TY - JOUR
T1 - Performance of PCR based syndromic testing compared to bacterial culture in patients with suspected pneumonia applying microscopy for quality assessment
AU - Andrews, Vigith
AU - Pinholt, Mette
AU - Schneider, Uffe Vest
AU - Schønning, Kristian
AU - Søes, Lillian Marie
AU - Lisby, Gorm
N1 - This article is protected by copyright. All rights reserved.
PY - 2022/7
Y1 - 2022/7
N2 - Syndromic testing for lower respiratory tract infections with BioFire® FilmArray® Pneumonia Panel Plus (BF) detects 27 pathogens with a turn-around-time of one hour. We compared the performance of BF with culture. Samples from 298 hospitalized patients with suspected pneumonia routinely sent for culture were also analyzed using BF. Retrospectively, patients were clinically categorized as having "pneumonia" or "no pneumonia." BF and culture were compared by analytical performance, which was evaluated by pathogen concordance, and by clinical performance by comparing pathogen detections in patients with and without pneumonia. The BF results for viruses and atypical bacteria were not included in the performance analysis. In 298 patient samples, BF and culture detected 285 and 142 potential pathogens, respectively. Positive percent agreement (PPA) was 88% (125/142). In patients with community-acquired pneumonia (CAP), clinical sensitivity was 70% and 51%, and specificity was 43% and 71% for BF and culture, respectively. In patients with hospital-acquired pneumonia, the corresponding numbers were 55% and 23%, and 47% and 68%. There was no significant improvement of performance, when only high-quality sputum samples were considered. Efficacy of both BF and culture was low. Both tests are best used in CAP patients for whom the diagnosis has already been clinically established. Indiscriminate use may be clinically misleading and a cause of improper use of antibiotics.
AB - Syndromic testing for lower respiratory tract infections with BioFire® FilmArray® Pneumonia Panel Plus (BF) detects 27 pathogens with a turn-around-time of one hour. We compared the performance of BF with culture. Samples from 298 hospitalized patients with suspected pneumonia routinely sent for culture were also analyzed using BF. Retrospectively, patients were clinically categorized as having "pneumonia" or "no pneumonia." BF and culture were compared by analytical performance, which was evaluated by pathogen concordance, and by clinical performance by comparing pathogen detections in patients with and without pneumonia. The BF results for viruses and atypical bacteria were not included in the performance analysis. In 298 patient samples, BF and culture detected 285 and 142 potential pathogens, respectively. Positive percent agreement (PPA) was 88% (125/142). In patients with community-acquired pneumonia (CAP), clinical sensitivity was 70% and 51%, and specificity was 43% and 71% for BF and culture, respectively. In patients with hospital-acquired pneumonia, the corresponding numbers were 55% and 23%, and 47% and 68%. There was no significant improvement of performance, when only high-quality sputum samples were considered. Efficacy of both BF and culture was low. Both tests are best used in CAP patients for whom the diagnosis has already been clinically established. Indiscriminate use may be clinically misleading and a cause of improper use of antibiotics.
KW - Bacteriology
KW - clinical microbiology
KW - diagnostic stewardship
KW - FilmArray
KW - microscopy & culture
KW - molecular microbiology
KW - pneumonia
KW - syndromic testing
UR - http://www.scopus.com/inward/record.url?scp=85130266362&partnerID=8YFLogxK
U2 - 10.1111/apm.13232
DO - 10.1111/apm.13232
M3 - Journal article
C2 - 35499302
SN - 0903-4641
VL - 130
SP - 417
EP - 426
JO - APMIS - Journal of Pathology, Microbiology and Immunology
JF - APMIS - Journal of Pathology, Microbiology and Immunology
IS - 7
M1 - 13232
ER -