Objectives: To identify the percentage of patients with Alzheimer's disease (AD) in a general clinic population who would be provisionally eligible for randomized clinical trials and the extent to which these patients represent the overall clinic-based population.
Background: Many randomized clinical trials have restricted enrollment criteria that may limit generalizability, i.e., AD patients who fulfill selection criteria for phase III clinical trials may not be representative of other AD patients in clinical settings.
Design and setting: Patients diagnosed as probable or possible AD from the nine clinical sites of the State of California's Alzheimer's Disease Diagnostic and Treatment Centers (ADDTC) were selected on the basis of their provisionally fulfilling the inclusion and exclusion criteria of two typical AD clinical trials at the time of their first visit (ECG and brain imaging criteria were not available).
Results: From a sample of 3470 subjects with possible or probable AD, overall, only 4.4% or 7.9% would have been provisionally eligible for each of two trials. Patients provisionally eligible were younger, relatively underrepresented by women, better educated, wealthier, and more likely to be white than ineligible patients. The major independent demographic predictors for eligibility were (1) income greater than $15,000 per year, (2) male gender, and (3) college education. More than 60% of probable AD patients were excluded because of significant behavioral problems; approximately one-quarter each were excluded because of significant medical or neurological problems. Allowing patients with probable or possible AD to enroll would have resulted in 10.6% being eligible.
Conclusion: Selection criteria for AD clinical trials result in a demographically and clinically constrained subgroup that is not representative of the overall clinic population.