Abstract
Introduction:
Birth asphyxia remains a major contributor to neonatal morbidity and mortality. Identifying clinical and biochemical characteristics linked to mortality may help improve care in resource-limited settings. This study examined the relationship between admission clinical and biochemical features and short-term mortality among asphyxiated neonates.
Methods:
An institutional-based cross-sectional study was conducted among 288 asphyxiated neonates admitted to public NICUs in Addis Ababa between January and December 2022. Data were extracted using a pre-tested tool, and associations were assessed using chi-square tests.
Results:
Overall mortality was 18.4%. Death was more frequent among neonates with severe hypoxic-ischemic encephalopathy, abnormal neurological signs, hypothermia, hypoglycemia, seizures, and hyperbilirubinemia. Non-survivors had higher potassium, liver enzymes, and creatinine and lower sodium and calcium levels. Median hospital stay was shorter among non-survivors.
Conclusion:
Several clinical and biochemical features at admission were associated with mortality among asphyxiated neonates, highlighting the importance of early risk identification in NICUs.
Keywords
Introduction
Birth asphyxia, defined as the failure to initiate and sustain adequate breathing at birth, remains a leading cause of neonatal death and disability worldwide, particularly in low and middle income countries (LMICs). While high-income countries have reduced birth asphyxia-related mortality to less than 0.1% through skilled intrapartum care and neonatal resuscitation, the burden in LMICs remains disproportionately high.1,2
According to the World Health Organization (WHO), approximately 3.6 million neonates in developing countries suffer from moderate to severe birth asphyxia, with nearly 840 000 deaths or cases of serious neurological sequelae. 3 In Ethiopia, birth asphyxia has been identified as the second leading cause of neonatal mortality after prematurity, with a high reported burden of 22.5%. 4
Beyond mortality, perinatal asphyxia is associated with significant short- and long-term morbidity. Survivors are at increased risk of neurodevelopmental impairments, including cerebral palsy, epilepsy, and cognitive delays.5-7 Additionally, affected neonates often develop acute complications such as feeding intolerance, necrotizing-enterocolitis, seizures, and metabolic derangements, including electrolyte imbalances and organ dysfunction.8-10
Despite its high burden, few studies in Ethiopia have comprehensively examined the combined clinical and biochemical factors associated with short-term mortality among asphyxiated neonates in-hospital mortality during NICU admission. Understanding which factors most strongly correlate with poor outcomes can guide early risk stratification, targeted monitoring, and timely intervention in resource-constrained neonatal intensive care units (NICUs).
Therefore, this study aims to identify early clinical and biochemical indicators assessed at admission that are associated with short-term (in-hospital) mortality among asphyxiated neonates in public hospital NICUs in Addis Ababa, Ethiopia. By concurrently evaluating routinely available signs, such as HIE stage, neurological reflexes, temperature, and glucose with basic biochemical markers (eg, electrolytes, liver enzymes, creatinine), this work addresses a critical evidence gap in low-resource settings, where few Ethiopian studies have integrated these domains. Crucially, our real-world, multi-center design reflects actual clinical practice enhancing the generalizability and policy relevance of findings. The results may have direct implications for strengthening national neonatal care protocols, enabling early risk stratification, and guiding context-appropriate interventions to reduce preventable asphyxia-related deaths across similar resource-constrained settings.
Methods
Study Design and Setting
We conducted an institutional-based retrospective chart-review-based cross-sectional study. Data were extracted from the neonatal intensive care units (NICUs) of 4 public tertiary hospitals in Addis Ababa, Ethiopia: Black Lion Specialized Hospital, Yekatit 12 Hospital Medical College, St. Peter Specialized Hospital and Gandhi Memorial Hospital. These facilities were selected by simple random sampling from the 11 public hospitals with NICUs in the city. Records of neonates admitted between January 1 and December 31, 2022, were reviewed during February 15 to March 15, 2023.
Study Population
The source population comprised all neonates diagnosed with birth asphyxia and admitted to NICUs in public hospitals in Addis Ababa during the study period. The study population included a randomly selected subset of these records from the 4 selected hospitals.
Eligibility Criteria
Neonates were eligible if they had a documented clinical diagnosis of perinatal asphyxia and a clearly recorded in-hospital outcome (discharge alive or death). Neonates with major congenital anomalies were excluded to avoid confounding mortality attributable to asphyxia with other causes. Additionally, records with incomplete clinical or biochemical variables were retained in the analysis and reported as missing, reflecting routine NICU practice where not all investigations are performed for every admission.
Sample Size and Sampling Procedure
The sample size was calculated using the single-population proportion formula, assuming a 22.5% prevalence of birth asphyxia-related mortality (based on a 2020 umbrella review), 4 a 95% confidence level, 5% margin of error, and 10% allowance for incomplete records. This yielded a final sample size of 295.
Using proportional allocation based on annual NICU admission volumes, we selected medical records from each hospital (Figure 1). A computer-generated random number sequence was used to select individual charts. Of the 295 sampled records, 288 (97.6%) were included in the analysis.
Schematic presentation of sampling procedure on clinico-biochemical profile and mortality outcomes of asphyxiated neonates, 2023.
Data Collection and Analysis
Data were collected by 2 trained graduate neonatal nurses using a pretested extraction tool adapted from Ethiopia’s Health Management Information System (HMIS) registration forms, standardized perinatal asphyxia follow-up charts, and validated instruments from peer-reviewed literature.8-13 The tool captured: maternal sociodemographic and obstetric characteristics; neonatal clinical and biochemical parameters; and treatment and outcome data.
Data Management and Analysis
Data were double-entered, cleaned, and coded using Epi-Data version 4.6 to minimize transcription errors and ensure data integrity, then exported to STATA version 16.0 for analysis. We used descriptive statistics (frequencies, percentages, means ± SD, or medians with interquartile ranges as appropriate) to summarize variables. Categorical variables were compared using Chi-square tests or Fisher’s exact test (where the expected cell counts were <5). A P-value <05 was considered statistically significant. The results are presented in tables and narrative form, with emphasis on the variables significantly associated with mortality.
Operational Definition
Perinatal asphyxia: A clinical diagnosis made by a physician based on the inability to establish spontaneous respiration at birth, low Apgar scores (<7 at 5 minutes), and the presence of signs suggestive of multi-organ dysfunction, as cord blood gas analysis is not routinely available.
Hypoxic-ischemic encephalopathy (HIE): Stage I, II, or III according to Sarnat’s clinical criteria as documented in the medical records.
Mortality: Death of the neonate during NICU admission, as recorded in the medical chart.
Results
Socio-Demographic & Obstetric Characteristics of Mothers
Of the 295 sampled records, 288 (97.6%) met the inclusion criteria and were analyzed. Most mothers (84.5%) were aged 20 to 34 years (mean: 26.5 ± 4.9 years). Over half (54.7%) of neonates who died were born to primiparous mothers. Fetal distress was documented in 18.4% of cases, and 27.8% of deliveries were by cesarean section. Notably, 26% of neonatal deaths occurred among infants born to mothers with pregnancy-induced hypertension (Table 1).
Socio-Demographic and Obstetrics Characteristics of Mothers of Asphyxiated Babies (n = 288).
Abbreviations: SVD, spontaneous vertex delivery; AVD, assisted vaginal delivery; C/S, cesarean section.
Neonatal Characteristics
Sixty-eight percent of the admitted asphyxiated neonates were male. More than two-thirds of the asphyxiated babies were term (78.1%) and had a normal birth weight (75.7%). When comparing the mean (±SD) age of babies at the time of NICU admission, it was longer for babies who died (4 ± 11.5 hours) compared to babies who survived (2 ± 3.8 hours). Eighty-six percent of the asphyxiated babies who died had a low Apgar score in the first minutes of life, compared to 77% of the babies who survived (Table 2).
Neonatal Characteristics of Asphyxiated Babies Who were Admitted at NICUs (n = 288).
Abbreviation: APGAR, appearance, pulse rate, grimace, activity, respiration rate.
Fisher’s Exact test.
Bold values indicate statistically significant associations (P < 0.05).
Clinical and Biochemical Characteristics
Mortality increased with HIE severity: 4% in stage I, 17% in stage II, and 49% in stage III (P < .001). At admission, 21.5% of the neonates had altered consciousness, 60.7% exhibited a depressed Moro reflex, and 76% had a depressed suckling reflex. Common complications during hospitalization included respiratory distress, hypothermia, meconium aspiration syndrome, seizures, hyperbilirubinemia, and sepsis.
Biochemically, non-survivors showed lower mean serum sodium (135.2 mmol/L vs 136.0 mmol/L), calcium (7.9 mg/dl vs 8.7 mg/dl), higher mean potassium (5.9 mmol/L vs 5.8 mmol/L), AST (321 U/L vs 218 U/L), ALT (100 U/L vs 70 U/L) and creatinine (1.17 mg/dl vs 1.07 mg/dl). Electrolyte values had high missingness (>50%) due to non-routine testing in these settings; analyses were conducted on available data without imputation or exclusion of records, as missingness was not associated with outcomes (P > .05 via chi-square). Significant factors associated with mortality included: Stage III HIE, depressed neonatal reflexes, hypothermia, hypoglycemia, seizures, hyperbilirubinemia, and abnormal hematocrit (Table 3).
Clinical and Biochemical Characteristics of Asphyxiated Babies [n = 288].
Abbreviations: nf, not fulfill chi-square assumptions; NEC, necrotizing enterocolitis.
Fisher’s Exact test.
Treatment and Outcome-Related Characteristics
In the first days of life, 186 (64.6%) of the asphyxiated babies received two-thirds of the total maintenance fluid. Two-thirds (66.7%) of the asphyxiated babies received a prophylactic dose of calcium-gluconate in their maintenance fluid on the first day of life. Among the survived babies, 71% received bag-mask ventilation, 21% received anticonvulsants, and 8% received aminophylline. Only 5 (27.8%) of the 18 asphyxiated babies who required advanced support (chest compression and drugs) survived. The median length of hospital stay was 4 days (IQR: 3.7) for the asphyxiated neonates who died, and 7 days (IQR: 5.10) for the surviving asphyxiated babies (Figure 2). The overall mortality was 18.4% (53/288).
Treatment characteristics of asphyxiated babies who were admitted at NICUs of public hospitals of Addis Ababa, 2023.
Discussion
This study highlights the contribution of birth asphyxia to neonatal mortality in public hospitals in Addis Ababa, Ethiopia, with a short-term mortality rate (ie, death during NICU stay) of 18.4% among asphyxiated neonates. Mortality increased dramatically with the severity of hypoxic-ischemic encephalopathy (HIE): 4% in stage I, 17% in stage II, and 48% in stage III, underscoring HIE staging as a powerful factors associated with the outcome.
Our mortality rate aligns closely with findings from Nigeria (18.4% vs 19.2%), 14 but is notably higher than rates reported in some Indian settings (8%-10%).8,9 These differences likely reflect disparities in neonatal care infrastructure, including access to continuous monitoring, therapeutic hypothermia (not available in our setting), and timely resuscitation. In Ethiopia, where most NICUs operate with limited staff and equipment, the early identification of high-risk infants becomes even more critical.15,16
We identified several readily assessable clinical factors of poor outcome at admission: depressed neonatal reflex, hypothermia, and hypoglycemia. These findings are consistent with global evidence17-21 and reflect the systemic impact of perinatal asphyxia on neurological and metabolic function. Notably, these signs require no advanced diagnostics, making them highly relevant for resource-constrained settings like ours.
Additionally, seizures and hyperbilirubinemia during hospitalization were significantly associated with mortality. Seizures indicate severe neuronal injury, whereas hyperbilirubinemia may result from hepatic dysfunction due to hypoxic insult or hemolysis due to polycythemia, both common in asphyxiated neonates.22,23
Biochemically, non-survivors exhibited lower serum sodium and calcium, elevated potassium, AST, ALT, and creatinine. These abnormalities reflect the cascade of anaerobic metabolism, ATP depletion, and failure of ion pumps during hypoxia, leading to cellular edema, organ dysfunction, and multi-system failure.24,25 Although many electrolyte values were missing (due to routine care limitations), the observed trends support the role of metabolic derangement in determining survival.
Strengths and Implications
This study provides real-world evidence from routine clinical practice in Ethiopia’s capital, highlighting modifiable risk factors that can be integrated into national neonatal care protocols. Clinically, these findings imply that simple, low-cost interventions at admission, such as reflex assessment, temperature stabilization, and blood glucose monitoring, combined with basic electrolyte monitoring could prioritize high-risk neonates for intensified care, potentially averting deaths in settings without advanced resources. Additionally, our findings reflect the actual conditions under which most Ethiopian neonates receive care, making them highly relevant for policy and practice.
Limitations
Our study has several limitations. First, we relied on clinical diagnosis and Apgar scores rather than objective markers such as cord blood pH or base deficit, which may have introduced misclassification. Second, critical variables such as time to spontaneous breathing, seizure type, urine output, bicarbonate, and lactate were not routinely documented, limiting deeper metabolic analysis. Third, the tertiary hospital setting may over represent severe cases, reducing generalizability to primary or secondary facilities. Fourth, time-to-event (survival) analysis was not performed because precise timestamps for key clinical events were inconsistently documented across medical records. In addition, the relatively small number of deaths limited the statistical power for robust survival modeling. Future prospective studies with standardized time recording are warranted to examine the temporal relationship between admission characteristics and mortality. Fifth, use of a national pooled mortality estimate for sample size calculation may have resulted in a conservative sample size and limited statistical power.
Conclusion
This study demonstrates that both clinical and biochemical parameters at admission are strongly associated with short-term mortality among asphyxiated neonates in Addis Ababa. Specifically, stage III hypoxic-ischemic encephalopathy, depressed neonatal reflexes, hypothermia, hypoglycemia, seizures, hyperbilirubinemia, and electrolyte imbalances (notably, low sodium and calcium, elevated potassium, and liver enzymes) were significantly more common among non-survivors.
These findings highlight the need for structured early risk stratification at the point of NICU admission using readily available clinical indicators and basic laboratory tests, particularly in resource-limited settings where advanced diagnostics are not routinely accessible. Systematic assessment of hypoxic–ischemic encephalopathy stage, neurological status (including reflexes and level of consciousness), body temperature, blood glucose, and selected electrolytes should be embedded into routine neonatal care protocols to guide early clinical decision-making, closer monitoring, and timely supportive interventions. Such an approach has the potential to improve prioritization of high-risk neonates, optimize use of limited resources, and reduce preventable in-hospital mortality among asphyxiated newborns in Ethiopia and comparable settings.
Footnotes
Acknowledgements
We would like to extend thanks to all staff members of the card room who work at the Black Lion Specialized Hospital, St. Peter Specialized Hospital, Gandi Memorial Hospital, and Yekatit 12 medical college hospitals and all the data collectors for their invaluable contribution to the realization of this study.
Ethical Considerations
This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board (IRB) of Addis Ababa University (Reference number: AAU-212/2023). Formal permission to access medical records was granted by the neonatal departments and medical records units of the four participating hospitals. All data were anonymized during extraction, and strict confidentiality was maintained throughout the study process.
Consent to Participate
Given the retrospective nature of the study and the use of anonymized data extracted from medical records after patient discharge, the requirement for informed consent from parents or guardians was waived by the Institutional Review Board.
Consent for Publication
Not applicable.
Author Contributions
FBG: conceptualization, methodology, software, formal analysis, and writing the original draft. AG: validation, data curation, writing reviewing & editing. EBW & ET: methodology, writing - review & editing. DBM & MAW: conceptualization, methodology, writing - review & editing. EBW: methodology, writing - review & editing, data curation. FBG & AM: methodology, writing - review & editing, data curation, and validation. Finally, all authors approved the manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.