Résumé
Background. External ventricular drains (EVDs) are essential to the clinical management and care of patients with neurosurgical complications; but EVD use is routinely associated with concomitant infection; sometimes resulting in mortality. Objective. To undertake an epidemiological study of ventriculostomy-related infections among paediatric neurosurgical patients at the Dr George Mukhari Academic Hospital; Pretoria; South Africa. Methods. Retrospective analysis was conducted on the clinical records of 92 children admitted to the neurosurgical unit at the hospital between 2010 and 2013. Records were included in the study only if they were complete; legible and accurate. Data were collected on the following variables: age; gender; frequency of catheter change; cerebrospinal fluid (CSF) sampling; use of prophylaxis; microbiology; Glasgow Coma Scale; glucose; chlorine; and other clinical; chemical and laboratory parameters routinely observed as part of patients' work-ups. Results. Two or more EVDs were placed on 45.7 (40) children; with a maximum of seven EVDs per child. Ventriculitis incidence was 28.3 (26 of 92). There was a significant association between the number of EVDs inserted and the incidence of ventriculitis (p=0.010). More frequent CSF sampling also increased ventricular-related infections (p=0.000); as did prolonged EVD retention (p=0.001). Using prophylactic antibiotics or impregnated catheters did not reduce ventriculitis incidence significantly. Conclusion. Evidence supports adherence to strict sterilisation protocols and techniques when inserting EVDs. Catheters should not be retained for extended periods; and CSF sampling can be limited to once in 3 days. Routine use of antibiotic-impregnated EVDs and antistaphylococcal prophylaxis is still recommended