By Christopher Guadagnino, Ph.D.

Published October 2007

• React to this article in the Discussion Forum.

DM: The Centers for Medicare & Medicaid Services (CMS) put out one of the most sweeping changes of the DRG system since it was created in the early 1980s. They have revised it significantly to move from 538 DRGs – patient classification categories – to 745 new Medicare Severity DRGs (MS-DRGs). This change, scheduled to go in effect Oct. 1, is something that will allow the Medicare inpatient prospective payment system to better address patient severity, and will hopefully better target patients by the severity and cost of their care. Unfortunately, this change was done along with a significant so-called "behavioral offset" – a 4.8 percent reduction to hospital payments spread over three years, which equates to $20.3 billion over the next five years.

There will also be a reduction in capital payments to urban and teaching hospitals totaling probably more than $2.5 billion. That cut comes in two policy decisions. The first was to eliminate the three percent add-on payment for hospitals that are in large urban areas. For teaching hospitals, CMS removed the indirect medical education adjustment – or payment add-on – that helps cover the indirect costs of running a teaching program.

The third major reimbursement policy change is related to hospital-acquired conditions. CMS has set aside eight conditions where, if they occur during the hospital stay, the presence of complications and comorbidities will no longer lead to a higher DRG payment. Those conditions are: object left in during surgery, air embolism, blood incompatibility, catheter-associated urinary tract infection, pressure ulcer, vascular catheter-associated infection, mediastinitis after coronary artery bypass graft, and hospital-acquired injury – which include fractures, dislocations, intercranial injuries, crushing injuries and burns.

DM: I think the rationale behind the MS-DRGs is to get a better sense of patient acuity, and to better target payments toward the severity of the patient’s condition and the complexity of their care. The rationale CMS is using for the behavioral offset cuts is that they believe payments will go up because of the way hospitals code their claims, rather than because of actual changes in patient severity. For capital, we believe CMS has moved forward with this policy strictly for budget purposes and has disregarded the high costs these hospitals experience, the significant changes that are needed to keep hospitals technologically advanced and up-to-date, and physician shortage issues – at a time when we need to invest more resources in the training of physicians. The primary impetus for the hospital-acquired conditions proposal was a requirement in the Deficit Reduction Act of 2005 that required hospitals to identify preventable complications of care that could cause patients to be assigned to a higher-paying DRG. The key concern here is whether these are preventable and whether there is an evidence-based practice that hospitals can use to prevent them.

DM: The move to MS-DRGs is meant to better target Medicare payments to where Medicare costs are. That’s something that the hospital community supports, if we move that way gradually in a way that makes sense. Unfortunately, CMS is moving too quickly. So, what we’re going to see this year is a lot of confusion among hospitals not being able to identify what they’re going to get paid, or even knowing what the payment for any particular case is when they submit a claim to Medicare. The reason is that many vendors and fiscal intermediaries that pay Medicare claims have not yet put together software packages that hospitals can use to group claims into those new MS-DRG categories. While CMS may be able to pull this off, and hospitals won’t have to bill differently, hospitals won’t know what they’re being paid under the new categories. They won’t know what their accounts receivable are, and it will be very difficult for them to identify cases in which CMS or the fiscal intermediaries have made a mistake in paying them.

As for the broader implications of the MS-DRGs, they will have varying impacts on different hospitals. As you redistribute Medicare payments based on patient severity, in essence you then redistribute payments to hospitals. In general, a hospital that has a lower patient severity mix – a smaller hospital with less complex patients – would see its Medicare payments decrease. Hospitals that have very complex patients should see their payments increase. These changes have been made budget-neutral, so to pay more for certain cases, Medicare has to pay less for other cases. A good example would be pneumonia. Before MS-DRG implementation, we have two buckets: pneumonia cases with complications and comorbidities, and pneumonia cases without complications and comorbidities. The new MS-DRG would split those cases into three categories: patients with no complications and comorbidities, patients with complications and comorbidities, and then patients with major complications and comorbidities. In many cases, the new system splits two DRGs into three, and reimbursement for the DRG without complications will be quite a bit lower than the existing DRG without complications and comorbidities. It is also possible that the DRG with complications and comorbidities will be less than the current DRG with complications and comorbidities because Medicare has taken out the major complications and created a third tier for those. On the whole, larger, urban and teaching hospitals will benefit under the MS-DRGs because the most complex and sickest patients are going into those tertiary centers.

The other big change is the behavioral offset: in 2008 Medicare is going to reduce the payment to every hospital in the country by 1.2 percent, which equates to $1.5 billion. They’ll also do two cuts of 1.8 percent on top of that — one in 2009, and another in 2010. We believe that there will be hospital coding changes that lead to higher payments, but the magnitude of change that CMS has estimated – 4.8 percent over the next three years – is way out of the ballpark. Right now they are guessing. When you look at the creation of the Prospective Payment System (PPS) back in 1983, the change in coding when we went from a cost-based system where coding didn’t matter at all to a brand new prospective payment system was less than what CMS is estimating under this new change. There is no way that we should expect a coding change after 20 years to be higher than when we first went to the PPS system. CMS’s estimate was based on a change that happened in Maryland: hospitals there went from a DRG system to "All Patient Refined" DRGs. The 4.8 percent number was the impact for two hospitals in Maryland – very high tertiary care hospitals. Maryland hospitals have a waiver from the Medicare program and are paid very differently from all of the other hospitals in the country – payment rates are set by the state – and CMS has no information to go on from other hospitals in the country.

We wanted a four-year transition to this new MS-DRG system, to allow hospitals to understand and plan for the types of services they may and may not be able to continue to provide in the future. CMS only went with a two-year transition to this system. We believe that CMS should make the changes for coding behavior on a retrospective basis, after they have actual impact information regarding changes in coding and changes in patient severity. To take $20 billion out of the system based on a guess is irresponsible.

DM: We did a lot of analysis of the new system and proposed its implementation over a four-year period. The key push that we’re going to be advocating is giving hospitals a year before implementation of the MS-DRGS, and removing that prospective behavioral offset cut. We want to see the capital payment cuts removed, as well. This is a final rule, so what we’re doing is pushing Congress to stop them from implementing it this year.

We’ll also be working with CMS on the hospital-acquired conditions. We supported inclusion of three conditions for non-payment: air embolism, objects left in during surgery and wrong blood type. The other ones, we believe, need a little more thought. There are certain places where case-by-case review is really necessary. There are certain conditions where the science of exactly how to prevent them all the time is not developed yet. There is also a new coding mechanism requiring hospitals to document whether each diagnosis that they put on a claim was present on admission or not. For a lot of conditions, the hospital hasn’t determined whether it was present on admission because they didn’t do a test for, let’s say, urinary tract infection when the patient came into the hospital for a heart condition. What we are going to see is a real change in the admission process, not just for the hospital, but for physicians who have to order these tests. What we need to do is look at how the coding for "present on admission" works over the next eight months and decide whether it really does make sense to use these for payment purposes beginning in October. That’s something we’ll be tracking very closely so that we can make comments to CMS while they’re developing a rule for 2009.