Article Text
Abstract
In a prospective, community-based study of 675 consecutive patients with a first-ever stroke, of whom over 90% had computed tomography (CT) and/or necropsy examinations, 129 deaths occurred within 30 days of the onset of symptoms, a case fatality rate (CFR) of 19%. The 30 day CFR for patients with cerebral infarction was 10% (57 of 545, for primary intracerebral haemorrhage 52% (34 of 66), for subarachnoid haemorrhage 45% (15 of 33) and for those of uncertain pathological type 74% (23 of 31). The CFR for patients who had been functionally dependent pre-stroke was 33% compared with 17% for those who had been independent pre-stroke. The age-adjusted relative risk of death for patients who had been functionally dependent pre-stroke was not significantly greater (1.8, 95% confidence interval 0 to 4.3). There was a significant trend for CFR to increase with age (Chi square for trend = 4.0, p less than 0.05). This relationship was found in those patients who had been functionally independent prestroke (Chi square for trend = 7.9, p less than 0.005) but not in those who had been dependent pre-stroke (Chi square for trend = 0.5, NS). The pattern of increasing CFR with increasing age amongst those who had been independent prestroke was seen particularly in patients with cerebral infarction (Chi square for trend = 8.6, p less than 0.005). The age-adjusted relative risk of death for patients with cerebral infarction who had been functionally dependent pre-stroke was 2.2 (95% confidence interval 1.2 to 4.1). Fifty three percent of all deaths within 30 days of stroke were due to the direct neurological sequelae of the stroke. Patients with primary intracerebral or subarachnoid haemorrhages were significantly more likely to die in this way than those with cerebral infarction (relative risk 4.1; 95% confidence interval 3.4-4.9) and 56% of such deaths occurred within 72 hours of onset. In patients with cerebral infarction, 51% of deaths were due to complications of immobility (for example, pneumonia, pulmonary embolism) and these were more likely to occur after the first week. These findings have implications for clinical practice and the planning of clinical trials.