Background: Ventilator-associated pneumonia (VAP) is the most frequent nosocomial infection ICU, accounting for almost 50% of all antibiotic prescriptions. The administration of inappropriate initial antibiotic therapy is associated with greater mortality and longer hospital stay. More precise and rapid microbiologic diagnostic approaches for suspected VAP are urgently needed to guide optimal therapeutic choices. In this study we evaluated the accuracy and the potential clinical impact of the Investigational Use Only (IUO) BIOFIRE® FILMARRAY® Pneumonia Panel (PNplus Panel), compared to the standard of care (SoC).
Materials/methods: From December 2018 to February 2019, in all consecutive adult ICU patients with suspected VAP, an endotracheal aspirate (ETA) was collected and sent to the laboratory for rapid microbiologic testing with PNplus Panel. The results were compared with the SoC, represented by quantitative cultures. For viral target detections, the decision whether to confirm the result with a different molecular assay was left to the judgement of the clinician. The potential clinical impact of PNplus Panel was evaluated registering the changes to the empiric antibiotic therapy that could have been possible based on its results.
Results: Fifty-seven samples from 37 patients were analysed. One test (1.8%) resulted invalid. A full concordance between the two methods was observed in 25/57 (43.9%) samples. PNplus Panel detected an additional bacterial target in 31/57 samples (54.4%) and a genetic marker for methicillin resistance in 2/10 samples positive for S. aureus, resulted methicillin susceptible by SoC. There were no cases of bacterial targets missed by PNplus Panel. Fourteen viral targets were detected from 13/57 (22.8%) specimens, with Influenza virus A being the most frequent. In one case, the detection of a viral target by PNplus Panel was not confirmed by SoC. Antibiotic adjustments to the empiric therapy could have been possible in 30/56 cases (53.6%): escalation in 13/56 cases (23.2%) and de-escalation in 17/56 (30.4%). In 5 cases (8.9%), escalation would have included an antiviral agent.
Conclusions: PNplus Panel allows for a rapid and reliable detection of common pathogens for VAP in lower respiratory tract specimens. The use of PNplus Panel in patients with suspected VAP could shorten the time to a targeted and effective antibiotic treatment, reducing the inappropriate or unnecessary use of antibiotics.