Instead of developing more antibiotics, can we reengineer these disease bacterias with a kill switch built in? Then release them into the wild to compete with non-engineered species and eventually displace them?
You would have to also engineer an evolutionary advantage - something that would make them likelier to survive in the wild than their non-engineered counterparts. Building in a 'kill switch' by its very nature gives engineered bacteria a huge disadvantage, meaning there'd have to be something else making them more likely to reproduce.
It would probably be a better use of time engineering the bacteria to be harmless to humans, negating the need for a course of drugs to kill it off.