Abstract
BACKGROUND: Molecular amplification techniques are suggested to be a useful adjunct in early detection of pathogens in septic patients. The aim was to study the feasibility of a polymerase chain reaction (PCR) assay compared to the standard microbiological culture (MC) technique in identification of pathogenic microorganisms from blood and non-blood samples in septic patients. METHODS: Samples for pathogen identification were taken during febrile septic episodes (SE) in 54 patients with sepsis and analyzed using both MC and PCR. Semi-automated multiplex PCR, provided by Philips Medical Systems, was able to detect nine different pathogens. The accuracy of pathogen identification using PCR vs. MC as well as the time-saving effect of PCR on the potential decision-making process for antimicrobial therapy was evaluated. RESULTS: In a total of 258 samples taken during 87 SE, both methods yielded more pathogens from the non-blood than blood samples (87 % vs. 45 %; p = 0.002). PCR identified more pathogens than MC in the blood samples (98 vs. 21; p < 0.0001), but not in other body fluids. In 35 SE, the potential decision on appropriate antimicrobial therapy based on PCR results could have been made 50 (median; interquartile range 35-87) hours earlier than decisions based on standard MC. CONCLUSIONS: In septic patients, multiplex PCR identified more pathogenic microorganisms isolated from the blood samples than the standard MC technique. In the non-blood samples, PCR was comparable to that of MC.