[False diagnoses of strokes in emergency departments]

Diagnosing a stroke can sometimes be difficult. There are a number of mimic conditions that can lead to false diagnoses.

To examine false diagnoses of acute stroke.

We reviewed the medical histories with diagnoses of acute stroke -i.e. ischaemic or haemorrhagic stroke and transient ischaemic attack (TIA)- for a three-month period. Alternative diagnoses were established in doubtful stroke cases (without meeting the World Health Organisation stroke criteria).

Altogether there were 358 patients: 110 TIA, 191 ischaemics and 57 haemorrhagics. In all, 65 false diagnoses were selected, which represented 18.2% of the total number (41.8% of the cases of TIA) and 31.8% of the strokes admitted in the emergency department. The subtypes of false diagnoses were: 46 TIA (70.8%), 18 ischaemics (27.7%) and one haemorrhagic (1.5%). The alternative diagnoses were the following: syncope/pre-syncope in 10.8% of cases (n = 7); confusional syndrome/disorientation in 21.5% (n = 14); lowered level of consciousness in 27.7% (n = 18); generalised weakness in 6.2% (n = 4); dizziness/vertigo in 3.1% (n = 2); isolated dysarthria in 10.8% (n = 7); epileptic seizure in 6.2% (n = 4); and others in 13.8% (n = 9). A total of 71.7% could be attributed to systemic causes. The mean age was 79 years and 64.6% were females (n = 42). Computerised tomography of the head was performed in 70.8% of the cases (n = 46). A neurologist assessed 7.7% of them (n = 5). The destination on being discharged was: primary care (53.3%), visit to neurology department (31.7%), visit internal medicine department (6.7%), hospitalisation in neurology department (1.7%), hospitalisation in other specialties (1.7%), transfer (1.7%) and death (3.3%).

False diagnoses of cerebrovascular diseases are common. In emergency departments almost half of the diagnoses of TIA may be wrong. Most false diagnoses refer to TIA (70%) and occur in elderly patients, can be attributed to systemic causes, have not been assessed by a neurologist and are referred to primary care. Hospital stroke registries that include emergency patients may be overestimated, especially in the number of cases of TIA.