The Audio Recorded Cognitive Screen (ARCS): a flexible hybrid cognitive test instrument.

Objective: To describe and characterise the Audio Recorded Cognitive Screen (ARCS), a novel instrument that uses an audio device to administer selected neuropsychological tests to unsupervised individuals who write their responses in a special booklet for later scoring.

Methods: The ARCS was administered to 733 individuals, comprising 550 from a normative community sample (mean age=59.14, range 18-84 years), 101 clinic patients and a separate validation sample of 82 community controls, who, together with the patients, underwent detailed neuropsychological assessments. These data were examined for the influence of demographic variables on ARCS performance, to establish normal performance and develop scoring routines, and to characterise the structure, reliability and validity of the instrument.

Results: Age, gender and education influenced ARCS performance. ARCS tests were generally reliable and correlated well with corresponding conventional neuropsychological tests. Factor analyses indicated that the ARCS measures executive functioning/attention, memory, language, verbal fluency and visuospatial functioning. The ARCS discriminated well between normal (n=82), impaired (n=33) and demented (n=25) individuals, and significantly better than did the Mini Mental State Examination (MMSE), including on a single, demographically adjusted, global QuickARCS score obtainable in about 3 min of the clinician's time. Receiver Operating Characteristic analyses confirmed the superiority of the ARCS over MMSE as a screen for mild dementia (AUC 0.98, 99% CI 0.95 to 1.00) or cognitive impairment (AUC 0.90, 99% CI 0.83 to 0.97).

Conclusions: The ARCS has good validity and reliability, has a sound normative base and measures functioning in multiple cognitive domains while imposing minimal time demands upon the clinician.