Overdiagnosis of TIA and minor stroke: experience at a regional neurovascular clinic

We compared the referral diagnoses of TIAs and minor strokes made by non-specialists with those of two consultant neurologists, in 565 consecutive cerebrovascular clinic patients, of whom 508 (90%) were referred with a diagnosis of any TIA or stroke. In 373 (73%), the neurologists felt the diagnosis of a cerebrovascular event to be correct. Agreement with the vascular syndrome (CVA vs. TIA) was significantly higher for patients with a referral diagnosis of stroke (136/176) (77%) than it was for patients with a referral diagnosis of TIA (200/332) (60%) (difference in proportions 17%, 95% CI 9-25). In 37 patients (7%) the neurologists confirmed the diagnosis of cerebrovascular disease but not the specific TIA/stroke diagnosis. Vascular surgeons were more likely to be correct in their referral diagnosis of carotid territory cerebrovascular disease (88% correct) than all other sources combined (63% correct) (difference in proportions 25%, 95% CI 11-39), but there was no significant variation in diagnostic accuracy between other individual groups. In 135/508 patients (27%) referred as any TIA or stroke, the diagnosis of cerebrovascular disease was undone. Alternative diagnoses included migraine (3%), epilepsy (1%), hyperventilation (1%), multiple sclerosis (1%) and a case of idiopathic Parkinson's disease, but many symptoms (8%) were unclassifiable. A strict comparison of diagnostic accuracy would have required assessment of patients not referred for specialist opinion, to estimate false-negative as well as false-positive diagnoses. However, in this patient group (which reflects current local practice) TIAs and strokes seem overdiagnosed.