Brought to mind http://slatestarcodex.com/2017/08/29/my-irb-nightmare/ . Maybe our standards for a)-style evidence are too high, and it needs to be easier to do evidence-based work within clinical practice?