L'analyse d'item des QCM a inclus la détermination des indices de difficulté et de discrimination. La fidélité des QCM et des TCS a été mesurée par le coefficient alpha de Cronbach. Les scores obtenus aux deux épreuves par les trois groupes ont été comparés (ANOVA avec correction de Bonferoni-Dunn après vérification de l'homogénéité des variances par un test de Levene). Ils ont été corrélés (test de

No abstract

Objectives: Accurate prediction of outcome in acute stroke would help in identifying subgroups of patients for therapeutic trials and intravenous thrombolysis. The purpose of this study was to prospectively test the hypothesis that brain SPECT, with ^99mTc-L,L-ethylcysteinate dimer (ECD), a tracer sensitive to cell function, performed in the first hours after stroke onset, adds predictive power to concomitant neurological evaluation. Methods: Twenty-four patients with a first-ever middle cerebral artery stroke were prospectively studied with ECD-SPECT within 12 h after stroke onset. Neurological evaluation was performed using Orgogozo’s scale at admission and 3 months later in order to calculate the percent Martinez-Vila evolution indices (EI%). Semiquantitative visual analysis of SPECT images was performed in 6 cortical regions relevant for carotid artery territory. Both the extent and the intensity of cortical reduced ECD uptake were calculated, leading to an ‘ischemia’ score, corresponding to the sum of regions of interest (ROI) where ECD uptake was between 40 and 80% of the contralateral healthy hemisphere, and an ‘irreversibly damaged tissue’ (IDT) score, corresponding to an uptake below 40%, and a total score (ischemia + IDT). Each patient was assigned to one of three patterns: (1) pattern I with severe ECD cortical uptake reduction defined by at least one ROI with uptake under 40%, (2) pattern II with moderate ECD cortical uptake reduction (40–80%) only and (3) pattern III with normal ECD uptake. Results: There were 11 patients (46%) with pattern I ECD-SPECT. This group had almost invariably (10/11 patients) a poor outcome. The 12 patients (50%) classified in pattern II had a variable clinical outcome, ranging from improvement to deterioration. The single patient with a normal SPECT (pattern III) had a full clinical recovery. Both total score and IDT score were strongly significantly correlated with neurological recovery EI% (respectively p = 0.006 and 0.004). Their predictive value was significantly higher than, and independent of, day 0 neurological evaluation. No patient had an increased ECD uptake. Conclusion: Our results show that the degree of ECD cortical uptake reduction, measured on early brain SPECT, is a strong predictor of neurological recovery. ECD-SPECT data have a higher predictive value than day 0 neurological evaluation. The apparently better predictive value of ECD over hexamethylpropyleneamine oxime may reflect this tracer’s brain retention mechanisms which are weighted more towards cell function than towards perfusion. ECD-SPECT is easily obtainable and may help in selecting out from therapy those patients who are likely to have either very good or very poor spontaneous outcome, and thus improve the assessment of acute stroke and the choice of therapeutic strategy.

No abstract