Respectful maternity care is a universal human right that is due to every expectant woman in every setting. Nonetheless, many women experience mistreatment during labor and childbirth in a health institution. Respectful care involves the appreciation of women's moral principles, emotional state, self-worth, choices, and favorites [1, 2]. Whereas, mistreatment is recognized as the violation of this fundamental principle of human rights [3, 4]. In the health facilities of several developing countries, women might encounter different types of mistreatment during labor and delivery, such as physical abuse, non-consented care, non-confidential care, non-dignified care, discrimination, abandonment of care, and detention at health facilities. Risk factors associated with mistreatment include lower socioeconomic status, normalization, provider prejudice, as well as the absence of independence, confidence, and strength 
As evidence revealed, the prevalence of maternal mistreatment during childbirth varies substantially from one country to another. For instance, it was reported as 15% in Tanzania , 20% in Kenya , 78% in Ethiopia , and 97.4% in Peru . Lack of respectful care during delivery is still a continuing cause of the untold suffering of women in the most vulnerable times of their lives. These traumatic experiences exert long-lasting negative effects on women all throughout their lives . Women who experience mistreatment during labor and childbirth may not choose the facility for their subsequent deliveries  and may not recommend it to other women . Mistreatment not only affects healthcare utilization but also violates women’s fundamental human rights .
The poor quality of care during childbirth has remained a major challenge and concern to the point that maternal health has been recognized as one of the three Sustainable Development Goals. In Ethiopia, access to health care services has improved, and free maternity services are introduced. However, the mere improvement of access is not sufficient to raise the uptake of skilled maternity care. According to the report of Ethiopian Demographic Health Study (2016), only 26% of mothers had health facility delivery and the rate of maternal death was found to be 412 per 100, 000 live births . Mistreatment continues to be a major threat in seeking health service delivery. However, it has not received the attention it deserves, as compared to other barriers, such as cost or distance. Therefore, the protection and appreciation of women's rights are of utmost importance which can be achieved by respectful care and psychological support [5, 15, 16].
In general, there is a paucity of inclusive evidence regarding the extent and associated factors of mistreatment in Ethiopia, especially in the study area. Nonetheless, it is absolutely essential to benefit from up-to-date information for the development of preventive strategies. It increases the number of women delivering at health facilities and by doing so decreases maternal mortality. Therefore, the present study aimed to determine the prevalence and risk factors associated with mistreatment amongst mothers during labor and childbirth at the hospitals of Silte Town, Southern Ethiopia.
Materials and Methods
This hospital-based cross-sectional study was conducted at Worabe Comprehensive Specialized Hospital, Kibet Primary Hospital, Tora Primary Hospital, and Alemgebeya Primary Hospital, Silte Zone, Southern Ethiopia within March 1-30, 2018. Our source population included all women who gave birth in selected Hospitals of Silte Zone in the study period. The study populations were sampled women who delivered in nominated Hospitals of Silte Zone in the study period. The inclusion criteria entailed the sampled women who delivered in nominated Hospitals during the entire study period. However, critically ill women during the data collection period were excluded from the study.
The sample size was calculated using the single population proportion formula via EPI INFO software (version 7). The subsequent conventions were applied: a confidence interval of 95%, the desired degree of precision 5%, women’s experiencing mistreatment during labor and delivery as 50%, and a non-response rate of 10%. Consequently, the sample size was obtained at 422. A systematic random sampling method was employed to enroll the respondents, while the probability-proportional-to-size method was used to determine the number of respondents from each study hospital. The first respondent from each hospital was selected randomly, and subsequent respondents were then selected from every third interval.
Data were collected by organized and pre-tested questionnaires. The data collection instrument was developed after the revision of the integrated tool kit of maternal and child health package of respectful care . It was designed to collect information regarding the characteristics of socio-demographic data, obstetric-related factors, and categories of mistreatment. Three midwives with relevant diplomas and two with bachelor’s degrees were recruited for the data collection and supervision.
Thereafter, the tool was initially prepared in English, and it was again translated into Amharic. In addition, the tool was translated into English to guarantee its standardization; finally, the Amharic version was employed to collect the data. The pre-test was conducted on 21 women at Butajira hospital one week prior to data collection. The tool was modified according to the outcome of the pre-test. The validity of the instrument was approved by the proper application of validity criteria (content validity). A reliability test was performed for the mistreatment items after the pre-test, and the internal reliability (coefficient alpha) of the instrument was found to be 0.78. Data collectors and supervisors received a two-day training on the content of the questionnaire, objectives, and techniques of data collection. It is worthy to note that the data collectors worked under strict follow-up of supervisors. In a similar vein, the supervisors and the investigators were verified for the completeness of the tools every day at the end of data collection. Additionally, the collected data was cautiously entered and checked for its cleanness before commencing the procedure of analysis.
Women were regarded to be mistreated during labor and childbirth when they answered yes to at least one of the itemized problems. They included physical abuse, non-confidential care, non-consented care, non-dignified care, abandonment of care, discrimination, and detention in the health facilities. Physical abuse was regarded as at least one of the following: slapping, forceful delivery, maternal-neonate separation after birth, culturally inappropriate care, non-use of anesthesia, and food or drink deprivation. Non-confidential care comprises at least one of the following: non-use of curtain and discussion about private health information. Non-consented care involved at least one of the following: care providers' refrain from introducing themselves, discouraging patient question asking, impolite responses, non-explanation of what is being done, absence of periodic updates, denied childbirth positions, and non-obtainment of permission for procedures. Non-dignified care involved care provider's negative comments and/or shouting. Abandonment meant being ignored when needed help and/or delivery without birth attendant. Discrimination included poorly treated due language and/or poorly treated due age. Detention implied being detained for failure to pay and/or being asked for informal payment.
The obtained data were entered into EpiData software (version) 3.1 and analyzed in SPSS software (version 23). First, bivariate logistic regression was conducted for the selection of candidate variables into multivariable logistic regression. In binary logistic regression, the variables with a P-value < 0.25 were transferred to the multivariable logistic regression model. It was performed to find the independent associated factors of the outcome variable and control probable confounders. An odds ratio was determined at a 95% confidence interval (CI). The fitness of the model was confirmed by Hosmer Lemeshow statistic test which has a p-value of 0.84.
The ethical approval letter was attained from the Institutional Review Board of Jimma University, Institute of Health, and Faculty of Health Science (Ref.no:IHRPGP/64/2018 Date: 16/03/2018). In addition, agreement letters were received from the Department of Silte Zone Health Bureau and the study health facilities. Finally, oral consents were obtained from all the respondents. Furthermore, respondents were assured that their non-participation in the study will not affect their treatment.
Among 422 invited participants, 409 mothers (96.9%) agreed to be interviewed. The mean age of mothers was reported as 28.1± 4.7(standard deviation) years; moreover, 169 (41.3%) of them fell within the age range of 25-29 years. A number of 408 (99.8%) mothers were married, 177 (43.3%) had primary school education and 85 (69.7%) of them were housewives. In addition, 353 (86.3%) women were Silte in ethnics, 363(88.8%) were Muslims, and 188 (46%) resided in rural areas. With regard to income, 228 (55.7%) earned a monthly income of less than 55 United State Dollar (USD) (Table 1).
A number of 350 (77%) mothers were multiparous, 141 (34.48%) had antenatal care visits for current pregnancy, and 284(69.4%) had a spontaneous vaginal delivery. 269 (65.8%) mothers did not stay in the health facility after delivery less than 24 h and 331 (76%) women did not report complications during delivery. In addition, 206 (50.4%) birth companions were present during labor and delivery, and 98 (24%) women encountered obstetric complications during delivery (Table 2).
Prevalence of mistreatment during labor and childbirth
Non-consented care 269(65.8%) and physical abuse 230(56.2%) were reported as the most commonly occurred types of mistreatment during labor and childbirth. The overall prevalence of mistreatment experienced by women delivering in health facilities was found to be 67.7% (Table 3).
Associated factors of mistreatment
The results of the binary logistic regression revealed the candidate variables for multivariable logistic regression. They included decision-makers on women’s health, having antenatal care (ANC) visits, unnecessarily longer stays at health facilities after delivery, presence of companions during childbirth, and complications during childbirth. The multivariate logistic regression finding indicated that having ANC visits, unnecessarily longer stays at health facilities, the presence of companions during childbirth, and the occurrence of childbirth complications were significant predictors of mistreatment at 95% CI and P
Similarly, the women who stayed unnecessarily longer at health facilities after delivery were about 2.6 times more susceptible to experience mistreatment, as compared to that of women who did not stay unnecessarily longer at health facilities after delivery (AOR=2.6; 95% CI: 1.34-5.18). At the same time, women who had birth companions in the delivery room had a 60% reduction in the likelihood of encountering mistreatment, in comparison to their counterparts (AOR=0.4, 95% CI: 0.21-0.57) (Table 4).
The current study aimed to consider the prevalence of mistreatment and its associated factors amongst women during labor and childbirth in public hospitals of Silte town, Southern Ethiopia. The overall prevalence of mistreatment experienced by women was reported as 67.7%; nevertheless, this prevalence was lower, as compared to the value reported as 78% in Addis Ababa . This can be attributed to more qualified health care experts and advanced equipment available in Addis Ababa, which is the capital of Ethiopia, as compared to the setting of the current study. Moreover, this prevalence was less, as compared to the value obtained as 98% in a study performed in Nigeria . This discrepancy can be justified by the variations in study time, sampling technique, and sample size. However, the frequency reported in the current survey was greater when compared to other studies conducted in Kenya  and urban Tanzania , in which 20% and 15% of women had experienced mistreatment, respectively. This contradiction may be due to differences in the implementation of maternal care policies and the training of providers on key aspects of respectful maternal care.
The existence of complications throughout childbirth was another significantly associated factor of mistreatment. The women who reported obstetric complications during delivery were 2.5 times more likely to experience mistreatment, as compared to their counterparts. Similar findings were also reported in a study carried out in Tanzania . which could be ascribed to the fact that women who experience complications during childbirth need extra support and encouragement. Unnecessarily long stay at health facilities was also another risk factor significantly associated with mistreatment. The women who stayed unnecessarily longer at health facilities after delivery were about 2.6 times more prone to mistreatment, as compared to that of women who did not so. This finding was also in line with a study performed in Tanzania . This association is due to prolonged contact with several health care provider which increases women’s chance of experiencing mistreatment.
The results of the present study revealed that having an ANC visit during pregnancy was associated with mistreatment. Women who had ANC visits were 50% less likely to experience mistreatment, as compared to those who had no ANC visits. The women who had birth companions in the delivery room had a 60% reduction in the likelihood of mistreatment experience, as compared to their counterparts. The results reported in the current study were in agreement with a study performed in Tanzania . It can be ascribed to the fact that the women who have ANC visit obtain health education and counseling that provide them golden opportunities to know their rights and responsibilities.
The presence of birth companions during labor and childbirth was also another preventive predictor of mistreatment. This finding was supported by the World Health Organization . This may be due to the fact that care providers feel more accountable for the provision of better care and counseling when birth companions are present during labor and delivery. Moreover, birth companions may provide women with emotional and physical support.
Mistreatment was measured by a validated and standardized questionnaire that was pretested and revised. In addition, the interviews were performed immediately after women's discharge from the maternity ward but in the study hospital, which enabled the women to vividly recount their delivery memories. Furthermore, the data collectors were selected from another health facility for the reduction of social desirability bias.
Every study has some limitations that should be addressed in the paper. In this regard, the result of the present study might not be generalized to the total population since the study was conducted in hospitals. Moreover, the current study adopted a quantitative approach alone.
As evidenced by the obtained results, a considerable proportion of women had experienced mistreatment during childbirth. Having ANC visit, presence of complication during labor, unnecessarily longer stays at health facilities, and the presence of birth companion were recognized as independent associated factors of mistreatment. All health institutions need to provide a mechanism for the prevention and elimination of mistreatment. This can be achieved by strengthening the continuous provision of education and counseling during antenatal care visits, allowing for the presence of birth companions, and minimizing unnecessarily longer health facility stays after childbirth. Further mixed methods studies are necessary to obtain more inclusive information on the multifaceted and interactive women–provider context of mistreatment.
Our sincere appreciation and thanks go to our data collectors, supervisors and the women who participated in the study, Jimma University, authorities and departments of Silte Zone Health Bureau, and the health care providers in the nominated hospitals.
Conflicts of interest
Authors declared no conflicts of interest.