Predictive Utility of the Hematological Scoring System (HSS) for Improved Neonatal Sepsis Diagnosis in an Indian Setting – A Single Centre Study

Background: Neonatal sepsis continues to be a major contributor to morbidity and mortality especially in resource-limited settings. Early diagnosis may be aided by the Hematological Scoring System (HSS). The present study was undertaken to evaluate the diagnostic performance of the HSS as a diagnostic tool for early-onset neonatal sepsis.Methods: This prospective observational study included 140 neonates with suspected sepsis. Blood samples were analysed by manual and automated methods to calculate HSS scores. Diagnostic performance metrics such as sensitivity, specificity, positive predictive value and negative predictive value (NPV) were assessed. Inter-observer agreement for manual counts was assessed using Bland-Altman Plots. Result: The HSS demonstrated significant association (<0.05) with sepsis diagnosis. Blood cultures yielded a low positivity rate (8.57%) with *Staphylococcus* species and Enterobacteriaceae family being the most commonly identified bacteria. C-reactive protein exhibited high sensitivity (100%), while the immature to total polymorphonuclear count ratio showed high specificity (98.44%). Manual methods generally outperformed automated methods in terms of sensitivity and NPV, particularly for immature polymorphonuclear count (IPMN count) and degenerative changes. Manual assessment improved diagnostic sensitivity, although inter-observer variability was observed. Manual HSS (with a smear review) exhibited superior sensitivity in comparison to automated HSS (without smear review), despite some missed diagnoses. Optimized smear preparation is essential for accurate HSS implementation.Conclusion: The HSS can be a useful supporting tool for early neonatal sepsis diagnosis, particularly in resource-limited settings. However, prospective interventional trials are required to confirm if implementing HSS can safely inform antimicrobial stewardship and minimize unnecessary empiric antibiotic use.