Aims: To analyse the validity of a set of neuropsychological and functional tests, and to study their value in detecting and diagnosing dementia through a pilot study.
Patients and methods: A total of 131 subjects (101 controls and 30 with dementia) were evaluated using a comprehensive neuropsychological and functional battery. Validity analyses were conducted using ROC curves in accordance with the definitions of diagnostic test validation. Finally, a discriminant analysis was performed with the tests that showed greater diagnostic validity in the study of the ROC curves.
Results: The case and control groups were not significantly different as regards age, sex and level of schooling. The ROC curves analyses showed the following to be the tests with the highest diagnostic validity: the MMSE, delayed recall of a short story, delayed recall of six pictures, the Spanish version of the S IQCODE (shortened) and Pfeffer s FAQ. The discriminant analysis evidenced the fact that the joint utilisation of all the foregoing tests, except delayed recall of six pictures, classified 96.55% of our sample correctly.
Conclusions: By combining direct cognitive evaluation of the subject and functional performance evaluated by a trustworthy informer, the vast majority of participants in the pilot study were correctly classified into patients with and without dementia. The high diagnostic validity of four relatively short tests lends support to their use in broader clinical or population studies.