Review Methods and Key Words
This scientific statement is intended for use by physicians and allied health personnel caring for patients with transient neurological symptoms resulting from brain, retinal, and spinal cord ischemia. A formal literature search was performed of the following Medline database: using the search strategy transient ischemic attack crossed with terms definition, epidemiology, incidence, prevalence, prognosis, recurrent stroke, diagnosis, imaging, magnetic resonance, diffusion, computed tomography, ultrasound, ECG, Holter, echocardiogram, and laboratory tests, covering the dates 1990 through June 2007. Writing panel members were each assigned topic areas and filtered the retrieved articles using the criteria identified in the Stroke Council’s Manual for Guidelines and Scientific Statements to identify high- or medium-quality studies of diagnostic tests and prognostic instruments. Data were synthesized through the use of evidence tables, meta-analyses, and pooled analysis of individual patient-level data. The American Heart Association (AHA)/American College of Cardiology/Stroke Council Levels of Evidence grading algorithm was used to grade each recommendation. Prerelease review of the draft guideline was performed anonymously by 3 expert peer reviewers, by the members of the Stroke Council’s Scientific Statements Oversight Committee, and by the members of the Stroke Council Leadership Committee.

TIA Epidemiology
Precise estimates of the incidence and prevalence of TIAs are difficult to determine mainly because of the varying criteria used in epidemiological studies to identify TIA. Lack of recognition by both the public and healthcare systems of the transitory focal neurological symptoms associated with TIAs also may lead to gross underestimates. Given these limitations, the incidence of TIA in the United States has been estimated to be {approx}200 000 to 500 000 per year, with a population prevalence of 2.3% that translates into {approx}5 million individuals.

Applying Classification of Recommendations and Level of Evidence

Table 2. Definition of Classes and Levels of Evidence Used in AHA Recommendations

TIA incidence rates have been projected from different study cohorts in the United States and abroad, ranging from 0.37 to 1.1 per 1000 per year. An overall TIA incidence rate of 1.1 per 1000 US population has been estimated on the basis of a review of the National Hospital Ambulatory Medical Care Survey among 2 623 000 TIA cases diagnosed in US emergency departments between 1992 and 2000. From the Greater Cincinnati/Northern Kentucky population between 1993 and 1994, the overall race-, age-, and gender-adjusted incidence rate for TIA was found to be 0.83 per 1000. Between the years 2002 and 2004, the Oxford Vascular Study determined that the overall incidence rate of TIA was 0.66 per 1000 per year.

Meanwhile, in rural and urban areas of Portugal, the crude overall annual incidence of TIA per 1000 population was found to be 0.67 and slightly higher in the rural region at 0.96 than in the urban area at 0.61. Comparable to stroke incidence, TIA incidence markedly increases with age and varies by race-ethnicity. Increased likelihood of TIA with advancing age was supported in recent UK studies, with 6.41 per 1000 for patients > 85 years of age. In the Greater Cincinnati/Northern Kentucky population, the greatest incidence of TIA occurred in black men ?85 of age at 16 events per 1000. The incidence of TIA increases exponentially with age regardless of race and gender. In addition, TIAs were found to be more common in Mexican Americans compared with non-Hispanic whites at younger ages (45 to 59 years) but not at older ages.

TIA prevalence rates vary, depending on the age distribution of the study population. For instance, the Cardiovascular Health Study estimated a prevalence of TIA in men of 2.7% for 65 to 69 years of age and 3.6% for 75 to 79 years of age. For women, TIA prevalence was 1.6% for 65 to 69 years of age and 4.1% for 75 to 79 years of age. In the younger Atherosclerosis Risk in Communities cohort, the overall prevalence of TIAs was found to be 0.4% among adults 45 to 64 years of age.

Among patients who present with stroke, the prevalence of prior TIA has been reported to range from 7% to 40%. The percentage varies, depending on such factors as how TIA is defined, which stroke subtypes are evaluated, and whether the study is a population-based series or a hospital-based series. In the population-based Northern Manhattan Stroke Study, the prevalence of TIAs among those who presented with first ischemic stroke was 8.7%. The majority of TIAs occurred within 30 days of the patient’s first ischemic stroke, with 41% of the TIAs lasting <1 hour.

Studies that have included patients with prior stroke such as the Harvard Stroke Registry and National Institute of Neurological Disorders and Stroke data bank have reported higher rates of TIAs as great as 50% among those with atherothrombotic stroke. In 2 population-based studies (Oxford Vascular Study and Oxfordshire Community Stroke Project) and 2 other randomized trials (UK TIA Aspirin Trial and the European Carotid Surgery Trial), the timing of a TIA before stroke was highly consistent, with 17% occurring on the day of the stroke, 9% on the previous day, and another 43% at some point during the 7 days before the stroke. In another population-based study that was biethnic with Mexican Americans and non-Hispanic whites, approximately half of the 90-day stroke risk for TIA occurred within the first 2 days, suggesting that in general TIA patients are at very high risk for a recurrent cerebrovascular event (see TIA: Short-Term Stroke Risk below).