Validation of the Intelligent Cognitive Assessment System (ICAS) for stroke survivors

Abstract

Primary objective: To investigate the internal consistency, test–re-test reliability of an Intelligent Cognitive Assessment System (ICAS) and its optimal cut-off score for stroke patients with or without cognitive impairment.Method: A prospective cohort study design was adopted. Sixty-six post-stroke patients of aged 60 or above were recruited. They were screened by the Chinese version of Mini Mental State Examination (MMSE-CV) and assessed by the Intelligent Cognitive Assessment System (ICAS) consisting of 65 testing items which could be presented at a level according to stroke patient's response. It was administered to examine the internal consistency and test–re-test reliability (by repeating within a 7-day interval). The optimal cut-off score to screen stroke patients having cognitive impairment was determined by the receiver operating characteristics (ROC) curve.Results: The internal consistency of the ICAS (Cronbach's α = 0.878) and its test–re-test reliability (ρ = 0.789; p < 0.001) were demonstrated. The cut-off score for the ICAS to determine cognitive impairment was found to be 3.02, with a sensitivity of 80.5% and specificity of 96%.Conclusion: Preliminary results suggested that ICAS was a valid and reliable cognitive screening tool for stroke survivors. The ICAS can be further developed by studying its norms for stroke patients. It is also programmable for potential application to different countries by changing ICAS to other language versions and including other culturally relevant content.