Well, you don't want to do that, necessarily, at the beginning, because all these drugs come with side effects. What you want to do, obviously, is to start with something where you have the best chances of success with the lowest number of side effects.
Your specific issue, though, is very similar to mine, in the sense that one thing we have a challenge with in treating drug-resistant infections is diagnostics. Oftentimes infections, like a bloodstream infection, show up as general symptoms as opposed to a specific bug giving you a specific infection, so a lot of treatment is empirical. As a result, you have this issue, in particular when it comes to drug resistance, trying to figure these things out.
As we move forward into the future—and we've heard this, I think, from almost all the speakers today—molecular diagnostics is the way of the future, at least in these kinds of situations. It's harder to do, obviously, in resource-poor areas, but certainly that's an area where there's tremendous opportunity for innovation and big data, and all that kind of stuff.