Meningitis in patients with a Gram-negative direct cerebrospinal fluid examination: the value of cytochemical markers for the differential diagnosis

We read with great interest the article by Viallon and colleagues [1] exploring the value of cytochemical markers for the differential diagnosis of meningitis in patients with a Gram-negative direct cerebrospinal fluid (CSF) examination. We have some concern as regards the confounding factors in their study. Firstly, although lumbar puncture was performed immediately on patient admission, the timepoint to obtain the CSF samples relative to the onset of disease was not uniform. For those patients with a Gram-negative direct CSF examination, the negativity may be attributed to a lumbar puncture at a very early stage. Secondly, the inclusion criteria are ambiguous. Meningitis was defined as a white blood cell count more than 5/mm³ in the CSF. Blood contamination due to the lumbar puncture procedure cannot be ruled out. Moreover, the white blood cell count is insufficient to localize infectious inflammation to cerebral pia mater. In terms of this insufficiency, it is not surprising to see the low incidence of nuchal rigidity [1]. Thirdly, the lack of a gold standard for viral meningitis is a vexing problem. An imperfect standard to evaluate a diagnostic test may lead to distortions in sensitivity and/or specificity [2, 3]. In this regard, a positive group and a negative group are indispensable to determine the sensitivity and specificity of either the serum or the CSF parameters in the differential diagnosis of meningitis. Aseptic meningitis caused by intravenous immunoglobulin has not been excluded, since the causative virus was identified in only 39% of the patients with viral meningitis.