Combination Therapies and the Theoretical Limits of Evidence-Based Medicine

The document from 2001 highlights the difficulties in determining the best combination therapies for conditions like Alzheimer's disease and ischemic stroke due to the vast number of potential treatment combinations. It points out that for Alzheimer's, testing all possible combinations would require 8,128 trials with over 4 million patients, while for stroke prevention, 496 trials with over 2.9 million patients would be needed. The authors propose that hierarchical, serial clinical trials could streamline this process to 127 trials with 63,500 patients over 286 years for Alzheimer's, and 31 trials with 186,000 patients over 155 years for stroke prevention. They conclude that the current clinical trial system is significantly limited in its capacity to test combination treatments and suggest alternative approaches such as preclinical testing, surrogate outcome measures, and factorial phase III trials to overcome these challenges. Additionally, they recommend that regulatory agencies might need to accept combination regimens that significantly reduce disease burden with tolerable side effects as sufficient, given the impracticality of exhaustive testing.