Background Labour companionship refers to support provided to a woman during labour and childbirth, and may be provided by a partner, family member, friend, doula or healthcare professional. A Cochrane systematic review of interventions by Bohren and colleagues, concluded that having a labour companion improves outcomes for women and babies. The presence of a labour companion is therefore regarded as an important aspect of improving quality of care during labour and childbirth; however implementation of the intervention is not universal. Implementation of labour companionship may be hampered by limited understanding of factors affecting successful implementation across contexts. Objectives The objectives of the review were to describe and explore the perceptions and experiences of women, partners, community members, healthcare providers and administrators, and other key stakeholders regarding labour companionship; to identify factors affecting successful implementation and sustainability of labour companionship; and to explore how the findings of this review can enhance understanding of the related Cochrane systematic review of interventions. Search methods We searched MEDLINE, CINAHL, and POPLINE K4Health databases for eligible studies from inception to 9 September 2018. There were no language, date or geographic restrictions. Selection criteria We included studies that used qualitative methods for data collection and analysis; focused on women’s, partners’, family members’, doulas', providers', or other relevant stakeholders' perceptions and experiences of labour companionship; and were from any type of health facility in any setting globally. Data collection and analysis We used a thematic analysis approach for data extraction and synthesis, and assessed the confidence in the findings using the GRADE‐CERQual approach. We used two approaches to integrate qualitative findings with the intervention review findings. We used a logic model to theorise links between elements of the intervention and health and well‐being outcomes. We also used a matrix model to compare features of labour companionship identified as important in the qualitative evidence synthesis with the interventions included in the intervention review. Main results We found 51 studies (52 papers), mostly from high‐income countries and mostly describing women's perspectives. We assessed our level of confidence in each finding using the GRADE‐CERQual approach. We had high or moderate confidence in many of our findings. Where we only had low or very low confidence in a finding, we have indicated this. Labour companions supported women in four different ways. Companions gave informational support by providing information about childbirth, bridging communication gaps between health workers and women, and facilitating non‐pharmacological pain relief. Companions were advocates, which means they spoke up in support of the woman. Companions provided practical support, including encouraging women to move around, providing massage, and holding her hand. Finally, companions gave emotional support, using praise and reassurance to help women feel in control and confident, and providing a continuous physical presence. Women who wanted a companion present during labour and childbirth needed this person to be compassionate and trustworthy. Companionship helped women to have a positive birth experience. Women without a companion could perceive this as a negative birth experience. Women had mixed perspectives about wanting to have a male partner present (low confidence). Generally, men who were labour companions felt that their presence made a positive impact on both themselves (low confidence) and on the relationship with their partner and baby (low confidence), although some felt anxious witnessing labour pain (low confidence). Some male partners felt that they were not well integrated into the care team or decision‐making. Doulas often met with women before birth to build rapport and manage expectations. Women could develop close bonds with their doulas (low confidence). Foreign‐born women in high‐income settings may appreciate support from community‐based doulas to receive culturally‐competent care (low confidence). Factors affecting implementation included health workers and women not recognising the benefits of companionship, lack of space and privacy, and fearing increased risk of infection (low confidence). Changing policies to allow companionship and addressing gaps between policy and practice were thought to be important (low confidence). Some providers were resistant to or not well trained on how to use companions, and this could lead to conflict. Lay companions were often not integrated into antenatal care, which may cause frustration (low confidence). We compared our findings from this synthesis to the companionship programmes/approaches assessed in Bohren’s review of effectiveness. We found that most of these programmes did not appear to address these key features of labour companionship. Authors' conclusions We have high or moderate confidence in the evidence contributing to several of these review findings. Further research, especially in low‐ and middle‐income settings and with different cadres of healthcare providers, could strengthen the evidence for low‐ or very low‐confidence findings. Ahead of implementation of labour companionship, researchers and programmers should consider factors that may affect implementation, including training content and timing for providers, women and companions; physical structure of the labour ward; specifying clear roles for companions and providers; integration of companions; and measuring the impact of companionship on women’s experiences of care. Implementation research or studies conducted on labour companionship should include a qualitative component to evaluate the process and context of implementation, in order to better interpret results and share findings across contexts.
