Diagnostic accuracy of clinical and laboratory characteristics in suspected non-surgical nosocomial central nervous system infections.

BACKGROUND: The diagnosis of meningitis in non-surgical hospitalized patients is often difficult and diagnostic accuracy of clinical, laboratory, and radiological characteristics is unknown. AIM: To assess diagnostic accuracy for individual clinical characteristics of patients suspected of non-surgical nosocomial central nervous system (CNS) infections. METHODS: In a prospective multi-centre cohort study in the Netherlands with adults suspected of CNS infections, consecutive patients who underwent a lumbar puncture for the suspicion of a non-surgical nosocomial CNS infection were included. All episodes were categorized into five final clinical diagnosis categories, as reference standard: CNS infection, CNS inflammatory disease, systemic infection, other neurological disease, or non-systemic, non-neurological disease. FINDINGS: Between 2012 and 2022, 114 out of 1275 (9%) patients included in the cohort had suspected non-surgical nosocomial CNS infection: 16 (14%) had a confirmed diagnosis, including four (25%) with bacterial meningitis, nine (56%) with viral CNS infections, two (13%) fungal meningitis, and one (6%) parasitic meningitis. Diagnostic accuracy of individual clinical characteristics was generally low. Elevated CSF leucocyte count had the highest sensitivity (81%; 95% confidence interval (CI): 54-96) and negative predictive value (NPV) (96%; 95% CI: 90-99). When combining the presence of abnormalities in neurological or CSF examination, sensitivity for diagnosing a CNS infection was 100% (95% CI: 79-100) and NPV 100% (95% CI: 78-100). CSF examination changed clinical management in 47% of patients. CONCLUSION: Diagnostic accuracy for individual clinical characteristics was low, with elevated CSF leucocyte count having the highest sensitivity and NPV.

Diagnostic accuracy of VIDISCA-NGS in patients with suspected central nervous system infections.

OBJECTIVES: Confirming the diagnosis in viral central nervous system (CNS) infections can be difficult with the currently available diagnostic tools. Virus discovery cDNA-amplified fragment length polymorphism next-generation sequencing (VIDISCA-NGS) is a promising viral metagenomic technique that enables the detection of all viruses in a single assay. We performed a retrospective study on the diagnostic accuracy of VIDISCA-NGS in cerebrospinal fluid (CSF) of individuals with suspected CNS infections. METHODS: Consecutive adult patients presenting to the Emergency Department or inpatients, who underwent a lumbar puncture for the suspicion of a CNS infection, were included if they were diagnosed with a viral CNS infection, or if a viral CNS infection was initially suspected but eventually a different diagnosis was made. A quantitative PCR panel of the most common causative viruses was performed on CSF of these patients as reference standard and compared with the results of VIDISCA-NGS, the index test. RESULTS: We included 38 individuals with viral CNS infections and 35 presenting with suspected CNS infection for whom an alternative aetiology was finally established. Overall sensitivity and specificity were 52% (95% CI 31%-73%) and 100% (95% CI 91%-100%), respectively. One enterovirus, detected by VIDISCA-NGS, was only identified by quantitative PCR upon retesting. Additional viruses identified by VIDISCA-NGS consisted of GB virus C, human papillomavirus, human mastadenovirus C, Merkel cell polyoma virus and anelloviruses. CONCLUSION: In patients for whom routine diagnostics do not yield a causative pathogen, VIDISCA-NGS can be of additional value as it can detect a broader range of viruses, but it does not perform well enough to replace quantitativePCR.