Women’s perspectives of mistreatment during childbirth at health facilities in Ghana: findings from a qualitative study

Abstract Mistreatment of women during childbirth at health facilities violates their human rights and autonomy and may be associated with preventable maternal and newborn mortality and morbidity. In this paper, we explore women’s perspectives on mistreatment during facility-based childbirth as part of a bigger World Health Organization (WHO) multi-country study for developing consensus definitions, and validating indicators and tools for measuring the burden of the phenomenon. Focus group discussions (FGDs) and in-depth interviews (IDIs) were used to explore experiences of mistreatment from women who have ever given birth in a health facility in Koforidua and Nsawam, Ghana. Interviews were audio-recorded, transcribed and thematic analysis conducted. A total of 39 IDIs and 10 FGDs involving 110 women in total were conducted. The major types of mistreatment identified were: verbal abuse (shouting, insults, and derogatory remarks), physical abuse (pinching, slapping) and abandonment and lack of support. Mistreatment was commonly experienced during the second stage of labour, especially amongst adolescents. Inability to push well during the second stage, disobedience to instructions from birth attendants, and not bringing prescribed items for childbirth (mama kit) often preceded mistreatment. Most women indicated that slapping and pinching were acceptable means to “correct” disobedient behaviours and encourage pushing. Women may avoid giving birth in health facilities in the future because of their own experiences of mistreatment, or hearing about another woman’s experience of mistreatment. Consensus definitions, validated indicators and tools for measuring mistreatment are needed to measure prevalence and identify drivers and potential entry points to minimise the phenomenon and improve respectful care during childbirth.


Introduction
Global efforts to improve maternal health care include strategies to reduce maternal mortality, increase the proportion of births attended by a skilled provider, and ensure access to quality reproductive health services for all women. 1,2 Addressing preventable maternal and newborn mortality and morbidity is a key component of the sustainable development goals. 1 Although there was a substantial reduction (43%) in maternal mortality from 1990 to 2015 globally, still 99% of maternal deaths occur in developing countries. 3 Ghana has recorded improvements in its maternal health indicators, with a decline in the maternal mortality ratio from 634 in 1990 to 319 per 100,000 live births in 2015 3 while the rate of This article has been republished with minor changes. These changes do not impact the academic content of the article. facility-based deliveries increased from about 42% in 1988 4 to approximately 73% in 2014. 5 However, a significant proportion of women still do not access health facilities for childbirth services.
Women's reluctance to give birth in health facilities may be partially attributed to a fear of mistreatment by healthcare providers. [6][7][8][9] In 2015, a systematic review by Bohren and colleagues proposed an evidence-based typology of what constitutes mistreatment during childbirth, including: physical abuse (hitting, slapping, pinching), verbal abuse, stigma and discrimination, and systemic issues in health facilities and systems that contribute to poor experiences of care. 7,9 Mistreatment during childbirth violates the human rights and autonomy of women 10 and may constitute a major disincentive for seeking maternity care services at health facilities. A recent WHO statement regarding this public health and human rights issue called for more proactive action, dialogue, research and advocacy to improve maternal health globally. 11 Mistreatment of women during facility-based childbirth in Ghana has been documented, albeit in few studies. [12][13][14] D'Ambruoso et al. presented women's accounts of their interaction with birth attendants during facility-based childbirth in semi-urban suburbs of Accra. 12 Their findings showed that women were deeply concerned about health provider attitudes and refused to attend health facilities where they were not treated kindly. A study among student midwives across Ghana also confirmed occurrences of mistreatment during childbirth. 14 Similarly, a study by Moyer et al. 13 in rural northern Ghana noted that mistreatment during facility-based childbirth was pervasive, and could serve as a disincentive for attending health facilities in the future. These studies did not specifically assess views of acceptability of mistreatment during childbirth, or the perceived factors influencing mistreatment. Previous studies relating to disrespectful care during childbirth tackled different aspects of the mistreatment of women in health facilities without recourse to standardised definitions and methodologies. This has resulted in different descriptions and estimations of the true burden of the problem with diverse propositions for eliminating the mistreatment of women during childbirth, and promoting respectful maternity care.
Although the occurrence of mistreatment of women and disrespectful care during childbirth in health facilities is increasingly well-documented in both resource-rich and resource-poor countries, consensus definitions including standardised categorisations and measurement tools are lacking globally. 9 The lack of consensus on definition, classification and specific indicators for measurement of disrespectful care during the birthing process impede accurate prevalence estimation and development of effective interventions. 7,9,15 WHO is coordinating a two-phased, multicountry, mixed-methods study on measuring the mistreatment of women during facility-based childbirth in four countries: Ghana, Guinea, Myanmar and Nigeria. 16 The overall objective of this study is to develop an evidence-based definition, identification criteria and two measurement tools to better understand and measure mistreatment, based on the experiences and perceptions of women, healthcare providers and administrators in maternity care settings.
The first phase of this study is a formative phase, including a multi-country qualitative study to inform the development and components of the two measurement tools (direct observations of labour and a community survey with women). These measurement tools will then be used to measure the burden of mistreatment in different settings, and permit standardised comparisons across settings and over time. The qualitative evidence from the first phase will also improve understanding of the context in which mistreatment during childbirth occurs, the contributing factors, and also identify potential entry points to reduce mistreatment and promote respectful care. This paper explores women's perspectives of mistreatment during facility-based childbirth in the Ghanaian context specifically.

Methods
The study was carried out in two different towns in Ghana, a country with a population of about 25 million in 2010 and a population growth rate of about 2.5% per annum. 17 Most women in Ghana give birth in government institutions where maternity care is free due to complete coverage by national health insurance. The health system in Ghana comprises primary, secondary and tertiary levels of care. Women with pregnancy-related complications are referred to a higher level of care based on clinical urgency and expertise.
The study was conducted from May to July 2015 in the towns of Koforidua and Nsawam in the Eastern region of Ghana which has a population of about 2.6 million. 18 Approximately 97% of pregnant women in the region attend antenatal care, and 68% of births take place in health facilities. 5 We employed an exploratory qualitative design using in-depth interviews (IDIs) and focus group discussions (FGDs) to gain comprehensive knowledge on how women are treated during facilitybased childbirth.
The study involved women of reproductive age (15-49 years) in Koforidua and Nsawam who had recently given birth in any health facility, as well as health care providers and health service administrators working in two public health facilities. The focus of this paper is to share findings of qualitative research on women of reproductive age only. The perspectives of healthcare providers and administrators will be published separately.
To explore individual experiences and perceptions regarding mistreatment during childbirth, IDIs were conducted with women who had given birth in any health facility within 12 months prior to the study. Secondly, to explore community norms regarding mistreatment during facilitybased childbirth, FGDs were held with women who had given birth in any health facility within the previous two to five years.
Recruitment, sampling and data collection procedures Participants for the study were purposively selected by the community health nurses and midwives who live in the communities and therefore had knowledge about the locations of the potential study participants. Women for the FGDs were contacted through face-to-face interaction with a community health nurse within their respective communities. Those for the IDIs were approached by midwives from child welfare clinics at their respective facilities. Following recruitment by the community health nurses and the midwives, research assistants then approached and informed them about the study. All the women who were approached agreed to participate in the study. They were given a date, time and location for the interviews. Participants were stratified according to age (Table 1).
Before each IDI or FGD, participants provided written informed consent and had their demographic information taken. Each IDI and FGD was conducted by two experienced research assistants; a moderator and a note taker. Field workers (research assistants) adept in qualitative interviews were recruited and trained for the purpose of the study. The social scientist in the team communicated constantly with the interviewers in the field and responded to any issues that arose during data collection. There were six research assistants per site; 12 in all. The research assistants were non-healthcare providers who worked mainly as research assistants and had social science backgrounds with at least a first degree from a recognised university. Non-healthcare providers were used to avoid response or social desirability bias in the data collection, as the research team hypothesised that people with clinical training may interact with the research participants in a different way, or that research participants may share their experiences with healthcare providers differently from with non-healthcare providers. None of the research assistants originated from the study areas, or had relatives or friends in communities or health facilities in the study area. Due to the sensitive nature of childbirth issues, all the research assistants were females. A training session was organised for the research assistants during which a mock interviewing exercise was conducted by each participant in English and the local language (Twi). Semi-structured IDI (Appendix 1) and FGD (Appendix 2) guides were used to guide data collection. The primary domains of interest were: (1) expectations of care during childbirth at health facilities (2) experiences and perceptions of mistreatment during childbirth (3) perceived factors influencing mistreatment (4) decision-making processes to give birth at a facility (5) views of acceptability of different scenarios of mistreatment during childbirth.
The IDIs and FGDs were conducted either in English or Twi and audio recorded. The moderator and assistant debriefed immediately after each interview and recorded the debriefing so as to capture their observations and comments on the discussion. The field notes were also incorporated into the debriefing. Each IDI and FGD lasted between 45 and 60 minutes and 60 and 90 minutes, respectively. All the interviews and discussions were conducted in a quiet community centre in the study sites, outside participants' homes and health facilities with only the study participant(s) and interviewers involved. Each interview was conducted only once.

Data management and analysis
The audio recordings were transcribed in English by the interviewers soon after each interview. The transcripts were checked for completeness and accuracy by the field supervisor and the social scientist in the research team. In preparation for the data analysis, the researchers (a social scientist, public health specialists and obstetricians) and research assistants participated in a three-day workshop on qualitative data analysis facilitated by researchers from WHO. A codebook was developed based on the interview guides, findings from a systematic review 9 and themes emerging from the data. During the coding process, each member read each transcript several times to ensure in-depth familiarity. Three of the research assistants who completed the qualitative training workshop independently coded different transcripts manually based on the thematic content. Thereafter, the codes and emerging findings were discussed as a group until consensus was reached. After coding and categorising the codes, a working analytic framework was developed based on consensus from all the researchers and this was applied to all the transcripts to obtain the final output. Thus, the framework method of qualitative research, with inductive and deductive analytic approaches, was systematically employed during the analysis of this study. In the inductive analytic approach themes were allowed to emerge from the data, whilst the systematic review and interview guides were used in the deductive approach. The final analysis and presentation of the results were performed by the social scientist in the team with contribution from all other authors. The data was enriched by triangulation of IDIs and FGDs. Differences among team members regarding interpretations of the findings were discussed until consensus was reached.

Technical and ethical considerations
Scientific and technical approval for the study was obtained from the World Health Organization Human Reproduction Programme (HRP) Review Panel on Research Projects (RP2). Ethical approval was obtained from the WHO Ethical Review Committee (Protocol ID, A65880) and the Ethical Review Committee of the GHS (Protocol ID GHS-ERC-13/01/15).
Written informed consent was obtained from the respondents just before the interviews. They were assured of confidentiality of any information provided. To ensure anonymity, no identifiable participant information was taken. The participants were not compensated for taking part in the study but were given a token of GHC 30.00 (USD 8.00) to cater for their time and transportation.

Findings
Overview Forty-one IDIs and 10 FGDs were conducted with a total of 110 women ( Table 1). None of the women approached declined participation in the study. Most women were 30-34 years (33) or 15-19 years old (29), married (75), and residing in urban communities (82). Almost all the participants were Christian (109) and most were of Akan ethnicity (71) ( Table 2).
Prominent themes which emerged on the subject of mistreatment included (1) experiences and perceptions of mistreatment during childbirth (2) perceived factors influencing mistreatment,  it was a phenomenon that participants commonly associated with facility-based childbirth, and they shared stories that they had heard from friends or family members.

Mistreatment during childbirth described by women
With regards to forms of mistreatment experienced, there were themes on verbal and physical abuse, as well as psychological stress resulting from neglect and lack of supportive care, along with the critical influence of mistreatment on preferences for choice of facility-based childbirth.

Verbal abuse
Women detailed experiences of verbal abuse, including shouting, yelling, insults and derogatory remarks from healthcare providers. They identified these experiences as demeaning, and explained that verbal abuse happened across the duration of their stay in the health facility, from the initial contact with the healthcare providers, through labour and childbirth, as well as during discharge.
I: How were you received by the health workers when you got to the hospital? R: When I got there, I went there with a sister and I was asked what I am coming to do and I told them I am in labour. I was actually in pains so I couldn't even talk hard and they were shouting at me that I should talk for them to hear. Why? Am I a baby? How they received me I even felt sorry for myself for going there. It made me feel so bad. (IDI, Married woman, 32 years, Nsawam)

I = Interviewer R = Respondent
Those who experienced shouting and yelling in the early stages of labour described it as scary and intimidating. Women acknowledged that health care providers often yelled at them because they did not arrive at the health facility with the required items for childbirth (e.g. sanitary pads, mackintosh, soap and baby dresses) or they did not assume the "required" position for childbirth (to lie on their back with legs apart).
When we were going to the hospital, I knew my pregnancy was eight months so I told my sister not to take my things for delivery. When we got to the hospital, the nurse we met said "villager, hmmmm … foolish girl … " I didn't talk because I was in pain. Then the nurse asked me to go and sit on a chair there, then the other nurse said she has to go and lie down. The other nurse said this villager sitting there which bed is she going to lie on, so the second nurse took a cloth and spread it on the bed for me to lie on. I didn't know when you lie down you have to lie on your right side, one nurse saw me turning myself and she said I'm stubborn.
(IDI, Married woman, 30 years, Koforidua) Some women recounted experiences of midwives shouting at them when they were ready to bear down during the second stage of labour. These derogatory remarks usually targeted the woman's inability to push as expected.
I: How was the shouting like? R: Hey Ewura (lady), you are not the only one to give birth here. You have given birth before. Are you coming to show your arrogance to me, the nurse? I: … at what time of the delivery process did this happen? R: During the pushing period before the delivery. When the baby was about to come I shouted "auntie nurse please it is coming oooh". But because she was angry, she shouted at me. (IDI, Widowed, 40 years, Koforidua) Although, verbal abuse was experienced by women across all age groups, abusive remarks directed at younger women (15-19 years) often touched on the issue of teenage pregnancy.
When you get pregnant and you go to the hospital they would insult you because you are a teenager, so when you are fifteen or fourteen and you go there you will never be happy until you deliver your child. They [healthcare providers] have an age when you can get pregnant, that is from 20-30. (FGD, Married woman, 19 years, Koforidua)

Physical abuse
Participants also recounted physical abuse they endured during childbirth at health facilities. Physical abuse included pinching and slapping of their thighs and sometimes their backs. In some instances, such acts were accompanied by derogatory remarks about their progress through labour or ability to push. I: Did you like how you were treated in the facility? R: Please no … When I was pushing there were three nurses around me, when I couldn't push one nurse pinched me and said I should force and push, the other also slapped my thighs and also said I should push but I didn't push, because of that I will not go there again. I wasn't happy; it made another madam insult me and even said that they are not here to help people like me. She even went further to use the word "nasty person".
(IDI, Single, 19 years, Koforidua) These experiences were evident among women of all ages, and often occurred during the pushing stage of the childbirth process. Women sometimes described these gestures positively, viewing them as corrective measures taken by the healthcare providers to ensure safe childbirth. Although women were displeased with being hit, they believed that such gestures were necessary due to their inability to effectively push: R: When I was due for labour and was asked to push, I couldn't push and the nurse beat me very well. She used a cane to whip me so I could push, but I told her I was tired but she insisted I should push. So she really whipped me with the cane and later used her hand to hit my thigh.
There I became conscious and was able to push. (FGD, Married woman, 35 years, Koforidua) Failure to meet professional standards of care (neglect and lack of supportive care) Women's experiences suggested psychological stress during labour and childbirth resulting from instances of neglect, lack of support, and health workers' unresponsiveness to their needs. Although women acknowledged that the number of skilled attendants was inadequate to meet the demands of women, they felt that attendants focused on other activities rather than showing interest in the women. In extreme cases women gave birth in the facility without the presence of a healthcare provider, and believed that this was due to poor monitoring and assessment of women during the labour process: I: Why did you think you did not get the kind of support you anticipated? R: The nurses were not enough … Also even though the nurses were not enough they sat by their table too much. If they could move around the labour ward, they will know who needs help. One woman delivered on the bed alone in the labour room because we are not well monitored. (IDI, Married woman, 30 years, Koforidua) Some women recounted their personal experiences of typical displays of disrespectful care and lack of clinical support during critical moments of maternity care. They described mistreatment by healthcare providers as psychologically disturbing and making them feel helpless and powerless, as enumerated below by a respondent in a FGD: In contrast, a few women believed that mistreatment during childbirth was acceptable under certain situations. For example, some women believed it was appropriate for clients who do not follow instructions given by health workers. For a few women "pinching" and "beating" are corrective measures to alert mothers to assume the right position during birth or to enable them to push": I: Will pinching ever be justified in some cases? R: Come to think of it, yes, because some of us women have difficulty in understanding things, and some because they behave outside just like they do at home sometime their behaviour calls for that. Sometimes when they are told to push, they don't push, they will be lying there doing nothing. So to avoid being tagged that those on duty caused the death of so and so number of children they will do whatever it takes … I: How would you feel if it happens to you? R: If it's for a reason to save my baby I will accept it, but if for no reason, I will take the matter up (IDI, Married woman, 26 years, Koforidua) Women implied that some acts of mistreatment were useful for addressing a woman's misbehaviours, misunderstandings, and miscommunications with health providers during the birthing process. In these instances, acts of mistreatment are not expected to be the "first line" of action used by the health provider, but could be applied as "supportive" practices to help mothers through the birthing process. R: Sometimes is not the fault of the health worker. When it is time for you to push the baby out and you don't push, that one it is not the fault of the health worker. If she doesn't slap you to push the baby out you will not [push]. Sometimes it is our fault. I: Would this be acceptable? R: It is not right. You don't need to slap too much because some people do not want to push when they are asked to push. When you slap a bit, she will push, so it is right. I: So it is acceptable to pinch and slap? R: Yes, it will force us to push the baby out. I: How would you feel if this happens to you? R: It will encourage me to push for the baby to come out. (IDI, Cohabiting woman, 28 years, Nsawam) In some cases, women felt that when the provider yelled or slapped a woman during labour, it would help to communicate the gravity of the situation and as a way of "guiding you as to when you should push and when you should breathe in between" (IDI, Married woman, 32 years, Koforidua). These inconsistencies, dilemmas and differences in opinions regarding acceptability were particularly evident during FGDs among women across all ages. Women's views on acceptability or non-acceptability placed the various forms of mistreatment on a spectrum and depicted some as more acceptable and less harmful than others. Most women believed it was unacceptable for a healthcare provider to refuse to help women during labour and childbirth, and some women even noted that such an act should be reported. I: Is it acceptable for a health care provider to refuse to help a woman in labour? R: It cannot be acceptable because it is their job and so even if I call you and it is not yet time for me to deliver I would want you to come and examine me and tell me that my sister endure the pain because it is not yet time for you to deliver. But if you refuse coming, then it is intentional. And so if you even come and say something I will feel that when the time is up the baby would come out. (FGD, Married woman, 30 years Nsawam) Influence of mistreatment on preferences for choice of facility-based childbirth Some women stated that health workers' attitudes, including mistreatment, are major reasons for non-utilisation of skilled birth attendance. Some women identified previous experiences of mistreatment from healthcare providers as influencing future decisions regarding the use of the facility for childbirth, in which case they prefer to give birth at home:

I:
Why do some also give birth in the house? R 1: Please sometimes it's because of the behaviour of some of the nurses that is why they give birth in the house … Shouting and the intimidation, they are not patient with them.

Discussion
This qualitative study has demonstrated that multiple forms of mistreatment are experienced by women during facility-based childbirth in Ghana. In this study, most women reported experiencing verbal abuse in the form of shouting, yelling, insults and derogatory remarks, which had a negative impact on their self-confidence. Physical abuse in the form of pinching and slapping was also reported. Other forms of mistreatment included abandonment, lack of support, neglect and unresponsiveness to women's needs. Although mistreatment occurred throughout the birthing process in the health facilities, women reported that it was most common during the second ("pushing") stage. This study has also shown that mistreatment can disincentivise facility-based childbirth in the future. Our findings are similar to the few previous studies in Ghana, which reported women experiencing physical abuse, scolding, shouting and abandonment during childbirth, as well as health facilities that are unresponsive to their needs, or unable to provide the necessary emotional and physical support during childbirth. [12][13][14]19,20 Failure to push, young age and inability to bring all items required for the birthing process were reported in our study as potential triggers for mistreatment, also echoing previous studies in Ghana. 13,14,20 Mistreatment during facility-based childbirth is increasingly recognised as a widespread problem. Comparable findings of mistreatment of women, in varied forms, during childbirth in health facilities have been reported in Tanzania, Kenya and Nigeria. 6,8,21 Other study sites in this WHO-led multi-country study (Nigeria and Guinea) have revealed similar experiences of mistreatment in health facilities. 22,23 Such mistreatment may have negative effects on future maternal health-seeking behaviour and choices regarding facility-based childbirth. A recent systematic review determined the existence of a wide range of disrespectful and abusive treatment of women during childbirth in Nigeria. 24 The authors suggested that female education and empowerment, strengthening of the health systems and refresher training of healthcare providers represent viable solutions to reducing disrespectful treatment of women during childbirth. 24 Disrespectful and undignified care during facilitybased childbirth also occurs in higher-income countries 25,26 but forms and patterns of mistreatment may differ.
These findings buttress the need for concerted efforts (both locally and internationally) to objectively measure and determine the prevalence of mistreatment and abuse during childbirth. This will serve as the necessary contextual information on which any evidence-based measures to eliminate mistreatment and promote respectful maternity care could be based. Elimination of such forms of mistreatment could improve facilitybased childbirth rates and ultimately maternal and perinatal health outcomes. Previous studies in Africa have demonstrated clearly the relationship between mistreatment and future childbirth in health facilities. 13,27,28 Eliminating all forms of mistreatment and ensuring that women's rights are respected hinges on extensive education, especially in low resource settings, on whether such acts are appropriate or useful. In this study, some women believed that in some contexts, some forms of physical and verbal abuse are needed to ensure the safety of the mother and baby during childbirth. Trainee midwives in Ghana 14 and healthcare providers and women in Nigeria 29 and Guinea 23 have also rationalised some mistreatment as ensuring good pregnancy outcomes. This may explain why mistreatment was reported to be more prevalent during the second stage of labour, the terminal and critical phase of the birthing process. Using mistreatment as a means to achieve better outcomes during childbirth may have dire consequences, given that mistreated women may opt for home births in the future, as attested to by this study and others. 13,[30][31][32] This could in turn derail gains made in reducing maternal mortality and improving quality of care. The cultural normalisation of some forms of mistreatment needs to be considered and addressed when developing interventions.
There is also a need to address systemic failures within health facilities and health systems that contributes towards mistreatment and abuse of women. Although low staff numbers (as indicated by women in our study) can affect the quality of care, women raised concerns regarding staff idleness. This made women feel powerless, and led to negative labour and childbirth experiences. Similar sentiments have been expressed by women in South Africa and Tanzania 32,33 and can erode the importance of attending health facilities for childbirth.
In Ghana, there are specific health system challenges which influence the optimal provision of maternal health care including delays in providing care, shortage of healthcare providers, logistics, laboratory supports and effective drugs. Some of these specific issues came up in the health workers' interviews which are not incorporated in this paper as its focus is on women's perspectives on mistreatment during facility-based childbirth. Health system-related issues are described in a forthcoming paper on health worker perspectives. A recent systematic review has highlighted persistent lack of adequate resources, among other factors contributing to the delay in providing emergency obstetric care in developing countries. 34 Despite these challenges, significant, although slow, progress has been achieved in key maternal health indicators in Ghana. A recent country-wide survey in 2015 indicated that 97% of pregnant women obtained antenatal care from a skilled provider compared to 82% in 1988. Health facility-based childbirth and supervised skilled-birth provision have increased from 42% and 40%, respectively, in 1988 to 73% and 74% in 2014. 35 However, a wide disparity in maternal morbidity and mortality persists compared with resourcerich countries. This variation in clinical outcome indicators could be attributed to differences in the quality of maternity care services where a significant proportion of women in the developing world do not have access to supervised facilitybased childbirth, despite remarkable improvements in antenatal care coverage.
The structure of the health system in Ghana is such that most uncomplicated maternal-related cases are managed at primary and secondary health facilities whilst complicated and severe cases are managed at tertiary or teaching hospitals. In cases of obstetric emergencies, patients are usually transferred to tertiary hospitals to access a more complete multi-disciplinary specialist level of care. Obstetric emergencies which present at lower-level facilities are transferred to a higher institution based on the preparedness of the facility for such complications. On the other hand, some clients obtain maternity services from traditional births attendants (TBAs) who have not had formal training in the provision of maternal health care. In Ghana, TBAs still conduct a significant proportion of deliveries, despite several attempts to channel all births to health facilities. The recent Ghana Demographic Health Survey indicated that 16% of births are delivered by traditional birth attendants. 35 The occurrence of a significant proportion of births unsupervised by trained providers in health facilities in the country might be related partly to the persistence of mistreatment of women during facility-based provision of maternity care services as recounted by respondents in the current study. The findings of this study are instrumental in the attempt by WHO to globally estimate the true burden of the phenomenon of mistreatment and its underlying factors using internationally validated measurement tools with subsequent identification of potential entry points to minimise mistreatment and improve respectful care.

Strengths and weaknesses
One of the strengths of this study is being part of a larger WHO multi-country study, which aims to define mistreatment during childbirth, and develop and validate two tools for measuring its occurrence (a labour observation tool and a community-based survey tool). The lived experiences and perceptions of Ghanaian women described in this paper have been synthesised with findings from other participating countries (Guinea, Myanmar and Nigeria), and stakeholder groups (healthcare providers and administrators) and integrated into the design and components of the two measurement tools. This will ensure that the tools address what is actually experienced by women giving birth in health facilities in different settings. For example, while the forms of mistreatment identified in our study have been documented in other settings, the timing, specific actions and terms used by Ghanaian women to describe mistreatment informed how corresponding measurement questions are framed and timed. Quantitative findings on the application of these tools in maternity care facilities in Ghana will be explored in a forthcoming paper.
An additional strength of this study is the inclusion of the views of women purposively sought from all age groups (adolescents as well as adults) in both urban and suburban settings. Grouping women into similar age groups for the FGD allowed the adolescents to express themselves without fear of reprisal from older women, which culturally is the norm. Also, we employed non-clinical research assistants to reduce the risk of social desirability/response bias and there were regular ongoing discussions with multi-country research teams about emerging findings. This was buttressed with regular monitoring by the social scientist in the team to respond to issues related to the data collection.
This study also has some weaknesses. It is possible that the nurses who recruited the study participants may have influenced their responses to favour providers. Different socio-economic factors, including educational level and women's marital status, that could affect their childbirth experiences, were not stratified. Despite this potential limitation, women in this study reported negative personal experiences or experiences of others within their social network. Given the qualitative methods used in this study, the findings may not be transferable to populations outside of the study settings. We did not compare the occurrence of mistreatment of women during childbirth at the facility to home deliveries by TBAs. However, our findings, indicating a significant burden of mistreatment of women during facility-based childbirth in various typological forms, are aligned with other studies conducted in sub-Saharan Africa. 23,[35][36][37] These findings, and forthcoming findings from our research group on healthcare providers' and administrators' views on mistreatment, can help to identify and implement strategies to address the problem of disrespectful facility-based maternity care. Our findings provide the evidence to encourage the Ministry of Health to engage stakeholders and formulate appropriate policies to improve respectful care of women during childbirth in health facilities. Strategies need also to engage pre-service training institutions, to address cultural norms around mistreatment at health facilities during childbirth and to sensitise staff on the occurrence and consequences of mistreatment of women during childbirth. Furthermore, collaborating with women's and community groups to raise their expectations and demands for high quality care will be critical to design and adapt health services to be more woman-centred. Provision of high quality of care by eliminating disrespectful treatment of women at the time of childbirth will improve maternal health in Ghana and ensure that all women have the right to access high quality maternal healthcare.

Conclusion
Mistreatment of women occurs during facilitybased childbirth in our study sites in Ghana, with negative impacts on women's care experiences. Conflicting opinions exist amongst women regarding its appropriateness and usefulness. Evidence-based consensus definitions, as well as validated indicators and tools for measuring mistreatment, are needed to help identify the underlying factors and potential entry points to minimise the phenomenon and improve respectful care during childbirth.
Thank the participant for his/her time. Remind them that the information will be kept confidential.
End time: