IntroductionVentriculostomy related infection (VRI) or ventriculitis is a common and serious complication related to the placement of an external ventricular drain. Numerous sets of diagnostic criteria for VRI have been reported. We sought to estimate the variation in the incidence of VRI in a cohort of patients according to published diagnostic criteria.Materials and MethodsWe conducted a retrospective cohort study. We included adult patients admitted to the Neuroscience intensive care unit with traumatic brain injury (TBI), subarachnoid haemorrhage (SAH) and intracerebral haemorrhage (ICH) who required an EVD. We estimated the incidence of VRI according to published diagnostic criteria. We compared the incidence to clinicians’ diagnoses of VRI. The primary outcome was the estimated incidence of VRI.ResultsThere were 190 study participants, median age (interquartile range) of 58 (48 - 72), 106 (55.8%) were female. Admitting diagnoses was ICH in 30 (15.8%), TBI in 49 (25.8%) and SAH in 111 (58.4%) of cases respectively. There were 158 (83.2%) who required mechanical ventilation for a median of 6 (2-13) days. There were 29 (15.3%) who were treated for VRI by clinicians, with 6 (3.2%) having a positive culture. Variation in the diagnostic criteria led to an estimated incidence of VRI that ranged from 1 (0.5%) to 178 (93.7%).ConclusionIn this critically ill cohort, the estimated incidence of VRI varied widely depending on which diagnostic criteria for VRI were applied. A comprehensive, consistent, objective and universal set of diagnostic criteria for ventriculostomy related infection is needed.