Why a TIA is an emergency, not a reassurance
A transient ischaemic attack is often called a mini-stroke, and the diminutive is dangerously misleading. By definition the symptoms resolve, usually within minutes, and the patient walks out of the episode feeling normal. But the days immediately after a TIA carry the highest risk of a completed, disabling stroke. In the cohorts that produced the ABCD2 score, roughly one in twelve patients with the highest scores had a stroke within two days. A TIA is therefore not a lucky escape; it is a warning with a short fuse, and it demands the same urgency as the stroke it may precede.
The practical problem is triage. Not every TIA patient can be assessed by a stroke specialist within hours, yet some need exactly that. A clinical prediction score that separates the highest-risk patients from the lowest-risk ones lets health systems direct urgent resources, same-day TIA clinics, immediate brain and carotid imaging, and cardiac monitoring, to the patients most likely to benefit. That is the job the ABCD2 score was built to do.
What the ABCD2 score is
The score was developed and validated by S. Claiborne Johnston and colleagues and published in The Lancet in 2007 (Johnston SC, Rothwell PM, Nguyen-Huynh MN, et al. Validation and refinement of scores to predict very early stroke risk after transient ischaemic attack. Lancet. 2007;369(9558):283-292). The team combined data from TIA cohorts in California and in Oxfordshire, England, first validating the four-item ABCD score proposed by Rothwell and then refining it by adding diabetes as a fifth item, producing ABCD2.
The five items are deliberately simple and available at the bedside: age, blood pressure, the clinical features of the episode, how long the symptoms lasted, and whether the patient has diabetes. Each contributes points to a total from 0 to 7. The score does not require blood tests or imaging, which is precisely why it is useful in the emergency department and in primary care at the moment the TIA patient first presents.
The criteria explained
- Age 60 years or older: 1 point. Older age is one of the most consistent predictors of early stroke after TIA. The cutoff is a single birthday: 60 or above scores the point, 59 or below does not.
- Blood pressure 140/90 mmHg or higher: 1 point. Either the systolic or the diastolic value reaching the threshold is enough; both do not need to be elevated. Use the presenting blood pressure, recognising that blood pressure fluctuates and that a single reading is what the score was validated with.
- Clinical features: unilateral weakness 2 points, speech disturbance without weakness 1 point. This is the heaviest-weighted item. Unilateral weakness means weakness of the face, arm, or leg on one side of the body. Speech disturbance without weakness means aphasia (trouble producing or understanding language) or dysarthria (slurred speech) with no weakness present. If neither feature occurred, no points are scored.
- Duration of symptoms: 60 minutes or more 2 points, 10 to 59 minutes 1 point, less than 10 minutes 0 points. Longer episodes predict higher risk, which fits the biology: a longer interruption of blood flow suggests a less stable underlying lesion. Ask carefully, because patients often misjudge duration; witness accounts help.
- Diabetes mellitus: 1 point. Diabetes was the refinement that turned ABCD into ABCD2. In the derivation cohorts it independently predicted early stroke, so its presence adds a point regardless of how well controlled it is.
Add the points for a total between 0 and 7. The calculator above does the arithmetic and reports the corresponding 2-day stroke risk band.
How to use this calculator
Enter the patient's age and presenting blood pressure as numbers, then select the clinical features, the symptom duration, and diabetes status. The calculator requires every item: if any is missing or invalid, it will name the problem rather than guessing. Press Calculate to see the total score, the items that contributed points, the risk band, and the 2-day stroke risk percentage. Use the Reset button to clear the form for the next patient.
Interpreting the result
In the Johnston 2007 validation, the 2-day stroke risk was 1.0% for scores of 0 to 3, 4.1% for scores of 4 to 5, and 8.1% for scores of 6 to 7. These bands guide urgency rather than dictating treatment. A score of 6 or 7 identifies a patient at high short-term risk who should be assessed emergently, ideally the same day, with prompt brain imaging, carotid imaging, and cardiac rhythm assessment according to local stroke pathways. A score of 4 or 5 indicates intermediate risk needing urgent assessment, usually within 24 hours. A score of 0 to 3 indicates lower risk, but 1.0% is not zero: these patients still need urgent assessment, just with slightly less immediacy than the higher bands.
Two points of interpretation deserve emphasis. First, the score estimates risk, it does not diagnose the cause of the TIA or select treatment. A high score does not by itself indicate thrombolysis or any specific drug; those decisions rest on imaging, the clinical syndrome, and specialist judgment. Second, the score is an adjunct to, never a replacement for, urgent referral. Every TIA, whatever the score, should be referred urgently through local pathways, and patients should be advised not to drive and to seek emergency care immediately if symptoms recur.
How accurate is the ABCD2 score?
The score's strength is its consistent ordering of risk across many independent validation studies in different countries and health systems: higher scores reliably carry higher early stroke risk, and the three bands have held up outside the derivation cohorts. The absolute percentages vary somewhat between populations, which is expected, since background stroke risk differs, but the gradient from the lowest to the highest band is the reproducible finding.
Its weakness is moderate discrimination. The score separates groups well but predicts individuals imperfectly: most patients with high scores do not have a stroke within two days, and a minority of strokes occur in patients with low scores. Later refinements such as ABCD3 and ABCD3-I added items including dual TIA episodes, carotid stenosis, and diffusion-weighted imaging findings to improve prediction, but they require information that is not available at the first bedside assessment, which is why ABCD2 remains the practical front-door tool. The honest reading is that ABCD2 is a triage aid with solid evidence, not a crystal ball.
Limitations
- Derived for TIA, not completed stroke. The score was built and validated in patients with transient symptoms. Do not apply it to completed stroke, where the diagnosis is already made and different management applies.
- A low score is not a clean bill of health. A 1.0% 2-day stroke risk is low, not negligible, and the score says nothing about longer-term risk. Low-score patients still need urgent assessment and secondary prevention.
- Important predictors are missing. The score does not include atrial fibrillation, carotid stenosis, imaging findings, or symptom fluctuation, all of which strongly influence real-world decisions. A patient with known atrial fibrillation or a crescendo TIA pattern needs urgent specialist assessment whatever the score says.
- TIA is a clinical diagnosis with mimics. Migraine aura, focal seizures, functional symptoms, and hypoglycaemia can all mimic TIA. The score assumes the diagnosis is correct; applied to a mimic, its risk estimates are meaningless.
- Blood pressure is a single snapshot. Blood pressure fluctuates, and the score uses the presenting reading. Do not game the score by re-measuring until the number looks better.
- No score replaces urgent referral. If the presentation suggests an ongoing stroke, or the patient deteriorates, activate emergency stroke pathways immediately rather than pausing to calculate.