AbstractBackground and Purpose—
The prevalence of poststroke neurocognitive disorder (NCD) has yet to be accurately determined. The primary objective of the present study was to optimize operationalization of the criterion for NCD by using an external validity criterion.Methods—
The GRECOG-VASC cohort (Groupe de Réflexion pour l'Évaluation Cognitive Vasculaire) of 404 stroke patients with cerebral infarct (91.3%) or hemorrhage (18.7%) was assessed 6 months poststroke and 1003 healthy controls, with the National Institute of Neurological Disorders and Stroke-Canadian Stroke Network standardized battery. Three dimensions of the criterion for cognitive impairment were systematically examined by using the false-positive rate as an external validity criterion. Diagnosis of mild and major NCD was based on the VASCOG criteria (Vascular Behavioral and Cognitive Disorders). The mechanisms of functional decline were systematically assessed.Results—
The optimal criterion for cognitive impairment was the shortened summary score (ie, averaged performance for action speed, executive functions, and language) because it was associated with the highest (P=0.0001) corrected true-positive rate (43.5%) and a false-positive rate ≤5%. Using this criterion, the mean (95% confidence interval) prevalence of poststroke NCD was 49.5% (44.6–54.4), most of which corresponded to mild NCD (39.1%; 95% confidence interval, 34.4–43.9) rather than dementia (10.4%; 95% confidence interval, 7.4–13.4).Conclusions—
This study is the first to have optimized the operationalization of the criterion for poststroke cognitive impairment. It documented the prevalence of poststroke NCD in the GRECOG-VASC cohort and showed that mild cognitive impairment accounts for 80% of the affected patients. Finally, the method developed in the present study offers a means of harmonizing the diagnosis of NCD.Clinical Trial Registration—
URL: https://www.clinicaltrials.gov. Unique identifier: NCT01339195.