DOI: 10.3390/jcm15166132 ISSN: 2077-0383

Differentiating Sepsis from Infectious and Non-Infectious Conditions in Patients with Elevated Procalcitonin Levels: Performance of Procalcitonin, CRP, and Lactate

Kamil Deveci, Necmi Baykan, Fatma Hançer Çelik, Fatma Ünlü, Esma Eryılmaz Eren, Davut Eren, Maruf Boran, Ömer Salt

Background/Objectives: Elevated Procalcitonin (PCT) levels are frequently encountered in clinical practice but may occur in both infectious and non-infectious conditions, making their interpretation challenging. This study evaluated the diagnostic performance of PCT, CRP, and lactate for differentiating sepsis, infectious conditions, and non-infectious conditions in patients with elevated baseline PCT levels. Methods: In this single-center retrospective observational study, 1072 adults presenting to the ED between June and December 2024 with serum PCT ≥ 0.51 ng/mL were included. Patients were classified into sepsis (n = 272), infection (n = 618), and non-infection (n = 182) groups according to the Sepsis-3 criteria and comprehensive clinical adjudication. ROC analyses and DeLong tests were performed to evaluate biomarker performance. Results: For identifying sepsis, PCT demonstrated better discriminatory performance (AUC 0.683) than CRP (AUC 0.619) and lactate (AUC 0.620), with significant differences between PCT and both CRP (p = 0.009) and lactate (p = 0.018). The optimal PCT cut-off was 2.35 ng/mL (sensitivity 65.8%, specificity 59.5%). For differentiating infectious from non-infectious conditions, CRP showed modest discrimination (AUC 0.640), whereas PCT (AUC 0.581) and lactate (AUC 0.439) performed poorly. Lactate showed the highest prognostic performance for in-hospital mortality (AUC 0.598), although overall prognostic accuracy remained limited. Conclusions: In ED patients with elevated baseline PCT levels, PCT showed better discrimination for sepsis than CRP and lactate, whereas CRP performed modestly in differentiating infectious from non-infectious conditions. None of the biomarkers demonstrated sufficient accuracy for standalone clinical use, supporting their role as complementary tools alongside comprehensive clinical assessment and the Sepsis-3 criteria.

More from our Archive