89Comparison Study of The Montreal Cognitive Assessment(MoCA-CN) and The Quick Mild Cognitive Impairment Screen (Qmci-CN) in Post-Stroke Patients

Background: Post-stroke cognitive impairment (PSCI) can be divided into post-stroke cognitive impairment no dementia (PSCIND) and post-stroke dementia (PSD) according to the degree of the cognitive decline. Several cognitive screening instruments are available but there is no evidence to support which to use in clinical practice. Further, PSCI is often not detected. Given this, it is necessary to screen for PSCI. The most commonly used instruments to assess for PSCI are the Montreal Cognitive Assessment (MoCA-CN) and the Mini-Mental State Examination (MMSE-CN). The Quick Mild Cognitive Impairment screen (Qmci-CN) is a short cognitive screen that has yet to be validated in stroke. Methods: We recruited post-stroke patients from a rehabilitation unit in a large university hospital; 11 with PSD, 15 with PSCIND, 10 with normal cognition (NC). The Qmci-CN, MoCA-CN and MMSE-CN were administered by the trained rater, blind to the diagnosis. Results: In total, 36 patients were available; 61% (22/36) were female. The median age of patients was 63 (interquartile +/−13) years and the median years in education was 12 (+/−4). The median Qmci-CN screen score was 49/100 (+/−16), the median MMSE was 23/30(+/−7), and the median MoCA-CN score was 20/30(+/−7). In this sample the Qmci-CN was less accurate compared to the MMSE-CN and MoCA-CN in discriminating PSCIND from NC, area under the curve (AUC), 0.551 compared to 0.676 and 0.881, respectively. It has similar accuracy in discriminating PSD from NC, (AUC of 0.924 vs 0.979 and 1.0), and in discriminating PSCIND from PSD, (AUC of 0.894 vs 0.854 and 0.909, respectively). Conclusion: This study suggested that the Qmci-CN has comparable accuracy to the MMSE and MoCA-CN in discriminating PSD but was less accurate in separating NC from PSCIND compared to the MoCA-CN. Further study with a larger sample is needed to confirm these findings.