A TIA is a warning delivered in the only language the brain has. The symptoms resolve; the risk does not. ABCD2 was built to size that risk in the minutes after the patient arrives, from things you already have — an age, a blood pressure, what the deficit was, how long it lasted, and whether they have diabetes.1

At a glance · what the score changes

  • 0–3 · roughly 1% stroke risk within 2 days in the original validation cohorts. Not a discharge criterion — later cohorts found the score discriminates inconsistently, and strokes occur in this band.
  • 4–5 · roughly 4%. 6–7 · roughly 8% — close to 1 in 12 at the top of the range.
  • At 4 or more, the patient meets the ABCD2 component of the high-risk definition CHANCE and POINT used to enrol patients for short-course dual antiplatelet therapy, started early after the event.
  • Course length matters. POINT showed major bleeding rises when the combination is carried on too long, which is why guideline-recommended courses are short rather than indefinite.
  • At any score, urgent brain imaging, imaging of the neck and intracranial vessels, and rhythm assessment are warranted. Carotid stenosis, atrial fibrillation and a cardiac clot source carry high recurrence risk wherever the number lands.

The five components

  • A — Age: 60 years or older = 1 point.
  • B — Blood pressure: systolic ≥ 140 mm Hg or diastolic ≥ 90 mm Hg at presentation = 1 point.
  • C — Clinical features: unilateral weakness = 2 points; speech disturbance without weakness = 1 point; neither = 0.
  • D — Duration: 60 minutes or longer = 2 points; 10–59 minutes = 1 point; under 10 minutes = 0.
  • D — Diabetes: present = 1 point.

Everything on that list is available within minutes of arrival — a history, a cuff, and a quick exam. That is why the score spread through emergency departments as fast as it did.

Risk by band

  • 0–3: roughly 1% stroke risk within 2 days.
  • 4–5: roughly 4% within 2 days.
  • 6–7: roughly 8% within 2 days.

The same climbing pattern held at 7 and 90 days.1 The shape of the curve is the useful part; the exact percentages come from cohorts that predate current secondary-prevention practice.

Pitfall

Using ABCD2 as an admit-or-discharge rule is where this score does harm. External validation has been inconsistent, and the causes that drive early recurrence — tight carotid stenosis, atrial fibrillation, a cardiac clot source — are exactly the ones the score does not measure. A patient with a score of 2 and a 90% carotid stenosis is not low risk.

Where it still earns its keep

Selecting candidates for short-course dual antiplatelet therapy. CHANCE and POINT both used an ABCD2 of 4 or more as part of the high-risk definition when enrolling patients with TIA or minor ischemic stroke, and both found that a brief course of aspirin plus clopidogrel reduced early recurrent stroke compared with aspirin alone.23 POINT also showed the cost of carrying the combination too long — more major bleeding — which is why current AHA/ASA secondary-prevention guidance frames it as a short course started early in selected patients.4

That is the score doing what it is good at: flagging a group, not making a decision. For sharper discrimination of 7-day risk, the Canadian TIA Score adds the mechanisms ABCD2 leaves out — atrial fibrillation, carotid disease, infarction on CT. The longer discussion of how ABCD2 has held up since 2007 is in the full ABCD2 review.

If you are a patient reading this

TIA symptoms that went away are still an emergency. Sudden weakness or numbness on one side, trouble speaking or understanding, a drooping face, vision loss, or sudden severe imbalance — even lasting only minutes — means calling emergency services (911 in the US), not calculating a score at home. Risk is highest in the first days, which is exactly when evaluation changes the outcome.