Background: Early-onset neonatal sepsis (EONS) is a systemic infection occurring in newborns within the first 72 hours of life. It represents a serious and potentially life-threatening condition requiring rapid recognition, empiric antibiotics and supportive management. However, true infection is confirmed in only a few cases, reflecting the clinical challenge of balancing timely treatment with avoiding unnecessary antibiotic exposure. This study describes the experience of the neonatology department of a Portuguese district hospital.
Method: A retrospective study was conducted in newborns with suspicion of EONS between June 2023 and June 2025. The data analysis was performed using IBM SPSS Statistics (Version 29).
Results: A total of 48 newborns were included. Most were full-term newborns (90%, median age 39 weeks), male (60%), with a median birth weight of 3320 g (IQR 581,25). Cesarean delivery occurred in 48% of cases. Symptoms typically appeared within a median age of 10 hours after birth (IQR 18.5).
The diagnosis of sepsis was clinically confirmed in 81% (n=39) of the newborns, although the culture exams were positive in only 6% (n=4) of the cases. All patients received antibiotics for a median duration of 7 days (IQR 2).
The presence of clinical signs such as respiratory distress (p=0.010) and lethargy/moaning (p=0.002) were statistically significant indicators of a sepsis diagnosis. 27% of these cases required ventilatory support. Comparative analysis between clinically confirmed versus suspected cases revealed no statistically significant differences in C-reactive protein values (median of 4.20 mg/dL; IQR 6.76).
Regarding risk factors, 65% of newborns had identified infectious risk. A positive Group B Streptococcus screen was found in 10% of the mothers, but showed no significant correlation with neonatal sepsis diagnosis (p=0.2). Similarly, prematurity had a negligible association with sepsis in the sample (ρ=0.01).
Conclusion: This retrospective study confirms that while a significant number of newborns are treated empirically for suspected EONS, definitive microbiological confirmation remains uncommon, a finding consistent with what is described in the literature. Our findings highlight the importance of clinical signs, which were statistically significant indicators of EONS. Overall, the results support the ongoing challenge of differentiating true infection from suspected cases, emphasizing the need for improved diagnostic tools to optimize antibiotic stewardship while ensuring effective treatment.
