The reduction of intracranial pressure using external ventricular drains (EVDs) is a life saving measure, but infection is a major complication of this procedure. Therefore, reducing procedural meningitis/ventriculitis is an important goal for healthcare infection prevention teams. This study determines the effectiveness of a newly implemented infection prevention bundle (IPB) and identifies possible risk factors increasing incidence of EVD-AIs.

This is a pre- and post-study performed at a single 440-bed academic medical center reviewing all EVD-AIs between January 2014 and October 2018 comparing the rates of EVD-AI before and after application of the IPB. Measured variables included procedure time, EVD placement location & duration, cerebrospinal fluid (CSF) specimen source, surgeon, and the presence of EVD-AI. Confirmed infection was defined as positive CSF culture, and possible infection was defined as high white blood cell count (>100 cells/ml) and/or low glucose levels (<40 mg/dL). Odds ratios (OR) of measured variables as related to EVD-AI presence were determined with a significance threshold of p-value <0.05.

Out of 313 procedures, there were 105 unique patients eligible for analysis. Seventy patients had CSF analysis performed; 37 had possible EVD-AIs (31 pre-IPB and 6 post-IPB) and 7 were confirmed EVD-AIs (5 pre-IPB and 2 post-IPB) with no statistically significant difference pre- and post-IPB. The EVD duration (OR 1.20), EVD placement procedure (OR 2.38), and CSF specimen source (OR 7.00) were associated with increased risk of infection.

Multiple risk factors are associated with EVD-AIs, including longer EVD duration, EVDs placed at bedsides, and CSF drawn from EVD collection bags. The decrease in EVD-AIs is not conclusively linked to the implementation of the IPB due to the small number of procedures and short duration of follow up. The application of effective evidence-based infection prevention measures for EVD-related procedural meningitis/ventriculitis is an essential intervention to reduce morbidity and mortality.