Methods: Bronchoalveolar lavage (BAL) specimens from 120 patients with suspected VAP, enrolled at four different French ICUs, during Jan. to Nov. 2013, were analysed by CC, following microbiological standard procedures, by BPP and MM. A total of 15 bacterial targets, commonly detected by the three methods, were analysed for concordance above an agreed threshold for positivity. While every step is fully integrated, from specimen-to-results (BPP), bacterial DNA was extracted from each sample on the NucliSENS easyMAG® Platform, and real-time polymerase chain reactions were run in an ABI 7500 Dx thermocycler (MM).
Results: A total of 117 different BAL specimens were processed. Positive culture was obtained for 65.8% of BAL, while positive detections were observed in 79.4% with BPP and 75.4% with independent MM. Fourteen different species were detected by the three methods, with majority of the bacteria being S. aureus, P. aeruginosa and H. influenzae. Overall concordance performance between BPP and CC was 89.0% (83.1%-94.9%) positive percentage agreement (PPA) and 95.9% (95.0%-96.9%) negative percentage agreement (NPA). Overall concordance between BPP and MM was 97.1% (93.8%-100.3%) PPA and 96.6% (95.6%-97.6%) NPA. Following discrepancy analyses overall performance increased to 95.3% (91.2-99.3%) PPA when comparing BPP to CC.
Conclusion: The new BioFire® Pneumonia Panel provides reliable quantitative microbiological data in BAL specimens, in only 65 minutes, which can lead to more appropriate management of VAP suspected patients in the ICU.
RUO products used in this study have not been evaluated by the FDA or other regulatory agencies for In Vitro Diagnostic use.
C. Weber, bioMÃ©rieux: Employee , Salary .
C. Alberti-Segui, bioMÃ©rieux: Employee , Salary .
C. Mousset, bioMÃ©rieux: Employee , Salary .
C. Ginocchio, bioMÃ©rieux: Employee , Salary .
M. Rogatcheva, BioFire: Employee , Salary .
O. Barraud, None
B. François, None
V. Moucadel, bioMÃ©rieux: Employee , Salary .
J. Yugueros-Marcos, bioMÃ©rieux: Employee , Salary .