A combination of tests for the diagnosis of dementia had a significant diagnostic value

• Published in: Journal of clinical epidemiology 2005-03, Vol.58 (3), p.217-225
• Main Authors: De Lepeleire, J., Heyrman, J., Baro, F., Buntinx, F.
• Format: Article
• Language: English
• Subjects: Adult and adolescent clinical studies; Age Distribution; Aged; Aged, 80 and over; Behavior; Biological and medical sciences; Cognitive ability; Confidence intervals; Cross-Sectional Studies; Dementia; Dementia - diagnosis; Dementia disorders; Diagnosis; Epidemiology; Family medicine; Family Practice - methods; Female; General aspects; General practice; Humans; Logistic Models; Male; Medical sciences; Mental disorders; Methodology; Neuropsychological Tests; Older people; Organic mental disorders. Neuropsychology; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Public health. Hygiene; Public health. Hygiene-occupational medicine; Regression analysis; Sensitivity; Sensitivity and Specificity; Specificity; Studies
• ISSN: 0895-4356; 1878-5921
• DOI: 10.1016/j.jclinepi.2004.07.005

To study the diagnostic parameters of a number of instruments for a diagnosis of dementia in general practice and the added diagnostic value of these tests. Cross-sectional diagnostic research in general practice. Participants: 152 persons aged 65 plus. The Mini-Mental State Examination (MMSE), the Clock Drawing Test, the ADMP scale, the Timed Up and Go Test, the Extrapyramidal Sign Scale, the Behavior Observation Scale, the Poon–Baro–Wens computer battery, and the Cognitive Drug Research Computerized Assessment System were evaluated against the Dutch version of the Cambridge Examination for Mental Disorders of the Elderly (CAMDEX-N). Diagnostic characteristics were calculated with their 95% CI. Using forward stepwise logistic regression analysis, a model was built with CAMDEX-N as the dependent variable and the tests under study as independent variables. Area under the curve was the main parameter for the comparisons. The main diagnostic gain results from age and ADMP, followed by the Clock Drawing Test. Subsequent addition of the MMSE and computer tests results in modest additional gain only. The final model including five tests has an area under the curve of 0.95. Sophisticated neuropsychological computerized tests have little added value in the diagnostic work-up of dementia in general practice. Basic clinical tests used in an appropriate sequence can be very valuable in establishing the diagnosis of dementia.