Comparison of different sepsis scoring systems and pathways: qSOFA, SIRS, Shapiro criteria and CEC SEPSIS KILLS pathway in bacteraemic and non-bacteraemic patients presenting to the emergency department

Background Bacteraemia is associated with high morbidity and mortality, with delayed antibiotic treatment associated with poorer outcomes. Early identification is challenging, but clinically important. Multiple scoring systems have been developed to identify individuals in the broader categories of sepsis. We designed this study to assess the performance of existing scoring systems and pathways-CEC SEPSIS KILLS pathway (an Australian sepsis care package), quick sequential organ failure score (qSOFA), systemic inflammatory response syndrome (SIRS) and the Shapiro criteria. Methods This was a retrospective cohort study performed in two metropolitan hospitals in NSW, consisting of adult patients (> 18 years) with positive blood cultures containing a true pathogen and patients matched by age without positive blood cultures. Performance (sensitivity, specificity, and mortality prediction) of recognised sepsis and bacteraemia criteria and pathways-qSOFA, SIRS, Shapiro criteria and CEC SEPSIS KILLS pathway in the first 4 h following ED triage was assessed. Results There were 251 patients in each cohort. Sepsis-related mortality was higher in the bacteraemic group (OR 0.4, p = 0.03). Of the criteria studied, the modified Shapiro criteria had the highest sensitivity (88%) with modest specificity (37.85%), and qSOFA had the highest specificity (83.67%) with poor sensitivity (19.82%). SIRS had reasonable sensitivity (82.07%), with poor sensitivity (20.72%). The CEC SEPSIS pathway sensitivity of 70.1% and specificity of 71.1%. The SEPSIS KILLS was activated on only 14% of bacteraemic patients. Conclusion The performance of all scoring systems and pathways was suboptimal in the identification of patients at risk of bacteraemia presenting to the emergency department.