The primary objective of this study is to determine if emergency department patients with Transient Ischemic Attack (TIA) that are managed using a TIA "accelerated diagnostic protocol", or "ADP", demonstrate a significant decrease in their index visit length of stay and cost, with comparable diagnostic and 90-day clinical outcomes relative to TIA patients randomized to traditional inpatient care.

The secondary objectives are to evaluate the potential role of a TIA risk stratification tool and to determine the time to a diagnostic endpoint in both groups.

Transient ischemic attack (TIA) is defined as a neurological deficit lasting less than 24 hours, with most lasting less than one hour, brought on by focal cerebral or retinal ischemia. TIAs are common, with an estimated 300,000 events occurring annually and an estimated 5 million Americans having been given the diagnosis of TIA. Furthermore, a TIA is essentially the "smoke before the fire" for these patients. Studies have shown that within 90 days of an emergency department diagnosis of TIA, 10.5% of patients will suffer a stroke, with most occurring within two days of the ED visit. Twenty one percent of these strokes are fatal and 64% are disabling. Additionally, 2.6% of TIA patients will die, 2.6% will suffer other adverse cardiovascular events, and 12.7% will have recurrent TIAs. Roughly 15% of patients who have had a stroke report a history of TIA. For many patients, stroke is considered to be a devastating event that is worse than death. In 1999 stroke was the third leading cause of death in the United States (National center for health statistics, U.S. dept of health and human services). The national direct and indirect cost of stroke is estimated to be $51 billion annually.7

The management and disposition of emergency department patients with TIA is not entirely clear. It is mutually agreed upon in guidelines written by the American Heart Association (AHA) and the National Stroke Association (NSA) for the management of TIA, and standard emergency medicine textbooks, that patients with symptoms suggestive of a stroke or TIA require urgent evaluation in a setting such as the emergency department. Furthermore, that this evaluation should include a history, physical, and ECG. AHA guidelines and emergency medicine texts recommend that the initial evaluation include appropriate blood testing based on the history and CT imaging of the brain. All agree that patients with noncardioembolic causes of TIA should receive antiplatelet therapy and that TIA patients with atrial fibrillation should receive anticoagulation. "Prompt" or "Urgent" imaging of the carotid arteries to detect stenosis greater than 70% is also agreed upon since urgent carotid endarterectomy is believed to be most beneficial in this group. However the optimal timing of endarterectomy in patients with high-grade carotid stenosis is unclear. NSA guidelines and emergency medicine texts recommend hospitalization of patients with new onset TIA if imaging studies, such as carotid doppler, can not be performed urgently. However in a separate review of TIA management, hospitalization was identified as an area of uncertainty and it was suggested that management in a setting such as an Emergency Department Observation Unit (EDOU) might be a more cost effective alternative.

Inclusion Criteria:

• Transient ischemic attack confirmed by emergency physician - resolved deficit, not a crescendo TIA.
• Negative HCT

Exclusion Criteria:

• Positive HCT scan for bleed, mass, acute infarct
• Possible embolic source - A. fib, paroxysmal.A.fib, cardiomyopathy, artificial heart valve, endocarditis, patent foramen ovale, recent MI
• Known carotid stenosis (>50%) or mural thrombus
• Persistent acute neurological deficits
• Crescendo TIAs
• Non-focal symptoms - ie confusion, weakness, seizure, transient global amnesia
• Hypertensive encephalophy / emergency
• Severe headache or evidence of cranial arteritis
• Fever
• Previous stroke
• Severe dementia
• Nursing home patient
• Social issues that make discharge or follow up unlikely
• Other acute medical problems requiring inpatient admission
• Patient unlikely to survive beyond study follow up period (90 days)
• History of IV drug use