Diagnostic Yield of Multiplex PCR Method in Cerebrospinal Fluid for the Diagnosis of Purulent Meningitis in Children

Abstract
Objective To evaluate the diagnostic yield of the multiplex polymerase chain reaction (PCR) method in cerebrospinal fluid (CSF) for the diagnosis of purulent meningitis (PM) in children. Methods PM was diagnosed according to the European Society for Clinical Microbiology and Infectious Diseases guideline (2016). Patients with PM between May 2015 and October 2018 were included. The multiplex PCR method was used to detect eight common identified bacteria in PM. Its sensitivity and specificity were compared with bacteria culture. Results A total of 106 cases were enrolled. Pathogenic bacteria were identified in 27 (25.5%) cases by culture and in 37 (34.9%) cases by multiplex PCR assay. The top three bacteria were Streptococcus pneumoniae, Escherichia coli K1, and Streptococcus agalactiae. When using culture as the gold standard, the multiplex PCR assay showed a sensitivity of 100, 88.9, and 75.0% for S. agalactiae, S. pneumoniae, and E. coli K1, respectively, and a specificity of more than 91.3% for all three bacteria. For detectable bacteria, the positive rate of the multiplex PCR assay (36.6%, 37/101) was significantly higher than that of the bacteria culture (21.8%, 22/101). When combining the two methods, etiology was identified in 42.5% (45/106) of the patients. Conclusion Streptococcus pneumoniae, E. coli K1, and S. agalactiae were the predominant pathogens causing pediatric PM. As a rapid method with high sensitivity and specificity, the multiplex PCR assay in CSF could be used as an adjunctive approach with bacteria culture for the pathogen identification of PM. J.L.Z. took part in data curation, investigation, formal analysis, and writing the original draft. C.Z.H. contributed in conceptualization, supervision, writing—reviewing and editing, final approval of manuscript. Y.P.X. took part in collecting resources, and writing—reviewing and editing. Y.X.P. took part in investigation, along with writing—reviewing and editing. B.F.H contributed in resources, along with writing—reviewing and editing. W.J.W. contributed in investigation, apart from writing—reviewing and editing. X.H. contributed in writing methodology, reviewing and editing of the manuscript. All authors approve the final manuscript and agree to be accountable for all aspects of the work. Received: 24 May 2020 Accepted: 30 September 2020 Publication Date: 19 November 2020 (online) © 2020. Thieme. All rights reserved. Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart, Germany