Abstract 18752: An Ultra-short Cognitive Screening Test to Detect Significant Post-stroke Vascular Cognitive Impairment

Abstract

Objective: The National Institute of Neurological Disease and Stroke – Canadian Stroke Network (NINDS-CSN) harmonization battery recommended a 5-minute bedside cognitive screening test drawn from the Montreal Cognitive Assessment (MoCA). However this 5-minute cognitive screening test consisted of (5 word registration, recall, recognition, orientation, animal fluency) was based on expert opinion and not empirical evidence. We therefore aimed to establish an ultra-short cognitive screening test for the detection of post-stroke vascular cognitive impairment (VCI). Method: Patients with ischemic stroke/ Transient Ischemic Attack received MoCA within 14 days after stroke and at 3-6 months later. Cognitive outcomes was defined by formal neuropsychological evaluation at month 3-6 and classified as either ‘none-mild VCI (impairment ≤2 domains)’ or ‘moderate-severe VCI (impairment >2 domains)’. Results: There were 327 out of 400 patients completed neuropsychological assessments 3-6 months after stroke onset. Of these, 250 (76.5%) had no-mild VCI and 77 (23.5%) had moderate-severe VCI. Most patients were Chinese (70.3%) and males (69.8%) with age of 59.8 ± 11.6 years and education of 7.7 ± 4.3 years. Random forest analysis showed 5 most sensitive items of MoCA for the ultra-short test (5-Min MoCA) at baseline and 3 months to detect significant VCI at 3-6 months. These 5 items were the clock, serial 7s, animal fluency, recall and orientation. Receiver operating characteristic curve analysis showed that MoCA at 3 months had the largest area under curve (AUC) (0.90), followed by 5-Min MoCA at 3-6 months (0.89), baseline MoCA (0.86) and baseline 5-min MoCA (0.83). At 3-6 months, 5-Min MoCA was equivalent to the MoCA and superior to 5-Min MoCA at baseline (p=0.17 and p=0.04, respectively). The cutoff points for 5-Min MoCA at 3-6 months and baseline was the same (cutoff: 12/13 vs 12/13, sensitivity: 0.70 vs 0.74; specificity: 0.87 vs 0.74; classification accuracy: 0.83 vs 0.74, respectively). Conclusion: The empirically derived 5-min MoCA differs from the NINDS-CSN 5-min test. The established 5-min MoCA at 3-6 months after stroke is equivalent to the MoCA and superior to the 5-min MoCA at baseline, therefore should be used for routine practice.