THE VALIDITY OF THE JUDGMENT OF GENERAL-PRACTITIONERS ON DEMENTIA

This study focused on two aspects of the validity of the judgement of general practitioners (GPs) on dementia: criterion and construct validity. The Cambridge Mental Disorders of the Elderly Examination (CAMDEX) was used as the gold standard to assess the specificity and sensitivity of the GP judgement. Construct validity was studied by comparing patient characteristics that were hypothesized to predict the GP judgement by predictive variables from our data, using logistic regression analysis. In addition, the patient characteristics that predicted a 'false positive' or 'false negative' GP judgement were determined. Thirty-six GPs participated in the study, as well as 475 community-dwelling patients. The overall agreement between CAMDEX diagnosis and GP judgement was 72%. The sensitivity and specificity of the GP judgement in making the distinction between 'normal cognitive functioning + minimal dementia' versus 'mild + moderate + severe dementia' were 52% and 94% respectively. Patient characteristics that predicted the GP judgement on dementia were: age, sex, cognitive functioning, IADL functioning (Instrumental Activities of Daily Living) and memory complaints. All these characteristics, except sex, were hypothesized to do so. Patient characteristics predicting a 'false positive' or a 'false negative' GP judgement were cognitive and IADL functioning. Both groups of patients ('overdiagnosed' and 'missed' by the GP) belong to the in-between group with some impaired functioning. The finding that impaired IADL functioning is so important in the GP's judgement reflects the significance of dementia diagnosis in general practice: a starting point for adequate care.