In-hospital stroke: a multi-centre prospective registry

Abstract

Background: in-hospital strokes (IHS) are relatively frequent. Avoidable delays in neurological assessment have been demonstrated. We study the clinical characteristics, neurological care and mortality of IHS.

Methods: multi-centre 1-year prospective study of IHS in 13 hospitals. Demographic and clinical characteristics, admission diagnosis, quality of care, thrombolytic therapy and mortality were recorded.

Results: we included 273 IHS patients [156 men; 210 ischaemic strokes (IS), 37 transient ischaemic attacks (TIA) and 26 cerebral haemorrhages]. Mean age was 72 ± 12 years. Cardiac sources of embolism were present in 138 (50.5%), withdrawal of antithrombotic drugs in 77 (28%) and active cancers in 35 (12.8%). Cardioembolic stroke was the most common subtype of IS (50%). Reasons for admission were programmed or urgent surgery in 70 (25%), cardiac diseases in 50 (18%), TIA or stroke in 30 (11%) and other medical illnesses in 71 (26%). Fifty-two per cent of patients were evaluated by a neurologist within 3 h of stroke onset. Thirty-three patients received treatment with tPA (15.7%). Thirty-one patients (14.7%) could not be treated because of a delay in contacting the neurologist. During hospitalization, 50 patients (18.4%) died, 41 of them because of the stroke or its complications.

Conclusions: cardioembolic IS was the most frequent subtype of stroke. Cardiac sources of embolism, active cancers and withdrawal of antithrombotic drugs constituted special risk factors for IHS. A significant proportion of patients were treated with thrombolysis. However, delays in contacting the neurologist excluded a similar proportion of patients from treatment. IHS mortality was high, mostly because of stroke.