Fast Microbiology: A Pilot Study Comparing Culture-Independent Digital PCR and Blood Culture for Pathogen Detection...

Bloodstream infections (BSIs) represent a major clinical challenge, particularly in patients with suspected sepsis, where rapid pathogen identification is critical for timely and appropriate antimicrobial therapy. Blood culture (BC) remains the reference diagnostic method but is limited by a long turnaround time and reduced sensitivity, especially in patients receiving prior antibiotics. This pilot study aimed to evaluate the diagnostic performance and clinical added value of culture-independent digital PCR (dPCR) performed directly on whole blood compared with paired BC. A total of 37 samples from 27 patients with suspected BSI admitted to a tertiary-care hospital were analyzed. dPCR demonstrated the ability to detect a broader spectrum of pathogens compared with BC (34 vs. 28 species), with improved identification of fastidious organisms and fungi, particularly Candida spp. Diagnostic performance analysis showed a sensitivity of 75.0% and a specificity of 53.8%, with a positive predictive value of 75.0% and a negative predictive value of 53.8%. Likelihood ratios indicated moderate rule-in capability (Likelihood ration positive 1.62) but limited rule-out performance (Likelihood ratio negative 0.46). Importantly, concordance analysis integrating microbiological and clinical data revealed that a substantial proportion of discordant dPCR-positive results were clinically plausible, suggesting detection of infections not captured by BC, particularly in polymicrobial settings, under antibiotic pressure, or in fungal infections. Longitudinal observations further highlighted the potential of dPCR for monitoring pathogen dynamics and treatment response. Overall, dPCR provides rapid and complementary diagnostic information, enabling earlier pathogen detection, identification of antimicrobial resistance markers, and improved characterization of complex infections. While not suitable as a standalone test due to limited negative predictive value, its integration with conventional microbiology may enhance diagnostic accuracy and support antimicrobial stewardship in the management of BSIs.