Abstract TP18: A 15-Minute Cognitive Screen Identifies Post-Stroke Cognitive Impairment

Abstract

Introduction: While comprehensive neuropsychological evaluations are a valuable part of post-stroke care, time-intensive assessments may not always be feasible or well-tolerated in certain clinical settings. There is a need for a brief cognitive battery that can detect post-stroke cognitive impairment while being efficient with regard to time and resources. Methods: As part of the parallel StrokeCog and Stroke-IMPaCT studies, 127 participants with chronic ischemic stroke completed comprehensive neuropsychological testing at Stanford. They underwent a cognitive battery consisting of 13 tests that encompassed 5 cognitive domains. Partial least square analyses identified a set of candidate tests that were most associated with cognitive impairment on this comprehensive battery, defined as 2 domain z-scores <-1 or 1 domain z-score <-1.5. We used these candidate tests to construct a brief battery of 5 tests (StrokeCog-Brief). Psychometric properties were compared amongst StrokeCog-Brief, the MoCA, and another established 30-minute battery (NINDS-30). Initial findings were cross-validated in an independent sample of 61 participants at University of Manchester. Results: The StrokeCog-Brief battery consisted of 5 measures of language, working memory, learning, and attention/processing speed (HVLT Immediate Recall, Trails A, Trails B, Digit Span Forward/Backward, Phonemic Fluency) that took approximately 15 minutes to complete. StrokeCog-Brief demonstrated high sensitivity (0.97) and adequate specificity (0.78) in detecting cognitive impairment on the comprehensive battery, outperforming both the MoCA (0.77, 0.73) and NINDS-30 (0.97, 0.35). In the cross-validation sample, StrokeCog-Brief similarly demonstrated high sensitivity (0.93) and adequate specificity (0.67), again outperforming both the MoCA (0.60, 0.86) and NINDS-30 (0.81, 0.08). Conclusions: We identified a brief, 15-minute battery of 5 tests that has high sensitivity in detecting cognitive impairment in people who have had a stroke. This brief battery covers multiple cognitive domains commonly impacted by stroke and represents an efficient means to identify chronic post-stroke cognitive impairment.