Effect of a bundle approach on external ventricular drain-related infection

Background Emergency placement of an external ventricular drain (EVD) is one of the most frequently performed neurosurgical procedures. EVD-related infection continues to be a major challenge causing significant morbidity and costs. Bundle approaches have been shown to reduce infection rates; however, they are still not widely used, and observation periods often were rather short. Methods The present study evaluated the effect of a multi-item bundle approach for EVD placement and care on the occurrence of EVD-related infection. A before/after approach was used to compare groups of consecutive patients over 5-year epochs to control for bias and secondary confounding variables. Results The number of patients in the group before implementation of the bundle approach was 141 and 208 thereafter. There were no statistical differences in demographic and other variables. While 41/141 patients (29.1%) had an EVD-related infection before, this was the case in only 10/208 patients (4.8%) thereafter (p < 0.0001). The EVD-related infection rate was reduced from 13.7/1000 catheter days to 3.2/1000, and the 50% probability of an EVD-related infection in correlation to the mean duration of EVD placement was significantly lower (p < 0.0001). Routine EVD replacement was not helpful to reduce EVD-related infection. EVD-related infection rates remained low also over the next 8 years after the study was finished. Conclusions The introduction of a multi-item bundle approach for EVD insertion and care resulted in a marked reduction of EVD-related infection. Long observation periods over 5 years and beyond confirm that short-term changes are sustained with continued use of such protocols.