Community Perspectives on the Creation of a Hospital-Based Doula Program

Objective: Racial and ethnic inequities in perinatal health outcomes are pervasive. Doula support is an evidence-based practice for improving maternal outcomes. However, women in lower-income populations often do not have access to doulas. This study explored community perspectives on doula care to inform the development of a hospital-based doula program to serve primarily low-income women of color. Methods: Four focus groups and four individual interviews were conducted with: (1) women who were pregnant or parenting a child under age 2 (n=20); (2) people who had provided support during a birth in the previous 2 years (n=5); and (3) women who had received doula training (n=4). Results: Participants had generally positive perceptions of doula services. Many aspects of doula support desired by participants are core to birth doula services. Participants identified ways that doulas could potentially address critical gaps in health care services known to impact outcomes (e.g., continuity of care and advocacy), and provide much-needed support in the postpartum period. Responses also suggested that doula training and hospital-based doula programs may need to be adapted to address population-specific needs (e.g., women with substance use disorder and younger mothers). Novel program suggestions included “on call” informational doulas. Conclusions: Findings suggested that women in racial/ethnic minority and lower income groups may be likely to utilize a hospital-based doula program and identified adaptations to traditional doula care that may be required to best meet the needs of women in groups with higher risk of poor maternal health and birth outcomes.


Introduction
Racial and ethnic inequities in maternal health and perinatal outcomes in the United States are pervasive. 1,2 Preterm birth, low birthweight, cesarean delivery, and severe maternal morbidity and mortality, are more common among Black, Latina, and Indigenous women than among White women. 1,[3][4][5][6][7] Services from a doula, a professional who provides physical, emotional, and informational support to women and families in the perinatal period, have the potential to reduce inequities in maternal health care and outcomes. 8 Doula support is associated with positive birth outcomes, including reduction in cesarean delivery rates, decreased use of pain medication, and greater satisfaction with birth experience. 9 Fewer studies have examined the potential benefits of extended postpartum doula services, but postpartum doula care could impact maternal mental health, maternal/infant bonding, [10][11][12] and provide connection to community and medical resources.
Only 6% of U.S. women report having labor support from a doula. 13 The lack of insurance reimbursement for doula services has resulted in limited access for women in lower-income populations, who are disproportionately women of color. 6,14 However, when women in these populations do receive doula services, there are substantial benefits, including lower rates of cesarean delivery, preterm birth, and low birth weight. 15,16 The community-based doula model has been developed to focus on the needs of women of color and low-income women, ideally delivered through organizations led by members of these groups. 17 Community-based doulas are part of the communities that they work with, and support typically includes more visits and connection to a wider network of services and referrals than the ''private pay'' doula model. 17 Community-based doulas can address the lack of racial and socioeconomic diversity in the current doula workforce, 18 which is important because racially/ ethnically and culturally concordant care may foster a positive relationship, including better-quality communication. [19][20][21][22] Women who receive services from community-based doula organizations have increases in breastfeeding duration and decreased chances of cesarean delivery, with larger improvements for Black women. 23,24 However, community-and hospital-based doula programs have faced challenges, including how to sustain a workforce how to ensure that training and services are tailored to the needs of women in the community. 23,25 Factors such as cultural perspectives and trust in the health care system may impact use of and experience with doula services. 26 Although several studies have explored doulas' perspectives on the value of the care they provide, 26-28 few U.S. studies have explored community members' perceptions of doulas and assessed how previously unmet support needs might be met through doula services. This study examined perspectives on doula care from members of a lower-income community of color with the goal of generating recommendations for a hospital-based doula program.

Setting, study population, and recruitment
This study was conducted in Springfield, MA, a city of *150,000, where 48% of the population is Hispanic or Latino, 18% is Black or African American, and 30% has a household income below the federal poverty line. 29 Springfield has higher rates of preterm birth, low birthweight births, and infant mortality compared with overall rates in Massachusetts. 30 The following populations were recruited to participate: (1) women who were currently pregnant or had given birth in the previous 2 years, including women in recovery from substance use disorders (SUD); (2) individuals who had provided support during a labor and birth in the previous 2 years; and (3) women who had recently received doula training through a hospitalsponsored initiative in partnership with a communitybased doula organization. Recruitment took place at an early childhood education center serving families with lower incomes and a prenatal clinic in a tertiary care teaching hospital. The trained doula focus group was converted to individual video interviews due to COVID restrictions. The study was approved by the Institutional Review Board of Baystate Medical Center.

Focus groups and interviews
Focus groups and interviews took place February-June 2020. Focus groups were conducted in a private conference room at the early childhood education center by a trained moderator using a semistructured interview guide developed by the research team. The interview guide included questions aimed at understanding community perspectives on doula care, identifying potential unique needs of this population, and eliciting recommendations for doula program implementation (Supplementary Appendix SA1). Each participant received a $50 gift card as compensation for their time. Focus groups and interviews were audiorecorded and professionally transcribed.

Analysis
Data were analyzed using conventional qualitative content analysis, a form of thematic analysis. 31 Two members of the research team (L.A. and S.G.) developed a preliminary codebook that consisted of broad categories that were anticipated to emerge in response to each of the interview questions. Three team members (L.A., S.G. and M.D.) performed line-by-line coding on the transcripts, iteratively modifying the codebook as new codes were identified and codes were changed or refined. Each transcript was coded by at least two team members. Coding was reviewed and disagreements resolved through discussion. Codes were then grouped into broader categories to identify emerging themes. An audit trail was created to track coding decisions. The analysis was performed using Dedoose. 32

Results
Four focus groups ranging in size from 4 to 9 participants (N = 23), and 4 individual interviews were conducted, for a total of 27 participants. Focus groups lasted an average of 90 min. Two focus groups included people who were pregnant or recently gave birth, one focus group included people who were pregnant or recently gave birth and identified as being in recovery, and one focus group included support people, two of whom had also recently given birth. Most participants (93%) were female (Table 1). About one-third of participants had education beyond high school. Participants overall were fairly evenly split among White, Black or African American, and Hispanic/Latino/a race/ethnicity. There was, however, some variability in the racial/ethnic composition by focus group (Supplementary Appendix SA2).
We identified six key themes and associated subthemes; each theme came up across all participant types unless otherwise noted in the text. Two themes were unique to interviews with trained doulas. Themes and illustrative quotes are presented in (Table 2 and  Table 3).

Theme 1: awareness and general perceptions of doulas
This theme included participants' prior awareness and perceptions of doulas, including general attitudes toward doulas and discussion of the circumstances where doulas might be helpful or not.
Enthusiasm amid limited prior awareness. Many participants had heard of doulas, although few had worked with one. There was some confusion about what kind of things is appropriate for doulas to do as part of that role. However, mothers overall had a positive reaction to the idea of doula support, generally endorsing the idea of having more support available. Several expressed a wish to have known about or had access to a doula in their recent birth. This subtheme did not come up in the interviews with trained doulas.
Good for others, not necessarily for me. A few participants did not see a benefit to doula support to them personally, although they felt that other women might benefit from doula care. Doula support was perceived as being more necessary for women without partner or family support, or with other challenging circumstances such as a fussier baby. Additionally, focus group participants were unaware of ways that doulas might be helpful during different kinds of births, particularly cesarean births. For women who had heard of doulas before, many had perceived doulas as working with women who were not like them, such as women who were wealthy and/or having a home or water birth. However, participants identified few specific downsides to working with a doula. This subtheme came up across participant types in the focus groups, but not in the interviews with trained doulas.
Mixed feelings from partners. Participants who had been a support person at a birth had more mixed feelings about the idea of doulas, and this subtheme was specific to that focus group. Several of them felt that they had provided all of the support that was needed for the person whose birth they attended and believed that having the close relationship of being a family member enabled them to better understand the birthing person's needs and provide comfort.

Theme 2: doula role
This theme included general views on the desired role of a doula.
Desired scope of work. Participants emphasized the need for support during pregnancy and in the postpartum months. Labor support was also seen as positive, but was less of an identified need. Participants noted a need to identify and discuss timing and frequency of visits and to negotiate expectations to avoid a mismatch of expectations. This subtheme did not arise in the interviews with trained doulas. Benefits of doula versus family support. Participants noted potential benefits of doula support compared with family support. Family members may bring their own expectations, judgments, and histories, whereas doulas are more ''neutral'' and familiar and comfortable with birth. Additionally, doulas are there for the birthing person specifically.
Theme 3: unmet support needs that doulas could address This theme included discussions of specific support needs that were currently lacking or insufficiently met that could potentially be met through doula services.
Core elements of usual doula services. Many aspects of doula support that participants desired are core elements of usual doula services, including emotional support, awareness of mental health needs, material support, providing information and guidance, and supporting family members or partners at the birth.
Gaps left by the health care system. Doulas can provide continuity of care throughout the perinatal period, which is not usually possible in the health care system, where women often see different clinicians throughout pregnancy and have typically never met the nurses that will work with them during labor. Advocacy within the health care system was also seen as a key role of doulas. Participants shared experiences in which their symptoms or pain were brushed off or minimized by clinicians. Women in the SUD-specific focus group related negative experiences of treatment in the hospital after giving birth, such as nurses expressing skepticism about whether a urine test would detect drugs, and noted the potential benefits of doula advocacy in such situations.
Population-specific support needs. Focus group respondents identified specific populations that might  also have particular support needs, such as women with lower incomes, young mothers, first-time mothers, and women with SUD. Young mothers and first-time mothers were perceived as needing more informational and emotional support than older women or women who had given birth before, and women with lower incomes were seen as potentially benefitting from connection to locations or organizations that could provide material resources such as free diapers. Women with SUD may be involved with the Department of Children and Families, and desired support in navigating the bureaucracy from someone who is on ''your side.'' Women with SUD also identified concerns with how to manage taking pain medications after a cesarean birth and the impact on recovery. Nonjudgmental doula support was desired, ''because that's something that you don't really want to talk to your doctor about.''

Theme 4: program recommendations
Many of the specific recommendations for a hospitalbased doula program are connected to the previous themes.
Priority populations. Participants were enthusiastic about the idea of doulas being available to all who want doula services, but also noted specific populations that could potentially be prioritized if there were not enough resources to make doulas available to all. These populations overlapped with the populationspecific support need identified in Theme 3, and included first-time moms, young moms, women who lacked other support, and women with SUD.
Matching and choice. Given the perceived variation in what might be important to women in choosing a doula (Theme 2), participants believed that a system should be created to match women with doulas and to allow women some choice of doulas, so that each individual woman could prioritize the characteristics that she felt were important in a doula. Some women felt that ideally there would be an opportunity for a woman and a doula to interact to see if they connected well in terms of personality, beyond attributes that could be identified through a questionnaire.
New models of doula services. Participants suggested new models of doula services that might be beneficial, including doulas who only provide phone or text support during pregnancy. Some participants also noted the overlooked needs of new fathers, and proposed ''doula dads,'' or experienced fathers/partners whose primary role would be to support the partner. This subtheme came up only in the support person focus group and in the interviews with trained doulas.
Connecting women to a doula program. Participants felt that for a new doula program to be successful, there would need to be efforts to inform women about the role of a doula and the ways that doulas can provide support for women in a range of circumstances. Most participants felt that it would be helpful if women were informed about and connected to the doula program through their prenatal care provider. Others felt that other places in the community that serve women in lower-income populations, such as the office for the Special Supplemental Nutrition Program for Women, Infants and Children (WIC).
Themes unique to doula interviews Theme 5: barriers and facilitators to becoming or working as a doula. Challenges associated with participating in doula training included finding a training program that addressed desired skills, and accessibility in terms of location and cost (Table 3). Doulas also discussed challenges with transitioning to work as a doula after training, including finding clients and attending first births without having had the opportunity to apprentice with an experienced doula. Employment could be a barrier or facilitator; some trained doulas had jobs that were compatible with attending births, or even helped them to find clients because they worked with pregnant women. For others, it was difficult or impossible to negotiate being on call with their employment.
Theme 6: interactions of doula and health care personnel. Interactions with health care providers, including nurses, physicians, and midwives, were an important aspect of doulas' experiences at births they had attended. Participants mentioned both situations where they were made to feel welcome and valued, and situations where they felt in the way or even felt hostility from health care providers. Some participants perceived that the environments that were less welcoming to doulas might be where advocacy was most needed. Participants discussed the importance of getting the buy-in of clinicians at all levels for the success of a hospital-based doula program.

Discussion
Many of the desired services identified by women in this study were consistent with a traditional privately-paid doula's role. Some of the potential benefits of extending doula support to this population would likely arise simply by making this successful model of support accessible to more women. However, we also found some differences in emphasis as well as unique needs for women in the community studied, compared with privately paid doula services.
Notably, participants emphasized support needs during pregnancy and the postpartum period over and above support during labor and birth. However, they did also identify additional ways that doulas could be helpful during labor. Interestingly, one specific example that arose was that a doula could help advocate for admission to the hospital, which seems to contradict one mechanism by which doula care may lead to lower intervention-delaying admission to the hospital until active labor. This seeming contradiction may reflect the difficulty in fully conceptualizing doula support among women who, for the most part, had not experienced it personally. It could also reflect a lack of knowledge about the labor process and the potential benefits of laboring at home for longer, which doula support could also address. The postpartum period was seen as a time when women's needs are extensive and family support might not be adequate, especially given the high prevalence and underrecognition of postpartum depression. The American College of Obstetricians and Gynecologists has new guidelines highlighting the importance of the postpartum period for maternal wellbeing 33 ; our findings indicate that doulas could be an important part of the postpartum support system.
Participants also identified ways in which doulas could help to address gaps or negative experiences in the health care system, through continuity of care and advocacy. Discrimination and disrespectful treatment are reported by up to 24% of birthing people, and may be particularly prevalent among women of color [34][35][36][37][38][39][40] ; doula advocacy could help buffer the effects of such treatment. 26 However, to optimally integrate doulas into a hospital-based program in an advocacy role, institutional commitments beyond the addition of doula services may be necessary. The doulas that we interviewed highlighted the importance of their interactions with clinicians in making them feel welcome or discouraged. Previous research has shown that clinicians may have negative perceptions of doulas if they have experiences interacting with them in an adversarial manner. 41,42 A recent survey of maternity care clinicians suggested that educating clinicians about doulas' training and creating clear guidelines for the role of doulas within the maternity care team could foster positive doula-clinician relationships. 41 The support needs identified by participants largely mirrored the model of community-based doula programs, which are designed to address the needs of underserved communities. While this study explored the creation of a hospital-based doula program, a community-based doula approach has successfully been integrated into some hospital-based doula programs. 43 For doulas to deliver this enhanced support, the allocation of adequate resources is necessary, and doulas must be fairly compensated for the work involved to avoid burnout and to promote sustainability 26 ; a hospital-based program must also take into account these resource considerations.
The focus group with women in recovery identified several ways in which tailored doula support could be particularly beneficial for this population. Stigmatization of patients with SUD is a general problem, and is particularly acute when the patient is a pregnant woman, [44][45][46] underscoring the importance of advocacy and the potential role of a doula for these women, as our focus group participants noted. This is very salient in the context of Springfield, where opioid-related deaths dramatically increased in recent years even as statewide opioid-related deaths fell. 47 Strengths and limitations Unique contributions of this study include examining partners' perceptions of doulas, an identified gap in doula research, 48 and exploring women's own perspectives of their support needs in the perinatal period and the role of doulas in addressing them. However, there are some limitations to consider. In developing the focus group guide, we did not solicit input from currently practicing doulas or members of the participant populations. It is possible that the definition of doula services influenced participant responses. Additionally, we purposefully asked open-ended questions about perinatal experiences to allow women to construct their own experience narratives, but the lack of explicit questions about mistreatment, disparities in treatment, and culture-matched care may also have influenced responses. We only included Englishspeaking women; a Spanish-speaking focus group was scheduled, but canceled due to COVID. Perceptions of doulas held by maternity care clinicians may be important, 41 but was beyond the scope of this study. While the focus group moderators encouraged an atmosphere conducive to open discussion, it is possible that responses were influenced by social desirability bias.

Conclusions
Findings suggested that women in racial/ethnic minority and lower income groups may be likely to utilize a hospital-based doula program and identified adaptations to traditional doula care that may be required to best meet the needs of women in groups with higher risk of poor maternal health and birth outcomes.