Derivation and validation of a Short Form of the Mini-Mental State Examination for the screening of dementia in older adults with a memory complaint.

BACKGROUND AND PURPOSE To validate a Short Form of the Mini-Mental State Examination (SMMSE) as a screening test for dementia in older ambulatory individuals followed in a memory clinic for a memory complaint. METHODS A total of 202 cognitively healthy individuals, 100 individuals with a mild cognitive impairment and 304 demented individuals sent for a memory complaint by their primary care physician to a memory clinic were prospectively included in this cross-sectional study. They were randomized into derivation (n = 303) and validation (n = 303) groups. The SMMSE score was built from six memory items of MMSE, with a score ranging from 0 to 6 (i.e. best performance). RESULTS The receiver operating characteristic curve showed an area under the curve of 0.98 for the derivation group and 0.97 for the validation group without differences between curves (P = 0.254). The cut-off between the sensitivity and the specificity of the SMMSE score for clinically diagnosed dementia was ≤4. The performance of the SMMSE for the diagnosis of dementia was high in the derivation and validation groups: sensitivity at 93.1% and 93.8%, specificity at 93.8% and 90.5%, positive predictive value at 94.3% and 90.1%, negative predictive value at 92.5% and 94.0%, likelihood ratio of positive test at 14.9 and 9.8 and of negative test at 0.07 and 0.07, respectively. CONCLUSIONS The Short Form of the Mini-Mental State Examination was a good screening test for dementia in older individuals followed in a memory clinic for a memory complaint. The next step should be the confirmation of its discriminative value in older primary care patients.