Materials and Methods: This cross-sectional-analytical study was conducted on 80 neonates (aged <28 days) with suspected sepsis admitted to Tabriz Children’s Hospital. Based on blood culture results, infants were divided into two groups: definite sepsis (n=3) and non-sepsis (n=77). Clinical symptoms and laboratory parameters (including CBC and CRP) were recorded, and the HSS score was calculated for each infant. The diagnostic accuracy of HSS was evaluated using receiver operating characteristic (ROC) analysis, and indices such as sensitivity and specificity were calculated.

Results: The mean gestational age and birth weight were significantly lower in the sepsis group (P<0.001). Gram-positive cocci, specifically Staphylococcus aureus (coagulase-positive), were the most common causative agents (n=2, 66.7%), followed by Klebsiella pneumoniae (n=1, 33.3%). Among clinical signs, lethargy, hypothermia, and cyanosis were significantly more frequent in the sepsis group (P<0.05). CRP, I/T ratio, and HSS scores were significantly higher, and platelet counts were lower in the sepsis group (P<0.001). The ROC curve for HSS showed an area under the curve (AUC) of 0.88. At the optimal cut-off point (≥4), sensitivity was 100% (95% CI: 43.9–100), specificity was 83.1%, positive predictive value (PPV) was 15.8%, and negative predictive value (NPV) was 100%.

Conclusions: In this cohort with a low prevalence (3.75%) of culture-proven sepsis, the HSS demonstrated high diagnostic accuracy, particularly with a perfect sensitivity and NPV within the limited sample, effectively ruling out sepsis when the score was low. Incorporating this simple, inexpensive tool into clinical assessment may improve decision-making regarding antibiotic initiation and reduce unnecessary antibiotic exposure. However, the precision of sensitivity and PPV estimates is limited by the small number of sepsis-positive cases.