[bhickey]

It depends.

There are extant phage that predate staph, like Phage K. There are some hurdles you'd need to overcome. Foremost among these is when compared to antibiotics, phage suck for treating disease. This is a big part of the reason everyone beat a path to antibiotics

Phage are extremely narrow spectrum. You can select for a high-fitness phage with serial passages: Plate the host bacterium along with some phage. Find the largest region of dead host (a plaque), pick it and repeat for several hundred generations. The good news is that you'll have a high-fitness phage. The bad news is that it will probably only infect the particular strain of bacteria you've been using to select it. I worked with a fitness optimized E.coli phage that wild-type E.coli probably would have laughed off. I haven't read the therapeutic phage literature, but my assumption would be that you may need to optimize

Phage have some issues with immune clearance. The immune system really loves to eliminate foreign looking viruses. The concentration of phage injected intramusucularly falls very fast. Again, this might be something you can overcome with serial passages. There was some work in rabbits where the authors found an enrichment for mutations in the capsid (the phage's head). I don't recall if they quantified the fitness impact of these mutations.

If I ever get out of software development, it'll probably be to work on phage.