Diagnostic Accuracy of Blind Bronchial Sample Testing by BioFire Pneumonia plus Panel in Pediatric Intensive Care Unit Patients and Its Impact in Early Adaptation of Antimicrobial Therapy: A Prospective Observational Study.

Abstract

Community-acquired and nosocomial lower-respiratory-tract infections in critically ill pediatric patients require early appropriate antibiotic therapy to optimize outcomes. Using blind bronchial samples, we assessed the diagnostic performance of the rapid-multiplex polymerase chain reaction (PCR) assay BioFire Pneumonia plus Panel vs. reference standard culturing with antimicrobial susceptibility testing. For this prospective observational study in a single pediatric intensive care unit, we included consecutive patients younger than 18 years admitted for suspected community-, hospital- or ventilator-associated pneumonia in 2021-2022. Sensitivity, specificity, positive predictive value and negative predictive value of the multiplex PCR assay were determined. The kappa coefficient was computed to assess agreement, and univariate analyses were done to identify factors associated with discrepancies between the 2 diagnostic methods. Of the 36 included patients (median age, 1.4 years; interquartile range, 0.2-9.2), 41.7%, 27.8%, and 30.5% had community-, hospital- and ventilator-associated pneumonia, respectively. The overall κ was 0.74, indicating good agreement. Overall, the sensitivity of the multiplex PCR assay was 92% (95% CI: 77%-98%) and specificity 95% (95% CI: 92%-97%), with variations across microorganisms. The median time from sample collection to antimicrobial susceptibility test results was 3.9 (2.5-15) hours with the multiplex PCR assay and 60.5 (47.6-72.2) hours with the reference technique. The BioFire Pneumonia plus Panel used to test blind bronchial samples had satisfactory diagnostic performance in critically ill pediatric patients. The rapid results provided by this test may improve the appropriateness of antimicrobial therapy and help minimize the use of antibiotics.