Abstract
BACKGROUND: External ventricular drains (EVDs) are frequently used in acute brain injuries for continuous intracranial pressure monitoring and cerebrospinal (CSF) fluid diversion. EVDs are associated with a 0–22% risk of ventriculitis. The evidence for antibiotic prophylaxis (AP) for ventriculitis prevention is not robust. This study aimed to delineate the incidence of EVD-related ventriculitis and causative organisms in patients receiving AP. METHODS: A retrospective chart review from 2013 to 2018 at Yale New Haven Hospital was performed. Patients were included if ≥18 years of age, admitted to the neurosciences intensive care unit (ICU), and had AP with cefazolin, vancomycin, sulfamethoxazole/trimethoprim, or clindamycin. Patients were excluded if they had a diagnosis of meningitis or ventriculitis prior to EVD placement, on multiple agents for AP, on antibiotics for indications other than AP, CSF leak, or skull fracture. The primary endpoint was the incidence of ventriculitis per 1,000 EVD-days. Secondary endpoints were causative organisms of ventriculitis, EVD duration, ICU length of stay (LOS), hospital LOS, and 30-day mortality. RESULTS: Five hundred ninety-nine patients were reviewed and 249 patients were included. Baseline demographics are noted in Table 1. Cefazolin was the most common agent for AP (98%). There were 7 cases of ventriculitis with an incidence rate of 2.8% (4 infections per 1000 EVD-days). All of the causative organisms were resistant to the prophylactic agents administered (Table 2). Patients with ventriculitis had a significantly longer duration of EVD placement (10 ± 3 vs. 7 ± 6 days, P = 0.03), hospital LOS (30 ± 19 days vs. 15 ± 12, P = 0.04), ICU LOS (22 ± 14 vs. 10 ± 7, P = 0.03). Two patients with ventriculitis (28%) died within 30 days of admission compared with 46 patients without ventriculitis (19%, P = 0.53) (Table 3). CONCLUSION: The rate of ventriculitis in our study was similar to previous studies that did not utilize AP. All of the causative organisms were resistant to the prophylactic agent. Patients who had ventriculitis had a longer duration of EVD placement, hospital LOS, and ICU LOS; however, 30-day mortality was not impacted. Based on our findings, the use of AP to prevent EVD-related ventriculitis should be reconsidered. DISCLOSURES: All authors: No reported disclosures.