BACKGROUND: Ventriculostomy-associated cerebrospinal fluid infection (VAI) is a major complication limiting the use of an external ventricular drain (EVD) in treating patients with intraventricular hemorrhage (IVH). Risk factors of VAI are still under wide discussion. METHODS: We performed a retrospective review of 84 patients with IVH who underwent EVD at our center between January 2012 and January 2017. Preoperative clinical parameters, surgeon status, number of catheters and catheter-days, subgaleal tunneling distance, frequency of urokinase flush, and prophylactic antibiotics were compared between the infective and noninfective groups. RESULTS: The overall rate of VAI was 31.0%. Univariate analysis showed a higher modified Graeb Score (mGS), higher proportion of bilateral catheters, and longer hospital stay in patients with VAI. Binary logistic analysis of all clinical factors identified high mGS (≥16) as an independent risk factor for VAI (odds ratio, 3.242; P = 0.026). Among operative and postoperative factors, the use of bilateral catheters significantly contributed to VAI (odds ratio, 4.211; P = 0.031), but a subgroup comparison showed an increased VAI rate only in the low mGS group (mGS <15). No VAI occurred in patients with a single EVD in the low mGS group. Catheter-days and multiple urokinase flushes were not related to VAI. CONCLUSIONS: Patients with a high mGS are vulnerable to VAI. Bilateral EVD may be an appropriate treatment option for patients with a high mGS, but might increase the risk of infection in those with a low mGS.
BACKGROUND: Ventriculostomy-associated cerebrospinal fluid infection (VAI) is a major complication limiting the use of an external ventricular drain (EVD) in treating patients with intraventricular hemorrhage (IVH). Risk factors of VAI are still under wide discussion. METHODS: We performed a retrospective review of 84 patients with IVH who underwent EVD at our center between January 2012 and January 2017. Preoperative clinical parameters, surgeon status, number of catheters and catheter-days, subgaleal tunneling distance, frequency of urokinase flush, and prophylactic antibiotics were compared between the infective and noninfective groups. RESULTS: The overall rate of VAI was 31.0%. Univariate analysis showed a higher modified Graeb Score (mGS), higher proportion of bilateral catheters, and longer hospital stay in patients with VAI. Binary logistic analysis of all clinical factors identified high mGS (≥16) as an independent risk factor for VAI (odds ratio, 3.242; P = 0.026). Among operative and postoperative factors, the use of bilateral catheters significantly contributed to VAI (odds ratio, 4.211; P = 0.031), but a subgroup comparison showed an increased VAI rate only in the low mGS group (mGS <15). No VAI occurred in patients with a single EVD in the low mGS group. Catheter-days and multiple urokinase flushes were not related to VAI. CONCLUSIONS:Patients with a high mGS are vulnerable to VAI. Bilateral EVD may be an appropriate treatment option for patients with a high mGS, but might increase the risk of infection in those with a low mGS.