Feasibility and Prevalence of Early Antibiotic De-Escalation in Patients with Viral Encephalitis

Abstract Corresponding author: Mary Kathryn Vance, 1 Children’s Way, Slot 304, Little Rock, Arkansas 72202, 501-364-1894, [email protected]. Alternate corresponding author: Holly Maples, 4301 W Markham St., Little Rock, Arkansas 72205, 501-364-1718, [email protected] Funding source: no funding to report. Conflict of interest: No conflicts. Background Central nervous system (CNS) infections are associated with significant morbidity and mortality. Due to concern for an increased risk of neurologic sequelae with delays in appropriate treatment for bacterial meningitis, broad antibiotic therapy that covers all likely pathogens according to age is typically initiated empirically. Utilization of rapid diagnostics such as the BioFire® FilmArray® Meningitis/Encephalitis (ME) panel can allow for faster identification of causative pathogens and targeted treatment or cessation of treatment as appropriate. Implementation of these tests does not always lead to rapid changes in antibiotic treatment though, particularly in cases of viral encephalitis where there is concern for bacterial co-infection. The purpose of this study was to describe the incidence of concurrent bacterial and viral CNS infections and identify opportunities for improvement in antibiotic de-escalation in patients with viral CNS infections. Methods This retrospective cohort study included all patients who had a virus identified on the ME panel at Arkansas Children’s Hospital from September 2022 to December 2023. No patients were excluded. The primary outcome of interest was the prevalence of bacterial meningitis in those with viral encephalitis. The total duration of antibiotic therapy and time to discontinuation of antibiotics from identification of viral infection in patients not receiving treatment for a confirmed bacterial infection were also assessed. Data on patient demographics, antimicrobial utilization, bacterial cultures, and infection characteristics were collected through manual chart review. Results One hundred and five adult and pediatric patients were included, 92 of whom were admitted for inpatient management. The median age of included patients was 28 days (range 3 days to 81 years). The most common virus identified was enterovirus (80%), followed by human herpesvirus 6 (11.4%) and parechovirus (3.8%). Seventy-nine patients received empiric antibiotic treatment, most commonly ampicillin and cefepime. Two of the 105 total patients (1.9%) had bacterial growth on CSF cultures. These were Streptococcus gallolyticus and Haemophilus influenzae and both grew within 27 hours. Additionally, six patients had other infections which included pneumonia, Salmonella enteritis, and two cases of urinary tract infections and impetigo each. Seventy-one patients received empiric antibiotic treatment and had no subsequent bacterial infection identified. The median duration of antibiotics targeting gram negative bacteria and gram positive bacteria in this group was 36 hours (IQR 24-48 hours) and 12 hours (IQR 7.5-25.8 hours) respectively. The time to discontinuation of antibiotics targeting gram negative and gram positive organisms was 31.5 hours (IQR 22.1-44.1 hours) and 11.6 hours (IQR 7.5-17.4 hours) respectively in this same group. Conclusions In this study, the incidence of concurrent bacterial and viral CNS infections was low with less than 2% of patients have concurrent infections. The median duration of empiric antibiotics was shorter than expected. However, there is room for further optimization with an aim to discontinue antibiotics around 24 hours after CSF culture collection when there is no growth and another bacterial infection is not suspected. Potential methods to help achieve this objective include provider education and antimicrobial stewardship intervention.