Clinical utility of an optimised multiplex real-time PCR assay for the identification of pathogens causing sepsis in Vietnamese patients.

Introduction: For the identification of bacterial pathogens, blood culture is still the gold standard diagnostic method. However, several disadvantages apply to blood cultures, such as time and rather large volumes of blood sample required. We have previously established an optimised multiplex real-time PCR method in order to diagnose bloodstream infections.

Material And Methods: In the present study, we evaluated the diagnostic performance of this optimised multiplex RT-PCR in blood samples collected from 110 septicaemia patients enrolled at the 108 Military Central Hospital, Hanoi, Vietnam.

Results: Positive results were obtained by blood culture, the Light Cylcler-based SeptiFast assay and our multiplex RT-PCR in 35 (32%), 31 (28%), and 31 (28%) samples, respectively. Combined use of the three methods confirmed 50 (45.5%) positive cases of bloodstream infection, a rate significantly higher compared to the exclusive use of one of the three methods (P=0.052, 0.012 and 0.012, respectively). The sensitivity, specificity and area under the curve (AUC) of our assay were higher compared to that of the SeptiFast assay (77.4%, 86.1% and 0.8 vs. 67.7%, 82.3% and 0.73, respectively). Combined use of blood culture and multiplex RT-PCR assay showed a superior diagnostic performance, as the sensitivity, specificity, and AUC reached 83.3%, 100%, and 0.95, respectively. The concordance between blood culture and the multiplex RT-PCR assay was highest for Klebsiella pneumonia (100%), followed by Streptococcus spp. (77.8%), Escherichia coli (66.7%), Staphylococcus spp. (50%) and Salmonella spp. (50%). In addition, the use of the newly established multiplex RT-PCR assay increased the spectrum of identifiable agents (Acintobacter baumannii, 1/32; Proteus mirabilis, 1/32).

Conclusion: The combination of culture and the multiplex RT-PCR assay provided an excellent diagnostic accomplishment and significantly supported the identification of causative pathogens in clinical samples obtained from septic patients.