1312. Evaluation of a Multiplexed PCR Pneumonia Panel in a Tertiary Care Medical Center

Abstract

Abstract Background Syndromic PCR testing for lower respiratory pathogens may give rapid, actionable results to aid in management decisions for suspected pneumonia cases. We sought to evaluate the performance of a multiplexed PCR pneumonia panel compared to routine microbiologic work-up in a tertiary care patient population. Methods Sputum and bronchoalveolar lavage (BAL) samples from Keck Medical Center (Los Angeles, CA) inpatients submitted for clinical microbiology work-up Dec 2019-Jun 2020 were tested by a multiplexed PCR panel (FilmArray Pneumonia Panel, BioFire Diagnostics). We compared panel results for typical bacterial pathogens to those of quantitative culture and susceptibility testing. We retrospectively determined the incidence of non-panel respiratory pathogens as detected by standard of care tests in this patient cohort. Results 68 of 180 samples yielded 80 positive bacterial PCR results: 34 were detected by both PCR panel and culture and 46 by PCR panel only, yielding a sensitivity of 100% (34/34) for pathogens detected and specificity of 73.1% (114/156) among negative cultures (normal flora or no growth). Concordant results had PCR Bin values ≥10^5 copies/mL whereas all 18 targets detected at 10^4 copies/mL were culture-negative. Among resistance gene targets, the panel detected 12 MRSA specimens, of which MRSA grew in only 4 cultures; E. coli and CTX-M in 1 specimen from which grew normal flora; and multiple gram-negative organisms and KPC in 1 specimen from which culture isolated carbapenem-resistant P. aeruginosa. Quantitation from positive BAL cultures (n=25) correlated weakly with PCR Bin values (R-squared=0.17). Non-PCR panel pathogens were detected in 22 of 180 (12.2%) specimens through routine methods (16 molds, 3 AFB, and 3 non-fermenter gram-negative bacteria). Conclusion The pneumonia panel had excellent sensitivity for its target bacterial pathogens, but results were often positive in negative cultures. This could be due to antecedent antibiotic therapy, differences in reporting threshold versus culture, or inability of PCR to discern results from normal flora. Non-panel pathogens were detected in a significant proportion in our population. The pneumonia panel should be implemented and interpreted carefully with consideration of antimicrobial stewardship. Disclosures All Authors: No reported disclosures