Predicting the Diagnosis of Bacteremia Using Procalcitonin With Systemic Inflammatory Response Criteria

Background: Sepsis remains a leading cause of morbidity and mortality in the US, and early recognition improves outcomes. Traditional diagnostic tools such as systemic inflammatory response syndrome (SIRS) criteria have limited specificity, prompting interest in biomarkers like procalcitonin (PCT). However, PCT interpretation may be affected by comorbidities including renal dysfunction and cancer. Methods: We conducted a retrospective observational review of inpatients ≥10 years old from 2014 to 2015 who had both a PCT level and blood cultures drawn. We assessed diagnostic accuracy of PCT, SIRS criteria, and white blood cell count for bacteremia, infection-related mortality, and all-cause mortality. Subgroup analyses examined patients with impaired renal function and active cancer. Diagnostic test characteristics, receiver operating characteristic curves, and logistic regression were used to compare performance. Results: We included 1014 patients. PCT at a cutoff of 0.5 ng/mL demonstrated superior specificity (65.1%) and sensitivity (66.0%) for bacteremia compared to SIRS criteria (53.0% and 50.9%, respectively). Substituting leukocytosis in SIRS with PCT improved specificity for bacteremia and infection-related mortality. In patients with renal impairment (estimated glomerular filtration rate <30 mL/min/1.73m 2 ), PCT was higher and PCT had markedly higher sensitivity (93.9%) but reduced specificity (36.9%). Cancer did not significantly alter PCT performance but reduced specificity of white blood cell count. No test reliably predicted all-cause or infection-related mortality. Conclusions: PCT at 0.5 ng/mL improves diagnostic accuracy for bacteremia over traditional SIRS criteria, both alone and in combination with SIRS, but adjustment for renal function is needed. Incorporating PCT into sepsis diagnostic pathways may improve early identification while acknowledging specific limitations.