BACKGROUND: Spontaneous intracerebral hemorrhage (ICH) commonly presents with intraventricular hemorrhage (IVH) and remains a highly disabling form of stroke. External ventricular drains (EVDs) are associated with decreased short-term mortality, but indications for use and outcomes benefit are controversial. METHODS: A multi-institutional, retrospective analysis of 563 patients with spontaneous ICH from 2010 to 2014 was performed with multivariate regression modeling. Primary outcomes were patient mortality and functional status with modified Rankin Scale score. To control for differences in patient and clinical characteristics influencing EVD utilization, a propensity score analysis was performed with patient-specific predicted probability of EVD use. RESULTS: The multivariable logistic regression model showed odds of EVD use increased with younger age, lower ICH volume, ICH located outside the brainstem, increasing IVH volume, and concurrent IVH; the model showed high discriminability for EVD use (area under the receiver operating curve 0.84, R2McFadden = 0.27). The use of EVD was associated with lower 30-day mortality in patients with ICH score of 4 (odds ratio = 0.09, P = 0.002), greater ICH volume (>11 cc, odds ratio = 0.47, P = 0.019), and lower initial GCS (<13, 0.38, P = 0.003) in propensity score-adjusted analyses, as well as a trend toward lower mortality in patients with IVH and greater modified Graeb score. There was no benefit to morbidity in patients receiving an EVD. CONCLUSIONS: Among a large, multi-institutional cohort, this statistical propensity analysis model accurately predicted EVD use in ICH. EVD use was associated with a trend towards decreased mortality but greater modified Rankin Scale score for functional outcomes.
BACKGROUND: Spontaneous intracerebral hemorrhage (ICH) commonly presents with intraventricular hemorrhage (IVH) and remains a highly disabling form of stroke. External ventricular drains (EVDs) are associated with decreased short-term mortality, but indications for use and outcomes benefit are controversial. METHODS: A multi-institutional, retrospective analysis of 563 patients with spontaneous ICH from 2010 to 2014 was performed with multivariate regression modeling. Primary outcomes were patient mortality and functional status with modified Rankin Scale score. To control for differences in patient and clinical characteristics influencing EVD utilization, a propensity score analysis was performed with patient-specific predicted probability of EVD use. RESULTS: The multivariable logistic regression model showed odds of EVD use increased with younger age, lower ICH volume, ICH located outside the brainstem, increasing IVH volume, and concurrent IVH; the model showed high discriminability for EVD use (area under the receiver operating curve 0.84, R2McFadden = 0.27). The use of EVD was associated with lower 30-day mortality in patients with ICH score of 4 (odds ratio = 0.09, P = 0.002), greater ICH volume (>11 cc, odds ratio = 0.47, P = 0.019), and lower initial GCS (<13, 0.38, P = 0.003) in propensity score-adjusted analyses, as well as a trend toward lower mortality in patients with IVH and greater modified Graeb score. There was no benefit to morbidity in patients receiving an EVD. CONCLUSIONS: Among a large, multi-institutional cohort, this statistical propensity analysis model accurately predicted EVD use in ICH. EVD use was associated with a trend towards decreased mortality but greater modified Rankin Scale score for functional outcomes.
Authors: Quincy K Tran; Sagar Dave; Daniel J Haase; Laura Tiffany; Shannon Gaasch; Wan-Tsu W Chang; Kevin Jones; Matthew J Kole; Aaron Wessell; Gary Schwartzbauer; Thomas M Scalea; Jay Menaker Journal: West J Emerg Med Date: 2021-01-12
Authors: Andrew D Warren; Qi Li; Kristin Schwab; Brenna McKaig; Alexa N Goldstein; Steven M Greenberg; Anand Viswanathan; Christopher Anderson; M Edip Gurol; Aman Patel; Joshua N Goldstein Journal: Int J Emerg Med Date: 2022-09-15