Related Experiment Video
Updated: Jul 22, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Screening for cognitive impairment in systemic lupus erythematosus: Application of the Montreal Cognitive Assessment
Emmanouil Papastefanakis1,2, Georgia Dimitraki1, Georgia Ktistaki3
1Department of Psychology, of Social Sciences, 381189University of Crete School of Social Science, Rethimno, Greece.
Background:
Cognitive impairment (CI) is one of the most frequent neuropsychiatric manifestations of systemic lupus erythematosus (SLE). Given that extensive neuropsychological testing is not always feasible in routine clinical practice, brief cognitive screening tools are desirable. The aim of this study was to evaluate the Montreal Cognitive Assessment (MoCA) as a screening tool for CI in SLE.
Methods:
Consecutive SLE patients followed at a single centre were evaluated using MoCA and an extensive neuropsychological test battery (NPT), including the Digits Forward and Digits Backwards, Rey Auditory Verbal Learning Memory Test, Trail Making Test, Stroop Colour-Word Test, Semantic and Phonetic Verbal Fluency tests and a 25-problem version of the General Adult Mental Ability test. The criterion validity of MoCA was assessed through receiver operating characteristic (ROC) analyses using three different case definitions: i) against normative population data, ii) and iii) against average performance of a comparison group of rheumatoid arthritis (RA) patients, to adjust for possible confounding effects of chronic illness and inflammatory processes on cognitive performance. The effect of patient-related (age, years of education, anxiety, depression, fatigue and pain) and disease-related (activity, damage, age at diagnosis, disease duration, use of glucocorticoid, psychotropic and pain medication) parameters on the MoCA was examined.
Results:
A total of 71 SLE patients were evaluated. MoCA significantly correlated with all NPT scores and was affected by education level (p < 0.001), but not by other demographic or clinical variables. The optimal cutoff for detecting CI, as defined on the basis of normative population data, was 23/30 points, demonstrating 73% sensitivity and 75% specificity. A cutoff of 22/30 points, using neuropsychological profiles of the RA group as inflammatory disease controls, exhibited higher sensitivity (100%, based on both definitions) and specificity (87% and 90%, depending on the definition). The standard cutoff of 26/30 points displayed excellent sensitivity (91-100%) with significant expenses in specificity (43-45%).
Conclusion:
The MoCA is an easily applied tool, which appears to be reliable for identifying CI in SLE patients. The standard cutoff score (26/30) ensures excellent sensitivity while lower cutoff scores (22-23/30) may, also, provide higher specificity.
