Objective: To characterize temporal trends in the use of endovascular treatment (EVT) for acute ischemic stroke at academic medical centers in response to recent clinical trials.
Background: Although endovascular devices for stroke were first cleared for marketing in 2004, initial clinical trials in 2013 failed to demonstrate efficacy and subsequent clinical trials beginning in 2014 were strongly positive. The impact of these data on practice patterns at academic medical centers, which perform most EVTs, is unknown.
Methods: We identified all acute ischemic stroke hospitalizations at academic medical centers that were members in the University HealthSystem Consortium from October 2009 to July 2015 using International Classification of Disease, 9th revision codes 433.x1, 434.x1, and 436 for stroke and procedure code 39.74 for EVT. We compiled quarterly data on the number and proportion of stroke hospitalizations using EVT and we used segmented log-linear regression to identify temporal trends and to evaluate changes in trends at prespecified time points corresponding to the quarter in which pivotal trials were first reported.
Results: From 2009-15, we identified 357,973 acute ischemic stroke hospitalizations at 161 medical centers. The proportion of stroke hospitalizations using EVT was 1.5% in 2009 and grew by 25% a year (95% CI 21% to 29%) to reach 3.1% in 2013. After negative results from the initial trials were reported in 2013, EVT use hovered between 2.5% and 2.7% (1% relative change per year; 95% CI -9% to +8%; p=0.004 for change in trend) until 2014 when the first positive trials were reported and EVT use jumped at a growth rate of 151% per year (95% CI 101% to 212%; p<0.001 for change in trend) to reach fully 4.7% of all stroke hospitalizations by 2015.
Conclusion: The previously steady growth in EVT flattened in 2013, coincident with the initially negative results from clinical trials, but has dramatically increased since positive trials were first reported in 2014.