Pure motor hemiparesis and sensorimotor stroke. Accuracy of very early clinical diagnosis of lacunar strokes.

Clinical differentiation of lacunar from nonlacunar strokes in the very early phase could help to exclude patients with lacunar stroke from pharmacologic trials designed for nonlacunar strokes, namely, those with thrombolytic agents. In a continuous series of acute ischemic stroke patients, we evaluated how accurately a clinical diagnosis of pure motor hemiparesis or sensorimotor stroke formulated in the first hours from onset predicts a lacunar stroke documented by cerebral computed tomography or by autopsy.We examined 517 patients (299 men, 218 women; mean +/- SD age, 67 +/- 10 years) within 12 hours (mean +/- SD, 6.1 +/- 3.2 hours) of the event. At hospital admission, we observed 151 (29\%) patients with pure motor hemiparesis and 68 (13\%) patients with sensorimotor stroke.Computed tomography or autopsy was compatible with a lacunar stroke (ie, detection of a lacune or permanently negative computed tomography) in 170 (33\%) patients, of whom 123 (72\%) had pure motor hemiparesis and 47 (28\%) had sensorimotor stroke. This led to a sensitivity of 72\%, a specificity of 72\%, a positive predictive value of 56\%, and a negative predictive value of 84\%. Overall positive predictive value of pure motor hemiparesis was 58\% (60\% for two areas and 58\% for three areas involved), and that of sensorimotor stroke was 51\% (87\% for two areas and 40\% for three areas involved). By separately evaluating the sides of lesions, we found a positive predictive value of 46\% for right-side infarcts and of 72\% for left-side infarcts. Right-side lesions constituted 51\% of lesions in lacunar syndrome patients with lacunar stroke, 76\% in those with nonlacunar stroke, 19\% in nonlacunar syndrome patients with lacunar stroke, and 31\% in those with nonlacunar stroke (P < .0001). During the first days of hospital stay we observed a deterioration of 21\% of lacunar syndrome patients with nonlacunar stroke and an improvement of 49\% of nonlacunar syndrome patients with lacunar stroke, with appearance and disappearance of symptoms of cortical involvement, respectively. The examination of these patients after the occurrence of these clinical changes would have led to a daily increase of the positive predictive value up to a maximum of 66\% at day 7.Pure motor hemiparesis and sensorimotor stroke diagnosed within 12 hours of the event are poorly predictive of lacunar strokes. Hence, the very early identification of these syndromes cannot be used for patient selection in therapeutic trials.