study-identifies-key-predictors-of-mortality-and-severity-in-newborn-brain-injury
Study Identifies Key Predictors of Mortality and Severity in Newborn Brain Injury

Study Identifies Key Predictors of Mortality and Severity in Newborn Brain Injury

A retrospective cohort investigation conducted at Jimma University Medical Center in southwest Ethiopia has identified critical clinical predictors that influence survival outcomes and condition severity among newborn infants diagnosed with hypoxic-ischemic encephalopathy following birth asphyxia. The analysis of clinical records spanning from April 12, 2019, to May 5, 2022, revealed that neonates presenting with more advanced stages of encephalopathy, acute kidney injury, lower five-minute Apgar scores, neonatal sepsis, and those referred from outside facilities faced markedly higher hazards of death. Hypoxic-ischemic encephalopathy remains a prominent contributor to neonatal morbidity and mortality across resource-constrained regions, where access to advanced supportive therapies such as therapeutic hypothermia is frequently hindered by infrastructure and resource limitations.

The investigated clinical cohort comprised 373 neonates who were admitted to the neonatal intensive care unit after experiencing perinatal asphyxia, defined by clinical criteria including a fifth-minute Apgar score below 7, arterial cord pH below 7, neurological status alterations, or manifestations of multiorgan dysfunction. Utilizing a census approach across the neonatal registry logbooks, investigators extracted maternal sociodemographic data, obstetric histories, and neonatal physiological parameters using a structured abstraction checklist. Patients over 28 days of age at admission and cases with missing medical records were excluded from the study. The analyzed population consisted of 197 male infants representing 52.8 percent and 176 female infants representing 47.2 percent, providing an institutional window into the operational realities of a tertiary referral center serving southwestern Ethiopia.

Encephalopathy severity was classified according to the Sarnat and Sarnat staging framework, which delineates the clinical progression of neurological injury into mild Stage I, moderate Stage II, and severe Stage III categories based on signs such as muscle tone, altered consciousness, and feeding difficulties. Within the study sample of 373 newborns, 121 neonates representing 32.4 percent were diagnosed with Stage I encephalopathy, 169 neonates representing 45.3 percent were diagnosed with Stage II encephalopathy, and 83 neonates representing 22.3 percent presented with Stage III encephalopathy. Overall, 34 infants died during hospital follow-up, resulting in an institutional mortality rate of 9.1 percent, whereas 339 infants representing 90.9 percent were censored upon being discharged alive or referred to secondary care facilities.

Mortality distribution demonstrated an escalating trend across advancing neurological injury classifications, recording mortality rates of 7.4 percent in Stage I, 6.5 percent in Stage II, and 16.9 percent in Stage III. Kaplan-Meier survival analysis yielded an overall median event-free survival duration of 30.6 days across the full cohort. Baseline survival probability commenced at 1.00 at day zero, remaining high at 0.98 with a 95 percent confidence interval of 0.98 to 1.00 on day one. Survival probability subsequently declined to 0.76 with a 95 percent confidence interval of 0.67 to 0.87 at day 20, and fell to 0.48 with a 95 percent confidence interval of 0.29 to 0.81 by day 40. The log-rank test established statistically significant differences in survival across the three encephalopathy stages, reporting a chi-square test statistic of 45.4 with two degrees of freedom and a p-value below 0.001.

Multivariate ordinal logistic regression modeling was utilized to determine physiological and environmental factors independently associated with increasing encephalopathy severity. The proportional odds regression model revealed that neonates who were born outside Jimma University Medical Center and transferred from external healthcare facilities experienced 1.50 times higher odds of being classified in a more severe encephalopathy category compared to infants born inborn within the tertiary center, with an adjusted odds ratio of 1.50, a 95 percent confidence interval of 1.00 to 2.26, and a p-value of 0.04. This association points toward logistical hurdles surrounding intrapartum recognition, immediate newborn stabilization, and transport times between peripheral clinics and tertiary medical infrastructure.

Systemic physiological complications also correlated directly with elevated encephalopathy severity. The presence of acute kidney injury, operationalized through the Kidney Disease: Improving Global Outcomes criteria as a creatinine elevation of 0.3 milligrams per deciliter, a 50 percent serum creatinine rise within 48 hours, or reduced urine output below 0.5 milliliters per kilogram per hour for over 6 hours, was observed in 89 neonates representing 23.9 percent of the cohort. Newborns with acute kidney injury had 1.93 times higher odds of presenting with greater encephalopathy severity compared to those without renal impairment, showing an adjusted odds ratio of 1.93, a 95 percent confidence interval of 1.21 to 3.08, and a p-value of 0.006, reflecting the extensive systemic impact of severe perinatal hypoxia.

Nutritional and metabolic factors similarly demonstrated significant associations with encephalopathy grade. Newborn macrosomia, defined as a birth weight equal to or exceeding 4000 grams and documented in 34 infants representing 9.1 percent of the study group, conferred 2.54 times higher odds of advancing encephalopathy severity relative to normal birth weights between 2500 and 3999 grams, yielding an adjusted odds ratio of 2.54, a 95 percent confidence interval of 1.29 to 5.07, and a p-value of 0.01. Furthermore, neonatal hypoglycemia, diagnosed in 124 neonates representing 33.2 percent of participants, was associated with an adjusted odds ratio of 1.59 and a p-value of 0.01 for higher disease severity. Maternal gestational diabetes mellitus was recorded in 35 mothers representing 9.4 percent and demonstrated a borderline association with severity, characterized by an adjusted odds ratio of 1.99, a 95 percent confidence interval of 0.97 to 4.10, and a p-value of 0.06.

To assess independent predictors of time to death, researchers utilized a multivariate Cox proportional hazards regression model. Statistical testing of Schoenfeld residuals noted that the proportional hazards assumption was violated for encephalopathy stage with a p-value of 0.00, demonstrating that the relative hazard of death varied across the timeline of observation. The multivariate model showed that progressing from Stage I to Stage II encephalopathy multiplied the hazard of mortality by an adjusted hazard ratio of 3.24, with a 95 percent confidence interval of 2.55 to 4.13 and a p-value of 0.00. Concurrently, acute kidney injury emerged as an independent predictor of fatal outcomes, presenting an adjusted hazard ratio of 3.27, a 95 percent confidence interval of 2.47 to 4.31, and a p-value of 0.00.

Newborn vitality scores at five minutes after delivery proved highly predictive of survival outcomes in the adjusted survival model. When compared against reference neonates with normal five-minute Apgar scores equal to or exceeding 7, who accounted for 131 cases or 35.1 percent of the cohort, newborns with scores of 3 or below faced an adjusted hazard ratio of 3.71, with a 95 percent confidence interval of 2.59 to 5.31 and a p-value of 0.00. Infants presenting intermediate five-minute Apgar scores between 4 and 6 exhibited an adjusted hazard ratio of 1.67, with a 95 percent confidence interval of 1.30 to 2.13 and a p-value of 0.00. Outborn transfer status independently elevated the mortality hazard relative to inborn birth, yielding an adjusted hazard ratio of 1.62, a 95 percent confidence interval of 1.29 to 2.04, and a p-value of 0.00, while neonatal sepsis, observed in 68 neonates, remained an influential clinical factor associated with survival outcomes.

The study authors highlighted several methodological considerations and constraints inherent to the retrospective investigation design. Because data extraction relied exclusively on routinely archived medical charts from the neonatal intensive care unit registry, some specific variables such as exact resuscitation timing and long-term neurodevelopmental follow-up parameters were unavailable. Residual confounding stemming from unmeasured factors such as intrapartum transport delays and variations in basic obstetric care quality could not be completely excluded. Furthermore, conducting the investigation at a single tertiary hospital means that findings reflect referral-level patient severity rather than the baseline presentations seen across lower-level district facilities. The researchers emphasized that strengthening regional referral networks, improving intrapartum monitoring, and instituting routine surveillance for metabolic and renal complications represent essential steps to reduce neonatal mortality from perinatal asphyxia in resource-limited healthcare environments.

Subject of Research: Predictors of neonatal mortality and disease severity in hypoxic-ischemic encephalopathy following birth asphyxia

Article Title: Predictors of mortality and severity of hypoxic-ischemic encephalopathy among neonates with birth asphyxia at Jimma University Medical Center, Southwest Ethiopia: a retrospective cohort study

Article References: Kulute, T. F., & Solbana, L. K. (2026). Predictors of mortality and severity of hypoxic-ischemic encephalopathy among neonates with birth asphyxia at Jimma University Medical Center, Southwest Ethiopia: a retrospective cohort study. Maternal Health, Neonatology and Perinatology, 12(1), Article 59. https://doi.org/10.1186/s40748-026-00309-z

Image Credits: AI Generated

DOI: 10.1186/s40748-026-00309-z

Keywords: hypoxic-ischemic encephalopathy, birth asphyxia, neonatal mortality, Apgar score, acute kidney injury, Jimma University Medical Center, Ethiopia, neonatal intensive care, retrospective cohort study, Cox proportional hazards, ordinal logistic regression, macrosomia, neonatal sepsis, hypoglycemia, outborn neonates, Sarnat staging