I like this idea at first glance. It needs a bunch of policy people to poke holes in it, of course. But it seems like it preserves the incentive to keep doing efficacy trials, while improving patient choice.