Abstract T P189: Time Trends in Transient Ischemic Attack Admission From the Emergency Department

Abstract

Background: Prior studies suggest transient ischemic attack (TIA) admission from the Emergency Department (ED) increased over time. We investigated how TIA admission trends may have changed with publication of TIA risk stratification rules involving the ABCD /ABCD2 scores. Methods: We used ED data on TIA from the National Hospital Ambulatory Medical Care Survey. We defined three time periods to account for the initial publication by Johnston et al., highlighting the risk of ischemic stroke following TIA (2000) and related reports on the ABCD (2005) and ABCD2 (2007) scores. We defined 2005 through 2007 as a washout period to allow for any practice change that occurred with the evolution of ABCD/ABCD2 scores. Admission trends were stratified across younger (18-64y) and older (65y+) patients. Results: There were 327,339 TIA visits yearly from 1994 to 2010 in adults 18+ years. We compared three distinct time periods (1994-1999; 2001-2004 and 2008-2010 to reflect the timing of publications above. We observed interaction between time and age with regard to TIA admission (Figure). TIA patients 18-64y showed a decreasing admission trend overall (p = 0.047). Patients 65y+ showed a similar decreasing trend up to 2005-2007, but after 2007 had a marked increase in TIA admissions (p = 0.021). A comparison of stroke admissions during the same time period did not show an increasing trend in the 65+ group. Conclusion: Contrary to prior reports, TIA admissions decreased over time among persons 18-64y. Among older persons 65+, there was a sharp increase in TIA admission after 2007 that was not due to an increase in admissions for stroke or stroke-related diagnosis. The increase in TIA admissions could be related to new tools for the risk-stratification of TIA patients, but a large percentage of TIA patients are discharged from the ED. More research is needed to ensure that all high risk TIA patients receive a rapid evaluation.