SIDS stands for Sudden Infant Death and used in coding does not corrolate to "crib death" but to unexplained death of an infant under 1 year. Coding is changing in many states, and the way things are coded makes these kind of statistics very difficult to gather. In Michigan, for instance, they have stopped coding as "SIDS" and they now code as positional asphyxia (suffocation). The few deaths that are not found to be suffocation are coded as "SUID" Sudden Unexpected Infant Death. Furthermore, the data is recorded into groups like "soft surfaces," "layover" etc. and that cannot be broken down to see what percentage is couches vs. beds, sober mothers vs. intoxicated non-relatives, etc. there is no way to break it down. The data you will find is that safe sleep related deaths are the number one cause of infant death one month to one year. Many of these tragically happen in cribs, couches and other non-family bed locations, but at least here, you can't break out the data, so there is no way, at least locally, to refute claims that it's due to co-sleeping or that co-sleeping can be done safely using the data collected by health departments.
My mouse isn't working properly and I'm having trouble copying in the links, but MDC's Sleep Environment Safety Checklist that is currently up on the main MDC page is fabulous! I love having resources like this to share with new parents and concerned family when we talk about safe sleep and making safe choices in the crib OR in the bed. Too many people assume that crib = safe and bed = unsafe, but even the most strident safe sleep educators I know (I work in public health) agree that this is not the case, cribs are not automatically safe.
Anyway, this is a bit of a soapbox issue for me. They COULD code the data to refute or prove the claims that co-sleeping can be done safely, but they have chosen not to.