Evaluating the Clinical Impact of Metagenomic Next-Generation Sequencing in CNS Infections: A Diagnostic Pathway and Resource Utilization Modeling Study.

Background: Diagnosing meningitis and encephalitis remains challenging due to nonspecific clinical presentations and the limitations of traditional microbiological methods. Metagenomic next-generation sequencing (mNGS) offers a broad approach to detect pathogens, but its real-world impact on clinical decision-making remains undefined. Methods: We used a cohort of patients with confirmed central nervous system infections and autoimmune encephalitis (AE) who underwent traditional microbiological cerebrospinal fluid testing at Columbia University Irving Medical Center. Using published sensitivity and specificity data for mNGS, we applied Bayes' theorem to calculate different etiology-specific pretest probabilities and model the potential impact in the diagnostic workflows including the number of lumbar punctures (LPs), additional etiologic tests potentially avoided, and time to diagnosis. Results: The cohort includes 54 patients in the infectious cohort and 29 patients with confirmed autoimmune encephalitis. In a modeled scenario, utilizing an mNGS test, such as Delve Detect, in patients with DNA viral infections (n = 23) could lead to a reduction of up to 88 microbiological tests, 145 days to diagnosis, and 2 LPs in total. For bacterial infections (n = 16), estimated impact included a reduction of 30 microbiological tests, 144 days to diagnosis, and 12 LPs (Table 1). Although fungal, RNA viral and parasitic infections were less common, with adjusted positive predictive values of 92.8%, 89.5%, and 84.6%, respectively. In the autoimmune cohort, a total of 2 LPs, 126 microbiological tests, and 297 days to diagnosis could have been avoided through the use of mNGS. Conclusions: Our analysis suggests that an mNGS test, such as Delve Detect, could potentially streamline diagnostic and treatment pathways in meningitis and encephalitis of infectious or autoimmune origin.