Less than one in four mothers get quality intrapartum health care services in Ethiopia

Intrapartum care is a platform of comprehensive healthcare for pregnant women that is designed to improve birth outcomes for mother and child. However, complications during the intrapartum period continued to be the leading cause of death for women of reproductive age and newborns. Therefore, the aim of this study was to assess the prevalence of quality of intrapartum care and its associated factors among mothers in Ethiopia. A community based cross sectional study was conducted among 4469 mothers who gave birth in the last 2 years. Quality of intrapartum care was analyzed based on the assessment of health facility delivery, skilled birth attendants and early initiations of breast-feeding. Stata version 14 software was used for data cleaning and analysis. A mixed effect multilevel logistic regression was conducted to determine factors associated with quality of intrapartum care. An adjusted odds ratio with 95% confidence interval and a P value of less than or equal to 0.05 was used for the identification of both individual and community level factors. Overall, the prevalence of quality intrapartum care in Ethiopia was 23.8% (95% CI 22.6, 25.13). Primary education (AOR = 1.46, 95% CI = 1.14, 1.88), rich household class (AOR = 1.48, 95% CI = 1.10, 1.98), history of ANC (AOR = 2.91, 95% CI = 2.18, 3.86), perceived distance to the health facility as not a big issue (AOR = 1.63, 95% CI = 1.30, 2.05), urban residence (AOR = 2.97, 95% CI = 1.93, 5.09), Tigray region (AOR = 5.01, 95% CI = 1.25, 20.59), community level poverty (AOR = 0.63, 95% CI = 0.41, 0.97), and having 2–4 children (AOR = 0.74, 95% CI = 0.56, 0.97) were significantly associated with quality of intrapartum care. The finding conclude that less than one in four mothers received good quality intrapartum care. In order to optimize the quality of intrapartum care, the government should empower women through extensive education. It is also recommended for the Ministry of Health to evaluate the health facilities and community health workers to increase coverage of ANC and provide financial assistance to rural residents and the poor household class.


Data collection procedures, tool and population
Even though Demographic Health Surveys (DHS) are conducted every 5 years, the recent available standardized data in Ethiopia is the 2016 Ethiopian Demographic Health Survey (EDHS) 30,31 .As a sampling frame, the 2007 Ethiopian population and housing census was used.The census was established into 645 enumeration areas: 202 urban and 443 rural.Then, a fixed number of 28 households per cluster was selected by equal probability systematic sampling technique 32,33 .All mothers who are members of the household and visitors who slept in the household the night before the survey were eligible for the interview 31 .The detailed of the methods for the survey is available elsewhere 34 .
Those mothers who were available during the survey in Ethiopia were the source population.Whereas, those mothers who were in the selected enumeration area were considered as study population and used for analysis.

Study variables
The dependent variable of the study was Quality of intrapartum care which was determined by the combination of the following three interventions: giving birth at the health facility, birth assisted by skilled personnel, and new born begin to breastfeed within 1 h of birth 4,35 .Skilled birth attendant (doctors, health officers, nurses, and midwifes) refers to provider of delivery during labour.When a woman received care from numerous providers, the most skilled provider was recorded 22 .Those mothers who reported the aforementioned skilled birth attendant, birth at health facility, and newborn begin breast feeding within 1 h were considered as getting quality intrapartum care and coded as 1 otherwise no quality of intrapartum care and was codded as 0.
All the independent variables were included after a thorough review of literatures [23][24][25][26][27][28] .Both individual and community level factors were considered for this study.Age: 15-24, 25-34 and 35-49 years; educational status: no formal education, primary education, secondary and higher education; current marital status: married, unmarried: individual level media exposure; yes (if the mother watches television, or listen radio, or reads a newspaper/ magazine at least once a week), otherwise no; history of ANC: yes, no; age at first birth: before 18, 18 or above were considered as individual level variables.Whereas, residence (urban vs rural); distance to the health facility (big problem vs not big problem), community level poverty, community level education and community level media exposure from the community level factors were considered for this particular study.Community level poverty was generated by the proportion of poorest and poorer wealth classes.Community level education was calculated by the proportion of primary, secondary and higher education.Similarly, community level media exposure was generated by the proportion of media exposure.The data for all the three community level variables were not normally distributed.Therefore, each variables were categorized as "low" if the proportion was below 50%, "high" for including median and above the median value 22,36 .

Data management and analysis
Stata version 14 software was used for data extraction, cleaning and analysis.Initially, the sample data to be representative, weighting was conducted.We used the individual weight for women (v005) and the individual sample weight was generated by dividing (v005) by 1,000,000 before analysis to approximate the number of cases 33 .Secondly, proportion of quality intrapartum care across each independent variable was analyzed.All the variables having a P value less than 0.05 were used for the multivariable analysis.Adjusted odds ratio with 95% confidence interval and a p value of 0.05 were used for identifications of associated factors.Finally, multilevel analysis having both the random (presented with ICC) and fixed effect (expressed by adjusted odds ratio) were done.
Demographic health survey data is hierarchical in nature where individual mothers are nested within the communities.As a result the assumption of independency might be violated in logistic regression.Hence, it could be better to assess both the individual and community level variables on the quality of intrapartum care 37 .Multilevel analysis having four models were analyzed.Deviance (− 2loglikelihoods) was used for model comparison.

Ethical approval
The ethical approval and permission to access the data were obtained from the DHS website www.measu redhs.com.All methods were approved by ICF International and an Institutional Review Board (IRB) in Ethiopia, in accordance with United states Department of Health and Human Services requirements for human subject protection.Ethical clearance was obtained by the Institutional Review Board of Demographic and Health Surveys (DHS) program data archivists after the consent manuscript was submitted to DHS Program/ICF International.Informed consent was obtained from all subjects and/or their legal guardian(s) of minors age below 16.No information obtained from the data set was disclosed to any third person.The study is not experimental study.Further explanation of how the DHS uses data and its ethical standards can be found at: http:// goo.gl/ ny8T6X.

Distribution of quality intrapartum care across individual and community level factors
Higher (26.32%) prevalence of quality intrapartum care was observed among mothers in the age group of 15-24 years.More than half (58.71%) of secondary educated mothers received quality intrapartum care.Of those mothers who had ANC history, 32.57% had reported quality intrapartum care (Table 1).

Distribution of quality intrapartum care across the community level factors
Of urban resident mothers, 61.41% reported quality intrapartum care.Among high media exposed mothers, 32.38% were reported the quality of their intrapartum care (Table 2).
In the final model, having primary education, rich household wealth index, having two or more number of living children, and history of ANC from the individual level factors and urban residence, distance as not big problem and community level education from the community level factors were significantly associated with quality of intrapartum care.
Accordingly, the likelihood of quality intrapartum care was 1.46 [AOR = 1.46, 95% CI = 1.14, 1.88] times higher among mothers who had attended primary education as compared to those mothers who had not followed formal education.
Those mothers from the rich wealth index class had 1.48 times higher likelihood of getting quality intrapartum care as compared to mothers from the poor household wealth status (AOR = 1.48, 95% CI = 1.10, 1.98).Similarly, the odds of quality of intrapartum care was 0.37 times lower among those mothers who were from population of high proportion of community level poverty (AOR = 0.63, 95% CI = 0.41, 0.97).
The odds of quality of intrapartum care was 0.26 (AOR = 0.74 95% CI = 0.56, 0.97) and 0.46 (AOR = 0.54, 95% CI, 0.36, 0.81) times lower among those mothers who had 2-4 and 5 or more children as compared with those mothers who had no children at all, respectively.
The likelihood of getting quality intrapartum care was 2.91 (AOR = 2.91, 95% CI = 2.18, 3.86) times higher among mothers who had history of ANC as compared to those mothers who had no history of ANC.
Vol:.( 1234567890 www.nature.com/scientificreports/Those mothers who perceived distance from the health facility as not a big problem had 1.63 times higher likelihood of getting quality intrapartum care than those mothers who perceived distance from the health facility as a big problem (AOR = 1.63, 95% CI = 1.30, 2.05).
Urban resident mothers had 2.97 times higher likelihood of getting quality intrapartum care (AOR = 2.97, 95% CI = 1.93, 5.09) as compared to their counterparts of rural resident mothers.
Compared with Afar region, the odds of quality of intrapartum care was 5 times higher among mothers who were from Tigray region (AOR = 5.01, 95% CI = 1.25, 20.59) (Table 3).

Random effects (measures of variation) results
The Intra class correlation coefficient for the null model was 0.4796, which means that 47.96% of quality of intrapartum care variation among mothers was due to differences in clusters.Final model which had lowest deviance was selected as the best fitted model.The heterogeneity of quality of intrapartum care was expressed by median odds ratio.The final model indicated that the odds of quality of intrapartum care was 3.01 times higher among mothers who were from higher risk of quality intrapartum care as compared with those mothers who were from lower risk cluster.Regarding the proportional change in variance (PCV), 53.46% of quality intrapartum care was explained by model four (Table 4).

Discussion
Findings of this research revealed that 23.8% (95% CI 22.6-25.13)mothers had experienced good quality of intrapartum care.The prevalence implies that more than three-fourths of the mothers did not received quality intrapartum care.The finding is lower than studies conducted at North Achefer, Ethiopia (27.3%) 38 , Tigray (29.2%) 39 , Jimma zone, Ethiopia (74.9%) 40 .The difference may accounted that the study setting where the current study was conducted based on country representative data.Whereas, the previous studies were conducted in limited geographical area at health facilities.The other possible reason for the difference may be due to quality assessment methods; the current study used three key variables which are found in the EDHS, whereas the previous studies used checklists which hold additional variables such as stages of labour, and availability of health services.
The finding is higher than study conducted in Jabi tehnan district (13%) 41 .The lower prevalence of good quality of intrapartum care might be differences in the assessment method where the study conducted in Jabi Tehinan district in North West Ethiopia, was evaluated the quality of intrapartum care by asking the women about their experience of resources, their cognition, respect, dignity and equity of care they received.
The likelihood of quality intrapartum care was higher among mothers who had learned primary education as compared to those mothers who had no formal education.The finding is similar studies conducted in Uganda 42 , Cameroon 43 , Nigeria 44 and Kenya 45 .More educated mothers had higher likelihoods of understanding the benefits of quality health care, being aware of pregnancy-related morbidities and mortality, perhaps because they are more likely to seek medical care 46,47 .In view of this finding, education is an indicator that needs to be improved to better enable women to access maternal health care.
Those mothers who were from the rich wealth class had higher likelihood of getting quality intrapartum care as compared to the poor wealth class 48,49 .Similarly, the odds of quality of intrapartum care was lower among mothers who were from population of high community level poverty.The finding is similar with studies conducted in Nepal 50 , Kenya 51 , and East Africa 52 .This is because low wealth class women can have specific worries and feels inadequate for service, which in turn lowers the minimum level of health care service 53,54 .Furthermore, the cost of travelling to distant health facilities contributes indirectly to the cost of quality intrapartum care, which can be easily covered by pregnant women from rich wealth class households 55 .Therefore, economic variations in obtaining maternal or reproductive health services plays a role in receiving quality health care services 56,57 .
The odds of getting quality intrapartum care is higher among mothers who were from the urban resident as compared with their counterparts of rural resident mothers.The finding is similar with elsewhere study in Bangladesh 46 .The reason for low quality of intrapartum care in rural area might be accounted due to barriers such as transportation, and availability of health facilities 47 .Conditionally mothers from rural area may face difficulties such as long distance to reach the health facilities and poor road conditions as compared with mothers from urban residence 58 .
The odds of quality of intrapartum care was lower among those mothers who had two or more children as compared with those mothers who had no children at all.In line with this, findings from Cameroon 43 and Sudan 59 revealed that those mothers who had more children were less likely to utilize health care services.Those mothers who had more children may sense overconfidence from their experience of past pregnancy 60 .
The likelihood of getting quality intrapartum care was higher among mothers who had history of ANC as compared with those mothers who had no history of ANC.The same finding was observed in Ethiopia 61 , and Rwanda 62 .The possible reason might be that those mothers who had history of ANC had higher likelihood of getting extensive health education sessions, skilled birth attendant and know quality of intrapartum care and report it positively 63,64 .
Those mothers who perceived distance from the health facility as not a big problem had higher likelihood of getting quality intrapartum care than those mothers who were perceived distance from the health facility as a  65 and Tanzania 66 revealed the same result.These findings indicated that improving access to health care services as well as distribution of health services, particularly in distant locations, should be prioritized.
The odds of quality of intrapartum care was higher among mothers who were from Tigray region compared with Afar region.The lower odds of quality intrapartum care in Afar region might be due to the reason that the region is highly affected by drought and disaster.Moreover the Afar region is lower socioeconomic status as compared to Tigray region 67 .This implies that strong attention is needed for disadvantaged areas of Ethiopia.
The study implied that out of the four mothers, only one was delivered at the health facilities, assisted by skilled birth attendants and initiated breast feeding within 1 h.This means that hundreds of mothers had not received all components of intrapartum care.Hence, to counteract the problem, the Ministry of Health and health facilities need to continuously evaluate and improve maternal health care services.

Strengths and limitations
Adequate sample size and nationally representative data was used.The study was considered both the individual and community level factors and accounted the hierarchical nature of the data.The study have the following limitations; first the cross sectional nature of the data causes to impossible to identify cause effect relationship.Some key factors such as labour, were not included in the analysis of quality intrapartum care where the level of quality intrapartum care may be low.

Conclusion
The finding revealed that less than one in four mothers had get quality intrapartum care which is minimal as compared to other studies.Individual level factors of mothers' primary education, rich household class, had more than two number of children were statistically significant factors of quality intrapartum care.Whereas, urban residence, distance as not a big problem and community level education from the community level factors were significantly associated with quality of intrapartum care.Therefore, to improve quality of intrapartum care, empowerment of mothers through extensive women education, and strengthening of ANC coverage, particularly for rural resident women are important recommendations.

Table 2 .
Distribution of quality intrapartum care across the community level factors in Ethiopia, EDHS 2016 (n = 4469).

Table 3 .
Multilevel analysis to identify individual and community level factors of quality of intrapartum care in Ethiopia, EDHS 2016 (n = 4469).Significant values are in[bold].Statistically significant at P value < 0.05.
AOR, Adjusted odds ratio; COR, Crude odds ratio; Null model, Without factors; Model 1, Adjusted for individual-level factors; Model 2, Adjusted for community-level factors; Model 3, Adjusted for both individual and community-level factors.

Table 4 .
A measure of variation for quality of intrapartum care in Ethiopia, EDHS 2016 (n = 4469).MOR Median odds ratio, PCV Proportional change in variance, ICC Intra class correlation coefficient.