OBJECTIVETo evaluate efficacy and optimal cut-off-point through cerebral vascular hemodynamic indexes (CVHI) examination to predict stroke.

METHODSA number of 20,333 people at 35 years old and over were checked by CVHI and accumulative score was calculated according to the value of detected indexes. Risk factors of stroke were investigated simultaneously. One hundred and sixty-eight patients with stroke occurred during 4-year following up. Typical syndromes and signs stroke were used as golden standard to evaluate screening efficacy of CVHI.

RESULTSScore of optimal cut-off-point of cerebral vascular hemodynamic indexes was under 75 in ROC curve analyses. Sensitivity, specificity, accuracy, positive and negative predictive values, positive and negative likelihood ratios as well as Youden's index for predicting stroke within 4 years after examination were found to be 87.50%, 67.70%, 67.86%, 2.21%, 99.85%, 2.71, 0.18 and 0.55 respectively. Sensitivity and positive predict values for predicting cerebral vascular thrombosis were superior to predicting cerebral hemorrhage. Positive predicting value in risk exposure population was higher than that of overall population. Coefficiency of variation of cerebral vascular hemodynamic examination was 4.03%. The agreement rate of examination between two physicians was 97.62% and Kappa value was 0.94.

CONCLUSIONThe score of optimal cut-off-point of cerebral vascular hemodynamic indexes examination was 75. Both Efficacy and reliability for predicting stroke seemed to be good, especially for predicting cerebral vascular thrombosis.