Blood transcriptomic for the diagnosis of nosocomial infections in critically ill patients: an observational proof-of-concept study.

Abstract

Purpose: The assessment of genetic activation has shown good diagnostic capacity in identifying sepsis in patients at the emergency department or upon admission to the intensive care unit. This study evaluates a gene expression diagnostic test for identifying nosocomial infections in critically ill patients and compares it with well-established tests used in routine clinical practice.

Methods: This was a prospective, observational, non-interventional study conducted in a single intensive care unit of a tertiary university hospital. Adult critically ill patients were enrolled if their attending physician suspected a nosocomial ICU-acquired infection, and a comprehensive microbiological study was performed. The genetic host response was assessed using the SeptiCyte RAPID assay at study inclusion, alongside microbiological, analytical, and radiological evaluations.

Results: Sixty-nine patients were enrolled, of whom 78.3% (n 54) received empirical antimicrobial treatment. ICU-acquired infection was confirmed in 35 patients (50.7%). A bacterial etiology was established in 32 cases (91.4%), and viral reactivation was diagnosed in 2 cases (5.7%). Diagnostic capacity was measured as follows: procalcitonin AUC 0.600 (95% CI 0.454-0.745), C-reactive protein AUC 0.703 (95% CI 0.564-0.839) and Septi-Cyte RAPID AUC 0.995 (95% CI 0.930-1.003). The optimal cut-off point for Septi-cyte score was 6.35, yielding a sensitivity of 91.4%, specificity of 73.5%, and positive and negative predictive values of 78% and 89.3%, respectively.

Conclusions: Genetic activation analysis in patients with suspected ICU-acquired nosocomial sepsis demonstrated good diagnostic capability, even surpassing traditional inflammatory biomarkers.