Screening scales for dementia: Toward reconciliation of conflicting cross‐cultural findings

Abstract A compendium instrument of five widely used dementia screening scales was applied to 550 subjects 65 years and older. Black, Hispanic and white groups were approximately equal in size. Recruitment was designed to produce an uneven and unknown mix of dementia cases and normals among these groups. A complex challenge to assessment and classification was thus contrived, approximating key aspects of the problem posed to cross‐cultural epidemiology. The published scoring systems of these dementia screens gave drastically conflicting results for absolute and culturally relative rates of cognitive impairment. The conflicts between scales appeared to be largely due to the varying sensitivities of the scales and, to a lesser extent, sociocultural bias. The conflicts were partially resolved through adjustments of the scale cut points.