Utilization of a Meningitis/Encephalitis PCR panel at the University Hospital Basel - a retrospective study to develop a diagnostic decision rule.

BACKGROUND: The Biofire® FilmArray® Meningitis/Encephalitis (ME) PCR panel covers 14 viral, bacterial, and fungal pathogens and has been implemented in many institutions worldwide. Post-marketing studies indicate a reduced sensitivity and overutilization underscoring the need for a more targeted usage. The aim of our study is to describe the utilization of the ME panel and to develop a diagnostic-stewardship based decision rule.

MATERIALS: Adult patients, who underwent CSF analysis with the ME panel between August 2016 and June 2021 at the University Hospital Basel, were included. Demographic, clinical, microbiological, and laboratory data were extracted from the electronic health record. Factors associated with a positive ME panel result were identified, and a decision rule was developed to potentially optimize the diagnostic yield and reduce the number of unnecessary tests.

RESULTS: 1,236 adult patients received at least one panel in the observed period, of whom 106 panels tested positive (8.6%). The most frequently observed pathogens were Varicella Zoster Virus (VZV, 27%), Streptococcus pneumoniae (19%), Enterovirus (16%), Herpes simplex Virus 1/2 (16%), and Human Herpesvirus 6 (HHV-6, 13%). Fever, vomiting, headache, and photophobia were more frequently present in test positive patients as were significantly higher CSF leukocytes and protein concentrations. When simulating a decision rule based on CSF leukocytes and protein concentration, only 35% of all patients would have qualified for a ME panel tests, thereby increasing the positivity rate to 22.7%. 10 of 106 positive ME panels would have been missed, only involving HHV-6 and VZV (6 and 4 cases, respectively). As these subjects were either severely immunocompromised or had clinical features of shingles we propose extending the testing algorithm by including those criteria.

CONCLUSION: The ME panel positivity rate at our institution was similar as previously published. Our results highlight the need for diagnostic-stewardship interventions when utilizing this assay by implementing a stepwise approach based on a limited number of clinical and laboratory features. This decision rule may improve the pretest probability of a positive test result, increase the quality of test utilization, and reduce costs.