RESUMEN
BACKGROUND: We aimed to examine the discriminant validity of a brief self-administered cognitive screening test, the Test Your Memory (TYM) and a brief neuropsychological test, the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), supplemented with executive and language tests (Color Trail Test [CTT] and modified Boston Naming Test [mBNT], respectively), in detecting cognitive impairment (CI) in a one-stop memory clinic in Singapore. METHODS: Ninety patients ≥50 years old with a diagnosis of no cognitive impairment, mild cognitive impairment, and mild Alzheimer disease were recruited from memory clinic. They received the TYM, Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), RBANS, CTT, mBNT, and a gold-standard formal neuropsychological test battery. RESULTS: The TYM had a significantly larger area under the curve (AUC) than MMSE (0.96 vs 0.88, P = .03) and was equivalent to MoCA in detecting CI (0.96 vs 0.95, P = .80). At the optimal cutoff points, the TYM (<38) was significantly more sensitive than the MMSE (<24) and MoCA (<20; P < .001). The RBANS had an AUC equivalent to the RBANS supplemented with CTT and mBNT (0.92 vs 0.86, P = .22) in detecting CI. The RBANS supplemented with CTT and mBNT was more sensitive than RBANS alone in detecting CI (sensitivity: 0.98 vs 0.93, P = .016) among patients screened negative using TYM. CONCLUSION: The self-administered TYM is superior to MMSE and equivalent to MoCA in detecting CI and could be implemented routinely. The RBANS supplemented with CTT and mBNT is more sensitive in detecting CI than RBANS alone therefore could be used for diagnostic purposes.