Impact of an external ventricular shunt (EVD) handling protocol on secondary meningitis rates: a historical cohort study with propensity score matching

Abstract

Background: External ventricular drainage (EVD) is frequently used in neurosurgical procedures for cerebrospinal fluid (CSF) drainage. It is, however, associated with high infection rates, namely secondary meningitis and ventriculitis. Based on a previous high prevalence of these infections among patients with EVDs, we have proposed and implemented a protocol in an effort to decrease the infection rate. The aim of this study was to measure the effect of hospital-wide implementation of the EVD handling protocol on secondary EVD infections. Patients and methods: We included 409 consecutive patients who received a new EVD for other indications than infectious pathologies from January 2000 until June 2012. Patients above 18 years of age were divided into pre- (n = 228) and post-protocol (n = 181) groups. Patient and disease demographics, as well as EVD data together with confounders for secondary meningitis were recorded in a database. Propensity score matching was then performed to create groups matched for sex, age, reason for drainage, type of shunt, time in situ and duration of surgery to place the EVD. Binomial logistic regression for confounder adjustment and regression discontinuity analyses were then performed on the matched cohort. Results: Infections occurred more frequently in the pre-protocol group (23% vs 9%, p < 0.001). The incidence of infection was 33/1000 drain-days pre-protocol and 9/1000 drain-days post-protocol. Regression analysis in a propensity score-matched cohort (n = 103 in the pre- and n = 178 in the post-protocol groups) showed that the pre-protocol period was independently associated with more infections (OR 2.69; 95%-CI 1.22–5.95, p = 0.01). Conclusions: The incidence of secondary EVD infections can be reduced significantly by the implementation of a strict hospital-wide EVD handling protocol.