Bacterial prevalence and inflammatory changes in positive blood culture in community-acquired neonatal sepsis in Jordan.

OBJECTIVES

This study aimed to identify the most common causative microorganisms of community-acquired neonatal sepsis and assess the diagnostic performance of C-reactive protein (CRP) and complete blood count as predictors for neonatal sepsis.

METHODS

This retrospective study examined 359 community-acquired septic neonates in Jordan from 2016 to 2022. Blood cultures, complete blood count, and CRP levels were collected.

RESULTS

Among 359 neonates with community-acquired sepsis, the mean age at presentation was 8.8 days; 61.6% were male, and 55.4% were full-term. Fever (54.6%) and respiratory distress (28.4%) were the most frequent presenting symptoms. Gram-positive and Gram-negative organisms were nearly equal (50.4% vs 49.6%), with Escherichia coli (17%) and Streptococcus viridans (16.4%) being the most common isolates. Mortality occurred in 11.1% of cases. Diagnostic analysis showed variable utility of inflammatory markers. CRP demonstrated limited discrimination with an area under the curve (AUC) of 0.64, moderate sensitivity (67.4%) and specificity (60.1%). White blood cell abnormalities showed weaker accuracy with AUC values of 0.54 for leukocytosis and 0.43 for leukopenia, with fair specificity (54.5%) but low sensitivity (39.8%). Neutrophil count abnormalities provided better diagnostic utility with AUC values of 0.63 for neutropenia and 0.57 for neutrophilia, yielding moderate sensitivity (64.1%) and specificity (51.1%). Thrombocytopenia (AUC 0.46) and thrombocytosis (AUC 0.41) demonstrated the lowest diagnostic performance, with low sensitivity and moderate specificity being 28.2% and 61.2%, respectively.

CONCLUSIONS

Early diagnosis and testing for inflammatory markers are essential for detecting community-acquired neonatal sepsis. However, because each test has limited sensitivity and specificity, combining these markers is necessary to improve diagnostic accuracy.