Observations made by Dr. Eric Topol, about getting rid of randomized trials because they are not in sync with personalized medicine, apply to the unethical use of placebo-controlled, one-size-fits-all studies. These are the studies that Medicare and health insurers use to determine whether or not to pay for a new technology.
Recently, Medicare said it would not pay for a test to detect Alzheimer's earlier than we usually do unless patients were enrolled in a clinical trial: http://www.reuters.com/article/2013/07/09/us-alzheimers-test-lilly-idUSBRE96804F20130709
This antiquated stance is based on the belief that such trials, which can take years to complete, provide the best evidence of what works.
As Topol points out in a Medscape article we republished here, randomized trials are expensive and often fail to tell us who will benefit from a test or treatment. And in the case of this Alzheimer's test, only a small group of patients with a particular genetic mutation would get it in the first place. So why is Medicare using a one-size-fits-all approach? To avoid paying for the test of course.
We could learn very quickly whether, and how, this test will benefit people. But as Topol points out, "That is not going to happen in the contrived clinical trial world. We need to get to the real world and into this digital world where we would have electronic surveillance of every single patient who is admitted and enrolled in a trial. Why can't we do that? Why can't we have conditional approval for a new drug or device or even a diagnostic test, and then monitor that very carefully. Then we can grant, if the data are supported, final approval.
I hope that we can finally get an innovative spirit, a whole new way of a conditional and then final approval in phases in the real world, rather than continuing in this contrived clinical trial environment. These are some things that can change in the rebooting or in the creative destruction, or reconstruction, of medicine going forward."