P3‐127: Identification of cognitive decline using a longitudinal telephone screening versus brief, one‐day computerized testing

and 100%; specificity: 82% and 87%; cut-off: 7.5 and 6.5) and AD from MCI (AUC: 0.89 and 0.91; sensitivity: 88% and 80%; specificity: 74% and 75%; cut-off 9.5 and 7.5). The test was less accurate to discriminate MCI from controls (AUC: 0.62 and 0.82; sensitivity: 75% and 95%; specificity: 54% and 50%; cut-off: 3.5 and 3.5). Conclusions: The SKT is subject to educational bias, and discriminates patients with AD from MCI and controls.