Interventions to support effective communication between maternity care staff and women in labour: A mixed-methods systematic review.

Only two eligible trials were found, and no qualitative studies were identified, so the review could not pool results or answer questions about women's and staff's lived experiences. Neither trial reported clinical birth outcomes, so the review says nothing about mode of birth, perineal trauma, newborn condition, breastfeeding, or pain relief. The women satisfaction result came from one trial with unclear randomisation of hospitals, uncertain outcome assessor selection, no blinding of participants, and only two of 2,000 women having a caesarean birth. The positive perception result came from simulated emergencies with patient-actors who were experienced midwives, not real women, and it was a before-and-after analysis with no comparison group; patient-actors also could not be blinded to whether ratings were before or after training. The UK study had low uptake and drop-outs: of 240 staff approached, 158 consented, 18 withdrew before the first evaluation, and of 140 participants, 136 attended training and 132 attended the post-training assessment. GRADE rated both outcomes very low because each depended on one study with methodological limitations. Contexts were specific: one trial was in crowded Syrian hospitals where relatives were often not allowed and eye contact with male providers was restricted; the other was in UK maternity hospitals. Usual practice was not described in either study.

This review was commissioned by the WHO Human Reproduction Programme as evidence for WHO intrapartum care recommendations. The WHO recommendations were financially supported by USAID. One author is WHO staff, and the paper states her views are not necessarily WHO's and that the paper represents the views of the named authors only.

Do not read the improved ratings from actors playing patients after training as proof that communication training improves women's care or birth outcomes. The only trial that asked real women found no significant difference in satisfaction, and the positive finding came from a before-and-after simulation study with no comparison group and very low-certainty evidence.