Abstract:
Background: Sri Lanka is a rapidly aging country, where dementia prevalence will increase
significantly in the future. Thus, inexpensive and sensitive cognitive screening tools are crucial.
Objectives: To assess the reliability, validity, and diagnostic accuracy of the Sinhalese version
of the Addenbrooke’s Cognitive Examination-Revised (ACE-R s). Method: The ACE-R was
translated into Sinhala with cultural and linguistic adaptations and administered, together
with the Sinhala version of the Montreal Cognitive Assessment (MoCA), to 99 patients with
dementia and 93 gender-matched controls. Results: The ACE-R s cutoff score for dementia was 80 (sensitivity 91.9%, specificity 76.3%). The areas under the curve for the ACE-R s, MiniMental State Examination (MMSE) and MoCA were 0.90, 0.86, and 0.86, respectively. The
ACE-R s had good interrater reliability (intraclass correlation = 0.94), test-retest reliability (intraclass correlation = 0.99), and internal consistency (Cronbach’s α = 0.8442). Conclusions:
The ACE-R s is sensitive, specific and reliable to detect dementia in persons aged ≥50 years
in a Sinhala-speaking population and its diagnostic accuracy is superior to previously validated tools (MMSE and MoCA).
