Disrespect and abuse in maternity care: individual consequences of structural violence

Abstract Disrespect and abuse of patients, especially birthing women, does occur in the health sector. This is a violation of women’s fundamental human rights and can be viewed as a consequence of women’s lives not being valued by larger social, economic and political structures. Here we demonstrate how such disrespect and abuse is enacted at an interpersonal level across the continuum of care in Tanzania. We describe how and why women’s exposure to disrespect and abuse should be seen as a symptom of structural violence. Detailed narratives were developed based on interviews and observations of 14 rural women’s interactions with health providers from their first antenatal visit until after birth. Narratives were based on observation of 25 antenatal visits, 3 births and 92 in-depth interviews with the same women. All women were exposed to non-supportive care during pregnancy and birth including psychological abuse, physical abuse, abandonment and privacy violations. Systemic gender inequality renders women excessively vulnerable to abuse, expressed as a normalisation of abuse in society. Health institutions reflect and reinforce dominant social processes and normalisation of non-supportive care is symptomatic of an institutional culture of care that has become dehumanised. Health providers may act disrespectfully because they are placed in a powerful position, holding authority over their patients. However, they are themselves also victims of continuous health system challenges and poor working conditions. Preventing disrespect and abuse during antenatal care and childbirth requires attention for structural inequalities that foster conditions that make mistreatment of vulnerable women possible.


Introduction
Tanzania has made slow progress in reducing maternal mortality, failing to achieve Millennium Development Goal 5. 1 Significant progress between 1999 and 2015, however, was achieved in increasing facility births (from 47% to 63%). 2 While this is a reason for optimism, over recent years several studies have reported evidence that raises concerns about the poor quality of care women receive in some of these Tanzanian institutions, including frequent experiences of disrespectful and abusive treatment by health providers during childbirth. [3][4][5] Disrespectful and abusive treatment during childbirth is a violation of women's fundamental human rights, can negatively influence birth outcomes and discourages women from seeking future care. 6 Numerous individual practices and behaviours of health care providers can be considered as disrespectful and abusive, depending on the definitions that are used. Examples range from behaviour being non-supportive (such as not providing information) to physically harmful practices (such as slapping or beating). 7 Mistreatment of women in health facilities is rooted in pervasive gender inequalities and power imbalance between health providers and women. 8 Therefore, disrespect and abuse can be viewed as a consequence of structural violence. 9 Structural violence refers to social forces that create and maintain inequalities within and between social groups, which make way for conditions where interpersonal maltreatment and violence may be enacted. 11,12 Although the term "violence" speaks to the physical nature of disrespect and abuse in childbirth, the essence of structural violence lies in the indirect, systematic and often invisible infliction of harm on individuals by social forces that disable individuals from having their basic needs met. 11 We may be tempted to analyse this phenomenon in a narrower framework, such as seeing women as "victims" and health workers as "perpetrators" of abuse. 13 However, the mistreatment of women in health facilities is systemic and requires a more structural analysis to look at the issue as a consequence of women's lives not being valued by larger social, economic and political structures. 14,15 Despite 30 years of action at the global level to improve care for women during pregnancy and birth, many countries, including Tanzania, have never been able to make the financial investments required. 14 Instead, expenditures for maternal health over the past decades have increasingly relied on household contributions. 1 In response to structural adjustment policies, the Tanzanian government introduced cost-sharing and decentralisation and reduced the already limited number of health workers and their salaries. Up until today, the human resource scarcity remains a major bottleneck. 16 At the same time, the population has doubled, increasing the burden on a fragile health system. HIV/AIDS and more recently non-communicable diseases have contributed to this fragility. 17 It is not surprising that increasing resource challenges and overload of health facilities have resulted in decreased health worker morale, lack of compassion, fatigue, and sometimes burnout, which are often reported to be underlying reasons for mistreatment of women. 18,19 Over a decade ago it was suggested that ensuring respectful, high-quality care for all women was a matter of political will to value the lives of women and newborns. 20 Nevertheless, the Safe Motherhood policy discourse remained focused on technical solutions and scaling up simple disease-specific interventions, particularly a focus on skilled birth attendance and access to emergency obstetric care. [21][22][23] Simultaneously, health system challenges, including limited resources, insufficient training and poor working conditions of health providers, continued and/or deteriorated even further. Disrespect and abuse during childbirth occurs in an impoverished social and political context, in which women's broader needs during pregnancy and birth have been systematically ignored or devalued. In this paper, we describe how and why women's exposure to disrespect and abuse in health facilities should be seen as symptomatic of structural violence.

Study setting
The study took place from September 2015 to February 2017 at two health centres and one district hospital in the Lake Zone in Tanzania. Facilities were selected based on our previous involvement in the district, ensuring familiarity with the leadership and health professionals. All three facilities were assessed in terms of basic infrastructure, staffing, resources and quality of service provision as part of a district-wide emergency obstetric care assessment. None of the facilities in the district performed in accordance with international guidelines, primarily influenced by lack of available resources and an insufficiently functioning health care system. 24 Some basic characteristics of the three health facilities are provided in Table 1.
Researchers positioning ASM (a medical doctor) and SP (a nurse) both speak Kiswahili and spent several years in the study area. Both authors were involved in setting up and managing a community-based project and volunteering at different health facilities in the study area. During the data collection period, both spent a total of 52 days at the antenatal care (ANC) clinics or maternity wards of these facilities, observing and participating in care provision. For ASM, this sometimes meant active participation in the form of providing ANC and assisting births. SP remained as an observer but also assisted with minor tasks. JS and TM supervised the study and both have extensive experience working in similar settings in sub-Saharan Africa. All authors were trained in a high-income setting and approached this study from a biomedical perspective. This study was performed with attention to respectful maternity care as defined by the World Health Organisation.
At many of the health facilities visited, the authors observed a lack of respectful maternity care. ASM and SP's long-term involvement in the study area revealed the challenging working conditions of health providers that compromised their ability to provide quality care. Many of the health facilities were in a state of collapse and the basic infrastructure allowed for little room to ensure patient privacy. Health providers frequently shared their struggles in terms of their working environment, underpayment and long working hours. With few exceptions, ASM and SP experienced that all health providers intended to provide good care, aiming for good outcomes and thus this paper, does not indicate health provider perspectives or intentionality of their behaviour.

Study population
Fourteen women were purposively selected with different obstetric backgrounds, age groups and poverty levels. All women had a vaginal birth and half of the women gave birth at home. They were followed up throughout their pregnancy, birth and post-partum period. Recruitment was done in a staggered way to ensure researchers did not follow more than four women at the same time. Women's characteristics are presented in Table 2. Socio-economic status was categorised based on a number of indicators including possession of assets (mobile phone, livestock, furniture) and living conditions (e.g. housing structure, electricity, type of water source). Additional details can be found in Supplementary File 1.

Data collection process
Following selection of women during observations at the ANC clinics, ASM and SP scheduled subsequent observations at the clinics for the expected days of women's return visit. In total, 25 antenatal visits of these women were observed. On some occasions, visits were not observed because women did not show up, did not receive services, or were attended to while the authors were unable to be present. Additionally, observation days were scheduled at the maternity wards for women's expected dates of delivery. Aspects of the birth process were observed for three of the seven women that gave birth in the health facility. In total, 92 indepth interviews were held with all women, scheduled 1-2 weeks after each of their clinic visits and after birth. Additional interviews were held if further clarification was needed. Interviews were conducted in Kiswahili, lasted 1-3 hours and took place at the women's home, or a location of their choosing. As a starting point, the focus of the interview was on women's perceptions and experiences related to their previous visits at the health facility, discussing both clinical and interpersonal aspects of care provision. Probing questions were asked based on the women's antenatal cards and on the observations. The way women define and explain events is influenced by their background and previous experiences, 25 therefore interviews included questions about women's childhood, their first pregnancy, marriage and subsequent pregnancy experiences, if any. Previous and current choices the women made in relation to care seeking or with regard to other major life events were discussed, providing information about women's perceptions of their self-efficacy, their social identity and the influence of their social networks.

Data collection tools
Observation of behaviour is highly subjective and challenging, particularly if conducted in a cultural setting different from the observers, since behaviour can be enacted differently across cultures. 26 However, health providers in Tanzania are expected to perform according to standards of professional conduct. 27 These standards include guiding principles that must be followed when caring for patients, such as ensuring to obtain patient consent before providing care and protecting confidential information. To reduce the influence of the author's personal judgment, observation guidelines were developed in line with these standards to provide some level of standardisation to the interpretation of what was observed. Few instruments exist for observation of interactions and behaviour of health providers in maternity care in low-income settings. 28 Considerably more work has been done in highincome countries, often limited to intrapartum care, or with reference to nursing care in non-maternity settings. 29 Based on existing literature reviews, 7,28,30-32 categories and sub-dimensions for both supportive (Table 3) and non-supportive behaviour (Table 4) were developed. The categories of disrespect and abuse as defined in previous studies have a tendency to be either too narrow, 31 or too comprehensive 7 for practical use. For these categories, complex concepts were avoided (e.g. nondignified care), potential overlap between categories was reduced (e.g. physical abuse, sexual abuse) and the total number of categories was limited.

Analysis
Analysis of observations and interviews occurred continuously throughout the data collection period. Detailed reports were written after each observation day. All interviews were recorded and transcribed in Kiswahili and translated into English by a research assistant. Transcripts and observation reports were synthesised and, in dialogue with the women, were placed in chronological order based on the timeline of women's lives. Through this, we developed detailed narratives of women's reproductive lives and interactions with the health facility during their current pregnancy. Narratives can be a tool to unravel the unconscious structures, conventions and norms through which people make sense of and cope with their lives. 33 For the purposes of this paper, we analysed the narratives in two phases. First, we performed a deductive thematic analysis of narratives, whereby we coded situations exemplifying supportive care and non-supportive care. Second, we looked at women's daily experiences through the lens of structural violence. We analysed women's exposure to non-supportive care in relation to the social context, deconstructing the categories of care and their meanings, forming overarching themes.

Validity
We took several measures to ensure the validity of the development and interpretation of the narrative text. First, the increased familiarity between the researchers and the women resulted in increased confidence and trust in the researchers. Women shared personal details they had left out initially and offered less socially desirable answers. Second, conducting several interviews allowed us to revisit previously discussed issues, gain clarification and further explore questions that arose during the writing of the narrative. The intervals between the interviews also allowed both the researchers and the women time for reflection. Third, the authors encouraged women to think Physical support Assists patient gently and in a culturally sensitive way during examinations. Birth specific: Offers, checks, encourages and assists woman to take fluids/food, go to toilet regularly, changes clothing and linen, showers or bathes. Provides pain medications, encourages relaxation or other ways of support (counter pressure, assists in walking, assuming different positions). Coaches through labour such as with breathing and relaxation or touch (holding hand, massage).

Effective communication
Gives explanations: Explains to woman when to contact the midwife, explains what needs to be done in case a complication occurs. Explains procedures or treatment, what is done and why and informs of findings. Gives information: Provides update on progress of pregnancy and birth. Gives instructions: Instructs woman what to do during pregnancy including how to cope with normal pregnancy symptoms. Informs woman where to go if supplies are not available. Birth specific: instructs patient during and after birth how to participate to improve outcome, information on how to cope with pain, coaching during pushing. Advises patient to change position, walk around, breath in and out for comfort. Involves: Provides woman with options and involves in decision-making. Asks woman if she has questions and encourages her to ask questions.
Nursing proximity Is accessible, comes quickly when woman or family member calls, expresses accessibility verbally, encourages woman to request assistance and express needs, faces the woman and position at the same level.
Privacy respected Ensures privacy and confidentiality: uses curtains, sheets, and positions to avoid exposure, discusses privately with client or with colleagues or with family, minimises interruptions.

Consent
Asks for permission before performing examination or medical procedures. more critically about the interpersonal behaviour of health providers in relation to norms and values of social interactions in daily life. We explored local perspectives on the interpretation of behaviour through discussion of the narratives with a small group of local health professionals including a male Tanzanian gynaecologist/obstetrician and a female midwife. The group also included a young mother (ICT specialist) with both positive and negative birth experiences. The group was consulted in relation to the observation guidelines mentioned above. As women were included gradually and data collection and analysis occurred simultaneously, discussions with the group guided our focus with women who were subsequently included.

Findings
All women were exposed to both supportive and non-supportive care, including instances of disrespect and abuse, throughout their pregnancy and birth. Half of the women described similar experiences during previous pregnancies and births. Tables 5 and 6 give an overview of both supportive  and non-supportive care that women were exposed to during their recent pregnancy and birth.

Normalisation of absence of care
Women and health providers often interacted in complete silence and care provision was frequently devoid of any form of verbal communication. Women were rarely greeted or welcomed and were not addressed beyond simple instructions such as "simama hapa" (stand here), "panda" (climb) or "kaa" (sit). This is a cultural deviation, as greetings are very important in all social interactions in Tanzania. Women were not always informed about the findings of examinations or results of laboratory tests and they rarely received information about the system of care provision. Additionally, women's concerns, opinions and knowledge were frequently ignored.
"They check and see what they see, they don't tell us whether it is positioned well or not, they don't say, they just measure." (Paulina, interviews)   Birth: Non-supportive care Other than being told birth was not until "later" she remained both at the waiting area and labour ward for a long time (total 24 h) without information about the progress of labour or condition of the baby or herself despite her own expressed worries about the amount of blood she had lost.
Jessica ANC: Non-supportive care, privacy violation, nonconsented care, psychological abuse During her first ANC visit she was not given explanation of the findings of laboratory investigation. Was ignored when she informed the nurse about her complaints (stomach ache). She was given a return date on a Sunday. During physical examination at her first visit, there were many interruptions by students coming in and out. Was reprimanded for coming too late to her second clinic visit and threatened she would not receive services.
HIV test was performed without counselling or asking for consent.
Angel ANC: Non-supportive care During her first ANC visit she was not provided with the ANC card and therefore had to buy and bring a notebook. She had been seen for ANC the morning of the day that she delivered a stillbirth. During this visit her stomach ache was dismissed as being part of normal pregnancy symptoms, she was not informed about her condition or that of the baby.
Flora ANC: Abandonment, nonsupportive care For her first attempt to visit the ANC clinic she was refused services because she did not come with her husband During her actual first visit she was not welcomed, no explanations were given about the service received, no safe space to express her symptoms of pain and cough. Birth: Psychological abuse, abandonment, non-supportive care Upon arrival at the facility for birth, the nurse responded annoyed and aggressive upon arrival at the facility, wanted to send her to another hospital for unclear reason. The nurse was angry at her for having squatted above the basin and said that she would have lost her job if the baby had been born in the basin. She was told to clean up after birth and was threatened not to get back ANC card or would not be allowed to go home if her sheets were not washed. Was not taken seriously when she told the nurse she felt to push. Was left alone for 4 hours without any check-up. Eventually, the head was born without nurse present.
Jane ANC: Non supportive care, abandonment. During her first ANC visit she was not believed when she told the nurse her last normal menstrual period. Her information was dismissed and the nurse gave her an estimated date of birth two months beyond what Jane believed to be true. No education or information provision on measurements or what to expect during pregnancy and birth despite it being her first pregnancy. Was refused services at her second clinic visit because this day was not for her village.
Birth: Psychological abuse, physical abuse, non-supportive care, abandonment, non-consented care Upon arrival for birth she was not welcomed at the health centre, instead she was ridiculed by the nurse because she was not adequately responding to the nurse's comments. The nurse had told her firmly not to drop her clothes on the floor. Was told her son almost died because she was not pushing well. Was told firmly to stop crying without attention for the reason why she was crying. Vaginal examination and interventions were done without consent and forced while restraining her if she resisted. She was told she was not relaxing, she was not doing what she was supposed to do. Received directions on what to do in a firm way "keep arm straight", "don't move", but no explanation on what was going on, why she received interventions and no consideration for her pains or her fear. Was refused to drink water as this was thought to reduce contractions. Was ignored when she arrived at the health centre until her family member provided the necessary gloves to the nurse. Vaginal examination and interventions including fundal expression and episiotomy without consent.
Tara ANC: Non supportive care During her second ANC visit, the nurse was ranting about her working conditions while performing services and scolded another woman because she said she could not hear her well.  Despite women expressing disapproval about how they were treated at facilities, women frequently referred to services being "kawaida tu" (only normal), "nzuri tu" (only good), because it is how it always is. Women routinely attended their scheduled visits at the clinic. They expressed that this was their responsibility and a necessity to know if everything was normal. Even though some women had performed a pregnancy test, it was not until a nurse at the clinic confirmed this that they embraced the full truth of being pregnant.

Justification of punishment and rewards
Younger, less experienced women were more likely to experience disrespect and abuse, mostly because they did not behave as they were expected to, for example, if they did not bring the necessary "vifaa" (supplies or materials) for birth, if they did not dress properly or if they did not follow the system of care provision. Sometimes women were reprimanded or scolded if they did not do as they were told. When Flora was admitted to the labour room, the nurse repeatedly told her to lie on her side and instructed her not to push, even though Flora felt it was already time. After birth Flora cleaned the bed and was instructed to clean her sheets, otherwise she would not get her ANC card back. Some women deliberately took precautions to avoid being confronted with disrespectful behaviour. Women said they would "keep quiet", refrain from asking any questions and make sure not to attract any unwanted attention. Particularly for younger women, acting as more confident and experienced could result in better treatment. For example, when Pili entered the ANC room, the nurses did not greet her but instead directly asked her for the name of her village. Pili responded swiftly and confidently while asking the nurse: "Did you forget?" During the interview, when asked how she presented herself in the clinic, she said: "I am entering there [at the facility] very confident, like a true woman […]. I am doing that because if you are scared you will feel they are bad but if you go in a charming way you are just like them. You see them they are good." (Pili, interviews) During health education sessions, stories about what happened to women if they did not behave well often resulted in laughter from both the nurses and the women. Not only was such behaviour by nurses considered "normal", it was a necessity because "some women don't know how to behave". Some women justified health providers' strict language, threatening behaviour and verbal or physical acts for disciplining, for example: Whose effort counts?
At times, several women were unable to receive services while attempting to attend the ANC clinic.
On some occasions women were refused services because they were too late, did not come with their husband or because their type of service was not available on that particular day. Sometimes clinics were closed unexpectedly due to lack of available staff, during national holidays or when health workers were receiving supervision or training. Often these closures seemed arbitrary, as we observed that attending to the pregnant women would have been possible. For example, when Bea was unable to attend her fourth visit, the following was observed: At 8 a.m. there were three women at the entrance of the ANC clinic, including Bea. At the reception two nurses were sitting and resting their head in their hand, another nurse was lying down with her head on the table.
One of the nurses approached the women and said there would be no service today because they were expecting to receive special education. Women were instructed to come back after the weekend.
[…] When walking back to the bus stand Bea said this was a bad situation and that she wasn't happy. She came with the 'daladala' [taxi bus] but now she needs to come back next week.
[…] The following hour and a half, while the nurses were waiting for the training to start, one more woman was told there was no clinic today, another woman was helped with measuring the weight of her baby and a pregnant woman was assisted to collect antiretroviral tablets. (Bea, observation notes) Bea was already far into her pregnancy and never managed to attend to a fourth visit because she gave birth at home the following day. When women were unable to receive services, there was rarely an empathic reaction or apology for the inconvenience. Women's efforts using their time and personal resources to come to the facility in vain seemed not to be valued. In contrast, the women nearly always appreciated nurses' efforts, even if this meant women needed to tolerate physical and verbal abuse. For example, the following events were observed during Jane's birth: Two nurses [medical attendants], walked towards Jane deciding to help her. Nurse Esther stood at the right side and Nurse Dynes stood at the left side of Jane. Dynes supported Jane's head while Esther actively spread Jane's legs and told her to push. 'We are using traditional methods now' she said. Esther and Dynes folded a 'kanga' [a local fabric] on the stomach of Jane like a belt and when Jane had a contraction they pushed the kanga down and screamed 'push!!' The head of the baby slowly became visible. Esther put her fingers in the vagina and said to the doctor who was present: 'Look, look there is space, mama is not pushing! There is a lot of space.' She moved her fingers around in Jane's vagina with force, around the head of the baby and repeated this several times. No one spoke with Jane, she gasped heavily, was sweating and looked tired but the nurses did not pay attention to her. Another contraction came and Jane pushed while Esther hung with her full body weight on Jane's abdomen to push the baby down. Esther screamed 'you are not pushing, mama push, you let us do all the work!' Dynes asked for a scissor, placed it at the perineum and made the cut. Jane was not informed and let out a piercing scream. Both Esther and Dynes took a part of the kanga at one side of Jane and, created a rhythm with their voices. 'Push, push, push, push, push' while pushing the kanga down. Jane looked exhausted. She was gasping for air with her eyes wide open. Every time she wanted to take a breath someone told her to push. 'You don't speak! PUSH', they said. […]. (Jane, observation notes) Jane explained later she was afraid her baby would die, she had been in pain, but was mostly worried about her child. She thanked God he survived.
"She [the nurse] was just giving me normal service, that is good service […] because the nurses worked at it, they attended me." (Jane, interviews)

Discussion
Women's narratives revealed how they were repeatedly exposed to disrespect and abuse in their interactions with health providers, during ANC, during childbirth, and from one pregnancy to the next. All women, regardless of their age or socio-economic status, experienced both non-supportive and supportive care (see Table 5), sometimes by the same nurse within the same setting. Women's experience of such conflicting treatment is further complicated by the manifestation of non-supportive care. Our findings reveal how normalised and legitimised non-supportive care has become over time, with women lacking power or opportunities to confront this experience.
The majority of women in our study grew up in poverty and were still living with grave economic insecurities. Many of them were pushed into early marriage due to teenage pregnancy and were unable to continue their education. Few women had an independent income. The majority of women therefore relied on their husbands to provide for the necessary expenses to access care. In Tanzania, many young girls and women experience abuse in school (Tanzania allows corporal punishment) 34 or are exposed to intimate partner violence. 35 Health care institutions reflect and reinforce dominant social processes in their society. 36 The way women are treated in health care settings correlates with their position in society and vice versa. It should not be a surprise that such frequent and normalised abuse in everyday life leads to equal normalisation of similar poor treatment in health care institutions. 37 For many women, their first experience of disrespect in a health facility is the absence of greeting by health providers and of a welcoming reception. This might seem of little relevance in the greater debate on abuse and disrespect during childbirth. However, the absence of greeting is a rejection of social rules that health providers outside the health institution abide by. In health institutions, women appear to lose their social identity, and "lose their right to be respected". 38 Women frequently expressed disapproval of such interpersonal behaviour but at the same time felt disempowered to change this. Normalisation of non-supportive behaviour is symptomatic of an institutional culture of dehumanised care. In such a context, women have to accept a deplorable physical environment, inadequate (human) resources, and to endure disrespectful and abusive treatment. 39 Repeated exposure to such non-supportive care ultimately weakens women's agency, including their self-esteem and sense of safety. 40 Regardless of low levels of education or socioeconomic status, women are aware that they deserve better, and do not simply submit themselves to poor treatment. 41 They were consistent in attending ANC, even if they were frequently disappointed or if their knowledge or opinion was dismissed. Women frequently expressed that they trusted nurses to know what was best for them. The active suppression of women's knowledge and women's firm belief in what nurses represent is referred to by Jordan as "authoritative knowledge". 42 Health providers may act in disrespectful or abusive ways, in part because they are in a powerful position and represent a powerful system. 32 Their level of education and technical biomedical knowledge confer superior social status 38 in relation to their female patients 43 and this power imbalance influences how they behave towards women. Women are expected to adopt behaviour imposed by the nurses and to abide by these rules when they come to the facility for services or to give birth. Consequently, if women don't comply, or are unaware, they are perceived to be disobedient, and are themselves held responsible for poor outcomes. To regain control, health providers can turn to abusive measures to force compliance. 43 Women justify this behaviour even though they fear exposure to it. 10 Our findings illustrate how women use tactics to avoid mistreatment and are proud if they are able to do so. Such submissive behaviour symbolises how women through their oppression have internalised the prescribed behaviour. 44 Addressing the mistreatment of women in health facilities is finally gaining momentum in the global field of maternal health, leading to the integration of respectful maternity care in critical guidelines. 45 But within the current global health culture of relying on metrics, 46 such guidelines risk oversimplifying individual women's needs. The search for universal definitions of disrespect and abuse in child birth, as well as clear typologies of what this includes, can result in misleading or narrow dichotomies which devalue the routine and often subtle nature of women's suffering and the complexity of what drives it. 10 Nurses are themselves confronted with hierarchical power structures within their work. Medical doctors or others in leadership positions can undermine nurses' authority and decision-making ability. 47 Predominantly female health providers have gone through the same abusive educational system and their ability to provide quality care is seriously compromised by a lack of resources and support, and the perceived threat of losing their jobs in case of poor outcomes. Similarly to the women they provide care to, they are unable to change their situation due to their perceived lack of voice, both within the nursing education system and within the health system as a whole. 47 Nurses may act as oppressors, while also being oppressed by the same social forces that maintain structural violence.
The global maternal health community needs to be more self-critical and reflect on how global health interventions may contribute to women's mistreatment. Examples include women being refused services if they come without their husband, or finding the clinic closed due to supervision visits or skills training. The lack of recognition of women's efforts to get to the health facility, often in vain, contributes to the complexity of this situation. Global statistics on antenatal coverage are a representation of services that are provided but do not reflect the true picture of women's care seeking. Women seek services, but do not always receive good quality care, nor are they always treated with respect. 39 Acknowledging disrespect and abuse of women in health facilities as a consequence of structural violence requires us to move beyond viewing disrespect and abuse as a primary problem during childbirth. Mistreatment of women should be holistically tackled across the continuum of care, through structural interventions. Preventing disrespect and abuse at its core requires an approach beyond improving health workers' skillsets and achieving organisational changes at institutions level. Societal conditions that keep women's status inferior must be addressed, 32 policy and funding priorities must be discussed, 21 and collective efforts are needed to establish accountability mechanisms whereby the appropriate authorities are held responsible for women's lack of access to respectful care. 14,48

Limitations
Although we attempted to keep much of the original wording of participants, the narratives are a product of our subjective interpretation of the situations and thus particular and incomplete. The knowledge generated can therefore not be generalised. 49 However, following Fathalla's story "Why did Mrs X die?" presented during the launch of the Safe Motherhood movement in 1987, 15 there are lessons which can be garnered from individual stories. Some authors argue that to determine if certain behaviour is "abuse", it needs to be subject to variation based on culture, context and personal expectation or experience. 5,21 Freedman et al 21 proposed that local consensus as to what constitutes disrespect and abuse helps to determine behaviour within local norms. 21 For this reason, we consulted with a local group of health professionals for the analysis. However, reflecting on behaviour based on local consensus risks ignoring that disrespectful acts can be invisible manifestations of inequality engrained in the fabric of society. 12 It is therefore possible that we interpreted situations as disrespectful or abusive, while these were not experienced as such, not intended as such and not considered as such by local standards.

Conclusion
In this study, all women experienced disrespect and abuse starting from their first obligatory and expected visit to the health facility for ANC and during birth. From the perspective of structural violence, non-supportive care is symptomatic of systemic gender inequality in society, which is manifested in health providers' interactions with women. Disrespect and abuse in health facilities has been normalised and legitimised as a consequence of women's lives not being valued. Health providers, however, are also victims of structural violence, even though at the same time they can be perpetrators of abuse. To achieve respectful maternity care for all, interventions to prevent disrespect and abuse cannot be implemented without recognition of structural inequalities that foster the conditions that make mistreatment of women possible.