Although there is no effective treatment for Alzheimer's disease, clinicians should be able to provide the patient and family with some idea of the natural history of the illness or an accurate prediction of what to expect. Since the rate at which Alzheimer's disease progress is variable, this in not possible. This project is designed to validate a predictor model based on our previous investigations of the natural history of Alzheimer's disease. The objective of clinical prediction rules is to reduce the uncertainty inherent in the management of Alzheimer's disease by defining how to use clinical information to make predictions about the course of disease. Predictive findings are well defined clinical signs that are not part of the diagnostic criteria, yet are relevant to the clinician and help predict disease outcomes. In our previous studies, specific clinical signs, such as muscular rigidity, myoclonus, and hallucinations or delusions were useful in predicting a selected outcome; patients with either sign reached a more severe stage of dementia earlier than patients without these findings. We also began to define a eries of outcomes that occur as a consequence of the disease process. In the current proposal we will validate and expand this predictor model of Alzheimer's disease. This study will e conducted at three study sites in a sample of 240 patients within 1 to 3 years of the onset of their illness. Standardized regular 6 month assessments will be initiated after all subjects are entered in the first year. We will utilize a series of outcomes that are clinically relevant and meaningful to the patient and family. The new cohort will be larger giving us greater power, and will be collected in a manner that will improve our ability to refine predictive techniques and improve their accuracy. The ultimate goal of clinical prediction in improved patient management. The prognostic implications of this study will be of great interest in the design of therapeutic trails in the future. The proposed studies will also help to improve the efficiency and accuracy of physicians' judgements.
Showing the most recent 10 out of 106 publications
