Changing Patterns of Transient Ischemic Attack Evaluation in US Emergency Departments

Importance: For US patients presenting to the emergency department (ED) with transient ischemic attack (TIA), guidelines recommend resource-intensive multimodal evaluation to optimize secondary stroke prevention. These resources are not available in all EDs, and national patterns of ED TIA care are not well described. Objectives: To evaluate national trends in ED TIA care, including discharge practices, imaging use, and associated costs, and to assess patient- and hospital-level factors associated with these patterns. Design, Setting, and Participants: This was a retrospective repeated cross-sectional study of ED visits using the 2013-2021 National Emergency Department Sample of the Healthcare Cost and Utilization Project, a weighted, nationally representative annual sample of ED visits in the US. TIA and stroke encounters during the study period were identified among patients aged 18 years or older by a first-listed International Classification of Diseases, Ninth Revision or Tenth Revision code for TIA or stroke diagnosis. Analyses were performed from October 2024 to April 2026. Exposures: Calendar year. Main Outcomes and Measures: ED discharge; imaging use among discharged TIA encounters; and inflation-adjusted ED cost per discharged encounter. Results: This study included 2 151 408 (95% CI, 2 107 855-2 194 960) weighted ED visits for TIA (1 194 630 [55.5%] female; mean [SEM] age, 69.8 [0.05] years). Adjusted models found the predicted probability of discharge increased from 45% (95% CI, 43%-46%) in 2013 to 66% (95% CI, 64%-68%) in 2021 (P < .001 for trend). During this time, mean ED charges increased from $1746 (95% CI, $1662-$1830) to $2755 (95% CI, $2674-$2836) per encounter (P < .001 for trend). Imaging use among discharged TIA encounters increased substantially, including computed tomography angiography (predicted probability, 4% [95% CI, 3%-4%] in 2013 vs 45% [95% CI, 42%-47%] in 2021; P < .001 for trend) and magnetic resonance imaging (predicted probability, 17% [95% CI, 16%-19%] in 2013 vs 53% [95% CI, 50%-55%] in 2021; P < .001 for trend). Rate of discharge was higher in rural encounters than in urban encounters for all years studied, and the rate of magnetic resonance imaging use was lower. Conclusions and Relevance: This cross-sectional study demonstrates a shift in TIA care over time that suggests increased use of the ED to expedite evaluation for TIA that previously occurred during a brief hospitalization. This approach has consequences in terms of cost and ED throughput and may not be as feasible in rural hospitals as it is in urban hospitals.