Michael A Silva1,2, Pui Man Rosalind Lai1, Rose Du1, Mohammad A Aziz-Sultan1, Nirav J Patel1. 1. Department of Neurosurgery, Brigham and Women's Hospital, School of Medicine, Harvard University, Boston, Massachusetts. 2. Department of Neurosurgery, Jackson Memorial Hospital, Miller School of Medicine, University of Miami, Miami, Florida.
Abstract
BACKGROUND: Arteriovenous malformation (AVM) rupture is highly morbid. Outcomes after AVM rupture differ from other types of brain hemorrhage. There are no specific widely used grading systems designed to predict clinical outcome after AVM rupture. OBJECTIVE: To develop an all-comers scoring system to grade patients with AVM rupture and predict clinical outcome more accurately than grading systems currently in use. METHODS: We retrospectively reviewed patients who presented to our institution with a ruptured AVM. Using change in modified Rankin Score (mRS) as our response variable, we generated an ordinal logistic regression model to test for significant predictor variables. The full model was sequentially condensed until the simplest model with the highest area under the receiver operating curve (AUROC) was achieved. RESULTS: A total of 115 patients who presented with ruptured AVMs were included in the study, with a mean follow-up time of 4 yr. The Ruptured AVM Grading Scale (RAGS) consists of the Hunt and Hess (HH) score (1-5), patient age (<35 = 0, 35-70 = 1, and >70 = 2), deep venous drainage (1), and eloquence (1). The RAGS score outperformed other neurosurgical grading scales in predicting change in mRS, with an AUROC greater than 0.80 across all follow-up periods. CONCLUSION: The RAGS score is a simple extension of the HH scale that predicts clinical outcome after AVM rupture more accurately than other grading systems.
BACKGROUND: Arteriovenous malformation (AVM) rupture is highly morbid. Outcomes after AVM rupture differ from other types of brain hemorrhage. There are no specific widely used grading systems designed to predict clinical outcome after AVM rupture. OBJECTIVE: To develop an all-comers scoring system to grade patients with AVM rupture and predict clinical outcome more accurately than grading systems currently in use. METHODS: We retrospectively reviewed patients who presented to our institution with a ruptured AVM. Using change in modified Rankin Score (mRS) as our response variable, we generated an ordinal logistic regression model to test for significant predictor variables. The full model was sequentially condensed until the simplest model with the highest area under the receiver operating curve (AUROC) was achieved. RESULTS: A total of 115 patients who presented with ruptured AVMs were included in the study, with a mean follow-up time of 4 yr. The Ruptured AVM Grading Scale (RAGS) consists of the Hunt and Hess (HH) score (1-5), patient age (<35 = 0, 35-70 = 1, and >70 = 2), deep venous drainage (1), and eloquence (1). The RAGS score outperformed other neurosurgical grading scales in predicting change in mRS, with an AUROC greater than 0.80 across all follow-up periods. CONCLUSION: The RAGS score is a simple extension of the HH scale that predicts clinical outcome after AVM rupture more accurately than other grading systems.