Mistreatment of women during childbirth and postpartum depression: secondary analysis of WHO community survey across four countries

Background Postpartum depression (PPD) is a leading cause of disability globally with estimated prevalence of approximately 20% in low-income and middle-income countries. This study aims to determine the prevalence and factors associated with PPD following mistreatment during facility-based childbirth. Method This secondary analysis used data from the community survey of postpartum women in Ghana, Guinea, Myanmar and Nigeria for the WHO study, ‘How women are treated during facility-based childbirth’. PPD was defined using the Patient Health Questionnaire (PHQ-9) tool. Inferential analyses were done using the generalised ordered partial proportional odds model. Results Of the 2672 women, 39.0% (n=1041) developed PPD. 42.2% and 5.2% of mistreated women developed minimal/mild PPD and moderate/severe PPD, respectively. 43.0% and 50.6% of women who experienced verbal abuse and stigma/discrimination, respectively developed minimal/mild PPD. 46.3% of women who experienced physical abuse developed minimal/mild PPD while 7.6% of women who experienced stigma/discrimination developed moderate/severe PPD. In the adjusted model, women who were physically abused, verbally abused and stigma/discrimination compared with those who were not were more likely to experience any form of PPD ((OR: 1.57 (95% CI 1.19 to 2.06)), (OR: 1.42 (95% CI 1.18 to 1.69)) and (OR: 1.69 (95% CI 1.03 to 2.78))), respectively. Being single and having higher education were associated with reduced odds of experiencing PPD. Conclusion PPD was significantly prevalent among women who experienced mistreatment during childbirth. Women who were single, and had higher education had lower odds of PPD. Countries should implement women-centred policies and programmes to reduce mistreatment of women and improve women’s postnatal experiences.

note with inputs from our WHO partners (HM and OT) based on the local need for research into maternal mental health which has not been given much attention and its association with mistreatment during childbirth.

How were local researchers involved in study design?
The overall study was designed with inputs from the whole study group. At the early stages of the project, a meeting was held at the WHO headquarters, Geneva in 2014. The country research teams (KA-B, EM-Ghana, TMM-Myanmar, MDB-Guinea and Prof Bukola Fawole-Nigeria) were involved in the deliberations to choose appropriate study designs based on their local context together with the WHO research team (HM, OT). Prof Bukola Fawole later passed on to glory before the completion of the study and was duly acknowledged by dedicating the primary paper of the second phase of the study to his memory. CG, KA-B and EM were the Ghanaian study site coordinators and conceptualized this analysis with inputs from HM and OT. The Ghanaian team led the analysis and writing of this manuscript on behalf of the whole research team. PAA joined the research team for this analysis and drafting of the initial manuscript based on her research needs as a PhD student at the University of Ghana, School of Public Health.

How has funding been used to support the local research team?
At the start of the study, a research capacity building plan was developed by the local research teams and their WHO counterparts, with support from the HRP Alliance for Research Capacity Strengthening. During the project, the WHO research team led by OT

How are research staff who conducted data collection acknowledged?
This paper is one of 18 papers (7 qualitative from the formative phase, 1 protocol, 1 methodological development, and 9 quantitative from the measurement phase) from our research collaboration. Among these 18 papers, research staff who collected data from all 4 study countries have contributed as co-authors on at least 1 paper. Each country's research team has led at least 4 papers with their teams. In all the papers, research assistants who collected data have been duly acknowledged.

Do all members of the research partnership have access to study data?
All members of this research partnership have full access to the data. This is evidenced by the number of published journal articles that have come out from this study with lead authors from all the partner countries including this paper.

How was data used to develop analytical skills within the partnership?
The research team has worked together in the data analysis throughout the whole study and the current paper. The data analysis workshops and the scientific writing workshops that took place as part of the research partnership helped to strengthen the analytic and writing skills for the research team. For this paper, the Ghanaian team (CG, PAA, KA-B and EM) led the analysis with support from HM and OT.

How have research partners collaborated in interpreting study data?
Throughout the research, all partners have been involved in interpreting the study data during data analysis. In particular, for multi-country papers emanating from our research partnership, discussions are held to understand the issues and develop the implications for each country's research, policy and practice.

How were research partners supported to develop writing skills?
The research team writing this paper is made up of all levels (senior, mid, junior) academics and clinicians. KA-B, PAA, AA and HM are currently working on their PhDs, and they are supported by OT and CG. The authors were supported by the WHO research team (HM OT) to develop and refine their writing skills through regular reviewing of the manuscript and providing constructive feedback. The scientific writing workshops also contributed towards developing the writing skills of the researchers.

All papers arising from this research partnership have been published in open access
journals. After the publication too, plans were developed for the dissemination of our findings within WHO and our individual institutions. Significantly, the standardized tools for measuring mistreatment during facility-based childbirth are also freely available to be used by other researchers.

How is the leadership, contribution and ownership of this work by LMIC researchers recognised within the authorship?
To ensure fairness in the leadership, contribution, and ownership of the work, a clear data use and authorship guidance were drawn and discussed with research teams from the study countries and WHO. Among the Ghanaian authors who led this work CG, PAA and KA-B are the first, second and third authors respectively. EM is the last and also the corresponding author. In addition, 9 out of the 11 authors are researchers from the study countries (Ghana, Guinea, Myanmar, Nigeria). The study teams include early career researchers and clinician researchers (CG, PAA, KAB, HM, OA, TAI, TMM, EM) within the authorship team. They contributed to the data collection, analysis plan, analysis, and writing. It is worth mentioning that, only 1 out of the 8 early career researchers is based in a high-income country. The remaining are based in lowincome countries (Ghana, Guinea, Myanmar, Nigeria).

How has gender balance been addressed within the authorship?
Gender balance has always been considered within the authorship. Four of the authors are females (PAA, HM, TAI, OT) and seven are males (CG, KA-B, AA, MDB, OA,TMM,EM).

How has the project contributed to training of LMIC researchers?
The overall study has also contributed to KA-B and TMM's PhD dissertations. TMM's PhD was made possible through a scholarship from the HRP Alliance for Research Capacity Strengthening). PAA, a PhD student, used the analysis and the initial drafting of the manuscript to improve on her research skills. The data analysis and writing workshops which were carried out as part of this research project also contributed to addressing the research training needs of the study countries.

How has the project contributed to improvements in local infrastructure?
This project has not directly contributed to improvements in local infrastructure. However, tablets that were procured for the study countries, have been used for other studies after this collaborative research.

What safeguarding procedures were used to protect local study participants and researchers?
The current study was adapted based on local context to safeguard the safety of our research team and study participants. For instance, in Myanmar labour observations were not made because it was considered inappropriate or unsafe to observe labour in the wards of public hospitals. We used all female data collectors due to the sensitive nature of the study. Again, the research team had a study site obstetrician and a senior midwife who were at hand to handle any issues arising. For instance, if a research assistant observes that a woman in labour (including those who are not study participants) is being abuse excessively, she will inform the obstetrician or senior midwife for immediate redress. Finally, there was regular debriefing and reflexive discussions between data collectors, country study teams and our partners from WHO which helped to address any teething issues during data BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)