Hemorrhagic complications of ventriculostomy: incidence and predictors in patients with intracerebral hemorrhage.

Object: Ventriculostomy--the placement of an external ventricular drain (EVD)--is a common procedure performed in patients with acute neurological injury. Although generally considered a low-risk intervention, recent studies have cited higher rates of hemorrhagic complications than those previously reported. The authors sought to determine the rate of postventriculostomy hemorrhage in a cohort of patients with intracerebral hemorrhage (ICH) and to identify predictors of hemorrhagic complications of EVD placement.
Methods: Patients with ICH who underwent EVD placement and had both pre- and postprocedural imaging available for analysis were included in this study. Relevant data were prospectively collected for each patient who satisfied inclusion criteria. Variables with a p < 0.20 on univariate analyses were included in a stepwise logistic regression model to identify predictors of postventriculostomy hemorrhage.
Results: Sixty-nine patients were eligible for this analysis. Postventriculostomy hemorrhage occurred in 31.9% of patients. Among all patients with intraparenchymal hemorrhage, the mean hemorrhage volume was 0.66 ± 1.06 cm(3). Stratified according to ventricular catheter diameter, patients treated with smaller-diameter catheters had a significantly greater mean hemorrhage volume than patients treated with larger-diameter catheters (0.84 ± 1.2 cm(3) vs 0.14 ± 0.12 cm(3), p = 0.049). Postventriculostomy hemorrhage was clinically significant in only 1 patient (1.4%). Overall, postventriculostomy hemorrhage was not associated with functional outcome or mortality at either discharge or 90 days. In the multivariate model, an age > 75 years was the only independent predictor of EVD-associated hemorrhage.
Conclusions: Advanced age is predictive of EVD-related hemorrhage in patients with ICH. While postventriculostomy hemorrhage is common, it appears to be of minor clinical significance in the majority of patients.