Okay. Well, the -- maybe a few things just from -- making sure there's a good baseline understanding of how we do our statistics to begin with. First of all, HIT incentive revenue is not included in patient revenue because of the way the regs are constructed. It says it's not for delivery of patient care, it's for completion of system implementations, and so we didn't put that into patient revenue, so it's not in our stats. Others may have it in their stats. I don't know. Also we don't include physician revenue on employee positions in our patient revenue for the sake of these stats. And of course, growth in Conifer is also not included because that's not patient revenue. So there's a number of elements of revenue which we have, which are growing, which are not in the stats, and that may be creating some apples-to-oranges. Beyond that, I can drill into greater detail on some of the things that I already talked about and quantify a little bit, but the things that people might not otherwise think of, as I said, we had the increase in or decrease, if you will, in our pricing of the uninsured in certain markets where we just felt like it was the right action from a competitive standpoint in a tough economic environment. And that was at a 0.5% or a 50 basis point move in the overall price per adjusted admit number. Once again, that's offset by lower bad debt expense, so it doesn't fall to the bottom line. I think the payer mix shift was worth about 0.5%. I would expect others have that as well or possibly even greater, but I don't have visibility. And then, I hate to go there, but the whole adjusted admission process is an estimation of how to combine inpatient and outpatient. And if you instead do a weighting of our visits, pricing growth and our inpatient pricing growth, you get a number about 0.4% difference. So that also is a driver of the stat. I mean, there's still -- everything's directionally correct, but it has some measurable effect as well. So that is some of the things, I think, that people otherwise wouldn't see. The Medicaid hits are apparent. The imaging volumes we've talked a lot about. So when you put all those together, it explains the difference for us. And as I said, some of that just really doesn't fall to the bottom lines, so it's more optics than reality.
John Rex - JP Morgan Chase & Co: That actually helps, to quote some of that. And then you had mentioned in the acuity topic, which has been the topic, you mentioned a decline in Medicare CMI. I don't think it was much, the way you characterized it, but were there just any specifics in terms of categories, and say, cardio's been getting spotlighted here in terms of where you might saw declines?