Frederick Cahn, PhD06.05.07
Medicare’s Plans for 2008 Inpatient Payments
Frederick Cahn, PhD
Frederick Cahn, PhD |
The basic concern is that because the existing DRG categories do not adequately account for variations in the severity of cases within a DRG category, there are incentives for profitable specialty hospitals to be established in order to “mine” the Medicare patients for the less severe cases within a DRG. However, the DRG system proposed for 2007 would have resulted in 10% to 15% reductions in payments to hospitals for many technology-intensive procedures, including use of stents and pacemakers.
In response to many objections and lobbying from the medical device industry, the CMS scaled back its proposal and made only modest changes to the DRGs by creating 20 new DRGs and modifying 32 DRGs to better capture differences in severity.
This year, the CMS is revisiting this issue and is proposing a less radical restructuring of its DRGs to better recognize differences in severity of illness; the agency calls its new system the Medicare Severity DRG (MS-DRG). As many commentators to the 2007 proposal suggested, the MS-DRG system is based on the existing DRGs rather than the completely different and proprietary system proposed last year. The existing DRGs have been refined over a period of years to differentiate resource utilization of different medical and surgical treatments.
The current DRG system already does have some severity adjustments. For example, 115 DRGs are split based on the presence or absence of “complications or comorbidities” (CC). For these DRGs, the presence of a CC assigns the discharge to a higher paying DRG. However, there are two problems: One is that the CC list is outdated; the other is that two levels of severity are not sufficient.
The current list of diagnoses designated as a CC (out of about 13,000 diagnosis codes) initially was created at Yale University in 1981. There are 3,326 diagnosis codes on that list. The problem is that our medical system has undergone many changes since 1981. Some are structural, such as reduced length of hospital stays and a shift of many services to outpatient or post acute care facilities. These structural changes largely have been driven by the inherent incentives of a fixed payment for a hospital stay, regardless of length of stay and the hospital’s cost for treatment. However, another factor has been a very large increase in the (accurate) coding of secondary diagnoses in claims, which is driven by the incentive to capture all of the “CC” conditions that may be applicable in order to receive a higher payment. The net effect of these factors is that today, nearly 80% of patients admitted to a hospital now have a CC in their claim.
Thus, the first component of the CMS’s changes to the DRGs was a comprehensive review of all of the 13,549 secondary diagnosis codes to determine which codes should be classified as a CC. The review was based both on statistical analysis of Medicare claims data as well as medical judgment based on current clinical practice. The revised CC list essentially is comprised of significant acute disease, acute exacerbations of significant chronic diseases, advanced or end-stage chronic diseases and chronic diseases associated with extensive debility. Compared to the existing CC list, the revised CC list requires a secondary diagnosis to have a greater impact on hospital costs. Thus, the CMS review reduced the number of diagnosis codes on the CC list from 3,326 to 2,583.
The agency then decided to further divide the secondary diagnoses by adding a “major complications and comorbidities” (MCC) category that represents a higher level of secondary diagnosis severity. In the new system, about 22% of patients will be classified as having a MCC, 37% as having a CC, and 41% with no CC.
The CMS then reorganized the DRGs. First, the agency combined the 115 pairs of DRGs that had been previously subdivided using the original CC list. Thirty-four of these “base DRGs” were eliminated by combining DRGs that had been split based on age, complexity, cardiovascular conditions and burn DRGs that had been split based on significant trauma; also, some low-volume DRGs were eliminated and merged into clinically similar DRGs; and some other DRGs with similar resource use were merged. DRGs relating to newborn, pediatric and maternity were left untouched because of their low volume in the Medicare population.
The second step was to provisionally split each of the resulting 335 “base MS-DRGs” into three subgroups—MCC, CC and non-CC—and then analyze the statistics of the three groups. Depending on the volume, charges and variance in the three groups, many of these provisional DRGs then were recombined. The end result was that some DRGs remained split into three severity levels, some were split into two levels (either MCC or CC vs. no CC, or MCC vs. non-MCC) and some were not split at all. The end result is 745 proposed MS-DRGs, compared to 538 in the current DRGs.
The agency’s analysis is that the impact of payments under the proposed MS-DRG system would be largely redistributive within each base MS-DRG, whereas the system proposed last year would have redistributed payments between different types of cases. It appears that payments for technology-intensive procedures are not as strongly affected as in last year’s proposal. Furthermore, the new system shifts more payment to urban hospitals, because patients treated in urban hospitals generally are more severely ill than patients in rural hospitals. It also favors hospitals that treat a more resource intensive patient mix. Both of these trends should tend to favor payments for procedures depending on advanced technology.
The MS-DRGs do not represent the end of the DRG revision process. The CMS has engaged the RAND Corporation to evaluate how well various commercial alternative DRG systems distinguish severity. RAND has released a preliminary study, with the final report due later this year. Thus, although the CMS is proposing to adopt the MS-DRGs for fiscal year 2008, this decision does not preclude the agency from adopting any of the systems being evaluated by RAND for fiscal year 2009.
Other, less sweeping changes in the 2008 IPPS also may have an impact on the medical device industry. The CMS is developing a plan to limit payments to hospitals for situations in which the increased severity of a case is due to causes that should be avoidable in the hospital (eg, bed sores, some hospital acquired infections) and to make specific revisions to the DRG assignments of some procedures involving medical devices, specifically neurostimulators, intracranial stents, cochlear implants, knee and hip replacements, and spinal fusions. In many cases, the new assignments will result in increased payment for these procedures.
Furthermore, the CMS is beginning to address “charge compression,” the practice of applying a higher charge markup over costs to lower cost than higher cost items and services. This has been a concern of the medical device industry, since this practice has the effect of lowering payment to hospitals carrying out procedures using a larger proportion of expensive medical devices. This topic will be explored in greater detail in a subsequent article.