Abstract
Objective: The aim of our study was to determine the strength of individual items in the Montreal Cognitive Assessment (MoCA) and then develop a short screening tool consisting of 2 items to detect subnormal cognitive performance based on the MoCA. Sample and Methods: The sample consisted of 540 cognitively healthy individuals 60 and more years old (average age 75.6 +- 9.1; years of education 12.7 +- 3.5, 54 % women).
We used the Czech version of MoCA. By using the Spearman correlation coefficient, we identified items with the strongest association with the total score of MoCA.
We performed ROC (receiver operating characteristic) analysis to find which combinations of the two items from the MoCA are the most effective in detecting individuals with performance less than one standard deviation below the norm in the MoCA. Results: Mean MoCA score was 24.69 +- 2.86.
Subnormal cognitive performance was psychometrically defined as one standard deviation from the average, a total score LESS-THAN OR EQUAL TO 22 in MoCA. Delayed recall, clock test and sentence repetition correlated most strongly with the total MoCA score: delayed recall (RS = 0.75), clock test (RS = 0.41), sentence repetition (RS = 0.37).
These pairs of items had the best discriminatory potential: delayed recall and sentence repetition (AUC = 0.869); and delayed recall and clock test (AUC = 0.859).Conclusion: The most effective item of the MoCA for the detection of individuals with subnormal performance was delayed recall. This subtest in combination with clock test may constitute a possible alternative to shorten /abbreviate the MoCA, allowing for a detection of individuals with subnormal cognitive performance.
Psychometric parameters of such an abbreviated screening test would need to be verified in a separate study.