Background/aims: Many cognitive screening instruments (CSI) are available to clinicians to assess cognitive function. The optimal method comparing the diagnostic utility of such tests is uncertain. The effect size (Cohen's d), calculated as the difference of the means of two groups divided by the weighted pooled standard deviations of these groups, may permit such comparisons.
Methods: Datasets from five pragmatic diagnostic accuracy studies, which examined the Mini-Mental State Examination (MMSE), the Mini-Mental Parkinson (MMP), the Six-Item Cognitive Impairment Test (6CIT), the Montreal Cognitive Assessment (MoCA), the Test Your Memory test (TYM), and the Addenbrooke's Cognitive Examination-Revised (ACE-R), were analysed to calculate the effect size (Cohen's d) for the diagnosis of dementia versus no dementia and for the diagnosis of mild cognitive impairment versus no dementia (subjective memory impairment).
Results: The effect sizes for dementia versus no dementia diagnosis were large for all six CSI examined (range 1.59-1.87). For the diagnosis of mild cognitive impairment versus no dementia, the effect sizes ranged from medium to large (range 0.48-1.45), with MoCA having the largest effect size.
Conclusion: The calculation of the effect size (Cohen's d) in diagnostic accuracy studies is straightforward. The routine incorporation of effect size calculations into diagnostic accuracy studies merits consideration in order to facilitate the comparison of the relative value of CSI.
Keywords: Addenbrooke's Cognitive Examination-Revised; Cohen's d; Diagnostic accuracy; Effect size; Mini-Mental Parkinson; Mini-Mental State Examination; Montreal Cognitive Assessment; Six-Item Cognitive Impairment Test; Test Your Memory test.