Background: Broad-range polymerase chain reaction (BR-PCR) detects infectious pathogens from clinical specimens using targets for bacteria (16S rRNA), fungi (28S rDNA), and mycobacteria (fluorescence resonance energy transfer and heat shock protein 65 gene) with reported diagnostic sensitivity and specificity ranging from 43% to 100% and 100%, respectively. We describe our experience when applying BR-PCR to clinical samples submitted for conventional infectious disease testing [conventional testing (CT)] from pediatric patients with concern for infection.
Methods: Retrospective analysis of clinical samples obtained from Nationwide Children's Hospital microbiology laboratory from January 2011 to December 2014 and sent for BR-PCR. Medical record review collected data on patient characteristics, clinical manifestations, laboratory results and antimicrobials prescribed, and a determination of clinical value of BR-PCR was assigned.
Results: There were 247 clinical samples from 163 patients identified; 71 (44%) patients were immunocompromised and 192 (78%) samples reflected pretreatment with antimicrobials. A clinically significant putative organism was identified for 59 samples (24%) between all diagnostic modalities. Conventional testing identified organisms in 41 (17%) samples, 17 of which were corroborated by BR-PCR. Broad-range polymerase chain reaction identified an organism in an additional 18 samples with negative CT results and was considered to provide additional important clinical information. Broad-range polymerase chain reaction detected a bacterial or fungal organism more frequently from tissue samples than from bronchoalveolar lavage or other fluid samples (P = 0.0096, χ).
Conclusions: In our cohort, BR-PCR was an important adjunctive diagnostic in identifying bacteria and fungi in complex clinical situations. Additional data are needed to define the optimal clinical circumstances and specimen type in which BR-PCR can provide the highest diagnostic yield.