Abstract WP57: Large-scale Reliability/Validity Assessments And Quantification Of Cognition Within The Imperial College Comprehensive Assessment For Cerebrovascular Disease (IC3): A Novel In-depth, Self-administered, Online Cognitive Tool

Background: Cognitive difficulties occur in up to 86% of acute stroke cases and persist in a third. There is a growing need for a scalable, cost-effective and in-depth assessment of cognitive function in patients, that enables longitudinal tracking of recovery and better understanding of predictors of long-term cognitive outcomes. This study aims to address this gap by presenting reliability and validity data on a novel self-administered online assessment tool that assesses cognition after stroke. Method: This self-administered web-based tool offers an in-depth cognitive profiling through 22 tasks spanning both domain-general (e.g. attention) and domain-specific deficits (e.g. memory, language, visual neglect), together with neuropsychiatric questionnaires. In light of these specifications, the battery has been named IC3 (Imperial College Comprehensive assessment for Cerebrovascular disease; https://ic3study.co.uk/ ) Results: We present results outlining the normative ranges for the assessment in 1000+ gender, age and education matched controls and in a modest sample of patients following stroke. Furthermore, we present results showcasing the reliability and validity of the assessment through specificity, sensitivity and test-retest analyses. Finally, we present a novel statistical approach that allows one to tease out motor-driven deficits during response selection, from true cognitive deficits, allowing for a more precise phenotyping of cognition in stroke. Conclusion: IC3 is the first assessment to offer in-depth unsupervised cognitive phenotyping of patients following stroke, while also allowing scalable and cost-efficient longitudinal monitoring. It fares well against various reliability/validation metrics and is able to capture cognition in a more precise and scalable manner than currently available assessments, making it an attractive tool for understanding the mechanisms of recovery in a plethora of vascular disorders.