Role of C-reactive Protein and Procalcitonin in Early Diagnostic Accuracy and Their Prognostic Significance in Sepsis.

Abstract

Sepsis is a serious condition that often results in high fatality rates, particularly in intensive care units (ICUs). Its nonspecific clinical characteristics makes early diagnosis and therapy difficult, despite how critical they are. The use of biomarkers like procalcitonin (PCT) and C-reactive protein (CRP) in the diagnosis and prognosis of sepsis has demonstrated encouraging results. In contrast to PCT, which is highly selective for bacterial infections, CRP is an acute-phase protein that reflects systemic inflammation. This study aimed to assess the diagnostic accuracy and prognostic significance of CRP and PCT in early sepsis detection and outcome prediction. This study was a retrospective cohort study that involved 90 patients in the ICU who met the criteria for sepsis-3. CRP and PCT levels, clinical data, and outcomes were obtained from electronic medical records. The diagnostic accuracy was tested using receiver operating characteristic (ROC) curves, while the prognostic relevance was analyzed by Kaplan-Meier survival analysis and Cox proportional hazards regression. The mean CRP level was 102.3 mg/L and PCT level was 5.4 ng/mL. ROC analysis revealed an area under the curve (AUC) of 0.78 for CRP and 0.82 for PCT, indicating better diagnostic performance for PCT. High levels of CRP and PCT were associated with poorer survival, with median survival times of 18 and 15 days, respectively, for high-level groups. Cox regression identified CRP and PCT as significant predictors of mortality, with hazard ratios of 1.50 and 1.68, respectively. Both CRP and PCT are valuable biomarkers for diagnosing and prognosticating sepsis. PCT, with its higher specificity for bacterial infections, demonstrates superior diagnostic accuracy compared to CRP. Elevated levels of both biomarkers are associated with increased mortality risk, highlighting their potential role in early sepsis management and outcome prediction.