Cognitive Screening Tools for Identification of Dementia in Illiterate and Low-Educated Older Adults, A Systematic Review and Meta-Analysis

International Psychogeriatrics

Paddick, S. M., Gray, W. K., et al. (2017).

International Psychogeriatrics, 29(6), 897-929.

This systematic review and meta-analysis investigates brief cognitive screening tools for dementia, specifically for adults age 45 and older, with low levels of formal education or literacy. Although speech-language pathologists do not diagnose dementia, findings from this review may be useful for identifying tools for detection of cognitive impairments in this population.

Not stated

Up to December 1, 2014

Peer-reviewed studies (not further specified)

45

Despite inadequately validation for use in individuals with low literacy, cognitive screening tools had a combined sensitivity estimate of 0.828 (0.789&ndash;0.862) and specificity estimate of 0.817 (0.717&ndash;0.887).The area under the ROC (AUROC) curve for these screening tools was 0.937 in low prevalence/community studies, 0.881 for clinic based/higher prevalence studies, and 0.869 for illiterate populations. Specific findings were as follows: <br /> <ul> <li><span style="color: #333333;">In lower prevalence/community studies, the most accurate tests were the Seven-Minute Screen (7MS; sensitivity= 100%, specificity= 95.1% in one study) , Prueba Cognitiva de Leganes (PCL; sensitivity= 93.9, specificity= 94.7 in one study with a cut-off of 22/32), and the Kimberley Indigenous Cognitive Assessment (KICA-Cog; sensitivity= 82.3-90.6%, specificity= 87.5-92.6% in one study) conducted in Spain and Australia, </span></li> <li><span style="color: #333333;">In higher prevalence/clinical studies, the most accurate were the Chula Mental Test (CMT, sensitivity= 100%, specificity= 90% in one study) and the Picture-Based Memory Impairment Screen (PMIS; sensitivity= 95.4%, specificity= 99.2% in one study) in Thailand and India. </span></li> <li><span style="color: #333333;">The Mini-Mental State Examination (MMSE) and its adaptations had an AUROC of 0.853, with variable sensitivity (82.8%; range= 67-97%) and specificity (81.7%, range; 60.3-98%&nbsp; seen across various cutoffs and modifications.</span></li> </ul> <span style="color: #333333;">Due to limited research and heterogeneity across studies, evidence is inadequate to support the use of any one particular test for screening for dementia in individuals with low education and literacy. Further research is indicated.</span>