A Systematic Evidence Review of the Signs and Symptoms of Dementia and Brief Cognitive Tests Available in Veterans Affairs
Kansagara, D., & Freeman, M. (2010).
Portland (OR): Health Services Research and Development Center, Veterans Health Administration, Department of Veterans' Affairs, (Evidence-based Synthesis Program. VA-ESP Project #05-225), 1-36.
This systematic review investigates the psychometric properties (e.g., sensitivity, specificity, time to completion) of brief cognitive tests (i.e., Blessed Orientation-Memory-Concentration Test, MiniCog, Montreal Cognitive Assessment, General Practitioner Assessment of Cognition, St. Louis University Mental Status Exam, and Short Test of Mental Status) for assessing cognitive function in adults without prior diagnosis of dementia.
Department of Veterans Affairs, Veterans Health Administration
Through July 2009
Cross-sectional studies
36 total; 18 addressing the relevant clinical question (#2)
"The [Montreal Cognitive Assessment (MoCA)] has the longest administration time among the 6 tests, and had low specificity in 2 of 3 studies (35-50%). The MoCA has been evaluated in a memory clinic population but has not been studied in a general practice setting" (p. 3).
The St. Louis University Mental Status (SLUMS) Exam demonstrated high sensitivity and specificity in a population of veterans. "The SLUMS takes longer to administer than other tests. It was developed more recently than the other tests and has not been widely studied" (p. 3).
The Mini-Cog has a shorter administration time than the other tests reviewed and has been validated in a large sample of the general population. It has sensitivity values ranging from 76% to 99% and specificity ranging from 83% to 93%.
The Blessed Orientation-Memory-Concentration (BOMC) Test was evaluated in a bi-racial population sample, and found to misclassify more black individuals than white individuals as impaired. The test demonstrated specificity values ranging from 38% to 94%, and sensitivity values ranging from 69% to 100%. "The inclusion of patients with previously diagnosed dementia might have inflated the sensitivity in 2 studies" (p. 3).
The Short Test of Mental Status (STMS) has been researched in the primary care setting and includes both patient and informant sections. Sensitivity values ranged fro 82% to 98%, but specificity was low (49-66%). "The specificity of the combined score and 2-stage method ranged from 77% to 86%" (p. 3).
