Community Based Doulas for Migrant Women in Labour and Birth in Sweden - a Randomised Controlled Trial

In Sweden, the proportion of births to migrant women has more than doubled over the last four decades, from 10% in 1973 to 28% in 2015. Migrant women are at increased risk of adverse pregnancy outcomes as low birth weight, infant small for gestational age , congenital malformations, stillbirth, neonatal morbidity and mortality and also obstetric interventions, such as induction of labour and caesarean section. Unfamiliarity with the health care system, language barriers, knowledge gaps and cultural issues are all obstacles that play a vital role in migrant women deprivation from achieving the standard maternal health care.

Migrant women have shown to consistently rate their maternity care more negatively in contrast with Swedish women, mainly due to lack of knowledge on how Swedish health care system is functioning and providing care which consequently cause communication difficulties, in addition to discrimination and prejudicial staff attitudes. Migrant women reported being left alone in labour, and feeling fearful, insecure and unsupported. Factors to be targeted to improve experiences of care and pregnancy outcomes include; breaking down language barriers and bridging cultural gaps, increasing familiarity with and enhancing understanding of Swedish maternity care, and empowering women's sense of safety and confidence in giving birth.

In Sweden, migration of women of childbearing age has increased most rapidly from Somalia and from a number of Arabic-speaking countries and also Eritrea and Russia. Somali women are known to be at highest risk of perinatal morbidity and mortality, they have problems engaging with Swedish antenatal care and because of communication and language barriers they reported to have the poorer experiences of care regarding labour and birth. Like Somali-women, many Arabic-speaking and Tigrinya-speaking women (from Eritrea) have relocated in Sweden after traumatic experiences of war and conflict. They, and the Russian women, constitute a growing group of women giving birth in Sweden with little knowledge of how Swedish maternity care operates and facing major communication barriers.

One initiative that has addressed communication problems between migrant women and the caregivers throughout childbirth was to provide language assistance and labour support , the Community Based Doula (Doula och kulturtolk) Project. CBDs are bilingual women from migrant communities trained by expert midwives to provide support for migrant women in labour and birth. CBDs meet twice with the migrant women prior to the birth, accompany her throughout the labour and provide emotional, physical support and communication; and language assistance; they meet again for follow up twice after the childbirth. Two small qualitative evaluations, conducted in the early years of the program, indicate high levels of satisfaction among supported women and midwives.

Current evidence supports the potential for CBDs to play an important role in improving continuity for migrant women during labour and birth, enhancing their experiences of birth and of care, as improving their pregnancy outcomes. Good physical and emotional support in labour is known to improve a range of outcomes; greater maternal satisfaction with care, less use of analgesia, shorter labours, lower rates of caesarean section and more spontaneous vaginal birth - with no adverse effects for women or infants.

Worldwide, no randomised controlled trials of doula support for migrant women have been reported, though two US non-randomised studies are relevant. One evaluated a hospital-based doula service provided to 123 of 348 Somali women giving birth in the study period. Lower rates of caesarean section (17% vs 26%) and greater satisfaction with care were found among the doula-supported Somali women17. Staff also felt more confident to care for Somali women when an English-speaking Somali doula was present. A second retrospective cohort study (n=11 471) evaluated a community doula program in an urban, culturally diverse setting and found a small but statistically significant reduction in caesarean section among women cared for by a midwife and a doula, compared with a midwife alone (15% vs 18%)18. Studies of midwives' views of working alongside labour companions have been positive14, 18.

In the Swedish setting, more robust and specific evidence is needed to support decisions whether to implement CBDs for migrant women at large-scale or not. Therefore, to evaluate the effectiveness of doula support to Somali, Eritrean, Arabic and Russian-speaking migrants we will here conduct a randomized controlled trial in Stockholm.

Aim To evaluate the effectiveness of community-based doula support for improving the intrapartum care experiences and postnatal wellbeing of Somali, Eritrean, Arabic and Russian-speaking migrant women giving birth in Stockholm.

Hypotheses

1. That women randomised to receive CBD support in labour will rate their care for labour and birth more highly and have better emotional wellbeing (lower mean scores on the Edinburgh Postnatal Depression Scale) two months after birth than women allocated to standard care.
2. That midwives and obstetricians caring for women randomised to receive CBD support will rate their provision of care more positively than when caring for women not supported.

Methods We plan a randomised controlled trial as a prospective evaluation of CBD support for migrant women in Stockholm, something not achieved when the original CBD program commenced in Gothenburg in 2008. The proposed trial will provide further and more rigorous evidence about the effectiveness of CBDs for improving migrant women's experiences of care and ultimately pregnancy outcomes for both mothers and infants. The study is designed to meet the CONSORT guidelines for reporting of randomised trials16 and will be registered in Clinical Trials.

Participants Migrant women from Somali, Tigrinya, Arabic and Russian-speaking countries will be recruited.

Inclusion criteria: Nulliparous and multiparous pregnant women between 25-36 weeks gestation, who are 18 years or older and from Somali-, Arabic-, Russian- and Tigrinya-speaking countries with no contra-indications for vaginal birth and cannot communicate fluently in Swedish.

Exclusion criteria: Women <18 years or women for whom a caesarean birth is planned and women not consenting to access to their birth records.

Recruitment: All Somali-, Arabic-, Russian and Tigrinya-speaking women cover the inclusion criteria and will participate in the study. They will be recruited with the assistance of interpreters during antenatal care visits between 25 and 36 weeks gestation.

Intervention - Community Based Doulas support Women allocated to receive CBD support will be contacted by an either a Somali, Arabic-, Russian- or Tigrinya-speaking CBD as appropriate, and arrangements made for them to meet twice prior to the birth to get to know each other and discuss the woman's wishes regarding support in labour and what the CBD can offer. Each woman will then contact her CBD when she goes into labour and the CBD will attend hospital with her and stay with her throughout her labour and birth, in addition to any other support people she may have, such as her partner.

CBDs will be recruited, trained and employed by non-profit organization MIRA, using well-tested processes. A total of 10 CBDs will be employed for the four migrant groups, ensuring the possibility of back-up when needed. Training content will cover the physiology of childbirth and strategies for providing effective continuous support in labour, as well as practical strategies for helping with communication/interpreting to enhance communication between women and midwives and discussion of CBD roles and boundaries. CBDs will be employed on an hourly basis to enable flexibility in timely provision of support to women in labour. Support for CBDs will be provided by an experienced midwife.

Comparison - standard intrapartum care Women allocated to the comparison arm of the trial will receive standard intrapartum care as provided at their chosen hospital of birth.

Women partner and/or other support people to accompany throughout her childbirth will be allowed, regardless of their trial allocation.

Randomisation and blinding Women will be randomly allocated to the intervention or control group utilising a computerised randomisation schedule. The randomisation ratio will be 1:1, CBD support to usual care, with block sizes of 4 or 6 distributed randomly. Allocations will be prepared in sealed, opaque envelopes accessible in a central location for antenatal recruiting midwives to access. Participants cannot be blinded in this study, however data collection and analysis will be undertaken blinded to group allocation.

Primary outcomes Women's ratings of care for labour and birth and maternal emotional wellbeing. Single item question from the Migrant Friendly Maternity Care Questionnaire (MFMCQ) will be used to assess women's ratings of care the Edinburgh Postnatal Depression Scale (the EPDS) will be used to assess maternal wellbeing.

Secondary outcomes Epidural analgesia, length of labour from admission and mode of birth. Data will be retrieved from hospital patient records will enable assessment of birth outcomes.

Data collection Trained bilingual researcher and research assistants will collect data on women's experiences of care by telephone or face-to-face interviews as appropriate, conducted at base-line and two months after birth. Birth outcome data for all trial participants will be extracted from patient records.

Sample size To detect an increase in women's ratings of intrapartum care from an expected 30% rating care as very good among those receiving usual care (based on estimates from studies of migrant women not fluent in the host country language) 22-24 to 53% in those receiving CBD support, (equal to Swedish speaking women in a national population based study) with 80% power and an alpha of 20%, 69 women in each group are needed. To have similar power to detect differences in mean scores on the Edinburgh Postnatal Depression Scale (a hypothesised reduction from a mean of 8.0 in the comparison group - similar to that found in studies of migrant women, to 6.0 in the intervention group - similar to that found in Swedish population-based studies), 63 women are required in each arm. Allowing for 20% loss to follow-up at the time of follow-up data collection with women two months postpartum, we expect to have outcome data on views of care and depressive symptoms for at least 100 women in each trial arm.

Statistical analysis First step of analysis will check the comparability of the groups, then the intervention group will be compared with the control group testing trial hypotheses and using intention to treat analyses. Odds ratios and 95% confidence intervals will then be estimated. If there are baseline differences on key participant characteristics (such as age, marital status, and medical conditions) between trial arms, logistic regression analyses will be performed adjusting for these differences. Comparison of means will be undertaken using t-tests where data are normally distributed, or medians compared using Mann-Whitney U tests used if not.