Patterns of brain arteriovenous malformation treatment: prospective, population-based study.

BACKGROUND AND PURPOSE The extent of variation in the interventional treatment of brain arteriovenous malformations (AVMs) is unknown, so we explored patterns of treatment at 4 neuroscience centers in one European country. METHODS We included every participant with an AVM in a prospective, population-based cohort study of adults aged >or=16 years residing in Scotland at the time of AVM diagnosis in 1999 to 2003. RESULTS Only 11 (5%) of the 229 adults were not managed at a neuroscience center. Adults who received interventional treatment were younger (median, 43 versus 54 years), more likely to have presented with hemorrhage (OR, 2.8; 95% CI, 1.6 to 4.9), and had smaller AVMs (median nidus diameter, 2 cm versus 3 cm; P=0.003) than those who did not. Adults seen at the 4 centers only differed in AVM Spetzler-Martin grade (P=0.04). The 4 centers did not differ in the proportion of adults with AVMs who received interventional treatment (P=0.16), but they differed in the Spetzler-Martin grade of the AVMs they treated (Grades III to IV, P=0.01) and the interventional treatments used (P=0.004). The 2 largest centers differed from each other in the likelihood of surgical resection (OR, 0.2; 95% CI, 0.1 to 0.6) and stereotactic radiosurgery (OR, 2.8; 95% CI, 1.3 to 6.1), and the choice of modality varied within some Spetzler-Martin grades. CONCLUSIONS Patient characteristics and patterns of AVM interventional treatment differ between neuroscience centers in the same population necessitating careful consideration of these factors when comparing one hospital's outcome with another.