Background: Early appropriate antimicrobial therapy reduce the risk of pneumonia-related morbidity and mortality in critical care patients. Nevertheless, current culture methods take at least 48 hours to obtain antimicrobial susceptibility results. The new BIOFIRE® FILMARRAY® Pneumonia plus Panel (PNplus Panel) offers advantage to rapidly detect and quantify in a single test, 18 bacteria, 9 virus, and 7 antibiotic resistance genes. This test was compared to conventional bacterial culture in 85 patients suspected of having hospital-acquired pneumonia (HAP).
Materials/methods: 207 respiratory samples from 85 patients [64 bronchoalveolar lavages (BALDS) and 74 endotracheal aspirates (ETADS) obtained at the time of HAP diagnosis, and 69 ETA obtained 2-3 days later (ETATT)], were analyzed in parallel by routine microbiology testing and PNplus Panel at the laboratory of the Nantes University Hospital. Results of PNplus Panel were not reported to clinicians. Routine microbiology testing were analyzed independently of PNplus Panel.
Results: 60/85 (70.59%) patients had a positive result for at least one bacterial pathogen at the time of diagnosis by culture [43/64 (67.19%) BALDS and 46/74 (62.16%) ETADS], compared to 71/85 (83.53%) for PNplus Panel [50/64 (78.13%) BALDS and 59/74 (79.73%) ETADS]. Among 53 patients with paired BALDS and ETADS, 77.36% had the same result for both samples with conventional method, compared to 66.04% for PNplus Panel. Among positive samples, polybacterial results were obtained for 23/43 (53.49%) BALDS, 19/46 (41.30%) ETADS, and 7/30 (23.33%) ETATT by culture vs 30/50 (60.00%) BALDS, 37/59 (62.71%) ETADS, and 32/49 (65.31%) ETATT for PNplus Panel. Haemophilus influenzae, the most frequent species detected by PNplus Panel (33 patients at diagnosis), was not always identified in culture (24 patients at diagnosis). A carbapenemase and/or an ESBL were detected with both methods in 7 patients.
Conclusions: Overall, PNplus Panel detected the main bacterial respiratory pathogens with greater sensitivity than culture. The concordance between both methods was higher for BAL, less contaminated with microbial flora. In case of polymicrobial specimen, PNplus Panel facilitated the detection of bacterial pathogens. Implementing this strategy on BAL should improve and accelerate clinical decision and isolation care. We are currently measuring the extent to which PNplus Panel results would have modified antimicrobial prescription and costs of care.