A qualitative analysis of obstetric violence in rural Madagascar

Abstract

In Madagascar, a country where maternal mortality remains high, the quality of obstetric care as perceived by users has been little explored. In this paper, we examine the perception of the quality of care in rural areas, by identifying women's experiences and expectations for basic and emergency obstetric care and how providers are meeting them. Data were collected in 2020, in three rural regions (Fenerive-Est, Manakara and Miandrivazo). 58 semi-structured interviews were conducted with women who had given birth in basic health centers or at home, and with other key informants including caregivers, birth attendants (known as matrones), grandmothers and community agents. 6 focus groups took place with mothers who had given birth at home and at a basic health centers and 6 observations took place during prenatal consultations. This article highlights the major dysfunctions perceived in the services offered and their influence on healthcare use. The women highlighted a lack of consideration of their expectations in obstetric care, with a defective caregiver/patient relationship, unforeseen costs and inadequate infrastructures incapable of guaranteeing intimacy. The women also complained of a lack of consideration of their fady (cultural prohibitions that can lead to misfortune) that surround pregnancy. These local practices conflict with the medical requirements of priority interventions in maternal care, and the respect of these practices by the women leads to reprimands and humiliation from caregivers. This obstetric violence, which emanates from the structure of society, gender relations and the biomedical practices governing pregnancy and childbirth in health facilities in Madagascar, constitutes an obstacle to the use of obstetric services. We hope that this description of the various dimensions of obstetric violence in Madagascar will make it possible to identify the structural obstacles limiting the capacity to provide quality care and to engender positive improvements in obstetric care in Madagascar.

1. Introduction

1.1. Research rationale: perceived quality of care

Over the last two decades, Madagascar has made significant progress in improving healthcare for vulnerable groups, including mothers and their children. However, maternal and neonatal mortality rates remain high, with 408 maternal deaths per 100,000 live births [1]. In some rural areas of Madagascar, up to 61% of women give birth outside of healthcare facilities, often without the assistance of health care workers [1].

Many researchers believe that the perceptions of the users of health services must be taken into account for true progress to occur [2,3]. In anthropological publications, the “perceived quality” of care is defined as a subjective and cognitive evaluation of the various dimensions that come into play when care is provided [4]. Certain characteristics have been shown to contribute to improvements in the perceived quality of care: the skill of the nursing staff [[5], [6], [7], [8], [9], [10]], positive interactions with staff [[11], [12], [13], [14]], the availability of material (15), infrastructure [16], drugs [12,17,18], and pain treatment and management [17].

Conversely, other studies on perceived quality of care have revealed various forms of obstetric violence (OV). For example, an analysis of relational aspects has highlighted violent interactions between caregivers and patients [12,14,15,18,19] and analyses of organizational aspects have revealed failings in the care offered, such as an absence of pain treatment [17,20,21], interventions without consent [[20], [21], [22]], a lack of intimacy [12,15,17], a lack of communication [11], and long waiting times for consultations [18]. A mismatch between local registers of disease interpretation and the biomedical or cultural resources on offer [17] has also been noted and defined as a form of violence. Finally, this body of research demonstrates that perceived quality contributes to the decisions of pregnant women as to whether or not make use of health facilities [11,12,15,[17], [18], [19]].

1.2. Definition of obstetric violence

In African countries, various terms have been used for practices deemed inappropriate and disrespectful, such as “mistreatment” or “abuse” [23] in contrast to Latino American countries who have mostly used the term obstetric violence [21,24]. This term was legally defined for the first time in 2007, in Venezuela, as: “the appropriation of women's bodies and reproductive processes by health personnel, expressed as dehumanizing treatment, abusive medicalization, and the pathologization of natural processes, resulting in a loss of autonomy and the ability of women to take their own decisions freely about their body and sexuality and having a negative impact on their quality of life” [25, p 201].

Here, we propose the use of the concept of “obstetric violence” as a conceptual framework in which to analyze the data collected on women's perception of the quality of obstetric care. This analytical framework makes it possible to highlight the negative experiences of women, which, according to the WHO, “not only violates the rights of women to respectful care, but can also threaten their rights to life, health, bodily integrity, and freedom from discrimination” [26]. In this sense, it is an essential theme to address in Madagascar given the near invisibility of the gender issue [27]. As many authors have pointed out, to discuss obstetric violence, it is necessary to acknowledge the unequivocal existence of violence against women which is part of the wider structure of gender inequity that produces, and normalizes, all types of gender violence [24,[28], [29], [30]]. By depriving women of their autonomy, states could be found accountable for violating numerous human rights, that is why it is important to understanding the human rights dimension of mistreatment of women during obstetric care [31].

1.3. Definition of structural violence

OV is also form of violence that intersects with other axes of structural inequalities which creates violent experiences of pregnancy and childbirth. Therefore, we think it may be also a valuable tool for exploring the origin of structural violence in maternal care. OV is better conceived from the standpoint of structural violence, which is attributed to the specific organizations of society that injure or harm individuals or masses of individuals, particularly in countries with stratified health care systems, as these are more likely to favor its incidence [32 p 1309]. Analyzing OV at the macrostructural level will allow us to recognize that it is a “systemic problem and therefore calls for systemic solutions” [33 p 62]. Typifying or categorizing disrespect and abuse during labor from an individual perspective is not enough [23,33,34]. Thus, structural violence may be linked to a lack of material and human resources within a healthcare system [23], in which case it also affects caregivers. This angle allows us to highlight the major structural dysfunctions in obstetric care delivery in Madagascar, potentially limiting the ability of caregivers to provide high-quality care and at the root of the structural violence observed.

1.4. Absence of OV concept in Madagascar

While those violent interactions have been widely explored in West Africa and in Anglophone countries with similar characteristics (economic, political and social context), it remains unexplored in Madagascar, where there is currently an absence of terms used universally, like obstetric violence, in an unequivocal manner.

Based on qualitative research conducted in rural Madagascar that aimed to document the different dimensions of obstetrical “quality of care” provided in basic health centers, this article analysis the perceptions of users of obstetric services, through the prism of gender, a concept that is too little addressed in Madagascar. Our contribution lies in underpinning the mutual frustrations that exists between patients and caregivers which are the result of structural dysfunctions analysis will help to. We will then formulate proposals for strategic solutions for improvements in the obstetric care offered to women in Madagascar.

The results are presented in two sections: mutual frustrations concerning the care framework and mismatch between local practices and registers concerning pregnancy and childbirth — defined here as the set of practices usually implemented at the local level concerning pregnancy and childbirth — and biomedical requirements.

2. Methods

This article is based on the results of an anthropological study carried out in three districts in Madagascar: Fénérive-Est on the east coast of the island, Miandrivazo in the western part of the island, and Manakara on the south-east coast. The study, entitled Preg-Eval (Pregnancy Evaluation), was conducted in 2020 by the Pasteur Institute of Madagascar in collaboration with the Ministry of Health (Family Health Department), with USAID funding, as part of the RISE (Research, Innovation, Monitoring and Evaluation) program. The objective of this study was to investigate the perception of the quality of basic (prenatal consultation, childbirth, postnatal care) and emergency (postpartum hemorrhage) obstetric care in women using these services. The districts were selected with the Family Health Department of the Ministry of Health based on their prevalence values for postpartum hemorrhage, to maximize our chances of finding women who have experienced postpartum hemorrhage, which were the highest nationwide.

The fokontany is the smallest administrative subdivision in Madagascar; each town or village includes several fokontany, and each fokontany has a “president”. This qualitative study was performed in three fokontany, satisfying the following criteria: at least 500 inhabitants, to ensure the availability of the various categories of people targeted by the project, and less than 5 km from a basic health center (BHC), such that geographic inaccessibility could be ruled out as the reason for the non-use of obstetric care. We interviewed four categories of people: mothers (with at least two children to collect several childbirth stories) who gave birth at home, mother (with at least two children) who gave birth at the BHC, mothers who survived a postpartum hemorrhage (PPH), and other key informants, including caregivers from the three regions, birth attendants (known as matrones), grandmothers and community agents. In our definition of these categories, we aimed to collect the widest possible range of diverse narratives concerning obstetric care.

An interview grid [35], was prepared for each group of people interviewed, essentially aiming to understand accounts of pregnancy and childbirth, the care sought or received during these periods, and postnatal care, the definitions of easy and difficult childbirth, knowledge about complications, including PPH, and expectations in terms of obstetric care. The interviewees were selected on the advice of the local administrative authorities, with the help of two community workers. In total, 58 semi-structured interviews were carried out. The interviews lasted a maximum of 1 h and generally took place in the respondents' homes (except for caregivers, for whom the interviews took place at the BHC).

In each on the three regions, we conducted a focus group with women who gave birth at BHCs and a focus group with women who gave birth at home. The six focus groups in total helped us to examine the consistency of informants' responses, focusing on the themes that emerged in the semi-structured interviews. The women interviewed in the focus groups were not the same as those who participated in the semi-structured interviews. As for in-depth interviews, semi-structured questions were used to allow the facilitator to direct and guide the discussions.

The interviews and focus groups were conducted in pairs. The researcher, working as a moderator, addressed questions, conducted the whole group discussion and an observer took notes and made sure that all the questions in the framework were asked. The conversations, dialogues, responses, and discussions were recorded, the interviews were transcribed before being analyzed.

In addition, six direct observations [36] structured by an observation grid were carried out in the BHC. More specifically, in each zone, two observation sessions of at least 4 h were held during the days allocated to prenatal consultations. The objective was to validate the data deriving from the in-depth interviews and the focus groups. Therefore, observations focused on the infrastructure and equipment of the BHCs, on payment and on the health care provider relationship.

It should be noted that our research had some limitations, we encountered biases related to the category of interviewees: difficulties in recruiting PPH mothers and mistrust of caregivers towards our research. Other limitations were more related to the context in which the survey was conducted: insecurity of the research areas and the covid pandemic which delayed data collection.

2.1. Analysis

All interviews were recorded, transcribed, and translated into French. They were then subjected to a thematic analysis with analysis grids prepared for each group targeted. The thematic analysis method involves codifying responses and then grouping them together into homogeneous categories [37]. A cross-sectional thematic analysis by filling in an analysis grid was carried out using the transcribed and translated data. Three analysis grids were designed for this study, one to process the interviews with the women, one for the interviews with the caregivers and another one for the remaining key informants. This method made it possible to highlight, for each theme, the recurrences and divergences between the interviews, which made it possible to highlight the main trends, while considering the particularities of the life courses or representations. This article focuses more specifically on the analyses of the following themes: the quality of care perceived and the expectations of women during pregnancy and childbirth, and during postnatal care. For the results, we selected several representative cases of the experiences and discourses collected during the interviews, which we highlight here. These analyses shed light on a certain number of sociocultural factors determining the perception of obstetric care.

2.2. Ethical considerations

The national ethics committee of the Ministry of Public Health approved this study. All interviews were conducted with the written consent of the interviewees. An information note was read to the interviewees which and explained to them in detail the importance of the study and the confidentiality of the interview. In accordance with the Helsinki declaration, interviewees were explained that they had the right to withdraw at any moment without any obligation to explain why. After all the necessary explanations concerning the study had been provided, the participants were asked to sign a consent form indicating their agreement to be interviewed.

2.3. Knowledge transfer

Knowledge transfer is a set of processes aiming to promote the use of knowledge from research to guide actions and political decisions. This process was carried out during and after the data collection. Our results were disseminated to the project partners and to all potential users (different departments of the Ministry of Health, international organizations, health professionals). Recommendations based on these results were developed in a participatory approach. This approach ensured that the recommendations were both feasible (considering resources, context, etc.) and relevant. To do this, the research team was accompanied by the Renard team from the University of Montreal. These recommendations are listed at the end of this article.

3. Results

The results are presented in two sections: OV that emanates from mutual frustrations concerning the care framework and mismatch between local practices and registers concerning pregnancy and childbirth and biomedical requirements.

We first describe the organizational aspects revealing the limited care framework reported by the mothers, their families and the caregivers. These include failings in the reception of the patient, a lack of communication, or of respect for patient intimacy during care, and the obligation of the patient and her family to clean the delivery room. We then describe the violence resulting from the mismatch between local practices and local registers concerning the management of pregnancy and childbirth and biomedical requirements, which leads to mutual frustration between caregivers and patients. The routine practices of the caregivers are carried out in a context of limited material resources (insufficient technical facilities), human resources (there may be one caregiver, working alone, for thousands of inhabitants), and isolation (caregivers are often sent to “the back of beyond”, far from their families and in remote areas). The various forms of OV presented in this research, that arises from structural dysfunctions, negatively influences the perception of service users, and their choice as to whether or not use biomedical care.

3.1. Mutual frustration concerning the limited care framework

3.1.1. Long waits, favoritism, absences, and overcharging

Although considered banal by healthcare personnel, certain attitudes considered unprofessional by the women have been determinant for their behavior. For example, waiting times for prenatal consultations were frequently mentioned, with patients often considering these waiting times to be too long. During our observations, PNCs (Prenatal Care) that were supposed to start at 8 a.m. started at least 2 h late. A second attitude considered unprofessional by the women was a system of “favoritism”, from which certain patients benefited, with certain women given priority for no apparent reason. In response to these observations, the caregivers interviewed said that they were obliged to see the most urgent cases, in addition to all-comers, and they also gave priority to their own relatives and acquaintances, according to a gift and counter-gift indebtedness system widely practiced in Madagascar (e.g., donation of eggs, chicken etc.). Another complaint of the mothers was that caregivers were too frequently absent. The reasons given by caregivers for such absences related to visits to their families and ongoing training provided several times per year in the national or district capital. Caregivers considered both these types of absence to be necessary, to improve their medical skills et alleviate their isolation. As BHCs are, at best, staffed by a paramedic and a doctor, but mostly by a paramedic alone, women in emergency situations were often faced with an empty BHC. This situation was reported by a woman who had PPH during a home birth and was unable to receive care because the BHC was empty. Finally, women often described a difference between the messages they had received indicating that obstetric care was free and the costs they actually paid to receive such care. The prices indicated by the women varied considerably for the same medicine or service. The women interviewed attributed such variations of price to caregivers deciding prices based on the “patient's appearance” mainly the way of dressing, which may reflect the economic means of the patient. They described practices of overcharging and “petty corruption”, jeopardizing the supposedly free services and the confidence of women in the services offered. In response, the caregivers said that they were regularly faced with shortages, and that this led them to having to charge for medicines that were supposed to be free. They then complained of the lack of materials for explaining in detail the care and medicines provided free of charge for pregnancy, childbirth, and the postpartum period. Finally, the caregivers complained of a lack of supplies for dealing with complicated deliveries, which, they claimed, often occurred due to delays in the referral of expectant mothers, and which were considered by the mothers to be included in “care related to childbirth”. The women made no distinction between the management of simple and complicated deliveries.

3.1.2. Lack of communication and intimacy

According to our observations and the statements of caregivers and mothers, PNCs lasted on average 6 min. The time range was between 3 and 10 min. During the PNCs, caregivers did not seek to understand the problems or needs of the expectant mothers and the dialog was extremely limited. The caregivers complain about this situation as much as the patients. The former indicate that this time is imposed on them by the number of consultations to be carried out in a given time frame and that there is a lack of understanding by the mothers of the information transmitted. The women said that they did not have time to ask questions during these consultations: a few brief questions were generally followed by equally brief answers in one direction: from the healthcare worker to the patient. As a result, women received little information about the usefulness of the drugs received and most could not name the drug they had been prescribed or to explain what it was for. They were aware only of the dose. For their part, caregivers confirmed that too little time was allocated to PNCs, mainly due to emergencies and the very small number of staff relative to the number of patients which affected negatively the quality of the service: A midwife (42 years of age) noted: “We don't have enough time. Sometimes there are 40, or 45 PNCs per day. We start at eight o'clock and expect to have 15 min per person. During a PNC, there is sometimes an emergency, such as childbirth. So, you have to take some time away from consultations. When you come back, there is too little time, especially when you bear in mind that they live far away and still have to go home. This is why a quarter of an hour per PNC is no longer respected”.

They also described the low level of education of the mothers, which is confirmed by the characteristics of the mothers interviewed. They felt it affected their understanding and management of their health. The caregivers deplore the absence of questions asked by the mothers, whether about their state of health, that of their unborn baby, or about the medication they received. As a result, the caregivers felt compelled to provide succinct and abbreviated information about prescribed medication or the state of health of the patient and rarely take the patient's opinion into account.

A second element linked to the structural conditions of the consultations was the lack of intimacy described by the mothers. Indeed, the preservation of privacy and confidentiality during PNCs was considered insufficient. An obvious example was provided in one of the regions, in which PNCs took place at the same time as other health examinations (for example, consultations for malaria), with many patients coming and going, including men. This implies that women must undress and undergo gynecological examinations and follow-up of their pregnancy publicly, without being able to ask intimate questions that concern them. According to caregivers, this lack of privacy is the result of a lack of space and an inadequate infrastructure. In this context, women complained of the use of BHC spaces by caregivers for personal purposes (e.g., housing family members, especially when the caregiver comes from another area than the one where he/she was assigned). Only one BHC ensured the confidentiality for PNCs, which took place in a closed room with closed doors.

“Either way, it's really annoying. Because there is a lot of noise, or if we want to talk or ask a question out loud, we can't. Instead, we whisper what we want to ask because there are people nearby. And for your secrets, you whisper them. We can't talk as we're doing now. Otherwise, the person next to you will hear it” - 24 years old, mother of two children born at home.

The preservation of privacy was also an issue for postnatal care, in which hospitalized women complained of a lack of space and having to share spaces with ill patients. All these factors led to a shortening of the period of postnatal care, as women preferred to return home quickly in view of the conditions of their stay in the BHCs (e.g., water dripping from the roof, no electricity, no beds, toilets out of order). Given the lack of space, the caregivers themselves complained that they sometimes had to ask women who had given birth and were “starting to feel better” to leave early so that they could take care of new occupants. They suggested that there should be more rooms in the BHCs to guarantee one room per department (e.g., one room for vaccination, one for family planning, one for outpatients, etc.). These various elements lead to shared frustration on the part of both the caregivers and the users, concerning the limited time for consultation and lack of privacy.

3.1.3. Obligation to clean the delivery room

Another aspect raised by the women interviewed concerned the cleaning of the room after childbirth, a task that falls on the women and their families. According to the interviews with the mothers, cleaning the hospital room is a condition for being able to leave the health facility. This implies that a woman must necessarily be accompanied during childbirth, which conflicts with the family sometimes being “under the way” in the eyes of caregivers. The women also complained that they lacked cleaning products to perform this task; the presence of one or more family members, in addition to the purchase of cleaning materials, therefore represents an indirect cost of delivery at the BHC.

3.2. Mismatch between local practices and biomedical recommendations for pregnancy management

The mothers reported various forms of discrimination, threats, and judgments about them from caregivers during routine consultations. Most related to physical characteristics: hygiene and appearance, which was often considered “negligent” by caregivers and was visible, according to caregivers, from the state of the health record booklet, the clothing of the newborn and the mother, and the materials used by the mothers at the time of birth. Socioeconomic characteristics (economic status, level of education, etc.) were also regularly the subject of comments or the cause of a lack of consideration. The feeling of shame felt by women with dilapidated material (for example, coming to give birth without a blanket or with a dirty blanket) garnered criticism of the entourage from the community, in addition to disapproving looks and comments from the caregivers. This is therefore a conflict between local health management practices and medical recommendations, as well as discrimination based on socio-economic status.

“With us, the doctor discriminates. If it's someone with clean clothes, he gets [medicine], but if not, he doesn't” and “Let's say, for example, that if you put on dirty clothes, that makes him angry too” - Woman, 32 years of age with two children born at home. In addition, women live under the threat of being held at the BHC if payments are not honored and can only, in rare cases, benefit from advances from nursing staff. The provision of advances is a frequent practice, but depends entirely on the “financial assets” of the caregivers. The monthly salary of a doctor or paramedic ranges between 600,000 and 1,000,000 million ariary (between 140 and 230 Euros). The timetable for repayments is discussed with the mother when this “credit” is given. This is a feature highly appreciated by users, as there are few caregivers who advance the costs of the care provided from their own pockets.

3.2.1. Differences between the need for discretion and the recommendations for prenatal consultations

In all the zones surveyed, and more widely in Madagascar, discretion about pregnancy is generally practiced. During the first few months, women do not want their pregnancy to be commented on or seen, to avoid causing jealousy and to avoid bringing misfortune upon oneself. Pregnancy is also seen as a fragile and non-immutable state during the early months, so women avoid talking about it. Only close family and friends are told initially, and a woman's pregnancy is discovered as her belly grows. However, this need for discretion surrounding the early stages of pregnancy is difficult to reconcile with the scheduling of the first prenatal consultation during the first trimester of pregnancy. Most PNCs are, therefore, performed when the woman's belly “has developed” and when the pregnancy can no longer be “hidden”, at around the fifth or sixth month of pregnancy.

However, many caregivers stress the importance of the first PNC occurring during the fourth month of pregnancy. In this context, the women described two types of caregivers, one type being considered “severe”, with a fear of reprimands mentioned on many occasions. The women described behavior marked by verbal violence on the part of caregivers. They said they were the victims of reprimands perceived as being humiliating concerning the “negligence” they showed in the management of their pregnancy and the “lack of regard” for the health of the baby that this reflected. These criticisms were addressed particularly to women who arrived late for their first PNC or without their health record booklet.

The violence described was also based on the threat of referral to the nearest hospital in the event of complications during childbirth. According to the mothers, these threats were made to ensure compliance with prenatal care. For example, one woman told us:

“She scares us. Because, as I had the operation the first time [caesarean], she said: “If you don't do your PNC, I won't help with your delivery. I'll send you straight away to Fenoarivo [Hospital in the town of Fénérive-Est] » - 35 years old women with two children born at home.

This statement, which might seem may seem anodyne, represented a threat for this mother, because being referred to a hospital entails a significant cost. According to the mothers, the arguments used, jeopardizing their economic security, are a form of intimidation and do not reflect the kind communication they expect. Some women attended PNCs with fear of reprimands but did so because attending PNCs provided them with the right to give birth at the BHC. Indeed, women mentioned their fear of not being accepted for delivery if complications arose at home.

As a means of respecting the need for discretion surrounding the first few months of pregnancy, some caregivers working in one of the study areas performed PNCs at times other than the days habitually scheduled for PNCs, which were known to all. Women appreciated being able to reconcile the constraints associated with the recommended early PNC with their need for discretion. This flexibility, highly appreciated by the interviewees, was mentioned several times as an understanding of local customs. Caregivers indicated that such flexibility is not always possible when the caregiver works alone, which is too often the case.

3.2.2. “Fady” opposing biomedical and traditional management of pregnancy and childbirth

According to the mothers, the healthcare staff disapproved of certain fady surrounding pregnancy. These fady can be defined as unfortunate types of behavior that should be avoided, or taboos in the shared knowledge of Malagasy society. One such fady concerns the prohibition of preparations for the birth, with women not allowed to provide themselves with baby wipes, diapers, and clothes in anticipation of the birth. According to local beliefs, the unborn child is considered to be a zavatra, literally a “thing”. The child is considered to be a “human being” only when it is visible on the day of its birth. Failure to respect this prohibition, in force in the three districts investigated, can lead to a series of misfortunes for newborns and their families, and even to the death of the newborn, according to the women interviewed. One of them explained:

“It is our ancestors who created this taboo, it comes from them. If you give birth and the baby does not survive then they (the baby's belongings) will not be used. That's why the parents refused it. Do not prepare the baby things, prepare the money to pay for them. When the baby is born and healthy in your arms, that's when you can buy the stuff” - 24 years old women with two children born at home.

Despite this fady, no measures are taken in national mother-child strategies to provide basic materials for the care of the newborn. The women therefore wrap the infants in their own blankets or leave with their naked newborns, sometimes exposing them to the cold and rain. The interviewees said that, from time to time, the BHC provides “delivery kits”, including new clothes for newborns and sanitary napkins for women, useful elements to make up for the shortcomings caused by this tradition. However, these kits are provided as part of initiatives supported by certain technical and financial partners of the Ministry of Public Health. Their content is not standardized, and they are not always available.

Women admitted to regularly receiving verbal criticism about their lack of preparation or of being judged. In this respect, one of the caregivers justified their approach as follows:

“It's really problematic because the baby is born, and we should keep it warm. So, there must already be a bonnet to cover its head. You see the baby comes from a warm place and arrives in a cold place and has nothing to welcome it. So that's problematic. And we find that very harsh. And this is where we get a little harsh on tradition. Because sometimes babies need resuscitation here when they have nothing to wear” - Head of a BHC, 38 years old nurse with nine years of experience.

By contrast to these “severe” caregivers, the mothers describe empathetic caregivers who try to reconcile biomedical recommendations concerning pregnancy and traditional practices, such as massages from the reninjaza (matrones) or being accompanied by family. This study revealed the existence of collaborations between medical staff and matrones (reninjaza). Caregivers at a BHC, for example, refer women to reninjaza for massages, for which only they have the know-how, especially when the child is not “in the right position”. Cooperation was also described in the opposite direction, with the reninjazas accompanying women, who seem to be in difficulty during childbirth, to the BHC.

Nevertheless, for caregivers, it appeared to be impossible to reconcile this taboo with newborns. They justify such reprimands concerning the cleanliness of the newborns and the need to keep them warm, in connection with the practice of fady, as a means of reducing the perceived risks of infant mortality. This unsuitable approach appears to widen the gap between caregivers and patients, leading both to the frustration of women at their cultural dimension not being recognized [and therefore a lack of confidence that can cause a breach in the care pathway] and a lack of understanding on the part of caregivers that can lead to negligent care.

In all three areas investigated, the individual characteristics of caregivers appreciated by service users — empathy, listening, and benevolence — therefore influenced the choice as to whether or not to seek care. The positive attributes of caregivers were seen as motivations to come to the BHC [for prenatal consultations or to give birth], whereas negative characteristics were seen as a reason to stay away. However, according to the interviews conducted with caregivers, the exercise of their profession in such a precarious context often leads them to develop feelings of frustration and powerlessness.

4. Discussion

Dysfunctional relationships between caregivers and those for whom they care during prenatal consultations and childbirth have been the subject of much anthropological research, highlighting mistreatment, a lack of empathy, and poor communication. The reception of patients, considered to be defective due to the lateness of caregivers, or inappropriate behavior, such as non-compliance with the working hours of the service, result in the users having to wait for care [18]. The quality of patient reception is also negatively affected by non-respect of the order of arrival. Indeed, these many authors have described such systems of “favoritism”, in which the “protected” individuals take advantage of privileged social relationships with caregivers [38. p 70, 39, 19]. All these elements reinforce the idea that the behavior of health professionals is a central element in patient assessments of health services [40,41].

On the other side, caregivers described themselves as feeling helpless in the face of extremely precarious situations that they cannot always deal with and over which they have no control. These include shortages of supplies, which increase the costs of care for mothers who have to obtain them elsewhere, or faulty or even absent equipment, considered to be an obstacle to the smooth running of the BHC (i.e. lack of a blood pressure monitor or scales) or childbirth (such as a vacuum extractor, as mentioned by caregivers). Those perceptions by healthcare workers that the health system is weak and under-supported has been documented by Oluoch-Aridi et al. in Kenya [39]. In addition, the lack of staff at the healthcare centers has also been investigated. Caregivers whose families cannot come to live in the area must reconcile their family lives with their professional responsibilities. As highlighted by Jaffré, in Benin and Burkina Faso, being assigned “to the bush” means separation from family, with consequences affecting the social identity of the caregiver. Thus, assignment to such posts has its share of difficulties due to the caregivers being foreign to the region and far from their families [42].

The obstetric violence described above, which can be attributed to the organization (lack of staff, lack of equipment, shortage of drugs, etc.) of the healthcare system, may occur without intent to harm [43 p 1]. This makes it possible to avoid pointing the finger directly at caregivers as the individual perpetrators of such violence, but to highlight their realities, as they themselves experience the deleterious effects of structural violence within the maternal care system. As Oluoch-Aridi et al. pointed out in their research in Kenya [44], the healthcare workers described a dysfunctional health system that created a context where the mistreatment of women seemed to be inevitable. Until today, the concepts of obstetric and structural violence are not used at all to define these experiences in Madagascar. Yet the use of the terminology “obstetric violence” is essential, not only to do justice to the users of the services but also to understand what, in structures and contexts, contributes to this form of violence. Oluoch et al. point out that OV is found worldwide and suggest there is a normalization of violence within the medical space that needs to be addressed [39]. Their research in Kenya and South Africa shows similar structural inequalities which are primarily linked to gender and lead to similar patterns of obstetric violence. According to Smith-Oka et al. [45], local infrastructures, policies and histories allow this violence to continue, especially in countries where violence against women is prevalent and this is unfortunately the case in Madagascar.

Our article sheds light on reprimands and threats directly linked to the individual judgments of caregivers concerning the practices and registers of local women. However, this individual factor does not mean that it isn't part of the broader spectrum of structural violence, here, health workers can be seen as agents of intertwined hierarchical, organizational and social violence. Such violence, in the dispensation of healthcare, could be better understood and dealt with, particularly if linked to the condemnation of the traditional approaches favored by women using maternal services. These local practices and registers for the management of pregnancy and childbirth, with which caregivers are not always familiar, are rarely in line with medical requirements concerning obstetric care. How, as anthropologists, do we reconcile the two realities? The women, who must respect the fady, wrap the infants in their own blankets if materials are not made available at the BHCs. Similarly, the requirement for prenatal consultation to begin in the first trimester of gestation is considered to be incompatible with discretion around the early stages of pregnancy. This requirement for discretion is not specific to Madagascar, as shown by the work of Beninguisse et al. [17] in Cameroon and Burkina Faso. The caregivers justify such social and professional verbal violence against women in terms of the perceived risk of the women's practices [42]. Thus, the threats used by caregivers (in particular, that of referring the patient to hospital if she does not attend her PNC) can effect changes in the behavior in the mothers, but these threats do not help women to understand fully the true reasons for which PNCs are important, and therefore do not encourage the long-term compliance of mothers with such consultations.

The lack of respect of women for local practices in pregnancy and childbirth management leads to regular sermons, judgments, and threats, on top of those relating to cleanliness and forgetting the BHC healthcare record booklet. Even if these situations depend largely on the characteristics of caregivers and their personal actions, they remain part of the broader spectrum of structural violence and gender violence that is rusting out the health care system.

Thus, women say they prefer to give birth at home rather than having to experience what they call a feeling of shame. This notion of shame was raised in Madagascar by Elise Huysmans [46], who explained that the social bond between caregivers and patients in Madagascar is characterized by a permanent stand-off, in which the question of materials and money is central, and in which the shame falls on those who do not have the richest food and the newest or cleanest clothes. Thus, shame, shyness, and humiliation, manifesting at the individual level, are the foundations of structural violence. The damage created by obstetric violence is difficult to measure since it relates to the fear, powerlessness, feelings of insecurity [32]. Giving birth at home is seen, in this context, as a way of protecting oneself from the judgment of others. Therefore, these damages directly influence women's individual decisions or future decisions. Disrespect and abuse are considered to be one of the reasons for the underuse of health facilities for childbirth, despite the provision of free services [47]. We argue that this could also be the case in Madagascar, where women make different choices, such as delivering at home instead of in the BHC, putting themselves at risk.

Nevertheless, questions remain concerning the discrepancies between what we call obstetric violence and what women say about it. Why is such violence tolerated? Although many women acknowledge having been abused, they are assured that childbirth at a BHC is preferable because it is “safer” biomedically. Several authors have investigated why abuse is accepted by the users of maternal services [31,48]. Those who use such services typically have expectations that boil down to what they already know, suggesting that women consider the abuse described above to part of the process of the care pathway. As Smith Oka et al. stated “it is important to keep in mind that the women experiencing this violence may not always be the ones at the front lines of activism; despite recognizing that what happened to them was not right, many seem resigned to the idea that things will remain as they currently are" [45 p 2716]. They may not always see OV as something that could be regulated, legislated and eradicated [45].

Concerning the healthcare system in Madagascar, Christine Bellas Cabane said: “collective indifference has made it possible, in certain situations, for the unbearable to become bearable and the pathological to become the norm” [49 p 20]. Freedman and Kruk [47] and Sadler et al. [24] argue that the prevalence of disrespect and abuse during childbirth are symptomatic of health institutions that devalue women and indicative of a crisis in institutional accountability. Thus, to break this circle and prevent interpersonal violence in the caregiver-patient relationship in Madagascar, attention should be paid to the inequalities and structural limits favoring the mistreatment of those who use these services.

5. Conclusions and recommendations

Based on the study that aimed to investigate the perception of the quality of basic and emergency obstetric care among women using these services, this article provides an unprecedented analysis of structural violence in the context of obstetrical care in Madagascar. Violent practices had previously been described without analysis under the prisms of obstetrical and structural violence. Indeed, Claire Mestre documented evidence of poor-quality care in Madagascar, including mistreatment, negligence and contempt, constituting what she referred to as a form of symbolic violence [50 p 68], a term coined by Pierre Bourdieu, which describes a type of non-physical violence manifested in the power differential between social groups. In her writings, Dolores Pourette describes the lack of communication between caregivers and cared-for women in the context of pregnancy and childbirth [27]. She also demonstrates that the care provided by matrons may be preferred by mothers because of their social proximity [27]. While some of the situations described in this article have been published before, the prism of obstetric violence has never been used in the Malagasy context. We hope that more research will highlight the voices of service users and caregivers who have little or no space to be heard.

This study sheds light on obstetrical violence affecting both patients and caregivers, which results from systemic dysfunctions, such as the impossibility of guaranteeing intimacy or ensuring optimal communication. In this context, we suggest encouraging the creation of adequate infrastructures to prevent overcrowding and the early departure of women who have given birth; the establishment of an accountability system to provide a voice for the providers (e.g., with the municipality or the reference hospital); the involvement of the accompanying individuals by giving them a badge formalizing their role; the management of activities, such as the supply of water and food and the cleaning of the delivery room during and after childbirth (a project already implemented in the region of Manakara by ACCESS), which would facilitate the work of the caregivers, whose performance of their duties is hampered by the lack of resources and available personnel.

Our study nevertheless reveals that some violence depends strongly on relational dimensions (quality of reception, empathy, etc.). Women rated the quality of the caregiver/patient relationship according to their own standards, formulated during individual and group interviews. From their point of view, improvements in relational aspects would involve the systematic presence of care staff at the BHCs; acceptance for all births (expectation formulated in view of the fear of being refused access if no PNCs occur beforehand); being greeted by polite, attentive, and caring healthcare staff who respect local pregnancy and childbirth management practices and registers and who do not discriminate on the basis of appearance; and being received according to an organizational system in which the order of arrival is respected, with the first person arriving being the first seen (with the exception of emergencies). In this context, we suggest setting up training courses for healthcare professionals on the humanization of care, the content of which should be standardized. This type of training would benefit from scaling up to national level and teaching in medical and nursing schools, with systematic monitoring and evaluation, making it possible to measure the impact of training and to identify difficulties in its daily application. But it is important to note that if individual OV is deeply embedded in structural forces, then trainings will not solve the core problem.

Regarding the discrepancies between local pregnancy and childbirth practices and biomedical requirements, we found that caregivers considered birth kits to be essential for the prevention of infections and that such kits were a motivation for women to come and give birth at BHCs, but that such kits were, unfortunately, not always available. We suggest that such kits, including baby clothes and hygiene items, should be made available, with a standardized content, at all BHCs throughout the country. Such kits would lessen the frustrations and problems arising from the cultural requirement not to prepare for the birth. Concerning the need for discretion in early pregnancy, we recommend more flexibility in the performance of PNCs, and we strongly recommend holding such consultations at any time of the week, rather than on specific days.

Our results also demonstrate a lack of transparency concerning the costs of services at the BHCs. We, thus, suggest displaying service costs at the healthcare facility, with the price of each drug displayed at the BHC pharmacy, for example, together with a list of free obstetric care services, or the price of a birth kit.

In conclusion, the obstetric violence experienced by women influences their individual decisions as to whether to use the healthcare system. The term OV provides a helpful tool in combatting some of the structural inequities that contribute to the incidence of abuse surrounding the period of pregnancy and birth and includes a denunciatory dimension. This dimension is necessary to raise the required collective awareness to urgently drive change and rethink obstetrical care in Madagascar. It does more justice to the experience of women, although it is not always referred to as such in published studies. Despite differences in the terms used, our results are not unique to the areas investigated in this study and relate more broadly to practices and the problems of seeking care for women throughout Madagascar.

Author contribution statement

Emilia Brazy-Nancy, M.D: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Chiarella Mattern, PhD: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Brigitte Irene Rakotonandrasana; Voahirana Ravololomihanta; Patricia Norolalao; Laurent Kapesa: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data.

Funding statement

Emilia Brazy-Nancy was supported by the USAID Research, Innovation, Surveillance and Evaluation Program (72068719CA00001).

Data availability statement

Data included in article/supp. material/referenced in article.

Declaration of interest’s statement

The authors declare no competing interests.

Appendix A. Supplementary data

The following is the Supplementary data to this article: