%0 Meta-Analysis
%T Diagnostic accuracy of rapid antigen tests in cerebrospinal fluid for pneumococcal meningitis: a systematic review and meta-analysis.
%A Someko H
%A Okazaki Y
%A Tsujimoto Y
%A Ishikane M
%A Kubo K
%A Kakehashi T
%J Clin Microbiol Infect
%V 29
%N 3
%D Mar 2023
%M 36503113
%F 13.31
%R 10.1016/j.cmi.2022.12.002
%X BACKGROUND: Streptococcus pneumoniae is a leading cause of bacterial meningitis worldwide. Conventional microbiological assays take several days and require the use of various drugs for empirical treatment. Rapid antigen tests in cerebrospinal fluid (CSF) may be useful to triage pneumococcal meningitis immediately.
OBJECTIVE: To elucidate whether rapid antigen tests in CSF are useful in the triage of pneumococcal meningitis.
METHODS: Data sourcesCochrane CENTRAL, MEDLINE, EMBASE, World Health Organization International Clinical Trials Registry Platform, and ClinicalTrials.gov databases were searched. Study eligibility criteriaAll types of cohort studies except multiple-group studies, where the sensitivity and specificity of rapid antigen tests in CSF compared with CSF culture can be extracted. ParticipantsPatients with suspected meningitis. TestsRapid antigen tests in CSF. Reference standardsOne or more of the following: blood culture, CSF culture, and polymerase chain reaction in CSF. Assessment of risk of biasThe methodological quality of the included studies was assessed using QUADAS-2. Methods of data synthesisWe used a random-effects bivariate model for the meta-analysis. We conducted a subgroup analysis by dividing studies into types of antigen tests, adults and children, low-income and high-income countries, and with or without exposure to antibiotics before lumbar puncture.
RESULTS: Forty-four studies involving 14 791 participants were included. Most studies had a moderate-to-low methodological quality. Summary sensitivity and specificity were 99.5% (95% confidence interval (CI), 92.4-100%) and 98.2% (95% CI, 96.9-98.9%), respectively. Positive predictive values and negative predictive values at the median prevalence (4.2%) in the included studies were 70.8% (95% CI, 56.6-79.9%) and 100% (95% CI, 99.7-100%), respectively. The diagnostic accuracy was consistent across the various subgroups, except for slightly reduced sensitivity in high-income countries.
CONCLUSIONS: Rapid antigen tests in CSF would be useful in triaging pneumococcal meningitis. Further studies are warranted to investigate the clinical benefit of ruling out pneumococcal meningitis based on the results of rapid antigen tests.