I have written a simple software based solution, it's already finished. There are 2 pieces of software. The first is run by the doctor. So when a patient comes in and needs 30 hydrocodone pills, they input the controlled drug they wish to prescribe, the patient's name, and quantity of that drug (and maybe other stuff like DEA #'s or other things I'm unaware of). Their software then generates a number that is not readable to humans based on that info (for example, 4r8934r984398r4983rj8349rj4938rj9834j). Then they can call/fax/email that number in to the pharmacy themselves, print it on a Rx or whatever. THe pharmacy fills scripts based on this number alone. No verfiable number, no script.
Once the pharmacy receives that number, they then input it to their part of the software, which then determines/verifies the patient's name, the drug they need, the quantity (possibly other info too). Each pharmacy software can be linked to other pharmacies to prevent doctor shopping, or be linked to a central database (ie "prescription monitoring program", but thus far I doubt this works well because it's allegedly already in place). The pharmacists then gives the patient what they need. This number is based on various algorhythms and will not be easy to forge, and that same math protects a lot of other sensitive info in the real world.
To recap, patient goes to doctor, doctor gives patient non human readable number, patient brings it to pharmacy, pharmacy gives patient medicine (if number checks out).
In this system, no stolen rx pads, no multi county doctor shoppers, no callign in bogus scripts, nothing. Even if the doctor's PC were compromised there are logging systems which would alert anytime a script is filled based on their software. This is the same concept used to prevent software piracy, verify login credentials, and much more. No system is 100% secure, but this one would have much higher success rates than the current, it doesn't cost a lot to implement.
The software is finished, it's written in C++ (that means it's not a slow, insecure webapp like most software) and works well. Profit would be nice, but my sheer hate for pain pills is why I did it. You factor in that with dealing with countless know it all doctors and pharmacists, and I'm not sure if it'd be worth it. Your answer, bidingmytime, shows that this is not centrally controlled by anyone and making all those groups come together and agree on something, well, you'd have to be a patient person to do that : ) I'll probably just make it open source and put it out there hoping someone gets use from it.