Background
Globally, three-fourths of neonatal mortality occurs in the first week of life [
1]. A high number of Early Neonatal Mortality (ENM) occurs among Low Birth weight (LBW) neonates [
2]. These neonates are 20 times more likely to die than normal birth-weight neonates [
3,
4]. Early Mortality among LBW neonates is continued to be a major public health problem, particularly in middle-income, and low-income countries [
5]. According to World Health Organization (WHO) report, in 2020, above 2.4 million neonates died annually, and 80% of them are LBW [
5,
6]. The first week of life is a critical period for the survival of LBW neonates. Of the total neonatal deaths, 75% occurs during the first weeks of life, with one-third of deaths occurring on the first day of birth [
1,
6]. Specifically in Southeast Asia, the mortality of LBW neonates ranges from 19 to 31% [
7], and in Sub-Saharan Africa, the mortality of these neonates ranges from 21% in females to 23.9% in males [
8].
The incidence of ENM among LBW varies across the country with the highest incidence of ENM occurring in low-income countries, for instance in Bangladesh it was 112 per 1,000 neonate day observations, in Brazil 52 per 1,000 neonates [
9], in India 21.22 per 1000 live births [
10], Burkina Faso 1.93 per 1000 neonate/day [
11], and in Bahir Dar, Ethiopia 22.16 per 1000 neonate day observation [
12].
Ethiopia aimed to reduce the Neonatal Mortality Rate (NMR) from 29 to 1000 live births in 2016 [
13] to 11 per 1000 live births in 2019/2020, including LBW neonates [
14]. Despite the government’s aim of reducing Neonatal Mortality (NM), paradoxically it had been increased to 33/1000 live births in 2019, from this, the mortality of LBW neonates is a major public health problem [
15]. In Ethiopia, the mortality of LBW neonates ranges from 11.0 to 37.8% [
12,
16‐
18]. Overall, early neonatal mortality in the Amhara region is 43 per 1000 live births, which is the highest figure in the country [
13].
These studies suggested that the predictors of mortality among LBW neonates are maternal age [
12], Antenatal care(ANC) follow-up [
16], neonatal age, residence, sex, type of pregnancy [
19,
20], birth weight [
9,
19,
21], sepsis [
12,
18,
22], perinatal asphyxia [
21,
23,
24], Respiratory Distress Syndrome(RDS) [
12,
18,
25,
26], prematurity [
12,
16,
25,
27‐
29], Necrotizing Enter Colitis(NEC) [
12,
26,
30,
31], 1st, and 5th minute (Appearance, Pulse, Grimace, Activity, Respiration) (APGAR) score [
19,
32], place of delivery, preeclampsia [
33,
34], maternal diabetes mellitus (DM) [
17,
25], and maternal Human Immunodeficiency Virus(HIV) infection [
13,
35]. Nevertheless, the incidence and predictors of mortality among LBW neonate, specifically in their first week of life is not fully addressed.
To tackle this problem, and increase the survival of LBW neonates, efforts have been made globally as well as nationally. For instance, WHO is working with the Ministry of Health (MoH) and other partners to expand the quality of services for LBW babies [
36]. Interventions like the provision of adequate nutrition for adolescent girls were done to reduce mortality among LBW neonates [
37]. In addition, the Sustainable Development Goal (SDG 3) focuses on reducing neonatal mortality to below 12 per 1000 live births by 2030 [
38]. Furthermore, Ethiopia accepted initiatives to decrease the mortality of LBW neonates [
39], and currently, planned to reduce neonatal mortality from 33 to 1000 live births to 21 per live births by the year 2024/2025 [
40].
Despite these efforts and strategies, neonatal mortality remains high [
35], and increased from 29 to 33 per 1000 live birth since 2016 [
13,
35], Despite this problem, the previous studies conducted in Ethiopia have not focused on the critical period of life before 7 days among LBW neonates, and, there is limited evidence on this subgroup of neonates in the critical period of life regarding the incidence of ENM and its predictors among LBW neonates in the study setting. Therefore, estimating the incidence, and identifying predictors among high-risk groups in the critical period of life are the cornerstones to improve LBW neonate survival, progress toward the national plan, designing appropriate interventions, and providing valuable clinical information for health professionals. Hence, this study aimed to assess the incidence and predictors of mortality among LBW neonates in the first week of life admitted to Comprehensive Specialized Hospitals of Northwest Ethiopiain.
Methods
Study setting, design, and period
An institution-based multi-center retrospective follow-up study was conducted using data of LBW neonates from March 21, 2020 to March 1, 2022 (the data were extracted from May 1 to June 26, 2022) in Neonatal Intensive Care Unit (NICU) of Northwest Ethiopia (West Amhara) Comprehensive specialized hospitals. Comprehensive Specialized Hospitals in the Northwest part of Ethiopia (West Amhara region) are the University of Gondar (UoG), Felege hiwot, Tibebe Ghion, Debre Tabor, and Debre Markos. UoGCSH is found in Gondar town, and this hospital has an average annual admission of 726 LBW neonates and 528 of them were admitted in the first week. Felege hiwot and Tibebe Ghion CSH are found in Bahir Dar city. Felege Hiwot CSH has average annual admission of 720 LBW neonates and 504 of them were admitted in the first week. Tibebe Ghion has an average annual admission of 690 LBW neonates, and 484 of them were admitted within the first week. Debre Tabor CSH is found in Debre Tabor town. This hospital has an average annual neonatal admission of 512 LBW neonates, and, 408 of them were admitted in the first week. Debre Markos CSH is found in Debre Markos town, and this hospital has an average annual admission of 612 LBW neonates and 444 of them were LBW early neonates. These Hospitals have NICU with mixed health professionals (neonatal and comprehensive nurses, general practitioners, pediatricians, and other staffs). The major services in the NICU include general neonatal care services, blood and exchange transfusion, phototherapy, and ventilation support such as Continuous Positive Air Pressure (CPAP).
Study population
All LBW neonates in the first week of life admitted to the NICU of Northwest Ethiopia Comprehensive Specialized Hospitals were considered as the source population, and all low birth weight neonates in the first week of life admitted to the NICU of Northwest Ethiopia Comprehensive Specialized Hospitals from March 21, 2020 to March 1, 2022 were taken as a study population.
Eligibility criteria
LBW neonates in their first week of life who were admitted to NICU from March 21, 2020, to March 1, 2022, were included, Whereas, Neonates with an incomplete chart (outcome status and charts with one of the following predictor is missed (gestational age, birth weight, neonatal age, sex, type of pregnancy, place of delivery, date of admission, and discharge)) were excluded.
Sample size estimation and sampling technique
The sample size was determined using STATA 14, using the Cox proportional hazard model by considering the following assumptions: confidence level (95%), Power (80%), a 28.35% Probability of an event, and 1.483 hazard ratio of the predictor (maternal age above 35) [
12], 5% Margin of error, and 10% probability of withdrawal (chart attrition). Accordingly, the final estimated sample size was 793 LBW neonates chart. Within two years, the hospitals had a total of 4752 LBW early neonate admissions. Accordingly, the total sample size was proportionally allocated to each hospital, and the neonate’s chart was selected by a simple random sampling technique using a computer-generated random number in Excel.
Variables of the study
The dependent variable of the study was mortality in the first week of life, whereas the independent variables include (1) socio-demographic variables such as (maternal age, sex of the neonate, age of the neonate, place of delivery, residence). (2) Maternal medical factors such as maternal DM, HTN, HIV/AIDS, TB, Anemia, and other illnesses (STI, UTI). (3) Maternal obstetric factors such as (type of pregnancy, gravidity, parity, PROM, APH, mode of delivery, preeclampsia, ANC follow-up, corticosteroid administration), and (4) Neonate’s clinical related variables such as (PNA, RDS, sepsis, jaundice, hypothermia, hypoglycemia, NEC, 1st & 5th minute APGAR score, breastfeeding, congenital anomalies, IUGR, birth weight, gestational age).
Operational definition
Data abstraction checklist and procedures
The data abstraction checklist was adapted by reviewing different relevant literature, and guidelines [
8,
12,
17,
19,
21,
42‐
44]. It contains four parts: socio-demographic characteristics of mothers and neonates, maternal medical disorders, obstetric, and gynecological factors, neonate’s clinical-related variables of LBW neonates, and outcome measures. The data were retrospectively collected by 5 experienced BSc nurses and supervised by 4 experienced BSc nurses. The medical registration number of LBW neonates who were admitted in the early neonatal period from March 21, 2020 to March 1, 2022 was first obtained from the FMOH registration book. Then, the required number of neonates’ medical charts was selected by a simple random sampling technique according to eligibility criteria.
Data quality control
The data abstraction checklist was evaluated by the expert researcher, and then a pretest was done on 5% of the sample size in a similar setting, and necessary amendments were done. One-day training was given to data collectors and supervisors focusing on the purpose of the study, the data collection tool, data collection methods, and ethical concerns during the data collection period. All the collected data were checked for their completeness and consistency by the data collectors and supervisors daily.
Data processing and analysis
The data were cleaned, coded, and entered into Epi-Data version 4.6.0.6, and the analysis was done using STATA version 14 statistical software. The outcome was dichotomized into death, coded as “1”, and censored coded as “0”. The variance inflation factor was used to assess multi-collinearity. The incidence rate of mortality was calculated for the entire follow-up by dividing the number of new cases of mortality of LBW early neonates by the total person-days of follow-up. Kaplan Meier failure curve was used to estimate the time to death during the follow-up. The log-rank test was employed to compare statistical differences between groups of independent variables. The Schoenfeld residual global test for proportionality assumption was checked, and its (P-value = 0.4282) was used as a suggestive of satisfying the assumption. The log-likelihood and Akaike Information Criteria (AIC) were applied to select the best-fitted model, and a model with minimum AIC was considered a fitted model. Based on this the Weibull regression model with the (AIC = 1196.251) value was the fitted model. The goodness of fit of the model was also checked using the Cox-Snell residual test, and it was close to the bisector. Variables with P-value < 0.25 in the bi-variable analysis were entered into the multivariable Weibull regression model analysis. Crude and adjusted hazard ratios with 95% CI, were used to determine the strength of the association. In multivariable analysis variables having a p-value < 0.05 were considered as predictors of mortality.
Discussion
The overall aim of this study was to assess incidence and predictors of mortality among LBW neonates in the first week of life admitted to the NICU of Northwestern Ethiopia Comprehensive Specialized Hospitals. This study showed that at the end of the follow up 32.46% (95% CI 29.12% -35.8%) of LBW neonates died. This finding was in line with a study conducted in South Africa (32%) [
22], and Brazil (29%) [
19]. The possible reason could be both of the study areas where referral hospitals in which different complicated neonatal cases can be referred from different corners of the country. So, due to a high number of complicated cases, the mortality could be increased. On the other hand, this result was higher than a study conducted in Tanzania (5%) [
45], and India 21.2% [
21]. This marked difference might be attributed to NICU setup, differences in study design, and inclusion criteria. Those studies from other countries were conducted in a cross-sectional study design. The other reason is those studies include only inborn neonates, but the present study included both inborn and outborn neonates. However, the result of this study was lower than a study conducted in Saudi Arabia (41%) [
46]. This difference might be due to the characteristics of the study participants, because they included only extremely low birth weight, and premature neonates [
21].
In this study, the overall incidence of early neonatal mortality was 75.63 per 1000 neonate day observations (95% CI: 66.76–85.67). This finding was lower than a study conducted in Bangladesh with 112 per 1000 neonate day observations with (95% CI, 91–136) [
47]. This variation might be due to the way of follow-up. Those studies start to follow from the time of birth and in their home up to the end of follow-up, but in the present study, the follow-up starts from admission to NICU and did not include neonates who died immediately after birth, and in their homes, so this may underestimate the overall incidence of ENM. In contrast to these, the finding of this study was higher than a study conducted in Brazil 52 per 1000 neonates [
9], India 21.22 per 1000 live births [
10], Burkina Faso 1.93 per 1000 neonate /day (95% CI, 1.2–3.1) [
11], and Bahir Dar 22.16 (95% CI14.14-34.75) per 1000 neonate day observation at the seventh day [
12]. The variation between Brazil, Burkina Faso, and the present study might be due to differences in inclusion criteria, quality of service provision, and NICU setup differences. Those studies included LBW neonates who were not admitted, whereas the present study included only admitted neonates. Additionally, hospitals in Brazil and Burkina Faso may have more advanced care and service provision, and their NICU setup may be more advanced than hospitals of our study setting.
In this study, the incidence of mortality on the first day was high, with an incidence of 105.12 (95% CI; 84.44 -130.88) per 1000 neonate days’ observation. This finding was nearly three times higher than a study conducted in Bahir Dar, Ethiopia 38 per 1000 neonate day observations [
12]. The possible reason for this discrepancy might be that, in the present study, the proportion of LBW neonates with RDS and LBW neonates from preeclampsia mothers was higher than a study conducted in Bahir Dar. Additionally, it could be due to differences in neonates’ age that include only neonates who were in critical time (the first week of life). The other possible explanation might be related to pregnancy and birth-related complications that may result in a delay in the diagnosis and management of neonatal problems among early neonates.
Concerning predictors of mortality, this study identified several predictors of mortality. Accordingly, LBW neonates born from a mother with preeclampsia were nearly two times more likely to die than neonates whose mothers did not have preeclampsia. This result was consistent with a study done in Greece [
34], and China [
33]. This is due to a decrease in uteroplacental blood flow, and then, the neonates develop hypoxia and IUGR which could increase the risk of dying [
48]. Additionally, preeclampsia can be a threat to maternal life, and in this case, terminating the pregnancy could be mandatory, and therefore, the neonate would be delivered with immature organ systems, which increases the risk of death.
In addition, LBW neonates with perinatal asphyxia (PNA) had a higher risk of early death as compared with neonates without PNA. This result was consistent with studies done in Brazil [
24], India, and Bangladesh [
21,
23]. The possible reason for this could be the PNA contribut of the deprivation of oxygen in the body, which leads to progressive hypoxemia, and hypercapnia resulting in the central nervous system and other end-organ damage [
42]. In the current study, the hazard of death among neonates with RDS was 1.76 times higher as compared with its counterpart. This finding was in line with a study conducted in Brazil [
49], Zimbabwe [
25], India [
21,
44], and Bahir Dar, Ethiopia [
12]. Since the majority of neonates are premature, the problem of lung immaturity (lack of adequate surfactant) is a common phenomenon that leads to lung collapse, and respiratory failure [
42]. Furthermore, it might be due to a lack of administration of antenatal corticosteroids to mothers before 37 weeks of gestation, lack of surfactant replacement therapy [
50], and due to inadequate number and lack of trained personnel in the use of Nasal Continuous Positive Airway Pressure (CPAP). Similarly, the hazard of death among neonates with necrotizing enter colitis (NEC) was 2.78 times higher than neonates who did not have (NEC). This result is supported by a study conducted in Brazil [
26], China [
30], South Africa [
31], and Bahir Dar, Ethiopia [
12]. The reason for this could be the majority of the study participants were preterm, and all of them were LBW,these groups of neonates have immature gastrointestinal tract and immature defense mechanisms, so they are more susceptible to infection, and as the intestinal tracts colonize by microbes it could easily perforate and develops peritonitis [
42]. In addition to the above reasons, neonates with NEC have intestinal necrosis and dehydration, and this may increase the chance of death. In the present study, neonates with birth weights, less than 1000 g and 1000–1499 g had three times and two times, higher risk of death as compared with neonates with a birth weight of 1500 to 2499.9 g respectively. This finding was consistent with studies conducted in Brazil [
19], India [
21], Bahir Dar, Ethiopia [
12], and southern Ethiopia [
16]. The possible justification for this could be, as birth weight decreases the risk of exposure to infection, hypothermia, and hypoglycemia could increase, and this could lead to the death of neonates. The other possible reason might be, due to immaturity of the lung which causes RDS, and it could lead to death. Particularly the immune system of LBW neonates may not be fully developed and may not fight of infections, therefore, it increases the hazard of death.
In this study, the other predictor of mortality was prematurity, which increases the hazard of death nearly two times as compared with those LBW neonates with a gestational age of 37 week and above. This finding was consistent with studies conducted in Brazil [
27,
51], Iraq [
29], Burkina Faso [
28], Zimbabwe [
25], and Southern Ethiopia [
16]. The possible reason for this could be premature neonates have difficulty in adaptation of the extra uterine environment, due to immaturity of the organ system, and they face many fatal neonatal problems. This problem could decrease the probability of survival [
8,
42]. Additionally, this may be due to improper care and inadequate availability of medical services and equipment like surfactant therapy, and mechanical ventilators unlike in middle-income and high-income countries.
Limitation
This study was conducted using a secondary data source, important variables, such as maternal and paternal educational level, occupational status, nutritional status of the mother, and monthly income were missed. In addition to this, the present study was conducted in hospitals, where neonates who were in the community were not included, which may underestimate the mortality. On the other hand, only hospitalized (high-risk) neonates were included, and it may overestimate the mortality rate.
Conclusion
In conclusion, the incidence of early neonatal mortality in LBW neonates was incredibly higher than e the overall early neonatal mortality in Northwest Ethiopia (Amhara region), and it continues to be a major public health problem. The highest number of death were recorded on the first day of admission. Preeclampsia, PNA, RDS, NEC, Birth weight, and prematurity were independent predictors of mortality among LBW neonates in their first week of life in Northwest Ethiopia.
Therefore, the stakeholders better to give strong emphasis to ELBW & VLBW neonates. The health care provider and other stakeholders shall give more emphasis and critical care on the first day of admission. Special care and support shall be given to LBW neonates who have fatal comorbidities like PNA, NEC, RDS, and neonates born from preeclampsia mothers. Further more, health care providers shall give appropriate ANC and PNC care and also they shall educate the mother and should give advice to the caregivers to perform KMC and breastfeed exclusively. A longitudinal prospective study in each low birth weight category will be better to address important variables that are missed in a retrospective study and to address the true effect of predictor variables.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.