Clinical Utility and Diagnostic Accuracy of the Tablet-Based Seoul Cognitive Status Test: Evidence for Scalable...

The Seoul Cognitive Status Test (SCST) is a brief (~30 minutes), tablet-based battery that yields domain-specific and composite scores and uses automated scoring for rapid result generation, reduced examiner burden, and scalability for large or longitudinal assessments. We evaluated the SCST's psychometric and diagnostic performance against multidisciplinary final clinical diagnoses derived from a multidimensional workup.

We conducted a cross-sectional diagnostic study of 777 outpatients (SCST-Seoul Neuropsychological Screening Battery [SNSB] n=639; SCST-Consortium to Establish a Registry for Alzheimer's Disease [CERAD] n=138) who underwent standardized diagnostic evaluation; multidisciplinary final diagnoses served as the reference. Participants completed the SCST and either the Korean version of CERAD (CERAD-K) or SNSB-second edition (SNSB-II). We assessed convergent validity of SCST subtests with traditional measures and compared SCST composite scores and classifications with reference diagnoses.

Pearson correlations between SCST subtests and corresponding CERAD-K/SNSB-II measures were all significant (p<0.001), with key domain correlations ≥0.7 (language, memory, executive). In the pooled sample, covariate-adjusted receiver operating characteristic area under the curves (AUCs) for the SCST composite ranged from 0.854 (cognitively unimpaired [CU] vs. mild cognitive impairment [MCI]) to 0.980 (CU vs. dementia); AUC for CU vs. cognitively impaired [CI] was 0.903. SCST-based classification sensitivity was 0.859 (CU vs. CI), 0.984 (CU vs. dementia), 0.758 (CU vs. MCI), 0.896 (MCI vs. dementia), and 0.882 (no dementia vs. dementia).

These findings support the SCST's clinical utility; further studies in broader community-based cohorts are warranted to confirm generalizability.