Related Experiment Video
Updated: Aug 22, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
[Validation of a screening test for age associated cognitive impairment, in Chile]
Pilar Quiroga1, Cecilia Albala, Gonzalo Klaasen
1Facultad de Medicina, Universidad de Concepción, Concepción.
Background:
The real prevalence of dementia in a given population must be determined through prevalence studies, using validated screening tests.
Aim:
To validate and determine cutoff points for a cognitive impairment screening test composed by the Folstein Mini Mental State Examination (MMSE) and Pfeffer Functional Activities Questionnaire (PFAQ).
Material And Methods:
Validation of the diagnostic test in a sample of 100 subjects over 65 years old (85 from the project "Age associated dementias" and 15 with a confirmed diagnosis of dementia). All were subjected to a complete neuropsychological test by a trained neurologist, that constituted the "gold standard" for the diagnosis of dementia. An independent interviewer applied the MMSE to the subjects and the PFAQ to a next of kin informer. Cutoff points were calculated using ROC curves. The points with the better equilibrium between sensitivity and specificity were selected, considering differences in results between groups with low and high educational level.
Results:
The cutoff point for MMSE was 21/22, with a sensitivity of 93.6% (95% CI 70.6-99.7%) and a specificity of 46.1% (95% CI 34.7-57.8%). The figure for PFAQ was 5/6, with a sensitivity of 89.2% (95% CI 70.6-99.7%) and a specificity of 70.7% (95% CI 58.9-80.3%). The combination of both instruments gave a sensitivity of 94.4% (95% CI 58.9-80.3%) and a specificity of 83.3% (95% CI 72.3-90.7%).
Conclusions:
This screening test, using MMSE and PFAQ, has a good sensitivity and specificity for the diagnosis of dementia in Chile. Being simple and of low cost, it can be applied in primary health care.
