A retrospective observational study of 1000 consecutive patients tested with the FilmArray® Meningitis/Encephalitis panel: clinical diagnosis at discharge and microbiological findings

Abstract

FilmArray® Meningitis/Encephalitis panel (FAME-p) is used to diagnose central nervous system (CNS) infections. In this study, we investigated performance of FAME-p compared to comparator assays (CA), and for the first time, clinical diagnosis at discharge (CDD). 1000 consecutive patients with a cerebrospinal fluid (CSF) sample analyzed with FAME-p were identified. As CA, culture, polymerase chain reaction and cryptococcal antigen test were used. Medical records of patients were obtained. A CDD of CNS infection was made in 139 of 1000 CSF samples. FAME-p was positive in 66 samples with 44 viral and 22 bacterial agents. Thirteen FAME-p findings were not confirmed by CA, with four discrepant results remaining after comparison with the CDD. Positive percentage agreement (PPA) calculated against CA was 100%. Negative percentage agreement (NPA) calculated against CA was 94.4–99.8% for Haemophilus influenzae, Listeria monocytogenes, Streptococcus agalactiae, S. pneumoniae and varicella-zoster virus (VZV). NPA calculated against CDD was higher (compared to CA) for L. monocytogenes, S. agalactiae and VZV (100%), and lower for Escherichia coli, enterovirus and herpes simplex virus 2 (50–83.3%). NPA of FAME-p for human herpes virus 6 was difficult to interpret. Eighty-four cases received diagnosis of CNS-infection despite negative FAME-p. The four most common non-infectious etiologies were primary headache disorders, cranial nerve palsies, neuroinflammatory disorders and seizure. Although FAME-p shows good performance in diagnosis of CNS infections, result of FAME-p should be interpreted carefully. Considering infectious diseases not covered by FAME-p as well as non-infectious differential diagnoses is important in this context.