Efficacy of Digital Screening to Detect Mild Cognitive Impairment and Early Dementias Validated 

BrainTest Inc., a medical software company that provides scientifically validated cognitive screening instruments on a wide range of tablets and mobile devices, has announced that the results of its Validity/Equivalency Study were presented at the Alzheimer’s Association International Conference 2017 (AAIC).
 
Data presented by Dr. Douglas W. Scharre, professor of Clinical Neurology and Psychiatry at The Ohio State University Wexner Medical Center demonstrates eSAGE performs similarly with neuropsychological batteries, MoCA, and MMSE and shows no scale bias compared to the validated SAGE. Meaning eSAGE has a high sensitivity (71 percent) and specificity (90 percent) in detecting cognitive impairment from normal subjects, with the added advantage of private at-home self-administration. Dr. Scharre’s results and data presentation will be available for the remainder of the AAIC at Location S8, P1-304.
 
Results of eSAGE Validity/Equivalency Study 
Of the 426 subjects screened, 66 completed the evaluation. eSAGE score correlation to a battery of neuropsychological tests was 0.73 (p < 0.0001) with no significant difference between the paper and digital format. Spearman correlation of SAGE versus eSAGE was 0.88 (p < 0.0001), and they are related by the formula: eSAGE score = –1.05 + 0.99 × SAGE score. Since the slope is very close to 1 (p = 0.86) there is strong evidence that the scaling is identical between eSAGE and SAGE, with no scale bias. Overall, eSAGE scores are lower by an average of 1.21 and the decrease is statistically significant (p < 0.0001). For those subjects familiar with smartphones or tablets (one measure of digital proficiency), eSAGE scores are lower by an average of 0.83 points (p = 0.029). With a score 16 and higher being classified as normal, eSAGE had 90 percent specificity and 71 percent sensitivity in detecting those with cognitive impairment from normal subjects.
 
Design of eSAGE Validity/Equivalency Study 
Over an 18 month period subjects aged 50 and over who had taken SAGE were recruited from community and clinic settings. Subjects were randomly selected to participate in a clinical evaluation including neuropsychological evaluations. SAGE and eSAGE were administered using a crossover design. Subjects were identified as dementia, MCI, or normal based on standard clinical criteria. Associations were investigated using Spearman correlations, linear regression, and sensitivity and specificity measures. The results of the study were recently published in Alzheimer’s Research & Therapy.