Prevalence and Associated Factors of Disrespectful and Abusive Care During Childbirth Among Women who Gave Birth in Health Facilities in Hawassa city, Southern Ethiopia. A Cross - Sectional Study

Background: There is growing evidence that shows phenomena of disrespect and abuse (D&A) occurs globally even though the degree and severity is different across countries. The problem is getting attention in recent years especially in developing countries like Ethiopia. However, there is a paucity of studies assessing the magnitude of disrespect and abuse. This study was undertaken to determine the prevalence and associated factors of disrespectful and abusive care during childbirth in health facilities of Hawassa city, Southern Ethiopia. Methodology: A facility-based cross-sectional study was conducted in Hawassa city from February 8 April 27, 2018. A total of 577 mothers from both public and private health facilities were randomly selected. Domains of D&A that were assessed were; physical abuse, verbal abuse, stigma and discrimination, failure to meet professional standards of care and poor rapport between women and providers. Multivariable binary logitmodel was used to examine the relationship between exposure and outcome variables. Adjusted odds ratio (AOR) with 95% condence intervals (CI) is used for summarizing the ndings of the analysis. Result: The mean age of the respondents was 26.8 (SD± 4.4) years. Overall 46.9% [95% CI: (42.8-51)] reported experiencing three or more forms of disrespect and abusive care during childbirth in health facilities. In the logit model; the adjusted odds of D&A among births in public health facility is 12.9 times higher than birth those in private facilities [AOR=12.94 (95% CI: 5.87, 28.50)],mothers who had total delivery four and above had 4.7 times increased odds of encountering D&A [AOR=4.67 95% CI: 1.69, 12.89)].In contrast to mothers who had spontaneous vaginal delivery mothers who had instrumental delivery had 2.6 times increased chance of encountering D&A [AOR =2.63 (95% CI: 1.05, 6.59)]. Conclusion: The prevalence of disrespect and abusive care in Hawassa health facilities during labor and delivery is high. Factors associated with D&A include the type of health facility, mode of delivery

Even though skilled maternity services in the world improved substantially during the new millennium era, there is a vital gap in the quality of maternity care which intended millions of women currently suffer from services that are neglectful, abusive and disrespectful (5).
WHO states compassionate and Respectful Maternal Care (RMC) as the right every woman has to attain the highest standard of health. This includes the right to dignity, compassionate and RMC to all childbearing women around the world throughout pregnancy, childbirth and postnatal period. Globally, many women experience D&A care while giving birth in health facilities. This D&A not only violates the rights of women to get respectful maternity care but also can affect their rights to life, their rights of health, their bodily integrity and freedom from discrimination (6).
Institutional delivery in which little attention has been given to the care and respect given during health facilities delivery could be why many women avoid giving birth in health facilities and their perceptions would have the most in uence in their decisions to use health facilities in the future (7,8).
Though every woman has the right to have respectful care, many women and adolescent girls do not receive the care that matches these terms. Caring, empathetic, supportive, trustworthy, respectful and empowering are among the identity that would describe the ideal skilled birth attendant (9).The promotion of RMC goes beyond relatively direct clinical quality improvement but also women's relationship and interactions with care providers (9,10). D&A is "interactions or facility conditions that local consensus seems to be humiliating or undigni ed and those interactions or conditions that are experienced as or intended to be humiliating or undignifying" (11). D&A is currently renamed as mistreatment and is categorized as physical abuse, sexual abuse, verbal abuse, stigma and discrimination, failure to meet professional standards of care, poor rapport between women and providers, and health systems conditions and constraints (5).
D&A during maternity care is violation of women's basic human rights (12). Evidence also shows that D&A at health facilities have a negative effect on women's decisions to seek care for delivery service even though the degree and severity is different across countries (5,12,13). The quality of care offered at maternity facilities not only affects pregnant women both emotionally and physically but also has an impact on the long-term health and survival of mothers and neonates (2).
Factors majorly contributing to D&A includes health system factors, provider level factor and personal factors like normalizing of abuse, individual knowledge on the right of health care service, level of education, socio-demographic factors and obstetric factors are some factor for D&A during facility-based childbirth (14,15).
While many interventions were aimed, the quality of relationships with caregivers during maternity care has received less attention (15). There are few researches conducted on D&A of women during childbirth in Ethiopia and no research at all was conducted speci cally in Hawassa city. Therefore, this study aims to assess the prevalence and factors associated with D&A during facility-based childbirth in Hawassa City, Southern Ethiopia.

Study design and setting
A quantitative cross-sectional study design was conducted in both public and private health facilities found in Hawassa city from February 8 -April 27, 2018. Accordingly, 4 health facilities (Hawassa University Comprehensive Specialized Hospital (HUCSH), Adare General Hospital, Millennium Health Center and Alamura Health Center) from public health facilities and 2 health facilities (Alathion General Hospital and Bushulo Health Center) from the privatehealth facilities were randomly selected. To measure experience of disrespect and abuse interviewer-administered questioner validated in Ethiopian context was used.

Study population
All mothers who have given birth in the selected study facilities and who are ready for an exit in a postnatal room were recruited to the study.

Sample size and sampling
The required sample size for the rst objective of the study was determined using single-population proportion formula with the assumption of 78.6% prevalence of D&A(16), 95% con dence level and 5% marginal error. The estimated sample size after considering 10% response rate and 1.5 design effect was 426.
Sample size for the second objective was calculated using EPiInfo version 7 software. The calculation was separately made for three potential predictors (mode of delivery (SVD/Instrumental delivery), number of ANC visits (<4visits/≥4visits) and delivery place (public/private health facility)) of D&A. The expected proportions for the aforementioned factors were extracted from a study conducted in Bahirdar, Ethiopia (17).Ultimately a highest sample size of 577 was calculated for factor associated with ANC visits. This sample size was reached at using two population proportions formula with speci cations of 80% power, 95% con dence level, 1.5 design effect, 10 % nonresponse rate, AOR of 1.97 and percent of D&A among unexposed of 61%.
Health facilities were initially strati ed into public and private health facilities. Afterwards Simple Random Sampling (SRS) technique was applied to select study health facilities among public and private health facilities. All primary, secondary and tertiary level of public health institutions found in the city were considered for the study. Accordingly, from public health facilities; HUCSH, Adare General Hospital, Millennium Health Center and Alamura Health Center were randomly selected. On the other hand, from the private health facilities; Alation General Hospital and Bushulo Health Center were randomly selected.
To allocate sample sizes for private and public health facilities we used a three-month delivery performance of respective health facilities of the preceding year within similar month of the data collection period. Thus 197, 188, 15 and 33 mothers from HUCSH, Adare General Hospital, Alamura Health Center and Millennium Health Center were sampled respectively. Additionally from private health facilities; 121 mothers from Bushulo Health Center and 23 mothers from Alathiyon Hospital were sampled. Respondents were engaged in the study one after the other.

Data collection
Levels of D&A during childbirth were measured using the 15-itemrespectful maternity care scale. The measurements were validated and reliable in Ethiopia context (18).The domains of disrespectful and abusive care that were assessed were physical abuse, verbal abuse, stigma and discrimination and poor rapport between women and providers.
The questionnaire was rst developed in English and then translated into Amharic. Seven data collectors and two supervisors who are BSc midwifery and nurses recruited from health facilities other than the study area were used. Other relevant variables (socio-demographic variables, obstetric characteristics, sex of service providers and type of health facility where the delivery was conducted) were added to the data collection tool as additional information.

Quality assurance techniques
Both the data collectors and supervisors were trained for one day on issues such as the techniques of approaching and introducing themselves to the respondents, data collection and how to ask questions in a neutral matter. Supervisors were also trained on how to check the completeness and consistencies of questionnaires lled by the data collectors to ensure the quality of the data. Pre-test was carried out on 28 mothers who gave birth at Leku Primary Hospital and Adare Health Center then after any unclear question was checked and corrected based on the result. Double data entry of questioner in EpiInfo was also performed on 28 samples to reassure the quality of the data.

Data analysis and interpretation
The collected data were checked for completeness and consistency. The data was coded and entered into EpiInfo version 7 and then exported to SPSS version 23 software for analysis. The outcome measurement was dichotomized into responses of "Yes" or "No" to identify reported events of D&A. We computed fteen questions and the mean score was used to identify the overall prevalence of D&A among mothers. Additionally to identify the domain of each category of D&A we computed the reported event in the respective category.
For categorical variables frequency and proportion and for continuous normally distributed variables; mean and standard deviation was applied to summarize information about dependent and independent variables. Potential confounders were identi ed by running Bivariable logistic regression analyses and then those independent variables that had a p-value less than 0.25 were further taken to multiple variable logistic regression analyses to control the effect of confounding variables. Adjusted odds ratio with 95% CI was used to summarize the ndings of the analyses. A p -value < 0.05 was used to declare signi cance of the association between independent and dependent variables. The goodness-of-t of the model assessed using Hosmer-Lemeshow statistics.

Socio-demographic characteristics of the study participant
Five hundred forty-eight mothers responded to the question, which makes a response rate of (95.0%). The mean age of the respondents was 26.8 (SD ± 4.4) years in which more than one-third of the respondents 243 (44.3%) fall in 25-29 years. More than seventy percent of the study participants 416 (75.9%) were from the urban. 257 (46.9%) were followers of protestant religion. Almost all respondents were married 526 (96.0%). Of the total respondents, almost half of the respondents are Sidama ethnic group 235 (42.9%) followed by Amhara 102 (18.6%). Most of the respondents attend primary education 224 (40.9%). Most of the respondents were Housewife 267 (48.7%) and the majority of the respondents have family average monthly income of greater than 1001 Birr (Table 1).

Obstetric History Of Mothers
Majority of mothers had history of ANC follow for their current pregnancy, 513 (93.6%). More than half of respondents 271 (53.1%) have four and above ANC visits. Among them,many of the mothers visited public health center 330 (64.5%). Majority of respondents have a total number of at least 1 delivery 236 (43.1%) and 33 (6.0%) of the respondents have four and above gravidity among those 180 (32.8%) of respondents deliver all their babies in health facilities and 261(47.6%) of respondents deliver at least one of their baby in health facilities. From the total respondents, 166 (30.3%) reported they faced complication during L&D. Majorly reported form of complication were prolonged labor 87 (15.9%). Most mothers get deliveries through spontaneous vaginal delivery 313 (57.1%) followed by episiotomy 113 (20.6%) ( Table 2).

Prevalence Of Disrespect And Abuse During Facility-based Childbirth
The overall prevalence of disrespect and abuse among mothers who experience three or more form of disrespect and abuse during childbirth in health facilities was found to be 257 (46. 9%) with 95% CI: (42.8-51.0) where the major share is from general hospital, which accounts 124 (48.2%) and the least belongs to private hospital 2 (0.78%) (Fig. 1).

Types Of D&a During Facility-based Child Birth
From the domain of physical abuse, 74 (13.5%) of women were beaten, slapped, kicked or pinched during delivery. The other most commonly experienced form of disrespect and abuse was poor rapport between women and the provider. Under this domain, all of the women 548 (100%) reported they were not allow to practice cultural ritual practice in the health facilities. By excluding this variable 241 (44%) of mothers report D&A in the domain of poor rapport between women and providers. The second domain of D&A care was failure to meet professional standards of care and from this domain the most reported form was, health worker does not respond to mothers needs 162 (29.6%). Additionally 145 (26.5%) of mothers reported they face verbal abuse (Table 3).

Women's Experience Of Disrespect And Abuse During Child Birth
Regarding the types of D&A mothers had experienced during labor and delivery; only one type of D&A was experienced by all mothers followed by two types (69%) of D&A and then more than three types of D&A (46.9%), no data was recorded for never abused (Fig. 3).

Disrespect And Abuse And Its Associated Factors
Binary logistic regression was performed to assess the association of each independent variable with D&A. Twelve factors that showed a p-value < 0.25 were further taken to the multivariable regression model. In a multivariable logistic regression analyses, mode of delivery, type of health facility, parity and sex of the main provider conducting delivery were signi cantly associated with D&A at P-value less than  A study conducted in Brazil found that; mothers who at least have one type of D&A were 18.3% and 5.1% of the mothers encountered two types D&A. The ndings of this study were much lower compared to the current study (19). This discrepancy might be due to socio-economic difference in which the majority of respondents from the Brazil study have nished class up to grade 12 but in our nding, most of the respondents have primary school educational background. Additionally in the Brazil study interview was conducted three months after they delivered and this may have lead recall bias. Similarly, the nding of the current study was higher than studies conducted in Tanzania 15%, Kenya 20% and India 28.8% (20)(21)(22).This inconsistency may be due to policy differences related to RMC and time and setting of the interview.
Contrarily, a study conducted in Southeastern Nigeria found 98% prevalence of at least one form of D&A during L&D which much higher than the nding of the current study (23). The nding of the current study is also lower than the study conducted in Addis Ababa, Ethiopia which showed the prevalence of D&A 78.6% (24). This discrepancy might be due to the operational de nition difference that has been used to identify experience of D&A where the study used experience of at least one form of D&A.
Relatedly, a community-based study on D&A which was conducted in Bahirdar, Ethiopia showed a higher prevalence (67.1%) of D&A than this particular research (17). Reasons for this discrepancy might be as the study was conducted in community; it might be less exposed to social desirability bias in addition the operational de nition used to estimate overall prevalence is different.
A cross-sectional study conducted in Addis Ababa, Ethiopia showed at least one form of D&A was experienced in 75.3% in health centers and 81.8% in hospital (24). Consistently our study showed same gure but with wide gap between experience of D&A in health center 22 (8.5%) and hospital 235 (91%). This discrepancy may be due to the operational de nition and the small sample size used in the Addis Ababa study than ours. This two studies are in line with a cross-sectional study conducted in Ethiopian public health facility which shows health centers had a higher RMC score compared to hospitals (25).
A study exploring Jordanian women's exposure to verbal abuse found that 37% were victims of verbal abuse (26) which is higher than the nding of the current study 26.5%. This may be due to literacy level of mothers in Jordan is higher compare to respondents in Hawassa health facilities.
Our study pointed out verbal abuse on mothers by health workers was 145 (26.5%) which is higher than a research conducted in Tanzania 19.48% and 28.21% which reported verbal abuse on exit and follow-up study respectively. Additionally it was also observed in Nigeria and Malawi 4.30% and 1.9%, respectively (23,27,28). This inconsistency may be due to socioeconomic background in which majority of educational background of respondents speci cally in Nigeria is above secondary level.
In the current study physical harm like slapping and kicking was experienced by 74 (13.5%) mothers which is three times greater compared to a study conducted in Tanzania, which is four point two percent (12). Differently lower when compared to study conducted in India and Enugu, Southeastern Nigeria 30.4% and 35.7%, respectively (23).
According to Bowser D. discrimination during childbirth is mentioned as discriminating a woman based on race, ethnicity, age, language, traditional beliefs and preferences, economic status and educational level (29). It is in line with this study as mothers reportedly not treated well with some of the reason related to their personal attribute 74 (13.5%). Moreover, their companions were also discriminated because of some of their personal attribute in which this gure is higher than a study conducted in Enugu, Southeastern Nigeria showed that D&A during childbirth from this discrimination accounts twenty percent (23).
In this study, type of health facility was signi cantly associated with D&A. Those mothers who deliver in public health facilities were 12.94 times more likely to be D&A as compared to mothers who delivered in private health facilities. This nding was consistent with a study conducted in Bahirdar, Ethiopia which showed that mothers who delivered in public health facilities were more likely to be D&A than mothers who delivered in private health facilities (17). Similarly study conducted in India showed that D&A are more common in the government health sector than private health sector (22).The reason behind this association could be sta ng shortages which may have led to longer wait times and neglectful and poorquality care. Lack of infrastructures like enough room and bed which may led to reduced privacy in the labor wards may also have contributed for increased D&A in public facilities. Shortage of medications can also create stressful working environments which might predispose health care providers to behave poorly (or even abusively) towards women in public health facilities (4,21).
Our study shows, those mothers who delivered four and above babies are 4.67 times more likely to be D&A than those who deliver only one baby. This nding is consistent with a study conducted in Kenya which showed that women of higher parity, were three times more likely to be D&A (12). This could be due to provider perception that multi-parous women already have previous birth experience.
This nding also showed that mode of delivery was associated with D&A. Mothers who had instrumental delivery were 2.63 times more likely to be D&A than those mothers who had spontaneous vaginal delivery. This nding is in line with studies conducted in Ethiopia (17,30). This might be related to the invasiveness of the procedure and health workers eagerness to ensure good health outcome since they used it to shorten the second stage of labor this will be perceived as painful experience for the mothers therefore they think that they are being D&A.
Our study revealed that the sex of the main health provider conducting delivery was signi cantly associated with D&A. Those mothers whose delivery was conducted by female health provider were 1.75 times more likely to be D&A than mothers whose delivery was conducted by male health provider. This nding shows consistency with a study conducted in Ethiopian public facility which reported male providers were observed engaging in RMC practices more frequently than female providers (25).
This particular study doesn't include non-consent, con dentiality in the domain which will be considered as limitation.

Conclusion
This particular study endeavored to provide the picture of D&A care that women are experiencing in Hawassa city health facilities. The study further highlighted that the prevalence of D&A in Hawassa health facilities found to be high 46.9%. The factors that are associated with D&A were found to be; types of facility, mode of delivery, sex of health care providers conducting delivery and total number of delivery.
After all, avoiding D&A care for women is not about luxury, it is about human right issue as a woman has the right to the maximum attainable standard of health, which includes sympathy and respectful health care.

Declarations
Ethical approval and consent to participate Ethical approval to start the study was obtained from Institutional Review Board of Hawassa University College of Medicine and Health Sciences. Additional, permission was also sought from Hawassa city health department and from the randomly selected health facility. Verbal informed consent was obtained from participants after a detailed explanation of the purpose and bene t of the study before the individual data collection.

Consent for publication
Not applicable.

Availability of data and materials
The datasets analyzed during the current study are not publicly available due institutional regulation but are available from the corresponding author on reasonable request.
The authors declare that they have no competing interests. Funding