Objective: To examine the psychometric properties, acceptance, and screening efficacy of the Mini-Mental State Examination (MMSE) in an internal medicine practice.
Material and methods: The MMSE was administered more than 4,000 times by 27 internists to 3,513 elderly patients (2,299 women and 1,214 men, 60 to 102 years old) who underwent general medical examinations. The efficacy of the MMSE for screening was measured in a subsample of age- and sex-matched patients with dementia (N = 185) and control subjects (N = 227). MMSE scores were correlated with age and education in the community sample. The attitudes of physicians about the MMSE were assessed with a 12-question survey. Sensitivity, specificity, and predictive values were calculated.
Results: Performance on the MMSE among persons older than 59 years was influenced by age and education but not by sex. During an interval of 1 to 4 years, a change of 4 or more points in the total MMSE score is needed to indicate substantial cognitive deterioration. Participating physicians considered the MMSE of little value for routine screening in unselected populations but wanted it available for use as a clinical test. The traditional MMSE cutoff score of 23 or less had a sensitivity of 69% and a specificity of 99%. Use of age- and education-specific cutoff scores improved the sensitivity to 82% with no loss of specificity. With use of typical base rates for dementia in a general medical practice, the positive predictive value was less than 35%.
Conclusion: The clinical utility of the MMSE and acceptance by physicians may be improved through awareness of the influences of age and education on the MMSE and by its application in settings with a high base rate of dementia. The MMSE is ineffective when used to screen unselected populations; it should be used for persons at risk of cognitive compromise.