Epidemiology of neonatal mortality: a spatial and multilevel analysis of the 2019 mini-Ethiopian demographic and health survey data

Neonatal mortality is a global public health concern [1] and the problem is strikingly high in developing countries. Low-and-middle income countries, particularly, sub-Saharan African countries are suffering from a high burden of neonatal mortality and showed slow progress in neonatal mortality reduction [2]. Neonatal mortality is also unequally distributed and almost 80% of newborn deaths in 2016 were from the two regions; southern Asia accounted for 39%, and similarly, sub-Saharan Africa 38% of all such deaths [3]. In Ethiopia, according to the 2016 demographic and health survey report, neonatal mortality was 29 per 1000 live births. It was only about a 24% decrease over 5 years [4]. Amhara is the leading region in the country that suffered from a huge burden of neonatal mortality. Most of the deaths during the neonatal period also unevenly occur in the first week of life [3].

Neonatal mortality, particularly early neonatal death is mainly attributed to complications of preterm birth, infection, low birth weight, and intrapartum-related events [3, 5‐7]. Some large-scale studies [6, 8‐10] also identified sociodemographic factors such as lack of education, obstetric factors like perinatal care none/low utilization, maternal morbidity, neonatal factors like sex, multiple gestations, gestational age at birth, and morbidity were significantly associated with neonatal mortality. At large, neonatal mortality is associated with the quality of maternal and fetal care predominantly around the time of childbirth and the death is preventable as well as treatable with well-known and cost-effective interventions [3].

Studies revealed that several interventions were found to decrease neonatal mortality. Maternal care before and during pregnancy namely, preconception care, antenatal care, prenatal nutrition counseling, and iron and multi-micro-nutrient supplementation are proven interventions. Likewise, helping babies breathe and basic resuscitation, cord care, kangaroo mother care, and exclusive breastfeeding are significant interventions to reduce neonatal mortality [11‐14]. Similarly, community-level interventions [13, 15, 16] like participatory women’s groups and community health education intervention as well as improving access to high-quality antenatal and postnatal care found to decline neonatal mortality. Moreover, facility-level interventions [15, 17, 18] such as the availability of comprehensive emergency obstetrics, quality-improving interventions, and availability of maternity waiting homes are among significant factors to reduce neonatal mortality.

However, based on the current trends, 30 million deaths of newborn infants would occur between 2017 and 2030. About 80% would occur in south Asia and sub-Saharan Africa [3]. In the past 5 years in Ethiopia, no significant change in neonatal mortality was reported between 2016 and 2019 [19]. Yet studies done in the country are mainly devoted to neonates admitted in hospitals [20‐22] and a small-scale communities [23]. These studies potentially lack national representativeness. In addition, other studies which analyzed the 2016 Ethiopian demographic and health survey data [8, 24, 25], didn’t consider the spatial analysis and are not timely for future planning. Thus, investigating geographic locations of a high burden of neonatal death and identifying contributing factors based on current data could be supportive of interventions to achieve the Sustainable Development Goal (SDG).

Neonatal mortality is a worldwide public health issue and one of the leading health concerns in low-and-middle-income countries [1, 3, 31]. This study detected significant hotspot areas and identified community and individual-level factors that contributed to neonatal and early neonatal mortality in Ethiopia.

Neonatal mortality in Ethiopia is geographically clustered in wide areas of the Amhara and Afar regions and some areas of the Somali region and eastern zones of the Oromia region. Similarly, a higher geographic variation of neonatal mortality in the 2016 EDHS was reported and it was revealed that the rate was high in Oromia and Afar regions next to Amhara and Somali regions [4]. Neonatal mortality was found to be predicted by maternal education attainment, less antenatal attendance [32, 33], adolescent pregnancy [33], and home births [5]. This attribution could be informative for the aforementioned regions and areas of high neonatal mortality in Ethiopia. As indicated in reports, in the clusters of high neonatal death, more than 70% of mothers were not accessing any form of media (> 80% in the Amhara region), more than 80% of mothers didn’t complete elementary education, > 65% didn’t attend at least four visits of antenatal care, > 70% gave birth at home, and only < 30% of births were assisted by skilled attendants [4]. Also, in the Amhara region, most neonates die in the perinatal period due to prolonged labor and its complication, ruptured uterus [34]. Though neonatal mortality in Ethiopia showed a 41% reduction from 49/1000 livebirths in 2000 to 29/1000 livebirths in 2016, the regional distribution has remained higher in Afar, Amhara, Oromia, and Somali regions with the persistently highest rate (47/1000 live births) in the Amhara region [4].

In addition, low birth weight, neonatal morbidity, and maternal morbidity were found to be predictors of neonatal mortality in Ethiopia [35, 36]. Maternal nutrition and morbidities are also directly associated with low birth weight among newborn babies [37‐40] and nutritional intervention is recommended to avert low birth weight thereby neonatal mortality [41‐43]. In Ethiopia, studies revealed that pregnant mothers are suffering from malnutrition due to inadequate food diversity and household food insecurity [44, 45], and these further result in low birth weight and preterm birth [40, 46, 47] as well as directly in neonatal mortality [48]. Food insecurity is common in Ethiopia as revealed by studies [49‐51] and significant food insecurity was observed in lowland and highland areas than in midland areas [51].

Neonatal death is statistically significant among teenage mothers in Ethiopia. Mothers whose age ranges between 15 and 19 years were about four times and those between 25 and 29 years were nearly two times more likely to have neonatal death than elderly mothers. Studies [52‐58] also revealed an association between maternal age and neonatal death. There is evidence that teenagers during pregnancy face many obstetric complications like anemia, urinary tract infection, pregnancy-induced hypertension, and preterm births [59]. Among the common obstetric complications that teenagers experience, preterm birth alone increases the risk of neonatal mortality in this age group of mothers [58]. Higher neonatal mortality among younger-age mothers in this study could reflect that developing countries are still having higher neonatal mortality in the late Millenium Development Goals (MDG) and Sustainable Development Goals (SDG) era and might continue to have it in the future [3].

Mothers who didn’t attend formal education and those who didn’t complete primary education had close to three times more likely risk of early neonatal and neonatal mortality in Ethiopia. Maternal level of educational attainment was also found a predictor of neonatal death in other studies [56, 60]. In Ethiopia, female education could be affected by the deep-rooted practice of early marriage, which in turn is associated with lower education and economic dependence [61, 62]. Educated and employed mothers are more empowered to decide using maternal and childcare services [63]. It is suggested that educating girls and accessing maternal health services are among the key interventions to reduce neonatal mortality [64, 65].

Neonatal and early neonatal mortality is higher among mothers who gave male birth than female birth. Death among males was found to be one and half times higher in Ethiopia. Several studies conducted so far in Ethiopia similarly identified higher mortality among male neonates [23, 66, 67]. Similarly, as supported by other studies [67‐69] done in sub-Saharan Africa, mothers who gave twin birth had about four and more times higher likelihood of death than singleton births in this study. Higher death amongst twin births could be related to significantly higher preterm birth and anomaly, and lower Apgar and birth weight [70]. Whereas, though the sex variation of mortality is inconclusive, some studies explored more susceptibility to infection, prematurity, and poor perinatal conditions in male neonates than in females [71‐74].

Although the government of Ethiopia executed different health-related goals of the Millennium Development Goal (MDG) and still implementing the Sustainable Development Goal (SDG), neonatal mortality in the country remained high [19]. In the health sector transformation plan [75], Ethiopia was proposed to achieve over 90% coverage of 4+ antenatal care, deliveries attended by skilled health professionals, antiretroviral therapy for HIV-positive pregnant mothers, and postnatal care to reduce neonatal mortality to the level of 10/1000 live births in 2020. The proposed time to reduce neonatal mortality was passed, yet the Ethiopian neonatal mortality estimate for 2019 showed 27/1000 live births [76]. As revealed by some studies, among the individual interventions believed to improve neonatal health, such as health facility construction campaigns [77] and institutional delivery [78] contributed a little. Countries like Ethiopia should adapt an integrated evidence-based continuum of care approaches than a fragmented individual intervention. Therefore, as revealed and suggested by a recent Live Saved Tools Modeling study to estimate the number of deaths that would be prevented by 2035, countries that account for the vast majority of the world’s maternal and neonatal deaths and stillbirths can potentially avert the problem substantially by installing midwife-led sexual, reproductive, maternal, newborn, and adolescent interventions [79].

In this analysis, large data was used and statistically significant clusters of neonatal mortality in Ethiopia were identified. However, this analysis is mainly focused on determining the lifetime experience of neonatal mortality among mothers in Ethiopia. To understand the current magnitude and determinates of neonatal mortality, analyzing the data from a recent 5 years period would be helpful.

Neonatal mortality in Ethiopia is common and actionable socio-demographic and obstetric factors were found to predict it. Also, it is geographically clustered and Amhara and Afar regions are widely affected areas in the country. Policymakers should focus on evidence-based recommendations such as; midwives-led community and facility-based continuum of care from preconception through prenatal, intranatal, postnatal, and neonatal periods to possibly reduce neonatal mortality.