Neonatal meningitis remains an important contributor to neonatal morbidity and mortality, particularly in low- and middle-income countries. Newborns are vulnerable because of immature immunity, and cases are commonly categorized as early-onset (≤72 hours) or late-onset (>72 hours). The burden, pathogen distribution, and resource constraints vary by setting, and there are few large, context-specific studies from Ethiopia. This study aimed to assess short-term treatment outcomes and identify factors associated with poor outcomes among neonates admitted with meningitis to public hospitals in Harar town, Ethiopia.
A facility-based cross-sectional review of medical records was performed using admissions to the neonatal intensive care units (NICUs) of Hiwot Fana Comprehensive Specialized University Hospital (HFCSUH) and Jugal General Hospital (JGH). Records spanned admissions between October 1, 2020 and October 31, 2024. HFCSUH and JGH are public hospitals in Harar that provide neonatal intensive care services.
All medical records of neonates with a confirmed diagnosis of meningitis admitted to the two NICUs during the study period were eligible. Excluded were records with incomplete critical information (for example, missing diagnosis, treatment regimen, or discharge summary), records of neonates who discontinued treatment within the first 72 hours, and those transferred or referred to other facilities.
Data extracted from charts included sociodemographic and perinatal variables (admission site, sex, age at admission, birth weight, gestational age, and mode of delivery), maternal and delivery complications (e.g., premature rupture of membranes, meconium-stained fluid, chorioamnionitis), pre-existing neonatal conditions (congenital heart defects, Down syndrome, hydrocephalus), clinical presentation (symptoms and presence of seizures), laboratory and CSF findings (CSF white blood cell count, CSF protein, CSF glucose, CSF culture results), isolated organisms where available, antimicrobial therapy choices, nutritional support, and management of complications.
The primary outcome was in-hospital treatment outcome at discharge or death, categorized dichotomously as good or poor. A good outcome was defined by clinical improvement and stable discharge status; a poor outcome included in-hospital death or documented adverse discharge status. The study assessed short-term in-hospital outcomes only.
Analyses included binary and multivariable logistic regression to identify factors associated with poor treatment outcomes. Adjusted odds ratios (AORs) with 95% confidence intervals (CIs) were reported. Statistical significance was defined as p < 0.05.
The study reviewed 506 neonatal meningitis records. Overall, 169 neonates (33%) experienced poor treatment outcomes by the time of hospital discharge. The source summary reports the prevalence of key clinical and laboratory abnormalities but does not provide the full table-level distributions in this summary. Details such as exact organism frequencies, age distribution, birthweight categories, and antimicrobial regimens were collected in the study but are not fully reported in the abstract-level summary.
On multivariable logistic regression, the following factors were independently associated with increased odds of poor treatment outcome:
These findings indicate that markers of severe central nervous system infection on CSF analysis (high WBC, high protein, low glucose, positive culture), clinical instability (seizures, early sepsis), certain delivery modes (assisted vaginal delivery), and feeding practices were associated with worse short-term outcomes in this cohort.
Severe CSF abnormalities and positive cultures likely reflect high pathogen burden or advanced disease and were among the strongest predictors of poor outcome, consistent with prior literature linking CSF markers to prognosis. Seizures—both at presentation and occurring during hospitalization—were strongly associated with adverse outcome, suggesting that neurologic involvement and complications substantially worsen prognosis. Early-onset sepsis likely represents concomitant systemic infection that complicates management. The association with assisted vaginal delivery may reflect intrapartum complications or increased risk of perinatal infection, although mechanism cannot be established from this observational dataset. Non-exclusive breastfeeding was associated with increased odds of poor outcome; the study does not provide causal inference but flags feeding practice as a potentially modifiable factor for further study.
In this review of 506 neonates with meningitis admitted to two public hospitals in Harar, Ethiopia, one-third experienced poor short-term treatment outcomes. Strong, independent predictors included indicators of severe central nervous system infection (positive CSF culture, high CSF WBC, high protein, low glucose), seizures, early-onset neonatal sepsis, assisted vaginal delivery, and non-exclusive breastfeeding. The authors highlight the need for early risk stratification and prompt management of high-risk neonates to improve outcomes. Context-specific protocols to identify and treat severe meningitis and coexisting sepsis may be warranted.
The de-identified minimal dataset is publicly available in a Zenodo repository. The authors reported no specific funding for the work and declared no competing interests. Ethical approvals or detailed consent procedures were not summarized in the abstract and main summary provided here.