselves, where over half of trained doulas considered transferring their acquired skills towards possible paid employment or towards careers in social or healthcare (67%).40 63

The themes about complementary support to the maternity care team was also reflected through the perspectives of women who reported feeling that their midwives and doulas had worked well together most of the time.40 Similarly, doulas reported feeling that they worked well together with midwives in labour most of the time.40

While community-based doula support as a means to provide culturally-responsive and respectful care was highly valued in the qualitative findings, there were limitations of this in the implementation of the quantitative studies. For example, in one study, only half of women were linguistically or culturally matched to their doulas, and the cultural and linguistic mismatch was reported as communication challenges by both women and doulas.63 In one study, 70% of doulas reported that having a shared background to women was not important, and almost all believed establishing positive relationships with women was more important.63

There was no quantitative evidence to support qualitative themes in sustaining the community-based doula programme (findings 20–21 absent) and workforce (findings 22–26 absent).

Discussion

Our review shows how community-based doulas can improve the experiences of migrant and refugee women resettled in HICs, particularly when they were trained, knowledgeable and experienced in providing support in labour and birth. Doulas complemented the maternity care team best when roles were clearly defined and boundaries were understood by both doulas and other maternity care providers. Community-based doula support bridged barriers to equitable access to continuity of care models.

Doulas provided culturally-responsive and respectful care to migrant women. These findings are supported by a recent systematic review exploring asylum-seeking and refugee women’s experiences of various perinatal social support interventions, which reinforced that community-based doula programmes were valuable in addressing existing structural challenges within maternity settings.76 Furthermore, some doulas’ drive for reproductive justice within their own communities resonated with those engaged in community-based doula work supporting other population groups.77–79

This review shows that there is a unique opportunity to increase the social capital of migrant women and their families through doula support, by signposting to support services and social connection within their communities. Doulas may also benefit through education and employment opportunities resulting from their experiences in these programmes.80 Despite these potential benefits, challenges existed including operating externally from hospitals, the demanding nature of doula work, limited pay, unfamiliarity and limited clarity of community-based doula roles. These challenges align with private-practicing doula research.26 47 81

This review highlighted the continued demand for doula programmes, often in the non-profit sector, where ongoing precarious funding arrangements, resource intensiveness, doula recruitment and retention issues may impact sustainability. These issues reflect the challenges in the broader non-profit community sectors dependent on volunteer workforces.82–84 The strengths of community-based doula programmes in this review were the sense of community and engagement between programme staff and doulas and professionalisation of programmes within maternity and community settings. These are promising strategies in improving motivation, recruitment and retention of the doula workforce. However, strategic partnerships within these settings may be needed to increase programme credibility, funding and long-term sustainability.

There were a few limitations with the included studies. Findings related to five studies40 61–67 73 may have included perspectives or programme specifics related to non-migrant women. One evaluation study by Spiby et al (2015) comprising of five papers40 63–66 may have skewed findings, however, assessment using the GRADE-CERQual approach accounted for this potential limitation. Similarly, as there was limited quantitative data available, meta-analysis was not conducted. Included studies in this review were predominately from urban settings in four HICs which have different models of maternity care, meaning that findings may not necessarily be transferrable to programmes in rural settings or where other models are dominant. We also acknowledge that migrant and refugee population groups are not homogenous and may have various circumstances as evidenced within included studies,39 60 61 69–71 all of which may impact their perspectives and experiences of community-based doulas and maternity care. Similarly, as there was limited disaggregated data for migrant gender diverse birthing people their perspectives and experiences may have been excluded. Furthermore, there was incomplete data on community-based doula programme characteristics on conclusion of this review which prohibited analysis of programmes’ structure (online supplemental appendix 8). One of this review’s key strengths was using Cochrane’s EPOC guidance and a GRADE-CERQual approach which enabled the use of systematic and rigorous methodology in synthesising and assessing our confidence in qualitative review findings.

Implications for practice

Despite doulas rising into prominence from the 1980s, this workforce remains a relatively new concept among providers as evidenced by the unclear and overlapping roles of community-based doulas, which resonates with private-practicing doula literature.26 47 81 85 Clearly defined roles need to be established for all stakeholders involved, especially when roles are shared between providers and doulas. In circumstances where interpreters were limited, bicultural doulas providing language support may face challenging situations due to potential competing expectations. This highlights the importance of doulas being non-judgemental and ensuring their priority is supporting their clients, and using professional interpreters unless bicultural doulas have received interpretation certification. This concept is similar to where it is recommended to use interpreters instead of families and friends, to minimise the risk of compromised care and overcome legal liabilities.86

From this review, trained community-based doulas provided non-judgemental continuous support in labour and birth for migrant women which reinforced the workforce’s professionalism. Additionally, findings highlighted the strong emotional and social bonds created between doulas and migrant women particularly in the context of social isolation in new countries. This may be problematic when inevitably the doula-client relationship will end as the primary role is to provide continuous support in labour and birth rather than longer-term social support. Therefore, clear expectations and roles of doulas must be established and reinforced to maintain professional interpersonal bonds for client-doula boundaries.

The demanding nature of community-based doula work was compounded by challenges often faced by the non-profit sector, despite the value evident from their services from all stakeholders. This highlights that securing financial and strategic partnerships with established community or maternity organisations may be warranted, in which monetary recompense would be symbolic of the recognition and value in doulas.62 This may include creating salaried community-based doula models73 or exploring hospital-based doula models offering continuity of care as seen recently in Norway.87 88 If working within a hospital-based model, migrant and refugee women must remain the foremost priority and clear boundaries of accountability and autonomy must be established.

Implications for further research

This review evidenced that community-based doulas may provide short-term social connections with migrant women through birth support or signposting, however, there is limited exploration into how and which organisation is accountable to facilitate this connection. Likewise, exploration into the experiences of migrant women and doulas ending their professional relationships is limited. There is a need for research exploring the potential impact of these services on the social capital and health outcomes of migrant women’s partners and families.

There are limited studies which explore the experiences of the demanding nature of community-based doula work,80 with the intersections of precarious funding arrangements and an often volunteer doula workforce. Exploration into these topics is recommended. A review on hospital-based doula programmes for migrant and refugee women may explore the prospective feasibility of hospital-based partnerships and allow for comparison of findings from this review.

It is imperative that these programmes are evaluated and quantitative evidence of the long-term impact of community-based doula programmes on migrant women, their families and doulas themselves strengthen the emerging evidence base.

Conclusion

The findings from this review can inform community-based doula organisations, maternity healthcare services and policymakers of the value community-based doula programmes may have in providing culturally-responsive care to migrant and refugee women in HICs. The factors impacting programme implementation and sustainability; engaging in strategic partnerships, and possible wider-reaching benefits, should be further explored. Although this review was focused on migrant and refugee women, this is the first systematic review on community-based doula programmes. We hope from this review, investigations into broader community-based doula programmes for underserved population groups are explored, in the pursuit of human rights and health equity in maternity care.