RESUMO
OBJECTIVE: Although cognitive impairment (CI) is common in multiple sclerosis (MS), it is difficult to suspect in patients with low disability and there is a lack of brief and effective CI screening tools with a define cut-off point to be used during routine clinic visits. This study aims to validate the Electronic Screening Cognitive Impairment in Multiple Sclerosis (SCI-MS) test for CI among MS patients. METHODS: Cross-sectional, observational study that included adult patients, diagnosed with MS, Expanded Disability Status Scale (EDSS) score ≤6.5, without relapses within the last 2 months and no depression symptoms. The SCI-MS test consists of two modules: questionnaire (SCI-MS-Q) and pictogram matching tool (SCI-MS-P) measured for score and time. At inclusion, patients completed the Beck Depression Inventory (BDI-II test), the Brief Repeatable Battery of Neuropsychological Test (BRB-N) and the SCI-MS. The SCI-MS feasibility, test-retest reliability and predictive validity were assessed. RESULTS: A total of 194 patients (59.3% female) were included: mean (SD) age of 42 (9) years, mean time since diagnosis of 10 (7) years, 89.7% relapsing-remitting MS, and median (Q1-Q3) EDSS of 2.0 (1.0-3.5). According to BRB-N, 26.8% of patients had CI. Internal consistency was high (Cronbach alpha: 0.97). The intra-class correlation coefficient was 0.88 for the SCI-MS-Q, 0.09 for the SCI-MS-P score and 0.48 for the SCI-MS-P time, corresponding to AUC of the ROC curves of 0.571, 0.574 and 0.714, respectively. For a clinically significant cut-off point of ≥60 seconds, the reached CI sensitivity of SCI-MS-P time was 0.75 and the specificity 0.51. CONCLUSION: SCI-MS showed good psychometric properties. SCI-MS-P time of pictogram completion had an acceptable diagnostic accuracy of CI in MS patients with low disability. SCI-MS-P time of pictogram completion tool is an easy and quick score that can help neurologists to early identify CI in MS patients that should be further assessed to confirm CI diagnosis and to describe its characteristics and mainly affected domains.
