Well, it’s true that honey is not sterile, but it is actually pretty close. In addition to putative antibacterial chemicals, the osmolarity due to the sugar content does prohibit bacteial growth. Botulinum spores are an interesting adaptation in that they are dormant elements encased in a protective coat and so are able to survive conditions which living bacteria could not. So honey is, as Doug points out, not necessarily perfectly safe, but there is also some logic to the idea that applying it to wounds could inhibit bacterial growth.
The logistical problem, and it’s a big one, is that tissue fluids dilute the honey and reduce the osmalarity, and then it becomes a perfect medium for bacterial growth! So honey soaked bandages have to be changes very often or they actually can make infection worse. There are far better options availabel now, so medical use of honey is really only a curiosity at this point.