Trends and Excess Risk of Place of Delivery for Under-five Mortality in Ethiopia: A Trend Analysis of the Ethiopian Demographic and Health Survey Data from 2000-2019

doi:10.21203/rs.3.rs-4828296/v1

Download PDF

Research Article

Trends and Excess Risk of Place of Delivery for Under-five Mortality in Ethiopia: A Trend Analysis of the Ethiopian Demographic and Health Survey Data from 2000-2019

https://doi.org/10.21203/rs.3.rs-4828296/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

The place of delivery plays a crucial role in the survival of neonates. Following the recommendation that giving birth at a health facility is safer than giving birth at home, health facility delivery increased from 26% in 2016 to 48% in 2019. However, neonatal mortality increased from 29 in 2016 to 33 per 1,000 live births in 2019 in Ethiopia. Hence, this study aimed to analyze the trends and disparities in risk differences according to place of delivery for mortality in children under five years of age in Ethiopia.

Methods

This study used five years of nationally representative demographic and health survey data. The Child and Birth Recode data files were accessed as electronic versions in STATA file format. A modified DHS_U5_rates shared code of chap08_CM was used to save the number of deaths and at-risk individuals for the eight age intervals. Then, we calculated the number of deaths, at-risks, and mortality rates for home- and facility-born children using the Stata command “collapse (sum)” for neonates, post-neonates, infants, children, and under-fives. The trends and disparities in mortality according to place of delivery are presented with line graphs and tables. The mortality difference (RD) at the place of delivery with a 95% uncertainty level was calculated using the “csi” STATA command.

Results

The excess risk difference of 12.7 under five mortality rates [RD=12.66, 95%CI: 3.27, 22.05] in 2016, 7.24 child mortality rate (CMR) [RD=7.24, 95%CI: 1.53, 12.95] in 2019, 14 infant mortality rate (IMR) [RD=13.95, 95%CI: 8.39, 19.52] for all survey years in total, 11 post-natal mortality rate (PNMR) [RD=11.22, 95%CI:6.44, 16] in 2016, and 7 PNMR [RD=7.34, 95%CI: 1.13, 13.56] in 2019 per 1000 live births might be attributable to home birth. However, the negative risk difference of -16.02 neonatal mortality rate (NMR) [RD= -16.02, 95%CI: -29.20, -2.84] in 2011 and -8.51 NMR [RD= -8.51, 95%CI: -16.23, -0.79] in 2016 per 1000 live births might indicate neonates born at home had reduced risk of death than those born at health facility.

Conclusion

The difference in the risk of excess U5MR, CMR, IMR, and PNMR per 1000 live births among children born at home might be attributed to their home delivery. These excess deaths would have been prevented if children born at home had the same risk of death as children born at health facilities. However, the difference in the risk of bias associated with NMR might indicate that neonates born at home have a lower risk of death than neonates born at health facilities; this possibility requires further investigation with contextual characteristics as well as access to emergency neonatal health care services.

Sexual & Reproductive Medicine

Trend

excess risk

place of delivery

U5MR

EDHS 2000–2019

Ethiopia

Globally, the mortality rate of children under five years old was reduced by 59%, from 93 in 1990 to 39 deaths per 1,000 in 2019[1]. However, children continue to face widespread regional disparities in their chances of survival. Regionally, sub-Saharan Africa continues to have the highest rate of under5 mortality in the world—estimated at 75·8 deaths per 1000 live births in 2019[2]. Newborn deaths now account for a greater and more recent share of all deaths occurring in children under five years of age [3].

Sub-Saharan Africa had the highest neonatal mortality rate in 2019, at 27 deaths/1,000 live births. The majority of all neonatal deaths (75%) occur during the first week of life, and 1 million newborns die within the first 24 hours [4]. Globally, child mortality declined by half between 2000 and 2019, but progress has slowed in neonates, and 65 (32%) of 204 countries, mostly in sub-Saharan Africa and southern Asia, are not on track to meet their SDG 3.2 target by 2030[5].

The place of delivery plays a crucial role in neonatal survival, as delivery outside a health facility is a risk factor for neonatal mortality, and giving birth in a health facility (not necessarily a hospital) is safer than doing so at home[6]. Moreover, current global health policies emphasize institutional deliveries as a pathway to achieving reductions in neonatal mortality in developing countries [7].

Despite this fact, according to the DHS comparative reports on the levels and trends in the use of maternal health services, Ethiopia (6%), Chad (13%) and Niger (18%) are the countries with the lowest numbers of births in health facilities [8]. In addition, births in health facilities are more common in urban areas than in rural areas. This difference is most apparent in sub-Saharan Africa. For instance, in Chad, Ethiopia and Niger, less than 10% of women in rural areas delivered their most recent child in a health facility[8].

According to the 2019 Mini-EDHS report, 48% of live births in the 5 years preceding the survey occurred in health facilities in Ethiopia. Despite such an impressive change in institutional deliveries, there was a slight increase in neonatal mortality from 29 in 2016 to 33 deaths per 1,000 live births in 2019 [9]. Referrals for complicated deliveries were low. Furthermore, among women who delivered at home and were referred for a complication, few were able to reach the next level of health facility[10], which might have an impact on the survival status of the neonates.

Evidence from different studies has shown no similar findings or conclusions with regard to the role of place of delivery in increasing or reducing the risk of mortality in children under five years of age. A study from Bangladesh reported that home as a place of child delivery played a significant role in decreasing underfive child mortality [11]. Moreover, home deliveries assisted by trained attendants do not significantly reduce the risk of early neonatal death compared with deliveries assisted by untrained attendants [12]. In support of these findings, studies from Ghana revealed that the risk of neonatal death was greater among hospital births than among home births[13], [20]. Additionally, studies from Nigeria [21] and Ethiopia [22] revealed that the risk of neonatal death was greater among neonates who were born at health facilities than among neonates who were born at home.

However, other studies have reported that when a mother delivers a child at home, the chances of neonatal mortality are greater than the mortality risks among children born at health facilities [7] [14] [15]. For instance, a systematic review reported that health facility delivery reduces the risk of neonatal mortality by 29% in low- and middle-income countries[15]. Similarly, another systematic review [7] and a study from Brazil [16] reported that health facility deliveries were significantly associated with reduced odds of neonatal mortality. In addition, a case‒control study from India reported that facility delivery combined with postnatal check-ups would have prevented approximately 33–34% of neonatal deaths, while facility delivery without postnatal check-up would have prevented less than 3% of neonatal deaths [16]. A study from China also revealed that the NMR, IMR and U5MR were negatively correlated with the hospital delivery rate [17].

Studies from different places also indicated that the magnitude and burden of U5M infection and neonatal deaths vary across regions for various associated reasons [23–26]. For example, a study from Nigeria revealed the existence of substantial regional disparities in childhood mortality [18]. A study from Kenya reported that there is significant regional variation in neonatal and postneonatal mortality [19]. Similarly, a study from Uganda revealed the presence of large regional variations in U5M rates ranging from 56–152 per 1000 live births [20]. A study also reported that region of residence was one of the determinants of underfive mortality in Chads [21].

Comparisons of infant and child mortality between populations, both within and between countries, are important because they highlight causative factors of mortality, which can be addressed by health policies and programmes seeking to reduce mortality at young ages[22].

In the past 25 years, researchers have built impressive frameworks for maternal and neonatal health in Ethiopia. This includes strategies such as the Making Pregnancy Safer, the Reproductive Health Strategy, the Adolescent & Youth Reproductive Health Strategy, the Revised Abortion Law & Free Service for Key Myomening & Child Health Services (Health Care Financing Strategy)[23–28] and Ccommunity-Based Newborn Care (CBNC) in Ethiopia, which is a national programme that aims to improve newborn survival through the Health Extension Programme (HEP)[10].

However, Ethiopia ranks 5th among the top 10 countries with the highest number of newborn deaths and newborn mortality rates in 2016, with 3% of all global newborn deaths and a 27.6 per 1000 live birth neonatal mortality rate [3]. Although under five and infant mortality rates declined from 123 per 1000 live births in 2000 to 55 in 2019 and from 77 in 2000 to 43 in 2019, respectively, the neonatal mortality rate increased from 29 in 2016 to 33 per 1000 live births in 2019[29].

Approximately 67% of all deaths of children under five years of age in Ethiopia occur before a child’s first birthday[9]. Infant mortality was 29%, and underfive mortality was 37% greater in rural areas than in urban areas (76 vs. 59 and 114 vs. 83 deaths per 1,000 live births, respectively). Infant mortality rates range from a low of 40 in Addis Ababa to a high of 101 in Benishangul-Gumuz, while underfive mortality rates range from a low of 53 to a high of 169 per 1,000 live births in the same regions [9, 30, 31].

The challenge facing maternal and child health programs in Ethiopia is accessing maternal and child health services, resulting in a nonhomogeneous place of delivery. Recognizing the regional and temporal variation in the excess risk of child mortality by place of delivery is more important than ever, given the regional and temporal disparities in underfive mortality in Ethiopia. This information is useful both for microtargeting future programmes and for identifying regions where progress has been made in investigating interventional efforts and sociodemographic and cultural conditions that may have contributed to the apparent progress.

Data source and sample

The data for the present study were obtained from five nationally representative consecutive Ethiopia Demographic and Health Surveys (EDHSs) conducted between February and May 2000 (EDHS, 2000), from April 27 to August 30, 2005 (EDHS, 2005), from December 27, 2010, to June 3, 2011 (EDHS, 2011); from January 18, 2016, to June 27, 2016 (EDHS, 2016); and from March 21, 2019, to June 28, 2019 (EDHS, 2019). The EDHS provides consistent and reliable estimates of fertility and childhood mortality levels; fertility preferences; use of family planning methods; maternal and child health; and other related indicators at the national and regional levels [9, 30–33].

The 2000 EDHS covered 14,642 sample households; 14,072 were interviewed, yielding a response rate of 99.3%, 15,716 eligible women were identified, and 15,367 were successfully interviewed, yielding a response rate of 97.8%[32]. The 2005 EDHS covered 14,645 sample households; 13,721 were successfully interviewed, yielding a response rate of 99%, 14,717 eligible women were identified, and 14,070 were successfully interviewed, yielding a response rate of 96%[33]. A total of 17,817 sample households were covered in the 2011 EDHS, 17,461 were successfully interviewed, yielding a response rate of 98%, 17,387 were eligible women in the 15–49 years age group, and 16518 were successfully interviewed, yielding a response rate of 95% [30]. The 2016 EDHS covered 18,008 sample households, and 16,650 were successfully interviewed, yielding a response rate of 98%, while 16,583 eligible women were identified; interviews were completed with 15,683 women, yielding a response rate of 95%[31]. The 2019 EDHS covered 9,150 sample households, and 8,663 were successfully interviewed, yielding a response rate of 99%, while 9,012 eligible women were identified; interviews were completed with 8,885 women, yielding a response rate of 99% [9].

A total of 25 years, starting from five years preceding the first 2000 EDHS year to the most recent 2019 survey year of analysis for this study, were used since Ethiopia started the EDHS (2000) in 2019 to obtain adequate data for trend analysis. Five full-scale DHS surveys were conducted in 2000, 2005, 2011, 2016 and 2019. Five questionnaires were used for the surveys: (1) the Household Questionnaire, (2) the Women’s Questionnaire, (3) the Anthropometry Questionnaire, (4) the Health Facility Questionnaire, and (5) the Fieldworker’s Questionnaire. These questionnaires, based on the DHS Program’s standard questionnaires, were adapted to reflect the population and health issues relevant to Ethiopia. The interviews were shortened substantially to collect data on indicators of particular relevance to Ethiopia and donors to child health programmes. The Women’s Questionnaire was used to collect information from all eligible women aged 15–49 years. The Ethiopian Central Statistical Agency (ECSA) used to collect the place of delivery, including age at death and date of child death within five years preceding the date of interview each survey year [9, 30–33].

Data processing and analysis

Data processing

The current trend analysis of the association between excess risk of delivery and underfive child mortality used primarily the birth record data file (ETBR41.dta to ETBR81.dta) and additionally the child recode data file (ETKR41.dta to ETKR81.dta). The Child and Birth Recode data files were accessed as electronic versions of the STATA file format, and unwanted variables and observations were dropped. This was done for each of the five surveys in line with the objective of this analysis. The Child Recode file was used to describe the trends and disparities in the place of delivery based on selected sociodemographic, maternal and child characteristics using a newly developed STATA do file. The do file was designed to produce weighted and unweighted frequency distributions of the place of delivery by the selected variables for the five consecutive survey years. To ensure that the birth record data are appropriate for the objectives of this analysis, the DHS_U5_rates shared code Chap. 08_CM (STATA do file) was obtained from the dhsmeasure of the gethub website[34]. Subsequently, the shared code (STATA do file) was modified to keep the variables of interest of the analysis, including place of delivery, child sex, place of residence of mothers, region of Ethiopia, educational status of mothers, year of interview, ANC visits and time of initiation of breastfeeding along with strata, cluster and weighting variables for each of the five consecutive survey years, separate from each other variables using the birth recode data file.

Data analysis

In the DHS_U5_rates shared code of the section that can be used to extract the total risk as a denominator for the specified age interval, we made important modifications to save the number (counts) of underfive child deaths and at risk for the respective eight age intervals, along with the variables of interest. Following the modification, execution of the code for the respective survey years generated a new dataset that was appropriate for extracting the number of deaths and at risk of neonates, postneonates, infants, children and underfive children by variables of analysis interest for each survey year. Subsequently, a new STATA do file was created and applied for calculating the number of deaths, number of at-risks and mortality rates for home- and facility-born children with the STATA command “collapse (sum)” by variables of interest for analysis using the newly generated EDHS data. The five individual survey datasets were then pooled into a single dataset using the Stata command to append the data. The data were manipulated to have a cross-sectional time-series or panel data format through sequential steps for the time series analysis.

Calculating the underfive mortality rates

During the calculation of the underfive child mortality rates, the procedure calculates eight conditional probabilities associated with the following age intervals, in months, given that the child survived to the beginning of the age interval.

i (age interval)	start_i (start age of the interval)	end_i (end age of the interval)	length_i (width)
1	0	1	1
2	1	3	2
3	3	6	3
4	6	12	6
5	12	24	12
6	24	36	12
7	36	48	12
8	48	60	12

In the section on extracting the total risk as a denominator for the specified age interval, we modified the STATA do file by adding the following command:

“local li=1

while `li'<=8 {

gen st_risk_`li'= stotal_risk_`li'

gen st_died_`li'= stotal_died_`li'

local li=`li'+1

save R_D_results.dta, replace

}”

This command assisted us in obtaining new STATA data concerning deaths and at risk. The ccomponent deaths and at risks are first saved. Then, the component deaths are

combined into the total number of deaths and at risk for each age segment. The component deaths and at-risks are calculated for age segments 0, 1–2, 3–5, 6–11, 12–23, 24–35, 36–47, and 48–59 months of completed age. Finally, to calculate

1. Neonatal mortality for home- and health facility-born neonates

2. Postneonatal mortality for neonates born at home and at health facilities

3. Infant mortality for home- and health facility-born infants

4. Child mortality for home- and health facility-born children

5. We used under five for home and health facility-born U5 children.

We used “Collapse (sum) died_HF risk_HF if age = = 1&Year = = 2000 [iweight = wt] according to the (V024)” STATA command.

Calculating the difference in mortality risk according to place of delivery

The risk difference (RD) or attributable risk (AR) is an absolute difference or excess risk measure. The underfive mortality risk difference (RD) or attributable risk (AR) at the time of delivery was calculated using simple arithmetic subtraction with the following formula programmed in STATA.

where RD (AR) = the risk difference (attributable risk) in place of delivery for mortality in children under five years of age

a = number of individuals who died from under five children and were born at home

b = number of at-risk individuals with five children and born at home

c = number of individuals who died from under five children and were born at a health facility

d = number of at-risk individuals under five children born at a health facility

Risk in the exposed = under five mortality rate in children born at home

Risk in unexposed = under five mortality rate in children born at health facility

Hence, the risk difference is calculated using the STATA command of “csi a b c d” or “csi number of died under five children and born at home number of died under five children and born at health facility number of at risk under five children and born at home number of at risk under five children and born at health facility”. As a result, one of the three possible results might be obtained:

1. A RD less than zero (negative risk difference), which indicates that exposed individuals (children born at home) have a lower risk of mortality than unexposed individuals (children born at health facilities);

2. When RD is equal to zero (zero risk difference), the exposure (home delivery) has no higher risk (has similar risk) than the unexposed (facility delivery) for underfive-child mortality and

3. A final RD greater than zero (positive risk difference) indicates that the exposed child (born at home) has a higher risk of mortality than the unexposed child (born at a health facility).

Calculating the attributable proportion of the place of delivery

The attributable proportion, also known as the attributable risk percent, is a measure of the public health impact of a risk/causative factor. The calculation of this measure assumes that the occurrence of mortality in the unexposed group represents the baseline or expected risk for that mortality. Furthermore, if the risk of death in the exposed group (children born at home) is greater than the risk in the unexposed group (children born at health facilities), the difference can be attributed to the exposure (home delivery). Thus, the attributable proportion is the number of deaths in the exposed group (mortality in children born at home) attributable to the exposure (home delivery). The risk profile represents the expected reduction in mortality if the exposure (home delivery) could be removed (or never existed) or if the patient had a risk profile similar to that of facility delivery. A measure of public health impact is used to place the association between an exposure and an outcome into a meaningful public health context. Whereas a measure of association quantifies the relationship between exposure and disease and thus begins to provide insight into causal relationships, measures of public health impact reflect the burden that an exposure (home delivery) contributes to the frequency of mortality in the population. One of the two measures of public health impact often used is the attributable proportion. It can be calculated using the following formula:

Trends in excess risk or mortality differences according to place of delivery

Temporal patterns are displayed by plotting in five-year intervals of home delivery and facility delivery under five mortality rates against consecutive survey years. The trend of mortality rates by place of delivery in the 25-year study period from five years preceding the first survey (2000) to the most recent survey (2019) was analyzed using the time series analysis graphic STATA command “twoway (tsline)”, subsequently declaring the data as a time series dataset. In this two-level model, underfive mortality rates by place of delivery for each survey year were referred to as level-1, and the level-2 mortality rate was considered for each region. The time-trend analysis was performed in two steps. In the first approach, all eleven subregions were pooled into one datum set. In the second step, all eleven subregions were analyzed separately. The trend of mortality rates for each place of delivery (home or facility) is presented with a line graph.

Ethics Statement

The study is based on publicly available data; therefore, no ethics statement is required for this work.

Trends in place of delivery -

Forty-eight thousand seventy-seven (48,077) underfive children, 84.79% (95% CI: 83.6, 85.91) of whom were born at home, were included in this stratified comparative trend analysis. Ten thousand eight hundred forty children, 94.98% of whom were born at home; 9,690 people, 94.69% of whom were born at home; 11, 505 children, 90% of whom were born at home; 10, 441 children, 73.69% of whom were born at home; and the remaining 5,601 children, 52.48% of whom were born at home, were from EDHS 2000, 2005, 2011, 2016 and 2019. The analysis of the trends and disparities in under five-child mortality by place of delivery indicated an increasing trend in institutional delivery from 5.02% in 2000 to 47.52% in 2019. As Fig. 1 clearly shows, there is a slowly increasing trend in the percentage of institutional delivery service utilization and a decreasing trend in home delivery, with a precise or narrow 95% certainty level, particularly for the five years preceding the first three consecutive (2000, 2005 and 2011) survey years. However, for the two consecutive (2016 and 2019) survey years, there was a significant increasing trend in the percentage of institutional delivery service utilization and a significant decreasing trend in home delivery, with a less precise or relatively wider 95% uncertainty level (Table 1 and Fig. 1).

Table 1

Trends in place of delivery with 95% uncertainty for U5 children, Ethiopia, in 2023.
EDHS Year	Home Delivery				Health facility delivery				Total
EDHS Year	N	Percent [95%CI]			N	Percent [95%CI]			Weighted (N)	unweight(N)
2000	11,625	94.98	[94.06	95.76]	615	5.02	[4.24	5.94]	12,240	10,840
2005	10,502	94.69	[93.84	95.42]	589	5.31	[4.58	6.16]	11,091	9,690
2011	10,627	90	[88.59	91.25]	1,181	10.00	[8.749	11.41]	11,808	11,505
2016	7,997	73.65	[70.72	76.39]	2,861	26.35	[23.61	29.28]	10,858	10,441
2019	2,842	52.48	[46.82	58.07]	2,574	47.52	[41.93	53.18]	5,416	5,601
Total	43,594	84.79	[83.6	85.91]	7,819	15.21	[14.09	16.4]	51,413	48,077

Underfive child mortality (U5M) (aged 0–4 years)

Trends and disparities in risk differences according to place of delivery for U5 mortality

The findings of the trend analysis revealed that home-born children had an increased risk of death compared to those born at health facilities for all survey years, except for 2005. The results showed that giving birth in a health facility could prevent additional deaths of 5.22, 4.95, 12.66 and 9.61 per 1,000 live births in the five years preceding the 2000, 2011, 2016, and 2019 survey years, respectively (Table 2 and Fig. 2).

On the other hand, approximately 4.2%, 7.05%, 21.44% and 17.44% of the U5M among home-born children might be attributable to their home birth, which would have been prevented if children born at home had a risk similar to that of children born at health facilities. The remaining 95.8%, 92.95%, 78.56% and 82.56% of the U5M in this group would have occurred during the five years preceding the survey years 2000, 2011, 2016 and 2019, respectively (Table 2 and Fig. 2).

Table 2

Trends and disparities in risk differences according to place of delivery for U5 mortality per 1000 live births, Ethiopia, 2023
EDHS Year	Home delivery			Facility delivery			Risk difference [95%CI]			Chi2	P value	Total
EDHS Year	died	at risk	Rate	died	at risk	Rate	RD	LB	UB	Chi2	P value	died	at risk	Rate
2000	1415	11416	124	71	597.6	118	5.22	-21.40	31.84	0.14	0.71	1996	12342	161.7
2005	950	10280	92.44	58	579.4	100.7	-7.75	-32.84	17.34	0.39	0.53	1382	11194	123.4
2011	730	10401	70.2	76	1165	64.95	4.95	-10.06	19.96	0.40	0.53	1051	11950	87.96
2016	464	7864	59.04	129	2783	46.2	12.66	3.27	22.05	6.26	0.01*	742	11040	67.21
2019	154	2796	55.18	115	2529	45.33	9.61	-2.12	21.33	2.55	0.11	329	5557	59.22
Total	3714	42757	86.86	448	7653	58.5	28.33	22.43	34.23	68.79	< 0.001**	5500	52083	105.6

*significant at P value < 0.05, ** significant at P value < 0.01

Child Mortality (Age 1–4 years)

Trends and disparities in risk differences according to place of delivery for child mortality

In this trend analysis, 47670 children aged 1–4 years were included. The analysis revealed that a significantly greater number of children died among those who were born at home than among those who were born at health facilities (833 versus 40) during the 25 years prior to the 2019 survey. The trend analysis revealed that home-born children had an increased risk of death compared to those born at health facilities for all survey years. In addition, birth at the health facility could prevent additional deaths of 13, 13, 14, 13 and 7 per 1000 live births in the five years preceding the 2000, 2005, 2011, 2016, and 2019 survey years, respectively. The excess risk of child mortality among home-born children was 28.89% (50 versus 37 per 1000 LB) in 2000, 41.27% (31 versus 19 per 1000 LB) in 2005, 66.82% (21 versus 6 per 1000 LB) in 2011, 70.8% (18 versus 5 per 1000 LB) and 62.42% (12 versus 4 per 1000 LB) in 2019 (Table 3 and Fig. 3).

Table 3

Trends and disparities in risk differences according to place of delivery for child mortality per 1000 live births, Ethiopia, 2023
EDHS Years	Home delivery			Facility delivery			Risk difference (95%CI)			Chi2	P value	Total
EDHS Years	died	At risk	Rate	died	At risk	Rate	RD	LB	UB	Chi2	P value	died	At risk	Rate
2000	363	7302	50	14	380.09	37	12.87	-6.71	32.46	1.28	0.26	809	10761.72	75
2005	205	6602	31	6	328.93	19	12.81	-2.24	27.87	1.74	0.19	529	10179.74	52
2011	145	6879	21	5	714.67	6	14.09	7.10	21.07	6.64	0.01*	349	11161.76	31
2016	102	5569	18	8	1496.40	5	12.97	7.86	18.07	12.94	0.00**	213	10395.15	20
2019	23	1982	12	7	1605.07	4	7.24	1.53	12.95	5.61	0.02*	71	5171.07	14
Total	838	28334.2	29.58	40	4525.16	8.80	20.74	17.37	24.10	64.51	0.00**	1971	47669.45	41
significant at P value < 0.05, * significant at P value < 0.01

Infant mortality (age 0–1 year)

Trends and disparities in risk differences according to place of delivery for infant mortality

The findings of the nonstratified comparative trend analysis revealed that health facility-born infants had an increased risk of death compared to those born at home for the first three consecutive EDHS report years. However, for the last two consecutive EDHS years, the results showed that giving birth at a health facility could prevent an additional 302 and 415 infant deaths per 100,000 live births in 2016 and 2019, respectively. The fraction of infant mortality among home-born infants attributable to home delivery was 65.5 per 1000 LB (46 versus 43 per 1000 LB) in 2016 and 88.5 per 1000 LB (47 versus 43 per 1000 LB) in 2019 (Table 4 and Fig. 4).

Table 4

Trends and disparities in risk differences according to place of delivery for infant mortality per 1000 live births, Ethiopia, 2023
EDHS Year	Home delivery			Facility delivery			Risk difference [95%CI]			Chi2	P value	Total
EDHS Year	died	at risk	Rate	died	at risk	Rate	RD	LB	UB	Chi2	P value	died	at risk	Rate
2000	1052	11415.78	92	56	597.58	94	-1.65	-25.63	22.33	0.02	0.89	1187	12341.88	96
2005	745	10279.80	72	52	579.43	90	-17.34	-41.16	6.48	2.42	0.12	852	11193.86	76
2011	585	10401.08	56	71	1164.86	61	-4.70	-19.13	9.73	0.43	0.51	702	11949.93	59
2016	363	7863.54	46	120	2782.58	43	3.03	-5.83	11.89	0.43	0.51	529	11039.95	48
2019	131	2796.35	47	108	2528.97	43	4.15	-6.96	15.26	0.53	0.47	258	5557.05	46
Total	2876	42756.55	67.26	408	7653.41	53.30	13.95	8.39	19.52	20.75	0.00**	3529	52082.67	68

*significant at P value < 0.05, ** significant at P value < 0.01

Post-Neonatal Mortality (aged 28 days to 1 year)

Differences in the risk of postneonatal mortality according to place of delivery

The findings of the nonstratified comparative trend analysis revealed that home-born postneonates had an increased risk of death compared to those born at health facilities for all EDHS reporting years except for the year 2000. The results showed that giving birth at a health facility could prevent additional postneonatal deaths of 2.80, 11.97, 11.22 and 7.34 per 1,000 live births in the 2005, 2011, 2016, and 2019 EDHS years, respectively. However, the difference in mortality risk was negative for the 2000 EDHS, indicating that neonates born at home had a reduced risk of death. The excess risk of postneonatal mortality among home-born postneonates was 75 per 1000 LB (37 versus 35 per 1000 LB) in 2005, 535 per 1000 LB (22 versus 10 per 1000 LB) in 2011, 564 per 1000 LB (20 versus 9 per 1000 LB) in 2016 and 441.8 per 1000 LB (17 versus 9 per 1000 LB) in 2019 (Table 5 and Fig. 5).

Table 5

Trends and disparities in risk differences according to place of delivery for postneonatal mortality per 1000 live births, Ethiopia, 2023
EDHS Year	Home delivery			facility delivery			Risk difference [95%CI]			Chi2	P value	Total
EDHS Year	died	At risk	Rate	died	At risk	Rate	RD	LB	UB	Chi2	P value	died	At risk	Rate
2000	495	10577.53	47	33	553.01	59	-12.87	-33.02	7.27	1.93	0.17	586	11697.80	50
2005	355	9571.89	37	18	525.03	35	2.80	-13.22	18.82	0.11	0.74	412	10658.72	39
2011	217	9702.90	22	11	1057.36	10	11.97	5.18	18.75	6.59	0.01*	255	11436.42	22
2016	153	7472.76	20	23	2578.08	9	11.22	6.44	16.00	14.87	0.00**	204	10720.06	19
2019	44	2647.56	17	22	2371.35	9	7.34	1.13	13.56	5.20	0.02*	75	5384.29	14
Total	1264	39972.64	31.63	107	7084.84	15.07	16.52	13.20	19.84	58.06	0.00**	1533	49897.29	31

*significant at P value < 0.05, ** significant at P value < 0.01

Neonatal mortality rate (younger than 28 days)

Differences in neonatal mortality according to pregnancy risk at the time of delivery

The trend analysis revealed that neonates born at health facilities had an increased risk of death compared to those born at home for all EDHS reporting years, except for the year 2000. The results showed that giving birth at home could prevent additional deaths—2078, 1602, 850 and 290 per 100,000 live births—in the 2005, 2011, 2016, and 2019 EDHS years, respectively. The difference in mortality risk (neonatal mortality rate among home-born children minus neonatal mortality rate among health facility-born children) was positive for the 2000 EDHS, indicating that neonates born at home had an increased risk of death. The percentages of neonatal mortality among neonates born at health facilities attributable to health facility delivery were 35.39% (59 versus 38 per 1000 LB), 31.11% (52 versus 35 per 1000 LB), 24.16% (35 versus 27 per 1000 LB), and 8.53% (34 versus 31 per 1000 LB) in the five years preceding each survey year of 2005, 2011, 2016 and 2019, respectively. For the 2000 EDHS year, the percentage of neonatal mortality among neonates born at home that was attributable to home delivery was 17.6% (49 versus 39 per 1000 LB) (Table 6 and Fig. 6).

Table 6

Trends and disparities in risk differences according to place of delivery for neonatal mortality per 1000 live births, Ethiopia, 2023
EDHS Year	Home delivery			facility delivery			Risk difference [95%CI]			Chi2	P value	Total
EDHS Year	died	at risk	Rate	died	at risk	Rate	RD	LB	UB	Chi2	P value	died	at risk	Rate
2000	556	11416	48.73	24	597.58	39.39	8.57	-7.65	24.79	0.91	0.34	601	12341.88	48.66
2005	390	10280	37.92	34	579.43	58.75	-20.78	-40.29	-1.28	6.31	0.01*	440	11193.86	39.32
2011	369	10401	35.46	60	1164.86	51.77	-16.02	-29.20	-2.84	7.53	0.01*	447	11949.93	37.40
2016	210	7864	26.71	98	2782.58	35.08	-8.51	-16.23	-0.79	5.29	0.02*	325	11039.95	29.46
2019	87	2796	31.00	86	2528.97	33.88	-2.90	-12.46	6.65	0.36	0.55	183	5557.05	32.89
Total	1612	42757	37.69	301	7653.41	39.35	-1.63	-6.34	3.08	0.47	0.49	1996	52082.67	38.32

*significant at P value < 0.05, ** significant at P value < 0.01

Trends and regional disparities in the mortality rate of children under five years of age (aged 0–4 years) stratified by place of delivery

As indicated by Fig. 2, the trend in U5CMR among children born at home was greater than that in children born at health facilities, particularly for the last three (2011, 2016 and 2019) survey years. However, the trend in U5CMR according to the place of delivery differed according to region and even according to the age of the underfive children, including neonatal, postneonatal, infant, child and underfive children (Fig. 2).

The trend analysis of underfive mortality by place of delivery showed that the U5CMR was consistently greater among those children who were born at home than among those who were born at health facilities for Tigray (except for the 2019 survey year), Harari, Addis Ababa and Dire Dawa throughout the five consecutive EDHS years. On the other hand, for the first three consecutive EDHS years, regions such as Afar, Amhara (except for the EDHS year 2000), Oromiya, and the SNNPR had higher U5CMRs among children who were born at health facilities than among those who were born at home. In addition, although child mortality had a declining trend overall for all regions, for most regions except for Addis Ababa, the mortality trend was either stable or slightly increasing for facility-born children, particularly for the last two consecutive EDHS years (Fig. 7).

Trends and regional disparities in the child (age 1–4 years) mortality rate by place of delivery

The trend analysis of child mortality by place of delivery showed that the CMR was consistently greater among those children who were born at home than among those who were born at health facilities for Tigray, Amhara, Oromiya, Harari, Addis Ababa and Dire Dawa throughout the five consecutive EDHS years. However, for the first three consecutive EDHS years, regions such as Afar, the SNNPR and Gambela had higher CMR among children who were born at health facilities than among those who were born at home. In addition, although child mortality had a declining trend overall for all regions, for the three regions, namely, Somali, Gambela and Harari, the mortality rate increased for facility-born children in the most recent EDHS year, 2019 (Fig. 8).

Trends and regional disparities in the infant (age < 1 year) mortality rate according to place of delivery

The trend analysis of infant mortality by place of delivery showed that the IMR was consistently greater among infants who were born at home than among those who were born at health facilities for Tigray, Harari, Addis Ababa and Dire Dawa throughout the five consecutive EDHS years. However, Afar, Amhara, Oromiya, Somali (except for the DHS 2000), and SNNPR (except for 2000 and 2019 EDHS) had higher infant mortality rates among those infants who were born at health facilities than among those who were born at home throughout the five consecutive EDHS years. In addition, although infant mortality had a declining trend overall for all regions, Somali, Benishangul-Gumze, Gambela and Dire Dawa had increased infant mortality trends irrespective of the place of delivery for the recent EDHS year 2019 (Fig. 9).

Trends and regional disparities in postneonatal (aged 28 days-1 year) mortality according to place of delivery

The trend analysis of postneonatal mortality by place of delivery showed that the PNNMR was consistently greater among those postneonates born at home than among those born at health facilities in regions of Somali Province, except in 2000; Afar, 2016 and 2019; Amhara, 2005; and Oromiya, 2000. According to this analysis, Benishangul-Gumze, the SNNPR, Harari, and Dire Dawa had higher postneonatal mortality rates among those postneonates who were born at home than among those who were born at health facilities throughout the five consecutive EDHS years. In addition, although postneonatal mortality has a declining trend overall in Ethiopia, postneonatal mortality has increased in Somali, Benishangul-Gumze, Gambela and Dire Dawa, irrespective of the place of delivery, in the recent 2019 EDHS (Fig. 10).

Trends and regional disparities in the neonatal (age < 28 days) mortality rate according to place of delivery

The regional disparities and trend analysis of neonatal mortality by place of delivery showed that the neonatal mortality rate was consistently greater among neonates born at health facilities than among neonates born at home in the Afar, Amhara, Oromiya and SNNPR regions. However, Addis Ababa had a higher neonatal mortality rate among neonates who were born at home than among neonates who were born at health facilities throughout the five consecutive EDHS years. The analysis also indicated that the neonatal mortality rates of neonates born at home and neonates born at health facilities exhibited similar trends in the Tigray and Somali regions. In addition, although neonatal mortality overall exhibited a declining trend in Ethiopia, except for Tigray, Addis Ababa and the SNNPR, all the other regions exhibited an increased neonatal mortality rate in the recent (2019) EDHS year, irrespective of the place of delivery (Fig. 11).

Trends and regional disparities in neonatal, postneonatal, infant, child and underfive mortality rates

The analysis of the regional disparities and trends in neonatal, postneonatal, infant, child and underfive child mortality rates clearly showed that all regions had similar declining trends, except for Somali, Ben-Gumuz and Gambella, which had increasing trends in recent (2019) survey years (Fig. 12).

The findings of this study showed that there was a slow increasing trend in the percentage of institutional delivery for the first three survey years, with a narrow confidence interval, and it rapidly increased with a wider confidence interval (low precision) for the last two survey years. The wider confidence interval for the point estimate of percentages of institutional delivery for the recent survey years might be due to the small sample size, particularly for the 2019 mini-EDHS report, compared to the previous survey years, which used larger sample sizes [9, 30–33].

The findings of the trend analysis revealed that children under five years of age born at home had an increased risk of death compared to those born at health facilities for all survey years, except for 2005. According to the findings of this study, 4.2%, 7.05%, 21.44% and 17.44% of the underfive mortality among home-born children might be attributable to their home birth, which would have been prevented if children born at home had a risk similar to that of children born at health facilities. The remaining 95.8%, 92.95%, 78.56% and 82.56% of the underfive child mortality in this group would have occurred during the five years preceding the survey years of 2000, 2011, 2016 and 2019, respectively. In addition, the analysis revealed that home-born children (aged 1–4 years) had an increased risk of death compared to those born at health facilities for all survey years. Furthermore, birth at the health facility could prevent additional deaths of 13, 13, 14, 13 and 7 per 1000 live births in the five years preceding the 2000, 2005, 2011, 2016, and 2019 survey years, respectively. The results also showed that giving birth at a health facility could prevent additional postnatal deaths of 2.80, 11.97, 11.22 and 7.34 per 1,000 live births in the 2005, 2011, 2016, and 2019 EDHS years, respectively.

The findings of this analysis for underfive (0–4 years old), child (1–4 years old) and postnatal (28 days–1 year old) mortality signify the implication of medium and long-term (distal) effects of being born at a health facility. Although underfive children are born at home, they can survive their early life (neonatal period) due to a normal pregnancy history and might miss essential child health care services, including nutritional counseling (for appropriate underfive child feeding practices), growth monitoring and immunization services. In addition, early screening and detection of vertically transmitted infections, including HIV/AIDS, might be unnecessary. Moreover, underfive children born outside the health facility might miss the opportunity for clean postnatal practices and follow up for any medical conditions that can be detected at birth or during the immunization period[35].

These findings support the findings of former studies reporting that home delivery increases the risk of neonatal mortality more than does health facility delivery [7] [14] [16], [18] and a systematic review reporting that health facility delivery reduces the risk of neonatal mortality by 29% [15]. However, these findings contradict those of a study from Nigeria on under age five survival times in which reported home-delivered children were less likely to die before age five than were their counterparts who were delivered in a health facility[36].

The findings of the analysis with regard to the neonatal mortality rate by place of delivery indicated that the percentage of neonatal mortality among neonates born at health facilities attributable to health facility delivery was 35.39% (59 versus 38 per 1000 LB), 31.11% (52 versus 35 per 1000 LB), 24.16% (35 versus 27 per 1000 LB), and 8.53% (34 versus 31 per 1000 LB) in the five years preceding each survey year of 2005, 2011, 2016 and 2019, respectively. This finding is supported by former findings that reported that home deliveries were associated with a lower risk of neonatal death than facility deliveries [37][20, 22].

This study revealed different scenarios regarding the trends and regional disparities in the risk of delivery for individuals under five years of age, children, infants, and neonates and neonatal mortality. For regions such as Tigray (except for the 2019 survey year), Harari, Addis Ababa and Dire Dawa, the U5CMR is consistently greater among those children who were born at home than among those who were born at health facilities throughout the five consecutive EDHS years. This might be due to the nature of the regions with respect to socioeconomic status and the urban‒rural mix of places of residence, which most of those regions are relatively urbanized. As a consequence, there is an expected improvement in socioeconomic status among urban residents and hence better outcomes at health facilities. However, even more urbanized individuals who gave birth at home still had a higher probability of underfive child mortality than did those who gave birth at health facilities. This, in turn, might indicate the presence of inequalities with respect to either socioeconomic status or access to maternal and child health care services. Reducing such inequalities and incorporating marginalized urban residents might prevent the excess risk of underfive child mortality due to giving birth at home.

On the other hand, during the first three consecutive survey years, the underfive child mortality rates were greater for those born at health facilities than for those born at home for regions such as Afar, Amhara (except for the EDHS year 2000), Oromiya, and SNNPR. This might be because these regions are relatively large regions with high numbers of rural dwellers, most of whom reach facilities in remote rural areas after maternal and fetal complications develop that could influence the survival of newborns. However, for the late two consecutive survey years, the reverse is true, which might have implications for better health service coverage and access to maternal and child health services relative to the first three survey years, including immunization for vaccine preventable diseases and early detection and treatment of childhood infections.

A different scenario was observed in the analysis of trends and regional disparities in neonatal mortality. The neonatal mortality rate is consistently greater among neonates born at health facilities than among neonates born at home in the Afar, Amhara, Oromiya and SNNPR regions. This finding supports the J-shaped correlation between age and mortality rate in Ethiopian patients written by Filature (an introductory epidemiology textbook). In general, the mortality rate at an early age, particularly at 0 to 4 years, is greater than that at 5 to 14 years, after which the death rate increases exponentially. In Ethiopia, surviving early life (early and late neonatal periods) irrespective of the place of delivery is a major challenge. Furthermore, with a rapid increase in facility delivery, especially during the most recent survey year, an increased risk of neonatal mortality might be expected among neonates born at health facilities. Evidence from different studies supports the finding that the neonatal mortality rate among neonates born at health facilities is greater for many reasons than that among neonates born at home[38], [39], [40, 41], and a study from Ghana strengthens the evidence that facility birth does not necessarily convey a survival benefit for women or babies and should be recommended only for facilities capable of providing emergency obstetric and newborn care and capable of safeguarding uncomplicated births[42]. A study from India also stated that institutional delivery was not protective against newborn mortality in districts with the poorest health system quality but was associated with decreased mortality in districts with higher quality [43]. Similarly, a study from Tanzania reported that overall hospital-based neonatal mortality rates increased from 2.6 in 2006 to 10.4 deaths per 1000 live births in 2015[44]. A study in sub-Saharan and Asian countries reported that widespread increases in facility delivery have not substantially reduced neonatal mortality in sub-Saharan Africa or South Asia over the past 2 decades. This may be due to the poor quality of care available in widely used primary care clinics [45].

According to this trend and regional disparity analysis of the association between excess risk at the time of delivery and underfive mortality, we found that the utilization rate of health facility delivery had an increasing trend, although it had a relatively low level of precision (wide 95% confidence interval), especially for the most recent (2016 and 2019) survey years. The sample size for those survey years, particularly for the 2019 mini-EDHS report, was very small compared to that of the previous survey years. Hence, central surgical authorities with ministries of health need to consider using a representative sample size for the next survey to improve the precision of institutional delivery service utilization estimates.

The difference in the risk of excess U5MR, CMR, IMR and PNMR per 1000 live births among children born at home might be attributed to their home delivery. These excess deaths would have been prevented if children born at home had the same risk of death as children born at health facilities. However, the difference in the risk of bias associated with NMR might indicate that neonates born at home have a lower risk of death than neonates born at health facilities; this possibility requires further investigation with contextual characteristics as well as access to emergency neonatal health care services.

The trend of underfive child mortality by place of delivery has an overall declining trend, with a higher rate of mortality in children born at home than in children born at health facilities throughout the five consecutive survey years except in the five years preceding the 2005 survey year. However, the trends and regional disparities in underfive mortality exhibited different trends. For regions such as Tigray, Harari, Addis Ababa and Dire Dawa, the underfive mortality rate among children born at home was greater than the mortality rate among children born at health facilities throughout the five years preceding the consecutive survey years. Hence, the respective Regional Health bureaus, along with the Ministry of Health, should design a facility delivery enforcing a maternal health service strategy with enhanced supportive supervision and follow-up.

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

G.H., conceptualization, methodology, formal analysis, and writing-original draft. G.A., and M.A. reviewed, edited and wrote the manuscript.

Acknowledgments

The authors would like to thank the measure DHS website (https://www.dhsprogram.com/data/) for allowing us to freely access the five consecutive years of Ethiopian demographic and health survey data.

Fund UNCs: Levels & Trends in child Mortality: Estimates Developed by the UN Inter-agency Group for Child Mortality Estimation. In.; 2019.
Sharrow D, Hug L, You D, Alkema L, Black R, Cousens S, Croft T, Gaigbe-Togbe V, Gerland P, Guillot M: Global, regional, and national trends in under5 mortality between 1990 and 2019 with scenario-based projections until 2030: a systematic analysis by the UN Interagency Group for Child Mortality Estimation. The Lancet Global Health 2022, 10(2):e195-e206.
UNICEF: Every Child Alive: The urgent need to end newborn deaths| UNICEF Publications| UNICEF. In.; 2018.
World Health O: Newborns: reducing mortality. Geneva: World Health Organization 2019.
Collaborators GBDU-M: Global, regional, and national progress toward Sustainable Development Goal 3.2 for neonatal and child health: all-cause and cause-specific mortality findings from the Global Burden of Disease Study 2019. Lancet 2021, 398(10303):870-905.
S Eswari VSaDB: Infant neonatal and post neonatal mortality in Madurai district, Tamil Nadu, India: A Geomedical study.
Chaka EE, Mekurie M, Abdurahman AA, Parsaeian M, Majdzadeh R: Association between place of delivery for pregnant mothers and neonatal mortality: a systematic review and meta-analysis. European Journal of Public Health 2020, 30(4):743-748%@ 1101-1262.
Wang W, Soumya Alva, Shanxiao Wang, and Alfredo Fort: LEVELS AND TRENDS IN THE USE OF MATERNAL HEALTH SERVICES IN DEVELOPING COUNTRIES. DHS Comparative Reports No. 26. In. Calverton, Maryland, USA: ICF Macro; 2011.
Ethiopian Public Health Institute (EPHI) [Ethiopia] and ICF. 2021. Ethiopia Mini Demographic and Health Survey 2019: Final Report. Rockville, Maryland, USA: EPHI and ICF. In.; 2019.
Berhanu D, Avan BI: Community Based Newborn Care Baseline Survey Report Ethiopia, October 2014. 2014.
Tabassum T, Bari W: Competing risks analysis of underfive child mortality in Bangladesh. Bangladesh J Sci Res 2014, 27(1):27-38.
Titaley CR, Dibley MJ, Roberts CL: Type of delivery attendant, place of delivery and risk of early neonatal mortality: analyses of the 1994–2007 Indonesia Demographic and Health Surveys. Health policy and planning 2012, 27(5):405-416%@ 1460-2237.
Dare S, Oduro A, Owusu-Agyei S, Mackay DF, Gruer L, Manyeh AK, Nettey E, Phillips JF, Asante KP, Welaga P: Neonatal mortality rates and characteristics of neonatal deaths in Ghana: analyses of data from two health and demographic surveillance systems. Global Health Action 2021, 14(1%@ 1654-9716).
Das U, Chaplot B, Azamathulla HM: The role of place of delivery in preventing neonatal and infant mortality rate in India. Geographies 2021, 1(1):47-62%@ 2673-7086.
Tura G, Fantahun M, Worku A: The effect of health facility delivery on neonatal mortality: systematic review and meta-analysis. BMC pregnancy and childbirth 2013, 13:1-9.
Fadel SA, Ram U, Morris SK, Begum R, Shet A, Jotkar R, Jha P: Facility delivery, postnatal care and neonatal deaths in India: nationally representative case‒control studies. PLoS One 2015, 10(10):e0140448%@ 0141932-0146203.
Zhang M, Qu H, Xia J, Hui X, Shi C, Xu F, He J, Cao Y, Hu M: Trends, influencing factors and prediction analysis of underfive and maternal mortality rates in China from 1991 to 2020. Front Public Health 2023, 11:1198356.
Antai D: Regional inequalities in under5 mortality in Nigeria: a population-based analysis of individual-and community-level determinants. Population health metrics 2011, 9:1-10.
Ikamari LDE: Regional variation in neonatal and postneonatal mortality in Kenya. African Population Studies 2013, 27(1%@ 2308-7854).
Nambuusi BB, Ssempiira J, Makumbi FE, Kasasa S, Vounatsou P: The effects and contribution of childhood diseases on the geographical distribution of all-cause underfive mortality in Uganda. Parasite epidemiology and control 2019, 5:e00089%@ 02405-06731.
Ahinkorah BO, Seidu A-A, Budu E, Armah-Ansah EK, Agbaglo E, Adu C, Hagan JE, Yaya S: Proximate, intermediate, and distal predictors of underfive mortality in Chad: analysis of the 2014–15 Chad demographic and health survey data. BMC Public Health 2020, 20:1-12.
Kibet MK: Comparative study of infant and child mortality: The case of Kenya and South Africa. African Population Studies 2010, 24(1-2%@ 2308-7854).
Health FDRoEMo: Strategy of Making Pregnancy Safer. In. Addis Ababa, Ethiopia: Ministry of Health; 2000.
Health FDRoEMo: Technical and Procedural Guidelines for Safe Abortion Services in Ethiopia. In. Addis Ababa, Ethiopia: Ministry of Health; 2006.
Health FDRoEMo: National Reproductive Health Strategy. In. Addis Ababa, Ethiopia: Ministry of Health; 2006.
Health FDRoEMo: National Adolescent and youth Reproductive health strategy. In. Addis Ababa, Ethiopia:: Ministry of Health; 2006.
Health FDRoEMo: Health Care Financing Strategy. In. Addis Ababa, Ethiopia: Ministry of Health; 2006.
Health FDRoEMo: Technical and Procedural Guidelines for Safe Abortion Services in Ethiopia. In. Addis Ababa, Ethiopia: Ministry of Health; 2013.
Ethiopian Public Health I, Icf: Ethiopia mini demographic and health survey 2019: key indicators. Rockville, Maryland, USA: EPHI and ICF 2019.
Central Statistical Agency [Ethiopia] and ICF International. 2012. Ethiopia Demographic and Health Survey 2011. Addis Ababa, Ethiopia and Calverton, Maryland, USA: Central Statistical Agency and ICF International. In.; 2011.
Central Statistical Agency (CSA) [Ethiopia] and ICF. 2016. Ethiopia Demographic and Health Survey 2016. Addis Ababa, Ethiopia, and Rockville, Maryland, USA: CSA and ICF. In.; 2016.
Central Statistical AE, Macro ORC: Ethiopia Demographic and Health Survey 2000. In. Addis Ababa, Ethiopia: Central Statistical Authority/Ethiopia and ORC Macro; 2001.
Macro O: Central Statistical Agency Addis Ababa, Ethiopia. Central Statistical Agency Addis Ababa, Ethiopia 2006.
Code Library, The DHS program [https://github.com/DHSProgram]
Blencowe H, Cousens S, Mullany LC, Lee ACC, Kerber K, Wall S, Darmstadt GL, Lawn JE: Clean birth and postnatal care practices to reduce neonatal deaths from sepsis and tetanus: a systematic review and Delphi estimation of mortality effect. BMC public health 2011, 11:1-19.
Egbon OA, Bogoni MA, Babalola BT, Louzada F: Under age five children survival times in Nigeria: a Bayesian spatial modeling approach. BMC public health 2022, 22(1):2207%@ 1471-2458.
Ijdi R-E, Tumlinson K, Curtis SL: Exploring association between place of delivery and newborn care with early-neonatal mortality in Bangladesh. Plos one 2022, 17(1):e0262408%@ 0261932-0266203.
Adjei G, Darteh EKM, Nettey OEA, Doku DT: Neonatal mortality in the central districts of Ghana: analysis of community and composition factors. BMC Public Health 2021, 21:1-14.
Mekonnen Y, Tensou B, Telake DS, Degefie T, Bekele A: Neonatal mortality in Ethiopia: trends and determinants. BMC public health 2013, 13:1-14.
Dare S, Oduro AR, Owusu-Agyei S, Mackay DF, Gruer L, Manyeh AK, Nettey E, Phillips JF, Asante KP, Welaga P et al: Neonatal mortality rates, characteristics, and risk factors for neonatal deaths in Ghana: analyses of data from two health and demographic surveillance systems. Glob Health Action 2021, 14(1):1938871.
Lohela TJ, Campbell OM, Gabrysch S: Distance to care, facility delivery and early neonatal mortality in Malawi and Zambia. PLoS One 2012, 7(12):e52110.
Gabrysch S, Nesbitt RC, Schoeps A, Hurt L, Soremekun S, Edmond K, Manu A, Lohela TJ, Danso S, Tomlin K et al: Does facility birth reduce maternal and perinatal mortality in Brong Ahafo, Ghana? A secondary analysis using data on 119 244 pregnancies from two cluster-randomized controlled trials. Lancet Glob Health 2019, 7(8):e1074-e1087.
Lee HY, Leslie HH, Oh J, Kim R, Kumar A, Subramanian SV, Kruk ME: The association between institutional delivery and neonatal mortality based on the quality of maternal and newborn health system in India. Sci Rep 2022, 12(1):6220.
Mangu CD, Rumisha SF, Lyimo EP, Mremi IR, Massawe IS, Bwana VM, Chiduo MG, Mboera LEG: Trends, patterns and cause-specific neonatal mortality in Tanzania: a hospital-based retrospective survey. Int Health 2021, 13(4):334-343.
Gage AD, Fink G, Ataguba JE, Kruk ME: Hospital delivery and neonatal mortality in 37 countries in sub-Saharan Africa and South Asia: An ecological study. PLoS Med 2021, 18(12):e1003843.

The authors declare no competing interests.

Download PDF

Version 1

posted

You are reading this latest preprint version

Trends and Excess Risk of Place of Delivery for Under-five Mortality in Ethiopia: A Trend Analysis of the Ethiopian Demographic and Health Survey Data from 2000-2019

Status:

Version 1

Abstract

Figures

Background

Methods

Data source and sample

Data processing and analysis

Data processing

Data analysis

Calculating the underfive mortality rates

Calculating the difference in mortality risk according to place of delivery

Calculating the attributable proportion of the place of delivery

Trends in excess risk or mortality differences according to place of delivery

Ethics Statement

Results

Trends in place of delivery -

Underfive child mortality (U5M) (aged 0–4 years)

Trends and disparities in risk differences according to place of delivery for U5 mortality

*significant at P value < 0.05, ** significant at P value < 0.01

Child Mortality (Age 1–4 years)

Trends and disparities in risk differences according to place of delivery for child mortality

Infant mortality (age 0–1 year)

Trends and disparities in risk differences according to place of delivery for infant mortality

*significant at P value < 0.05, ** significant at P value < 0.01

Post-Neonatal Mortality (aged 28 days to 1 year)

*significant at P value < 0.05, ** significant at P value < 0.01

Neonatal mortality rate (younger than 28 days)

*significant at P value < 0.05, ** significant at P value < 0.01

Trends and regional disparities in neonatal, postneonatal, infant, child and underfive mortality rates

Discussion

Conclusions

Declarations

Competing interests

Authors' contributions

Acknowledgments

References

Additional Declarations

Status:

Version 1