Here's the NEJM piece: http://content.nejm.org/cgi/content/full/357/4/328 They should have do a proper study. It would be hard to exclude bias, but they should try. The above is just a story.