Background/Purpose: Cognitive dysfunction (CD) is among the most common neuropsychiatric manifestations in SLE. However, the diagnosis is oftentimes delayed and occasionally overlooked because of the lack of a sensitive and practical screening test. Several methods have been used to ascertain CD in patients with SLE, including traditional neuropsychological testing (NPT) and the Automated Neuropsychologic Assessment Metrics (ANAM), a computerized battery of symbol based tests measuring many of the same cognitive domains as NPT. Both are time-consuming, relatively costly, and not readily available. The Montreal Cognitive Assessment Questionnaire (MoCA) is a one-page performance-based screening test that has been validated to identify mild cognitive impairment in the elderly and, in preliminary work done by us, appeared to hold promise in SLE as well. This study aims to further evaluate the MoCA as a screening tool for the diagnosis of CD in SLE.

Methods: Patients with SLE fulfilling the American College of Rheumatology criteria were recruited. All subjects were administered the ANAM test and the MoCA questionnaire. MoCA scores were compared to the total throughput score (TTS) (correct responses/time for the responses), a standard measure of performance of the ANAM. Individual MoCA questions were also compared to TTS to identify the MoCA questions with the strongest correlation with CD. The classification of normal or abnormal by the ANAM was compared to that of the MoCA using various cutoffs. Sensitivity, specificity, likelihood ratios and predictive values were also computed at the various cut-offs for the MoCA.

Results: In total, 74 patients were evaluated. Of these, 15 (20 %) were identified by the ANAM as having cognitive dysfunction in comparison with 33 (44.6 %) by the MoCA using the standard cutoff of 26. The scores were significantly correlated (r= 0.51, p <0.001). Six out of the 10  MoCA questions showed significant correlation with the ANAM score with the assessment of visuospatial and executive function being the most highly correlated with CD (p= 0.000199). Using the standard cutoff of 26, the sensitivity of the MoCA was 93%, specificity 68%, and negative predictive value 98%. All other cut-offs were inferior.

Conclusion: The MoCA appears to be a promising and practical screening tool for identification of patients with SLE at risk for cognitive dysfunction.

Table 1.  MoCA Performance Characteristics

^aPositive likelihood ratio

^bNegative likelihood ratio

^cPositive predictive value

^dNegative predictive value