Abstract
Early sepsis diagnosis is crucial for implementing adequate antibiotic therapy and for patient survival. This study investigated whether using multiplexed PCR for detecting microorganisms in critical septic patients affects initial antibiotic treatment and compared it to microbiological culture. It also explored scenarios where PCR is more effective in clinical practice. One hundred nineteen specimens (83 blood and 36 respiratory specimens) belonging to 93 patients were analyzed. Multiplexed PCR determinations were performed using the FA-BCID Panel (bioMérieux) for blood samples and the FA-Pneumo for respiratory samples. The mean turnaround times were 1.7 h for the FA-BCID and 1.5h for the FA-Pneumo. Conversely, they were 96.1 h for blood cultures and 72.3 h for respiratory cultures. FA-BCID showed a mean sensitivity of 97% and specificity of 100%. FA-Pneumo showed a sensitivity of 100% and specificity of 90%. However, the positive predictive value was only 39%. Discrepancies were common in polymicrobial samples. Based on the PCR results, initial empirical treatment should have been changed in 71% of patients with bloodstream infections and 61% with respiratory infections. We conclude that multiplexed PCR improves the response time in identifying germs with a high degree of coincidence for blood cultures and moderate for respiratory cultures. These results highlight the importance of PCR in choosing an appropriate antibiotic therapy.