WE NEED STRONGER PREDICTORS OF MAJOR VASCULAR EVENTS IN PATIENTS WITHA RECENT TRANSIENT ISCHEMIC ATTACK OR NONDISABLING STROKE

Background It has been proposed that most prognostic factors in patien ts with transient ischemic attack or nondisabling stroke are weak and consequently that patients at high risk of recurrent major vascular ev ents cannot be reliably identified. Methods In the Dutch TLA trial, a multicenter, double-blind study of low-dose versus medium-dose aspirin , 3127 patients were included within 3 months after onset of a transie nt ischemic attack: amaurosis fugax, or nondisabling stroke. In a prev ious analysis, we developed a prediction model by means of Cox proport ional hazards regression for the composite outcomes of fatal or nonfat al stroke and for myocardial infarction, stroke, or vascular death, ba sed on clinical and demographic Information as well as on the results of ancillary investigations. We assessed the discriminatory power and the calibration of the prediction models. Results The median numbers o f prognostic factors for stroke, myocardial infarction, or vascular de ath outcome and for stroke alone were 3 and 4, respectively. The propo rtion of patients with a predicted probability exceeding 30% was less than 5% for bath models; here the calibration of the models was poor. Only four of the patients with stroke, myocardial infarction, or vascu lar death were assigned a probability of greater than 50% for thar out come, and only one of the patients with stroke was given such a high p robability. The models' discriminatory ability was a little disappoint ing (areas under the curve of 0.73 and 0.75, respectively). Conclusion This analysis indicates that we need stronger predictors of recurrenc e risk in patients with a transient ischemic attack or nondisabling st roke.