Abstract
Background Worldwide over 800 women lose their life each day from complication in pregnancy and child birth. Inequity by urban–rural residence is one of the most pronounced challenges in maternal health service coverage with women living in rural areas at a greater disadvantage than other women. This study aims to assess the magnitude and factors affecting the utilization of health facility delivery for the most recent live birth among women of reproductive age in rural Ethiopia.
Methods This is a cross-sectional study based on a data from Ethiopian Mini Demographic and Health Survey 2019 dataset with a total weighted sample of 2900 women of reproductive age group in rural Ethiopia. Data cleaning, coding and labeling were done using STATA version 14 software. Multilevel mixed effect logistic regression model was employed to identify associated factors.
Result Only 44% of reproductive-age women in rural Ethiopia gave their most recent live birth in health institutions. In the multivariable multilevel binary logistic regression analysis educational status, wealth index, marital status, household family size, attending 4+ANC, had ANC in the first trimester of pregnancy and ANC by skilled provider were found to be statistically significant factors associated with health facility delivery.
Conclusion In a rural part of Ethiopia, the prevalence of institutional delivery is low. Special emphasis should be given for mothers with no formal education, with poor household wealth index, married women, and mothers with more than five household family sizes. Furthermore implementing public health programs that target to enable women to have early first trimester Ante Natal Care, ANC by skilled provider and more frequent Ante Natal Care follow-up may increase the number of health facility deliveries.
Background
Based on recent evidence there were decline in number of women and girls who lose their life each year related to complications of pregnancy and childbirth, with a decline from 451,000 in 2000 to 295,000 in 2017. But Still, we are losing over 800 women each day in death from complications in pregnancy and childbirth(1). Despite all other reasons, low institutional delivery is one of the root causes of high maternal and newborn mortality(2).
Even though reducing global maternal mortality ratio (MMR) to lower than 70 per 100,000 live births is one of the Sustainable Development Goals(SDG) to be accomplished by 2030, maternal mortality mainly attributed to obstetric hemorrhage is still one of Africa’s leading public health challenge(3, 4). Despite there was good progress in reducing maternal mortality in Sub-Saharan African countries, there are the most off-track achievements of region-based maternal deaths, where the burden is still highest in rural women as compared to those urban women(5). Over two thirds (68%) of all maternal deaths globally occurs in Sub-Saharan Africa with around 200,000 maternal deaths a year or 533 maternal deaths per 100,000 live births (1).
In low-income countries, most newborn deaths occur at home(6), and in rural Ethiopia, nearly one in every ten (11%) of neonates die before celebrating their first month of life, mainly during the first week(7). Institutional delivery is one of the key strategies for reducing maternal mortality and for ensuring safe birth by reducing and intervening in any complications that will occur to the mother and her newborn during delivery and up to 24 hours postpartum (8, 9).
Even though addressing people who are more disadvantaged and have lower levels of health service utilization is one of the key parts of achieving SDG, inequalities by urban–rural residence are one of the most pronounced challenges in maternal health service coverage with women living in rural areas at a greater disadvantage than other women(10). For achieving the 2030 development goal of health facility delivery in sub-Saharan Africa narrowing the gap or inequity between the rural and the urban areas is one of the ways forwarded (11). Studies highlight that the urban-rural difference in institutional delivery was higher in East Africa especially the disparity is worst in the case of Ethiopia (12, 13). Although the Federal Ministry of Health of Ethiopia initiated a free delivery service policy in all public health facilities to encourage mothers to deliver in health facilities, utilization of institutional deliveries remains minimal with a pooled prevalence of 31% (2, 14). In Ethiopia, based on the most recent Ethiopian Mini Demographic and Health Survey(EMDHS) 2019 report seventy percent of live births in the 5 years before EMDHS2019 from urban women were delivered in a health facility while only forty percent of live births from rural women were delivered in a health facility(15). Therefore, highlighting important factors for the designing and implementing tailored public health interventions for improving institutional delivery in rural Ethiopia is needed.
Previous research has shown the magnitude and factors associated with institutional delivery in Ethiopia (14, 16-21), but as per the knowledge of the author, no study in Ethiopia investigates the determinants of health facility delivery of reproductive-age women in rural Ethiopia using nationally representative data. The very few studies conducted previously were either based on a small sample or a small segment of the population of rural Ethiopia. Therefor this study aimed to fill this gap by assessing the magnitude and factors affecting the utilization of health facility delivery for the most recent live birth among women of reproductive age in rural Ethiopia using data from the most recent EMDHS.
Methods
Study design, Data source, and Setting
This is a cross-sectional study conducted from October 10, 2022-December 29, 2022 using data extracted from the latest EMDHS 2019. The data were obtained from the Demographic and Health survey (DHS) website (https://dhsprogram.com/data/) after submitting a request justifying the aim of the study. The 2019 EMDHS is the second EMDHS and the fifth DHS conducted in Ethiopia from March 21, 2019, to June 28, 2019. The survey was implemented based on a nationally representative sample that provided estimates for the urban and rural areas at the national and regional levels. 8,885 women of reproductive age (age 15-49) were interviewed from a nationally representative sample of 8,663 households(15). Ethiopia is a country in the Horn of Africa with a total area of 1,100,000 km2 and lies between latitude 3° and 15° north and longitude 33° and 48 east(22). During the time of the survey, Ethiopia had nine ethnic-based and politically autonomous regional states and two cities (Addis Ababa and Dire Dawa).
Population and sampling procedure
This study used all women of childbearing age (15–49 years) with a live birth in the five years preceding the survey in rural Ethiopia. The most recent birth was considered for women with two or more live births during the five-year period. EMDHS 2019 used a two-step stratified cluster sampling method, in which sample households were selected in cluster enumeration areas (EAs). In the first stage, 305 EAs were selected (93 in urban areas and 212 in rural areas) with probability in proportion to EA size. In the second stage, a fixed number of 30 households in each cluster were selected. A household listing operation was carried out in all selected EAs from January through April 2019.Further information related to the population, study area, data collection, sampling procedure, and questionnaires used in the survey were detailed in the 2019 EMDHS Report(15). In the current analysis, as shown in the figure (Fig 1), a weighted total of 2900 mothers who resides in a rural part of Ethiopia were included.
Study variables
An outcome variable is place of delivery, which is dichotomized as a “health facility” (if a woman gives birth in public, private, or NGO health institutions) and a “Non-health facility” (if a woman gives birth either in home or any other places) (23).
The potential covariates considered to have an association with health facility delivery were chosen based on prior literature and based on the presence of the variable of interest in the 2019EDHS dataset (14, 19, 21, 24). These variables were woman’s age, woman’s educational status, wealth index, religion, household family size, sex of household head, mass media exposure, Had ANC,Visiting skilled providers during Ante Natal Care (ANC), history of giving birth to a boy or girl who was born alive but later died, frequency of ANC, and the timing of ANC.
Description and measurement of independent Variables
Age of respondents
the age of the women was re-coded into three categories with values of “1” for 15–24, “2” for 25-34, and “0” for 35 and above
Educational status
this is the minimum educational level a woman achieved with a value of “0” for no education, “1” for primary education, “2” for secondary education, and “3” for higher education.
Wealth index
The datasets contained a wealth index that was created using principal components analysis coded as “poorest”, “poorer”, “Middle”, “Richer”, and “Richest in the EMDHS data set.” For this study, it was recoded into three categories “poor” (includes the poorest and the poorer categories), “middle”, and “rich” (includes the richer and the richest categories)
Marital status
This was the marital status of women at the time of survey and recoded into three categories with a value of”0” for never in union,”1” for those married or living with partner, and “2” for those widowed, divorced and no longer living together/separated
Religion
The variable religion was recorded as orthodox, Muslim, protestant, catholic, traditional, and other in the dataset and we used it without change.
Sex of household head
the variable sex of household head was recorded as male and female in the dataset and we used without change
Having a son or daughter died
A composite variable obtained by combining if a woman has a son or daughter died with a value of”0” if a woman didn’t have a son or daughter died, and “1” if a woman has a son or daughter died.
Media exposure
A composite variable obtained by combining whether there was a radio and /or TV in the respondent’s household with a value of”0” if a woman didn’t has either TV or Radio in her household and “1” if a woman has access to either of the media
Household size
The family size of the women’s household re-coded into two categories with values of”0” for a family size greater than 5, and “1” for a family size of less than or equal to 5.
Region
Geopolitical features of regions were grouped in to three categories: Metropolitan for Harrar and Drie-Dawa, Large central for Amhara, Oromia, South Nations and nationalities and Tigray, and Small peripheral for Afar, Benishangule, Gambella, and Somalia.
Had ANC; if women reported any prenatal care, it is recorded as “Yes” if not “No”
Frequency of ANC visit
The number of ANC visits during pregnancy were categorized into two groups and recoded as 1 “yes “if a woman have greater than or equal to four ANC, and 0 “No “if a woman didn’t have greater than or equal to four ANC visit for the most recent live birth
Timing of ANC visit
The timing of ANC visits were categorized into two groups and recoded as 1 “yes “if a woman have ANC visit in the first trimester of her pregnancy to the most recent live birth, 0 “No “if a woman didn’t have ANC visit in the first trimester of their pregnancy to the most recent live birth.
ANC by skilled providers
A composite variable recoded as 1 “yes” if a woman received care from skilled providers, such as doctors, nurses/midwives, health officers, and health extension workers, and 0 “no “if she didn’t receive care from either of these professions during her pregnancy of the most recent live birth
Data management and analysis
After extracting the data from EMDHS 2019, further coding and descriptive analysis were done using STATA version 14. The DHS data was weighted using sampling weight, primary sampling unit, and strata before any statistical analysis to restore the representativeness of the survey and to tell the STATA to take into account the sampling design when calculating standard errors to get reliable statistical estimates. Due to the hierarchical nature of EMDHS data, women within the same cluster may be more similar to each other than women in the rest of the country. This violates the assumption of independence of observations and equal variance across clusters. This implies the need to use advanced models considering the between-cluster variability. Due to the dichotomous nature of the outcome variable logistic regression and mixed effect Logistic regression were fitted. Model comparison was done using Akaike’s information criterion (AIC) value, Bayesian information criterion (BIC) value, and Deviance Information Criteria (DIC)(25). A Mixed-effect model with the lowest AIC, BIC, and DIC were chosen (Table 1).
Furthermore, the Intra-cluster Correlation Coefficient (ICC) value was 0.47 which is in support of choosing mixed effect Logistic regression over the basic model. Variables with ≤0.2 p-values in the bi-variable analysis were fitted in the multivariable model to measure the effect of each variable after adjusting for the effect of other variables. Adjusted Odds Ratio (AOR) with a 95% Confidence Interval (CI) and p-value < 0.05 in the Multivariable model was declared as determinant factors associated with health facility delivery for the most recent live birth among women aged 15-49 in rural Ethiopia who had a live birth in the 5 years preceding the 2019 EMDHS. Multi-collinearity was also checked using Variance inflation factor (VIF), and a value of 10 was used as cut off.
Ethical Consideration
Since this study is based on publicly available secondary data, formal ethical approval was not required. The author obtained the data from Demographic and Health Surveys (DHS) website (https://dhsprogram.com/data/) after an online request to access the data set.
Ethical procedures were done by the institutions that funded, commissioned, and managed the surveys, and no further ethical clearance was required. ICF international approved that all the DHS surveys follow the U.S. Department of Health and Human Services rules for respecting of human subjects’ rights. Any personal identifiers like names, house numbers, and phone numbers were not included in the dataset. Furthermore as the study is secondary data analysis, gaining participants consent is not applicable to this study.
Result
Characteristics of study populations
This study includes a weighted number of 2900 reproductive aged women in rural Ethiopia, who gave birth in the last 5 years preceding the 2019 EMDHS, and was interviewed for their most recent live birth. The majority of the study participants (28.64) were between the age group of 25-34, and most of them (58.96 %) didn’t have formal education. Furthermore, only (23.9 %) of them had media exposure to (TV or radio). The household wealth quintiles of (52.54%) of women were poor and below. Regarding their marital status and religion most of them were married /living with partner s(94.88%), and orthodox follower (36%) in religion. More than half (54%) of the participants had a household family size of more than 5 individuals, and around 89% of them were from households headed by males Table 2.
The Magnitude of health institution delivery, and ANC Visit of the study population’s ANC visit for their most recent live birth
Only 44% [38.00, 50.15] of reproductive-age women in rural Ethiopia gave their most recent live birth in health institutions. Even though most of the participants 67.94% [62.78, 72.70] had started ANC visit in their first trimester of pregnancy, the majority of them 62.52% [58.51, 66.37] didn’t attend four or more ANC visit Fig 2.
Factors associated with Health Facility Delivery for the Most Recent Live Birth
Variables like had ANC at some point during pregnancy were excluded from the analysis because of multi-collinearity with having ANC from skilled provider. Variables like the sex of the household head, and having a son or daughter died were excluded from multivariable analysis since their p-value was greater than 0.2 at bi-variable analysis. In the multivariable multilevel binary logistic regression analysis educational status, wealth index, marital status, household family size, attending 4+ANC, had ANC in the first trimester of pregnancy, and ANC from skilled provider were found to be statistically significant factors associated with health facility delivery for the most recent live birth among women of reproductive age in rural Ethiopia.
The odds of giving birth at a health facility for the most recent live birth among women of reproductive age in rural Ethiopia adolescent girls with the educational status of primary, secondary and higher education were 1.80 (AOR = 1.80, 95% CI: 1.41 - 2.29), 3.67 (AOR = 3.67, 95% CI: 2.22 - 6.07), and 13.71 (AOR = 13.71, 95% CI: 4.35 - 43.22) times higher than women of reproductive age in rural Ethiopia with no formal education.
The probability of giving birth in health facilities increased as the household wealth index increased. The middle wealth quintiles were 1.51 (AOR=1.51, 95%CI: 1.14 - 2.00) times more likely to give birth in a health facility than those in the poor wealth quintiles. The rich wealth quintiles were 2.72 (AOR=2.72, 95%CI: 1.94 - 3.80) times more likely to deliver their most recent live birth in a health facility than those in the poor wealth quintiles.
Women who married/living with partner lower odds of giving birth in a health facility by 69% (AOR = 0.31; 95% CI: 0.10 - 0.99) as compared with never in union.
The odds of giving birth at a health facility was 1.28times (AOR = 1.28; 95% CI: 1.02 - 1.60) higher among women with household family size of less than or equal to five compared to women with household family size of above five.
Looking at the frequency of ANC visit women made for the most recent live birth in 5 years preceding the 2019 EMDHS, women who had more than four ANC visit had 2.12 times (AOR = 2.12; 95%CI: 1.68 - 2.69) higher odds of giving birth at a health facility as compared to their counterparts.
Mothers who had ANC visit in the first trimester for the most recent live birth were 1.45 times more likely to give birth in a health facility than women who didn’t have an ANC visit in the first trimester of their pregnancy (AOR=1.45.95%CI=1.13 - 1.85)
Moreover, the odds of giving birth at a health facility was 3.05 times (AOR = 3.05; 95% CI: 2.36 -3.93) higher among women who had ANC from skilled provider compared to women who did not had ANC from skilled provider Table 3.
Discussion
This study aimed to assess the magnitude and factors affecting the utilization of health facility delivery of the most recent live birth among women of reproductive age in rural Ethiopia using data from the most recent EMDHS 2019. According to this study, only 44% of reproductive-age women in rural Ethiopia gave their most recent live birth in health institutions. This is consistence with a study conducted in different parts of Ethiopia(19, 26, 27), and rural Haiti(28).This magnitude of institutional delivery is lower than a study conducted in northwest Ethiopia(8, 29), women in rural Ghana(30-32), and rural women in Nepal(33), and it’s higher than a study conducted in Nigeria(34). This might be because of the studies conducted in northwest Ethiopia were based on a small sample or small segment of a population of rural Ethiopia while the current study is based on representative data of the whole women of reproductive age in rural Ethiopia. And also it might be due to the differences in socio-cultural characteristics as well as difference in utilization of maternal health services like ANC service. In this study Majority of study populations 62.52% didn’t attend four or more ANC visits for the most recent pregnancy whereas a study conducted in Ghana reports that 67.9%, and 75% of women attend four or more ANC visit during their recent pregnancy (31, 32).
In multivariable multilevel logistic regression analysis educational status, wealth index, marital status, household family size, attending 4+ANC, ANC in the first trimester of pregnancy, and ANC by skilled provider were found to be statistically significant factors associated with health facility delivery for the most recent live birth among women of reproductive age in rural Ethiopia. Consistent with different studies conducted in Ethiopia (19, 35), Bangladesh (36), Ghana (31, 37), and Senegal (38)the probability of delivering in a health facility increases parallel with increasing women’s educational status. Women with Primary, secondary, and higher educational status had higher odds of giving birth in a health facility compared with women with no formal education. This might be because women with good educational status might have better information processing skills and improved cognitive skills that enable them to understand the purpose of health facility delivery and the risk of home delivery, which will result in the confidence to choose health facilities as a place of delivery (35).Moreover, women with good educational status might have a high chance of reading and understanding information about health facility delivery(37).
In this study wealth index is another most important variable significantly associated with giving birth in a health facility for the most recent live birth among women of reproductive age in rural Ethiopia. That is women with middle and higher household wealth indexes were more likely to report institutional delivery as compared with women with poor household wealth indexes. This finding is consistent with a study conducted in Ethiopia (19, 35, 39), Uganda (40), India (41), Ghana (37, 42), and Cambodia (43). Such discrepancy associated with wealth status might be due to the cost of transportation and any other extra cost associated with giving birth in health institutions (39).In addition, women with poor household wealth status might have a low educational status that in turn affects their decision to give birth in health facility.
The household family size had a significant association with giving birth at a health facility in which women from household family size of less than five have higher odds of giving birth in a health facility. The potential justification for this discrepancy might be as a consequence of a small number of family size women might have a good economy to cover the transportation and related costs while traveling to health facilities.
Moreover, in this study the frequency of ANC visit women made for the most recent live birth was significantly associated with the place of delivery, meaning that women who had more than four ANC visit had a higher probability of giving birth at a health facility as compared to their counterparts. This is in line with the previous study conducted in Ethiopia (8, 24, 35), and Ghana (31). This might be due to the exposure of women with frequent ANC to repeated counseling about birth preparedness and complication readiness that can encourage mothers to deliver at a health facility(8).
Furthermore, Mothers who had ANC visit in the first trimester for the most recent live birth were more likely to give birth in health facility than women who didn’t have ANC visit in the first trimester. This is consistent with a study conducted in Ghana (32), this might be due to the opportunity women got to have frequent contact with health professionals that will enable them to get adequate information on the benefit of giving birth in a health facility to themselves and their newborn’s health, as well as the women might acquire good awareness about the possible complications related with home delivery (32).
One of the strengths of this study was its trial to fill the gap of equity by addressing rural women by using large population-based data with large sample size, so it can be generalized to all women of reproductive age group in rural Ethiopia, and it will help as a baseline information to provide audience specific/tailored public health interventions in rural Ethiopia. Furthermore, the use of advanced statistical methods capable of accommodating the hierarchal nature of DHS data is also strength. This study might have limitations. First, since the author used secondary data some potentially important predictors were not available like distance from a health facility, knowledge, and attitude towards health facility delivery. Secondly, EMDHS 2019 was a questionnaire-based survey and asked women about their live births for the past five years before the survey, so recall bias might be the other limitation, but the author tried to minimize this by considering only the most recent live birth with in the past five years of the survey.
Conclusion
In a rural part of Ethiopia, the prevalence of institutional delivery is low. Health facility delivery among reproductive age women of rural Ethiopia was significantly associated with educational status, wealth index, marital status, household family size, attending 4+ ANC, having ANC visits in the first trimester of pregnancy and ANC by skilled provider. Thus, especial emphasis should be given to those mothers with no formal education, with poor household wealth index, and mothers with more than five household family sizes. Furthermore implementing public health programs that target to enable women to have early first trimester ANC with more frequent ANC follow-up, and ANC by skilled provider may be an effective way to increase the number of health facility deliveries.
Data Availability
The data underlying the results presented in the study are available from https://dhsprogram.com/data/
Acknowledgments
The author would like to extend his acknowledgment to the measure DHS for providing the data.
Footnotes
This version of the manuscript has been revised to update the categorization of the Independent variable "Had ANC in the first trimester of pregnancy", and the model was re-fitted after fixing the problem associated with this variable.Following this the author would like to inform readers of this paper to consider some important changes through out the whole manscrpit.Also,Table 1,Table 3 and Figure 2 has been revised.
Abbreviations
- ANC
- Ante Natal Care
- AOR
- Adjusted Odds Ratio. CI: Confidence Interval
- COR
- Crude Odds Ratio
- DHS
- Demographic and Health Survey
- EMDHS
- Ethiopian Mini Demographic and Health Survey
- SDG
- Sustainable Development Goal