A quality dementia-screening tool is required for older remote Aboriginal Australians who have high rates of dementia and limited access to appropriate medical equipment and clinicians. The Kimberley Indigenous Cognitive Assessment (KICA Cog) is a valid cognitive test for dementia in Aboriginal and Torres Strait Islander peoples. The KICA cognitive informant questionnaire (KICA Carer) had yet to be analyzed to determine validity alone or in combination with the KICA Cog.

The KICA Carer was completed by nominated informants of 349 remote-living Aboriginal Australians in the Kimberley region, Western Australia. Validity was assessed by comparing KICA Carer with Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) and International Classification of Diseases (ICD-10) consensus diagnoses based on a blinded specialist review. KICA Carer and KICA Cog were then compared to determine joint validity.

A KICA Carer score of ≥3/16 gave optimum sensitivity (76.2%) and specificity (81.4%), area under curve (AUC) 0.89 (95% CI = 0.85, 0.94) with positive predictive value (PPV) of 35.8%, and negative predictive value (NPV) of 96.2%. A KICA Cog score of ≤33/39 gave a sensitivity of 92.9% and specificity of 89.9%, AUC 0.96 (95% CI = 0.94, 0.98), with PPV of 55.6% and NPV of 98.9%. Cut-off scores of KICA Cog ≤ 33/39 and KICA Carer ≥ 2/16 in series indicate possible dementia, with sensitivity of 90.5% and specificity of 93.5%. In this setting, PPV was 66.5% and NPV was 98.6%.

The KICA Carer is an important tool to accurately screen dementia in remote Aboriginal Australians when the KICA Cog is unable to be used for a patient. It is readily accepted by caregivers.

• • For the best practice in the cognitive assessment of an Aboriginal Australian aged over 45 years, KICA Cog should be utilized.

• • In cases where Aboriginal patients are not assessed directly, KICA Carer should be conducted with an informant. A cut-off score of ≥3/16 should be used (these tools can be downloaded from www.wacha.org.au/kica.html).