The importance of aspirin, catheterization accuracy, and catheter design in external ventricular drainage-related hemorrhage: a multicenter study of 1002 procedures

Background: External ventricular drainage (EVD) is the commonest neurosurgical procedure performed in daily neurosurgical practice, but relatively few studies have investigated the incidence and risk factors of its related hemorrhagic complications.

Methods: This was a multicenter retrospective review of consecutive EVD procedures. Patients 18 years or older who underwent EVD and had a routine postoperative computed tomography (CT) scan performed within 24 hours were included. EVD-related hemorrhage was defined as new intracranial hemorrhage immediately adjacent or within the ventricular catheter trajectory. The volume of hemorrhage and the position of the catheter tip were assessed. A review of patient-, disease-, and surgery-related factors including the ventricular catheter design utilized was conducted. The Bonferroni correction was applied to the alpha level of significance (0.05) for multivariable analysis.

Results: Nine hundred sixty-two patients underwent 1002 EVD performed by neurosurgeons in the operating theater. Sixteen percent (154) of patients were on aspirin before the procedure. Thirty-four percent (333) of patients had intracerebral hemorrhage, 25% (251) had aneurysmal subarachnoid hemorrhage and 16% (158) had traumatic brain injury. The mean duration from EVD to the first postoperative CT scan was 20 ± 4 h. EVD-related hematomas were detected after 81 procedures with a per-catheter risk of 8.1%. Mean hematoma volume was 1.2 ± 3.3 ml. Most were less than 1 ml (grade I, 79%, 64), 1 to 15 ml (grade II) in 20% (16) and a single clot larger than 15 ml (grade III, 1%) were detected. Clinically significant hemorrhage that resulted in catheter occlusion occurred in 1.7% (17) of procedures. Most catheters (62%, 625) were optimally placed, i.e., its tip being within the ipsilateral frontal horn or third ventricle. Three non-antibiotic-impregnated ventricular catheter designs were used with 55% (550) being the 2.2-mm Integra™ catheter, 14% (137) being the 2.8-mm Medtronic™ catheter, and 31% (315) being the 3.1-mm Codman™ catheter. Independent significant predictors for EVD-related hemorrhage were the preoperative prescription of aspirin (adjusted OR 1.94; 95% CI 1.10-3.44), catheter malposition (aOR 1.99; 95% CI 1.22-3.23), and use of the 2.8-mm Medtronic™ catheter (aOR 4.22; 95% CI 2.39-7.41).

Conclusions: The per-catheter risk of hemorrhage was 8.1%, but the incidence of symptomatic hemorrhage was low. The only patient risk factor was aspirin intake. This is the first study to evaluate and establish an association between catheter malposition and catheter design with EVD-related hemorrhage.