| The Canadian population suffering from dementia has been estimated at 8% and the cost associated with it is remarkably high. Simple, validated cognitive screening tools, such as the Montreal Cognitive Assessment (MoCA) and Mini-Mental State Examination (MMSE) are used in clinics to evaluate cognitive impairment with a total score. Just as test scores can identify patients with global cognitive impairment, we hypothesized that individual test item scores can identify cognitive impairment in different domains. In the present study, individual MoCA items were tested for their validity and clinical utility for prediction of domain-specific cognitive impairment. A total sample size of 185 patients who were tested on both the MoCA and neuropsychological tests were extracted from data collected at two geriatric outpatient clinics. Domains assessed by the neuropsychological tests were identified using a Principal Component Analysis. Validity of the individual MoCA items was demonstrated by comparing scores obtained from MoCA items and neuropsychological tests using bivariate correlations followed by multiple stepwise regressions. Significant, but weak-to-moderate correlations were seen between the MoCA items and neuropsychological tests. The strongest association (r = .46, p < .01) was seen between MoCA 5-word recall and performance in the memory domain. Some of MoCA items were significantly correlated to multiple cognitive domains. Predictive utility of the MoCA items for domain-specific cognitive impairment in clinical setting was tested using accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Items with a satisfactory level of accuracy (>70%) were: Date for memory, Repeat Sentence 2 for processing speed, Serial 7s Subtraction for visuospatial, and Repeat Sentences and Clock tasks for language. Most of the MoCA items were sensitive, but not specific. Additionally, all the items showed high NPVs, but poor PPVs. These findings suggest that the MoCA items that patients failed were not as informative as items that they passed. |
