Rational Use of Microbiological Tests in the Diagnosis of Central Nervous System Infections Using Restrictive Criteria: a Retrospective Study.

Central nervous infections, mostly represented by meningitis and encephalitis, remain a diagnostic challenge despite substantial advances in microbiological tools in recent years. Meanwhile, extensive microbiological workups, which often prove to be irrelevant retrospectively, continue to be processed on a large scale, therefore leading to unnecessary costs. The main goal of this study was to evaluate a systematic approach enabling more rational use of microbiological tools in the setting of community-acquired central nervous system infection diagnosis. In this single-center descriptive study, the modified Reller criteria were retrospectively extended to all neuropathogens tested in cerebrospinal fluid (CSF) samples with the FilmArray meningitis/encephalitis panel (BioFire Diagnostics, LLC) and bacterial culture. The inclusion period was 30 months. In total, 1,714 fluid (CSF) samples analyzed from 1,665 patients over 2 and a half years were reported. According to the retrospective application of the modified Reller criteria, microbiological testing was considered unnecessary in 544 CSF samples. Fifteen positive microbiological results were found among these samples, interpreted either as inherited chromosomally integrated human herpesvirus 6 (HHV-6), a false-positive result, or a true microbial detection without clinical relevance. No CNS infection case would have been missed if these analyses were not carried out, while about one-third of all meningitis/encephalitis multiplex PCR panels would have been saved. Our retrospective analysis suggests that the modified Reller criteria could be safely applied to all microbiological tests performed in CSF, thereby saving substantial costs. Microbiological testing in general and in the setting of central nervous system (CNS) infection in particular are often excessive, leading to superfluous laboratory work and costs. In this regard, restrictive criteria, named Reller criteria, have been developed to reduce unnecessary CSF herpes simplex virus 1 (HSV-1) PCR testing when suspecting encephalitis. These criteria were then adapted for increased safety to become the modified Reller criteria. This retrospective study aims at evaluating the safety of these criteria when applied to CSF microbiological testing in general, including multiplex PCR, direct examination, and bacterial culture. The postulate was that a CNS infection can be excluded if none of these criteria is present. According to our data set, no CNS infection would have been missed if the modified Reller criteria would have been applied to save microbiological tests. This study therefore proposes a simple way to reduce unnecessary microbiological testing in the context of CNS infection suspicion.