Poster Presentations

Session Title: Alzheimer's Disease (AD) including Non-Cognitive Aspects
Presentation Date: Friday, March 14 – Saturday, March 15, 2009

ADDENBROOKE'S COGNITIVE EXAMINATION-REVISED: CROSS-SECTIONAL AND LONGITUDINAL USE FOR COGNITIVE ASSESSMENT

A.J. Larner
Walton Centre for Neurology and Neurosurgery, Cognitive Function Clinic, Liverpool, United Kingdom

Objective: To report use of the Addenbrooke's Cognitive Examination-Revised (ACE-R) over a 3-year period (2005-2008).
Methods/setting: Cognitive Function Clinic based in a regional neuroscience centre.
Results: ACE-R was used 261 times in 243 patients (M:F = 135:108, age range 24-85 years, dementia prevalence 35%). ACE-R was highly correlated with MMSE (r = 0.90; kappa = 0.72). Maximal accuracy for the diagnosis of dementia (0.89) was achieved with ACE-R cutoff of 73/100, with sensitivity, specificity and positive predictive value of 0.87, 0.91, and 0.83. Comparable values for the most accurate MMSE cutoff (24/100; 0.82) were 0.70, 0.89, and 0.77 respectively. Area under the receiver operating characteristic curve was 0.94 for the ACE-R, 0.91 for MMSE. In the 17 patients assessed for a 2^nd or 3^rd time with ACE-R over 6-36-month follow-up, four were eventually diagnosed with dementia, of whom 2 declined and 2 were stable (≤ 5-point change) on ACE-R, whilst of the 13 non-demented patients 8 were stable and 5 improved.
Discussion/conclusions: ACE-R proved easy to use and showed impressive metrics for the diagnosis of dementia in this clinical cohort, better than those for MMSE.