We aim to evaluate the accuracy of the broad-range 16S polymerase chain reaction test in the diagnosis of bacterial meningitis through a systematic review and meta-analysis.

We searched MEDLINE, EMBASE, and the Cochrane Controlled Trials Registry, using the Medical Subject Headings terms “polymerase chain reaction,” “RNA, ribosomal, 16S,” and “bacterial meningitis.” For our primary analysis, we examined the 16S polymerase chain reaction in culture-proven bacterial meningitis. In ancillary observations, we included studies of culture-negative presumed bacterial meningitis, in which there was high clinical suspicion for bacterial meningitis despite negative cerebrospinal fluid culture results. We extracted information necessary to calculate sensitivity and specificity and used bivariate hierarchic modeling meta-analysis methods to obtain pooled statistics. We also estimated potential sources of error and bias such as between-study heterogeneity and publication bias.

Fourteen of 299 studies met inclusion criteria for culture-proven bacterial meningitis; 448 (16.1%) of 2,780 subjects had positive cerebrospinal fluid culture results. Pooled analysis demonstrated a sensitivity of 92% (95% confidence interval [CI] 75% to 98%), specificity of 94% (95% CI 90% to 97%), positive likelihood ratio of 16.26 (95% CI 9.07 to 29.14), and negative likelihood ratio of 0.09 (95% CI 0.03 to 0.28) for culture-proven bacterial meningitis. The polymerase chain reaction test result was also positive in 30% of cases of culture-negative presumed bacterial meningitis. There was significant heterogeneity between studies.

This meta-analysis supports the role of 16S ribosomal ribonucleic acid polymerase chain reaction as a diagnostic tool in bacterial meningitis. With further refinements in technology, the polymerase chain reaction test has the potential to become a useful adjunct in the diagnosis of bacterial meningitis in the emergency department.