Timely diagnosis of neuro-infections remains a significant challenge, with the etiology of up to 50% of encephalitis cases and 68% of meningitis cases remaining unidentified despite recent diagnostic advances. Given the global burden and high morbidity and mortality of neuroinfectious disease, understanding factors contributing to efficient diagnosis of these pathogens is essential to limiting spread, providing treatment, and preserving public health.

To identify patterns and characterize limitations of current neuroinfectious disease diagnostic approaches.

Retrospective analysis of 111 hospital admissions of confirmed neuroinfectious disease made between 2010-2018 at Columbia University Irving Medical Center. Demographics and diagnostic studies were collected via chart review.

Viral pathogens comprised a majority of confirmed neuro-infections (64.9%, 72/111 patients), followed by bacterial (23.4%, 26/111 patients) and fungal/parasitic (11.7%, 13/111 patients). Viral and fungal pathogens were predominantly detected with CSF PCR/antigen (91.6%, 53.8% respectively), while bacterial infections were largely detected via CSF culture (38.5%). Viral neuro-infections were negatively associated to length of hospitalization (9.4 days +- 11.6 SD) and time to diagnosis (2.4 days +- 5 SD). Fungal/parasitic infections were associated with greatest length of stay (24.9 days +- 19.8 SD), while bacterial infections were associated with longest time to diagnosis (4.9 days +- 8.4 SD). Longer hospitalizations were positively associated with immunosuppression status regardless of infectious etiology. Viral infections required significantly fewer tests sent for diagnosis (24.1 +-11.7 SD) compared to bacterial (32.8 +- 19.2) and fungal/parasitic (37.5 +- 13).