Objective: To evaluate dipstick rapid diagnostic tests (RDTs) for meningococcal meningitis in basic health facilities.
Methods: Health facility staff received a one-day training. During the meningitis season, they performed RDTs on cerebrospinal fluid (CSF) specimens from suspected cases of meningitis. A frozen aliquot of CSF was later tested using polymerase chain reaction (PCR) to establish the reference diagnosis. RDTs used in health facilities were archived to allow checking the concordance between reported diagnosis and observed results. Reported diagnosis was also compared to PCR diagnosis. A second RDT was performed on each CSF specimen at the reference laboratory.
Results: Using RDTs, health facilities reported 382 negative results (73.9%), 114 NmA (22.1%), 12 NmW135 (2.3%) and nine uninterpretable results (1.7%), the latter corresponding to the misuse of a reagent by three agents. The agreement between reported diagnosis and archived dipsticks was excellent (kappa = 0.98). The agreement between PCR diagnosis and reported RDTs results was strong (kappa = 0.82). In health facilities, the sensitivity of RDTs for N. meningitidis A was Se = 0.91. The kappa coefficient measuring the agreement between RDTs operated in the reference laboratory and RDTs operated in health facilities was kappa = 0.78.
Conclusion: We confirmed that dipstick RDTs to identify N. meningitidis serogroups A, C, W135 and Y can be reliably operated by non-specialized staff in basic health facilities. RDTs proved very useful to recommend vaccination in NmA epidemics, and also to avoid vaccination in epidemics due to serogroups not included in vaccines (NmX).