Functional outcome and recurrence rate were evaluated retrospectively
in a series of 566 consecutive patients with a cardioembolic TIA or st
roke. Seventy-five patients had a TIA/RIND, 163 a minor stroke, 238 a
major deficit, and 90 a fatal event. In a logistic regression analysis
the probability of a disabling or fatal outcome was significantly inf
luenced by age (p = 0.0023), a previous stroke (p = 0.0262), concomita
nt carotid artery disease in Doppler sonography (p = 0.0210), and the
underlying heart disease (p = 0.0001). Fatal or major strokes predomin
ated among patients with nonvalvular atrial fibrillation, prosthetic h
eart valves, acute or chronic myocardial infarction. Kaplan-Meier esti
mates revealed a cumulative risk of recurrent cerebral embolism of 2.9
% within three weeks. In a proportional hazards model only the underly
ing heart disease was significantly linked with the risk of recurrent
embolism (p = 0.021).