Background
This study examines the experience of disrespect and abuse during labor and delivery in Ethiopia through the juxtaposition of in-depth interviews with midwives, midwifery students, and women who have recently given birth. Disrespect and abuse of women during labor and delivery has become an increasingly recognized phenomenon over the past decade. Global public health norms now explicitly condemn such practices, acknowledging them as both a violation of a woman’s rights and also, instrumentally, as a deterrent to the use of life-saving facility-based labor and delivery services [
1‐
3] .
Low levels of facility-based delivery are one of the drivers of maternal deaths, and delivery with a skilled birth attendant can significantly reduce maternal mortality [
4‐
6]. While multiple factors explain low health service utilization, there is increasing recognition that many women are reluctant to use reproductive, maternal, neonatal, and child health (RMNCH) services because of poor service quality and fears of provider mistreatment. Numerous studies demonstrate that women’s perceptions of how they will be treated at health facilities can strongly influence their choice about where to deliver, and deter women from accessing services in a timely manner, or at all [
7,
8]. Unfortunately, disrespect and abuse of patients, particularly during childbirth, persists globally and is prevalent throughout East Africa [
7,
9‐
15].
Despite the Ethiopian Ministry of Health’s prioritization and vigorous support of efforts to reduce maternal and child mortality [
2], underutilization of RMNCH services remains a problem in Ethiopia and contributes to the country’s high maternal mortality rates: 420 women die for every 100,000 live births in the country and maternal deaths constitute 21% of all deaths to women ages 15–49 [
16]. Currently, only 28% of women receive skilled health care services at delivery [
17]. This low utilization in Ethiopia has been shown to be associated with women’s education levels, residence, ethnicity, parity, autonomy and household wealth, among other factors [
18‐
20]. Studies in Ethiopia also show that perceptions of poor quality of care such as lack of privacy and lack of psychosocial support, are significant factors in a woman deciding whether or not to give birth at a health facility [
21,
22]. Furthermore, recent studies reveal evidence of disrespect and abuse in Ethiopian facilities [
15,
23‐
25]. For example, findings from women and providers in health facilities in two regions found that 21% of post-partum women surveyed reported disrespect and abuse, non-consented care (17.7%), lack of privacy (15.2%), and non-confidential care; and 82% of providers cited occurrences of disrespect and abuse in their facilities [
26]. Nonetheless, Ethiopia has enshrined promotion of women’s rights and status in its constitution and subsequent national policies [
27], and has supported the core United Nations General Assembly resolutions and other international agreements that acknowledge the rights of childbearing women to respectful maternity care [
1,
26‐
33]. However, individual patients are unlikely to know about, much less use, any mechanisms to address rights violations.
In an effort to reduce maternal mortality, Ethiopia’s government has both expanded health care infrastructure and coverage and has undertaken initiatives to make care more hospitable. These include expanding numbers of midwives trained and posted in rural areas (matched with their region of origin); operationalizing a Women’s Health Development “Army” to conduct health outreach to rural women, and providing traditional foods to women who give birth in rural health centers [
34,
35]. A distinctive feature of the expansion of the midwifery profession is the growing proportion of male midwives (22%) due to new, exam-based selection criteria [
36].
1 Relatively understudied challenges of this scale-up are ensuring the quality of these services and understanding women’s readiness to use them. This study aims to help address this knowledge gap.
Theoretical framework
Mistreatment of women during labor and delivery has persisted across time and geography and has been given numerous names including “obstetric violence” and “dehumanized care”. There is growing global commitment to addressing this challenge, which has been buttressed by policy statements from the World Health Organization [
2], the Lancet [
42] and notably the White Ribbon Alliance’s 2011 facilitation of the Respectful Maternal Care Charter [
1], a global consensus statement on a positive vision for respectful maternity care with a definition of disrespect and abuse and the corresponding rights (see Table
1 below). This Charter has been translated into eight languages and shared among providers, health managers, and advocates [
3], and is anchored in United Nations and other international commitments signed by most national governments.
Table 1:
Respectful Maternity Care: Charter on the Universal Rights of Childbearing Women
Physical abuse | Freedom from harm and ill treatment |
Non-consented care | Right to information, informed consent and refusal, and respect for choices and preferences, including companionship during maternity care |
Non-confidential care | Confidentiality, privacy |
Non-dignified care (including verbal abuse) | Dignity, respect |
Discrimination based on specific attributes | Equality, freedom from discrimination, equitable care |
Abandonment or denial of care | Right to timely healthcare and to the highest attainable level of health |
Detention in facilities | Liberty, autonomy, self-determination, and freedom from coercion |
Systematic reviews of studies of disrespect and abuse during labor and delivery highlight both structural and individual drivers. They find that abuse is not limited to a few individuals or institutions, but rather is reflective both of systemic failures as well as of deeply embedded provider attitudes and beliefs [
10,
11,
14,
43‐
45]. Structural factors identified include provider shortages/heavy workloads, poor physical infrastructure, lack of supplies and equipment, and lack of supervision [
46]. Such conditions are particularly prevalent in sub-Saharan Africa, and are often associated with individual-level drivers of abuse such as provider stress, overwork, low motivation and stigmatizing attitudes [
10,
14,
47]. General lack of supportive care and poor communication between providers and patients are also often considered forms of disrespect and abuse. Systematic reviews show that patients view both intentional and unintentional mistreatment as abusive [
10].
Our study makes two contributions to this emergent literature. It is one of the few to triangulate findings and contrast provider and patient perspectives by studying patients as well as both practicing and student providers. It also offers a new examination of the relationship between disrespect and abuse, providers’ knowledge of patients’ rights, and their behavior in clinical scenarios.
Methods
The overall goal of this cross-sectional, qualitative study was to examine the nature of disrespect and abuse in midwifery care during labor and delivery in the Debre Markos area. The specific research aims were to:
1.
Examine women’s experiences of care from midwives during labor and delivery, including any disrespect or abuse;
2.
Explore midwives’ understandings of patients’ rights and patient-centered care;
3.
Describe midwives’ experiences of patient abuse and disrespect;
4.
Identify patient and midwife recommendations for strengthening the quality of labor and delivery care.
Setting
This project took place in Debre Markos, a city in Amhara region, located 5 h northwest of Ethiopia’s capital. A joint team of researchers from Debre Markos University’s (DMU) Department of Public Health, Touro University California’s Public Health program and the Bixby Center for Population, Health, and Sustainability, at the University of California, Berkeley, conducted this research. Data collection took place on the DMU campus in the School of Midwifery, at Debre Markos three main public health centers (the Hidase, Gozeman, and Debre Markos Health Centers), and at the Debre Markos Referral Hospital in February and March, 2015.
Sample
The study examined two populations: women who had recently given birth and midwifery professionals (both students who were providing care and practicing midwives). A convenience sample of 23 women over the age of 20 who had given birth attended by a midwife within the past year, was recruited from health facilities to take part in an open-ended interview. Three women who had given birth at home were recruited by Health Extension Workers and interviewed. The study also conducted in-depth interviews with fifteen randomly selected (93% response rate) third-year bachelor’s degree midwifery students from DMU and four purposively sampled practicing midwives from the study health facilities. Patients were recruited as they left the postnatal or well-baby clinics at Debre Markos health centers or the Debre Markos referral hospital on a first-come first-served basis.
Both the provider and patient samples were stratified by the gender of the midwife provider. Midwifery students were drawn from gender-separated, numbered class lists and were randomly selected using SPSS’s random number generator. Recruiters selected two males and two female practicing midwives from the study health centers and screened patients during recruitment to ensure that at least one-third of the patients had had male midwives at their last delivery.
Interview guide
Interview guides and surveys (Additional file
1) were developed in English, reviewed with colleagues in the DMU Department of Public Health, and then translated into Amharic by a professional translator of reproductive health materials. The instruments were pre-tested for length and comprehensibility with a sample of five women at the Debre Markos Health Center and five midwifery students at the DMU School of Midwifery. A project investigator debriefed interviewers and the project coordinators after pre-testing and confirmed the faithfulness of the survey and transcript translation.
All interview guides contained questions on respondent demographics and socio-economic characteristics. To address our first research aim, patients were asked about the quality and content of their care during pregnancy and labor and delivery, the quality of their interactions with healthcare providers, their satisfaction with the care they received, and about their knowledge of the quality of other women's experiences during labor and delivery. They were also asked directly about common forms of mistreatment that might not be perceived as abuse, such as the denial of food and drink during labor, refusal of accompaniment, and not being able to give birth in their desired position.
In order to gather provider perspectives on disrespect and abuse of patients (research aim two), we asked midwives and midwifery students about their experiences of provider-patient interactions, and their observation and/or awareness of patient mistreatment. Our third research aim was to examine provider’s understanding of patients’ rights in order to see whether gaps in their knowledge could contribute to patient disrespect and abuse. Midwifery students and midwives were, therefore, also asked questions on the coverage of patients’ rights in midwifery training. In addition, they also responded to a short self-administered survey on their knowledge of patients’ rights.
To enable us to contrast midwives’ understanding of patients’ rights with the degree to which they might observe these rights in practice, we questioned them about their knowledge and comfort with service provision in two clinical scenarios where official policy, correct medical practice, and respectful medical care likely conflict with prevailing cultural beliefs, leading to poor quality of care. One scenario gauged their willingness to provide contraception to an unmarried adolescent, the other, their comfort with providing abortion care services. Although the Ethiopian government has actively promoted access to contraception and has liberalized its laws on abortion [
48], Ethiopia remains culturally conservative with 67% of the population regarding abortion as “never justifiable” [
49] and premarital intercourse for women relatively rare and culturally discouraged [
50].
To address our fourth research aim, both patients and providers were asked about the reasons they thought abuse occurred and their recommendations for improving the quality of labour and delivery care.
Data collection
The study used four masters-level interviewers, recruited from the DMU Department of Public Health who had carried out survey- or interview-based data collection previously. We chose interviewers who were outside the departments that train health professionals (midwives, pharmacists, physicians) in order to reduce age and power differentials with study participants, as well the chances of interviewees knowing interviewers.
The interviewers participated in a three-day workshop covering the motivation for the study, a refresher on research ethics, project data collection and management procedures, an overview of qualitative methods, and practice using interviewing techniques.
Patient interviews took place at coffee stands near to the health facility or in private rooms in the health facility, depending on availability and the patients’s choice. For the three patients who gave birth at home, interviews were conducted in their homes. Student interviews took place in private rooms on the Debre Markos campus.
All interviews were conducted in Amharic, audio-recorded, and then simultaneously translated and transcribed into English by a single professional translator conversant with the reproductive health field. The project coordinator held weekly debriefing sessions with interviewers to discuss experiences and surprises encountered during the interviews and to refine the interview guide. In addition, project investigators reviewed interview debrief memos and interview transcripts as they were translated, and provided ongoing feedback and suggestions for making the interviews more open and consistent.
We developed a codebook (Additional file
2) with a priori codes guided by the framework of Bowser and Hill, the categories of disrespect and abuse and rights defined in the Charter of Respectful Maternity Care [
1,
14], and review of the disrespect and abuse literature. We performed deductive and inductive thematic content analysis of interview transcripts: using a priori codes for initial rounds of analysis and adding new codes to reflect themes emerging from the data.
Coding was conducted separately for the two population groups. One investigator was responsible for coding responses of midwives and students, another for patient responses. After completing coding for providers and patients, coders shared results and noted common themes and divergences both within and between the samples. This approach, keeping the samples separate, may have limited the tendency for coders to expect, and therefore find, codes in their sample based on the responses found in the other study groups. It did, however, prevent us from conducting tests of inter-rater reliability. The coding and analysis was conducted using the HyperResearch version 3.73 qualitative data analysis software.
Discussion
This study examines the experiences of disrespect and abuse in maternal care from the perspective of providers and patients. We find that mistreatment of patients during labor and delivery—particularly verbal abuse—is relatively common and that this abuse has the potential to reduce patient demand for services. Our findings are largely consistent with those from recent international studies of patient mistreatment in maternity services both in terms of the extent of abuse they describe and the triggers for abuse they identify [
43,
46]. However, unlike studies conducted in other East African countries, we find no reports of inappropriate demands for payment from midwives, or of detention for non-payment. In addition, reports of abandonment and refusal of care were relatively rare in comparison to other sub-Saharan African studies, and when such abuse was reported, it was not linked to ethnic discrimination or concerns about payment as cited elsewhere [
53]. In addition to finding no instances of discrimination based on specific attributes, or of detention in facilities, several themes emerge here that are noted in few other studies, and point to the utility of comparative research on these phenomena between Ethiopia and other contexts.
First, there is an observed discordance between patients and providers in the types of abuse most frequently mentioned. While providers consistently report witnessing physical abuse, patients gave only indirect, anecdotal reports of physical mistreatment. This discrepancy may be in part due to differences in the structure of our interview questions, as providers were asked more directly about abuse than patients. The fact that providers witness more deliveries and spend more time in facilities than patients might also explain differences in reports of physical mistreatment. Finally, although patients seemed open and vocal in their discussions of the quality of care and their experiences of verbal abuse, it could be that they were reluctant to talk about physical abuse, either for fear of retaliation at their local health facility, or because they were uncomfortable discussing traumatic or embarassing events with strangers.
Another area of discordance was the differential patient/provider reporting of abuse stemming indirectly from the poor functioning of the health system rather than the direct actions of providers. Patients frequently reported long wait times, lack of privacy on crowded wards, rushed or abrupt care, and long periods of being left alone during labor as problems that were a form of mistreatment. However, such problems were rarely mentioned by providers as examples of patient mistreatment. Nonetheless, both providers and patients identified “good care” in general as a patient right. It is also notable that both patients and providers did identify sub-standard clinical practices, such as episiotomy without anesthesia or prescription of improper medications, as a type of abuse. Poor clinical care, whether intentional or not, is not a phenomenon that fits neatly in the categories of abuse defined in the Respectful Maternity Care Charter.
Another notable finding here is the role that abuse may play in health-seeking or health-enhancing behaviors. Patients reported that verbal abuse reduced their compliance with instructions, influenced their choice of health facility, and, more importantly, was a factor in their decision-making about whether to give birth at a facility at all, as found elsewhere in Tanzania and Ethiopia [
15,
43,
53‐
57].
A third key finding concerns an area of commonality between patients and providers. Both groups expressed confusion and ambivalence about whether accompaniment and choice of birth positions constitute abuse or a violation of patients’ rights. Patients frequently reported denial of accompaniment and lack of choice in birth positioning, but few identified it as a form of abuse or mistreatment directly. This may reflect a lack of empowerment among patients. Similarly, no providers mentioned these violations of patient autonomy as a form of abuse. This points to an area that may have to be strengthened in future healthcare ethics and patients’ rights training for midwives, as disagreements over positioning can lead to other kinds of abuse.
Our study findings suggest that professional ethics training should be strengthened. We find that ethics and patients’ rights are covered unevenly in the midwifery curriculum. While subjects of privacy and confidentiality are well discussed, issues around respect, patient choice, and autonomy are less thoroughly reviewed. It is also not clear that providers are given the tools to communicate with patients effectively (particularly rural women) or to cope with tense situations where patients resist provider direction. Encouragingly, providers were responsive to additional training on topics of patients’ rights, but mainly because they thought that greater respect for patients’ rights would result in more women coming to health facilities to deliver.
Two final finding of note concern gender and stigma. We find no differences in reports of disrespect and abuse or in knowledge of patients’ rights by gender of the midwife. Moreover, amongst our small sample of patients, we find that male midwives are well accepted, if not preferred, as they were perceived to be more empathetic, in contrast to studies conducted elsewhere [
58]. This finding was somewhat unexpected. While reports have indicated that male midwives in Ethiopia are well accepted by patients, it was surprising to find that that male midwives are often seen as being
more sympathetic and less abusive than female midwives. Our relatively well-educated patient sample may obscure the possibility that rural women with less education might be more receptive to female birth attendants.
3 Future research on respectful maternity care in Ethiopia should explore gender dynamics in more detail, in particular the relationships between age, patient education levels, perceived lack of provider authority, and gender. For example, it would be interesting to examine whether patients perceive
young male midwives as having greater legitimacy and authority than their young female counterparts, and whether young female midwives therefore feel a greater need to exert authority in ways that manifest as abuse.
Our study was one of the few to incorporate clinical scenarios involving stigmatized services into an examination of disrespect and abuse during maternity care. The responses to the clinical scenarios indicate that respect for patient autonomy and the right to information and timely care might vary depending on the degree of stigmatization of the service. Further research on the prevalence of disrespect and abuse in stigmatized populations and services would be useful. To date, research on this subject in sub-Saharan Africa has almost exclusively focused on stigma in HIV/AIDS care.
It is important to understand the study results in the context of a health care system that is seeking to expand and improve access to maternity care, but where shortages of staff, facilities, and supplies remain. Our findings suggest that these health system weaknesses are associated with abuse and are seen in themselves as abuse by women. The interviews suggest that abuse most often occurs when harried providers encounter patients who are non-compliant. Patients’ birth stories reveal an undercurrent of provider impatience and haste or rushing to provide care due to over-crowding and heavy workload. Therefore, addressing health systems and structural issues around provider workload should complement any training initiatives on disrespect and abuse of patients. Further operations and evaluation research on the feasibility and effectiveness of these interventions would be required.
Although our findings point to deficiencies in the health system, they also highlight several successes. Notable examples include the high level of knowledge on basic patients’ rights demonstrated by providers, and the absence of reports of corruption and bribery by patients. It is also noteworthy that most of the women interviewed seemed to not be intimidated by health care providers even when abuse occurred (although this might be due to our patient sample being more educated than the national average). They were clear about what they wanted in regards to their care and did not hesitate to point out and criticize lapses. Several patients knew what the obligations of midwives were. As noted above, some also chose facilities based on their reputations for providing quality care, often with the input of health extension workers. This knowledge and strength is a crucial element for building health system accountability, and its presence is an encouraging sign that recent health education and outreach initiatives may have had some success in raising expectations and conveying to women what services they have the right to receive from their local health facilities.
Limitations
The main study limitations are as follows: our study sample is small and limited to a single geographic region, which makes it difficult to generalize findings to other Ethiopian or African contexts; in addition, we were not able to directly observe the provision of care, so our reports of disrespect and abuse are indirect. The use of clinical scenarios, however, gave us an idea of how providers approach care of patients who might be prone to receiving substandard care and allowed us to gauge whether this care would have involved abuse or an abrogation of rights. Secondly, the use of a single translator limited our ability to conduct a systematic quality assurance of transcript translations.
Further, patients interviewed had a disproportionately high level of education. This is both a limitation and an asset. Educated respondents may be more empowered to speak out about abuse (either experienced or observed) than their less-educated counterparts, so this unusual sample may provide a more accurate picture of the extent of disrespect and abuse among patients than a more representative sample would have. Our less educated patients, however, were far less forthcoming than those reported in other similar studies. This suggests that we may be missing important information on the perspectives of rural, less-educated women in this study. This is a significant limitation, even though we did find that some of our more educated patients seemed inclined to speak up on behalf of those who are least empowered.
Acknowledgements
Our profound thanks to the busy patients, midwives and midwifery students in Debre Markos who took the time to participate in this research. We thank Ms. Nicole Daoud who assisted with interviews and helped to create data analysis templates. We are grateful to our interviewers—Daniel Bekelem, Ayenew Negesse, Yshimharg Shita, Tigist Kefale—for their valuable feedback on our data collection instruments, and to Dr. Belete Tafesse for his assistance with translation and analysis. Finally, we appreciate reviewer comments that strengthened this research.