While there is obvious value in prospective, randomized, clinical trials, this study is an elegant demonstration of how, sometimes, there are experimental questions that cannot be answered by randomization and must be controlled for in some other experimental way.