The New York Islands AVM Study: Detection rates for brain AVM and incident AVM hemorrhage

Abstract

P159 Background: Population-based data on the incidence of brain AVM hemorrhage is scarce. We studied detection rates of brain AVM and AVM hemorrhage in a defined population. Methods: The New York islands (Manhattan, Staten Island, and Long Island) are a ZIP-code defined area comprising a 8,898,000 population. In a retrospective investigation, major NY islands hospital centers retrieved the number of patients who lived in the study area and were discharged with a diagnosis of brain AVM between 1996 and 1999. Starting March 15, 2000, all major NY islands hospitals prospectively reported data on consecutive NY islands patients with a diagnosis of brain AVM and whether or not the patient had suffered AVM hemorrhage. Referral patients living outside the study area were excluded from the sample. Results: The retrospective detection rate of brain AVMs in patients living in the study area (Table 1) was estimated to be 1.2/100,000 person-years (95% CI:1.1–1.4). As of July 14, 2000, 37 prospective AVM patients were encountered leading to a calculated AVM detection rate of 1.25/100,000 pers-yrs (95% CI: 0.9–1.7) with an estimated incidence for first-ever AVM hemorrhage (n=6) of 0.2/100,000 pers-yrs (95% CI:0.1 to 0.5). The prevalence of AVM hemorrhage (n=14) was 0.5/100,000 pers-yrs (95% CI:0.3 to 0.8). Conclusions: Our preliminary data suggest similar prospective and retrospective AVM detection rates. More than one third of AVM patients may suffer intracranial hemorrhage.