Clinical trials in Alzheimer's disease. A report from the Medical Research Council Alzheimer's Disease Clinical Trials Committee.

Recent advances in Alzheimer's disease imply a need for adequate clinical trials of new treatments which require careful design. The disorder is progressive and shows clinical heterogeneity. While large-scale trials of elderly subjects are appropriate in relation to assessment of drugs or other treatments designed to prevent progression of the disorder, the outcome measurements in such biological treatment trials require careful planning. Studies of individual patients are relevant for answering certain specific questions. Relatively short cross-over trial designs may be appropriate to some pharmacological studies. The choice of neuropsychological instruments for measuring change is critically important, particularly in excluding test/retest artefact and in avoiding floor and ceiling effects. Test scales designed for assessment of specific neuropsychological deficits, or forming part of standard IQ assessments are unlikely to prove robust. Tests can be selected and developed for individual patients, but generalisation of the results of such experiments to the disease as a whole is not inevitable. There is a need to develop psychological instruments for measuring change that are robust and relevant to the clinical problem of progressive dementia.