Screening for Mild Cognitive Impairment: Comparing the SMMSE and the ABCS

• Published in: Canadian journal of psychiatry 2005-01, Vol.50 (1), p.52-58
• Main Authors: Molloy, D William, Standish, Timothy IM, Lewis, David L
• Format: Article
• Language: English
• Subjects: Age; Aged; Aged, 80 and over; Cognition & reasoning; Cognition Disorders - diagnosis; Cognition Disorders - epidemiology; Cross-Sectional Studies; Education; Female; Humans; Male; Mass Screening - methods; Medical screening; Mental disorders; Middle Aged; Neuropsychological Tests; Observer Variation; Psychiatry; Reproducibility of Results; Sensitivity; Sensitivity and Specificity; Severity of Illness Index; Surveys and Questionnaires
• ISSN: 0706-7437; 1497-0015
• DOI: 10.1177/070674370505000110

Objective: To compare the sensitivity and specificity of the AB Cognitive Screen (ABCS) with the Standardized Mini-Mental State Examination (SMMSE) to differentiate normal cognition from mild cognitive impairment (MCI), especially when educational level and age are taken into account. Method: This cross-sectional study took place at geriatric outpatient memory clinics. Participants were community-dwelling adults, aged 55 years or over, referred from primary care settings (a minority of participants were referred from specialists) for assessment of memory loss and age-matched control subjects with no complaint of memory loss. Each participant had the ABCS and the SMMSE administered in random order on the same day. Results: Participants included 124 patients diagnosed with MCI and 111 with normal cognitive function. The ABCS showed a statistically significant difference between normal cognition and MCI (ABCS score 111.7 and 104.6 points, respectively, P < 0.001) for the whole group. This difference was significant with the ABCS, regardless of participants' age or education. There was a significant difference between normal cognition and MCI for SMMSE scores (SMMSE score 27.8 and 27.2 points, respectively, P = 0.040), but the differences were not significant when age and education were taken into account. Age and education were shown to affect the scores of both instruments except for the ABCS scores of MCI subjects, which were not significantly affected by education (P = 0.059). Conclusions: The ABCS is more sensitive than the SMMSE in differentiating normal cognition from MCI. The ABCS appears to be less influenced by education. It has improved clinical utility with a wider range of scoring gradations, reduced ceiling effects, and shorter scoring and administration times.