Background The Face Arm Speech Test (FAST) is used in public education to help people recognise stroke symptoms. It was designed to identify major stroke, however it is uncertain how well the test identifies those with TIA and minor stroke, in whom acute prevention is required.

Methods In a population-based study, we categorised consecutive patients with TIA and stroke (minor stroke=NIHSS≤5) using FAST in relation to delay in seeking medical attention and early recurrent stroke risk.

Results Of 2183 patients (1000 TIA,775 minor stroke,408 major stroke), 1432(66%) were FAST positive. Of patients with TIA or minor stroke, 127 had a recurrent stroke within 7-days (7.2%, 95% CI 6.0 to 8.4). A positive FAST did not predict early recurrent stroke (7-day risk=7.6% vs 6.5%, p=0.4; sensitivity=62.2%, 53.1–70.7, specificity=41%, 39.1–43.9). The test also had limited potential to improve access to care, being positive in only 130/309 (42%) of TIA patients who delayed seeking medical attention >24-h and in only 35 (38.6%) of a further 81 patients who did not seek medical attention until after a recurrent stroke. Moreover, there is a potential for false reassurance in FAST negative patients, 435/685 of whom did seek attention within 24-h.

Conclusion FAST does not identify patients with TIA or minor stroke and therefore has limited potential to improve secondary prevention. One unintended consequence of the FAST campaign is that patients with TIA and minor stroke may be falsely reassured by a negative test and may delay seeking medical attention or not present at all.