My answer is mostly a rehash of anongoodnurse's general idea, but here we go:
If you make it ontopic, make it a Skeptics.SE model. Only answers explicitly backed by peer-reviewed research should be allowed for medical questions. In the most literal way, if an answer wouldn't pass a muster on Skeptics, it shouldn't pass here either. Up to an including asking Skeptics moderators to arbitrate (in case of major disagreement, or to help Parenting mods get into the groove - not to imply that they need handholing :)
This approach would place extra high load on moderators, FYI, AND require active user side moderation.
Also, please see if there's a huge scope overlap with http://health.stackexchange.com/
Any time a question gets answered with popular but unreferenced advice, we can always easily take that advice as a "notable claim", and turn it into an excellent Skeptics.SE question - this allowing us to close the original one here if needed without leaving the original poster with no avenue of info.