«DEPARTMENT OF HEALTH AND HUMAN SERVICES Centers for Medicare & Medicaid Services 42 CFR Parts 409, 424, and 484 [CMS-1560-F] RIN 0938-AP55 Medicare ...»
LUPA mis-application and also for the real need severity CMS determined exists. This commenter wrote that the combination of LUPA incidence, an outlier rate below 5 percent, changing the single therapy threshold to multiple therapy thresholds, and the increased incidence of high therapy cases constitutes more than 100 percent of the observed increased in the average case-mix weight.
Response: Based on a 10 percent random beneficiary sample, our data show the LUPA incidence rates from 2001 to 2007 were the following: 15.06 percent, 14.11 percent‚ 13.35 percent‚
12.53 percent‚ 12.12 percent‚ 11.16 percent‚ 10.54 percent. We note that LUPA incidence rates, while higher than the forecasted 5 percent, continue to decline. LUPA episodes were not used in the measurement of case-mix change in either our analysis or in the Abt Associates model of real case-mix change. We have no evidence that LUPA episode assumptions caused agencies to be underpaid; in fact, margin analysis shows PPS payments have been adequate. It should be recognized that we proposed to adjust the episode national standardized payment amount to be consistent with an outlier expenditure proportion of less than 5 percent of total outlays. This upward adjustment is a continuation of the
adjustment is simply to provide for a lower rate of outlier expenditures than the 5 percent assumption we have traditionally used. We made this proposal in conjunction with the proposal to cap outlier payments at 10 percent on an per-agency basis. We have no basis to change payment rates on account of the refinement of the therapy thresholds. Even if agencies return to more clinically based therapy treatment plans, resulting in a new distribution of therapy visits per episode and reduced total expenditures, we would not make any payment rate changes in isolation from other issues, such as the change in the mix of visits since the original PPS final rule, and change in the total number of visits in a 60-day episode. Similarly, we do not believe it is appropriate to adjust payment rates for the deviation of LUPA episodes from the forecasted 5 percent, in isolation from other issues, such as addressing the issue of lower visits per episode existing today, as compared to the number of visits per episode on which the HH PPS rates were originally based. We believe that the appropriate time and place to deal with any re-estimates, in these multiple areas, is if and when a rebasing for the rates were to take place.
Comment: A commenter wrote that the elimination of the
incentives with patient care needs. This commenter felt the case-mix change primarily reflects growth in therapy utilization. A different commenter asked CMS to clarify how going from single to multiple therapy levels did not constitute a “double dip” penalty. This commenter wrote that the multilevel therapy equation model HHRG modifications may have lowered the relative value for all higher therapy cases, but the commenter couldn’t confirm this since CMS did not release the data. The commenter stated that “re-jiggering” of service factors was likely directed toward lowering reimbursement rates and having therapy services delivered in a more clinically driven manner. The commenter added that the relative loss of aggregate case-mix weight under the 4-Equation model equals measured case-mix weight change, which is tantamount to a “double dip”. Another commenter wrote that the data he analyzed showed that 95 percent of case-mix growth was a direct result of higher levels of service domain in care delivery under PPS. He added that when PPS was originally proposed, and again in 2007, CMS acknowledged that it did not have good data to measure or apply case-mix based on patients’ service needs, yet CMS stated that it believed that the multi-level therapy thresholds was an
of total episodes under HH PPS. Ten to twelve therapy visits, a range that would generally be most profitable to agencies, grew the most, and by 2007 such episodes accounted for about one quarter of all the episodes that had a therapy visit. These episodes, of course, also were among those with the highest case-mix weights and had a minimum case-mix weight of 1.4847.
One goal of the case-mix refinements was to better match payments with agency cost experience under PPS; thus we used 2005 data for estimating the final case-mix model that was used for the 153-group system. Changing to multiple therapy thresholds with a gradual increase in payment better aligns costs and payments and avoids incentives for providers to distort patterns of good care that would occur at each proposed therapy threshold. As a disincentive for agencies to provide more care than is appropriate, we proposed that any per-visit increase incorporate a declining, rather than constant, amount per added therapy visit. It should be understood that the refined case-mix methodology redistributed the resource costs expended in 2005 to the new set of 153 groups we defined from the severity levels developed from the four-equation model generating OASIS item scores. Instead of a single high therapy
10 or more therapy visits (including those with the very highest number of therapy visits), the refined system had multiple therapy ranges, with the payment addition for therapy being based on all episodes with therapy visits in the stated range.
Therefore, the right tail of the distribution (that is, cases with the highest therapy visits and thus the highest resource costs for therapy) is not figuring into the payment increment until the 20+ therapy visit level is reached. Thus, it was our intention to have lower payments for episodes with 10 to 12 therapy visits, so as to better align costs and payments.
The redistribution of resource costs among the new 153 groups resulted in some lowering of case-mix weights, as just described, but all the resource costs expended in 2005 were accounted for in the payment system. The final case-mix change adjustment addresses nominal case-mix change and is applied across all case-mix groups in a similar manner. Therefore, the final case-mix adjustment is completely separate from the realignment of payments to the 153 groups, and thus there was no double-dipping. In sum, the multiple therapy thresholds and the case-mix change adjustment are unrelated and do not doubly adjust the rate as each adjustment is clearly warranted by the
We do not have enough information to verify the commenter’s finding that 95 percent of case-mix growth was a direct result of higher levels of service domain in care delivery under PPS.
Comment: Several commenters wrote with suggestions or alternatives to the case-mix analysis. One commenter wrote that CMS should continue to work on developing post-acute care national assessment tool for use across all settings, which would allow CMS to better determine what settings were appropriate for patients based on acuity. It would also allow CMS to understand how changes in home health case-mix are affected by the type of patient admitted to home health. Some wrote that CMS should allow implementation of OASIS-C before any further case-mix reductions are made. A commenter suggested that we fully analyze and compare information within OASIS-C with the development and testing of the Continuity Assessment Record and Evaluation (CARE) instrument. Another commenter felt that the data from OASIS-C would be helpful to CMS in determining real changes in case-mix rather than those stemming from coding or documentation improvements.
A number of commenters felt that the proposed 2011 adjustment was too steep, particularly given low or negative
another suggestions that we consider the impact on low-margin agencies before finalizing the rule. Some commenters suggested that the complexities of the case-mix methodology warranted making relevant CMS staff and contractors available to respond to questions regarding the assessment methods prior to expiration of the comment period. Additionally, these commenters suggested that CMS make all data used in the analyses available, and provide a 120-day comment period to allow time for expert analysis to evaluate the methodology and findings. A different commenter was strongly opposed to reductions for 2011 until more analysis of medical necessity of the care provided was complete. This commenter encouraged us to reduce or eliminate the creep attributed to the shift to provision of higher therapy services unless clear evidence existed that the therapy services were not medically necessary. This commenter suggested we make a distinction in the application of creep between therapy and non-therapy HHRGs, and recommended that physical and occupational therapists be added to MAC review departments with mandatory education and experience as qualifications for medical review.
Response: We thank the commenters for these thoughtful
continuing our work associated with the post-acute care demonstration. We are currently in the early stages of data analysis of the assessment data and resource data which has been collected to date. We will finish data collection by the end of calendar year 2009. We remind the commenter that the analysis of these data is a multi-year project, and that the analysis will consider the data collected via the CARE instrument, the validity and reliability of those data, and the strength of the items as payment predictors. CMS plans to present the analysis of the data collected during the demonstration and associated recommendations to Congress in the summer of 2011. Regarding the commenters’ suggestions that we wait to make further casemix reductions until we assess the OASIS-C data, we remind the commenter that the OASIS-C revisions did not significantly change payment items. We believe that the commenter may be suggesting that CMS analyze OASIC C non-payment items to assess whether these new items would enable CMS to better identify the health status of the patient, and whether these new items might be more reliable in assessing real patient acuity change versus that which is unrelated to real changes in acuity (nominal). It is important to note that because we are just beginning to
analysis would only be possible after several years of OASIS-C data collection. We may consider the suggestion that we account for increases in nominal case-mix over a longer period of time, in future rulemaking. In this final rule, we are not accounting for additional changes in nominal case-mix which we identified from current data analysis. Rather, we are maintaining the policy, finalized in CY 2008, to reduce CY 2010 base episode payments by 2.75 percent. With regards to the suggestion for a 120-day comment period, we are unfortunately unable to adopt such a comment period given our rulemaking timeframes, but we will continue to make every attempt possible to share our analyses with the public in as timely as possible. Regarding the commenter’s suggestion that CMS should assess the medical necessity of therapy visits before applying up-coding reductions, as we described in an earlier comment, we find this suggestion impracticable. With finite resources, it would be challenging to perform a medical review on every claim which includes therapy.
Again, as a point of clarification, the 2.71 percent reduction for CY 2011 is not a proposed adjustment. That percentage reduction was promulgated in the CY 2008 final rule
Comment: A commenter stated that while he did not assess changes in home health case-mix, an increase in case-mix unrelated to severity in 2007 confirms the need for continuing review of annual case-mix change. The commenter noted that nominal changes in case-mix had been found when major revisions were implemented in other payment systems, suggesting particular scrutiny of the 2008 changes in case-mix was warranted. The commenter wrote that if additional nominal case-mix change was indicated, CMS should adjust payments as appropriate. The commenter further recommended that we combine the planned reductions for 2010 and 2011, and reduce payments in 2010 by 5.5 percent, and that payments should be rebased to a level equal to average costs in 2011.
Response: We thank the writer for these comments. We agree with the commenter that we need to continue to analyze current data as they become available to us and update our identification of nominal case-mix using these more current data. We are currently analyzing 2008 data to assess the impact of our CY 2008 refinements, and determine the effect these refinements may have had on nominal case-mix growth and will address the need for additional reductions to the HH PPS rates
Comment: Another commenter wrote that CMS uses MedPAC’s reports of strong profit margins and high levels of new entrants to bolster the view that access will be unaffected after the full creep cutbacks are implemented. This commenter wrote that an industry association disagrees with MedPAC’s methodology, and concluded that one-third to one-half of HHAs would lose money when creep reductions are fully implemented. The commenter questioned MedPAC’s use of a sample of HHA cost reports representing less than 60 percent of HHA visits. This commenter asked that the full information from MedPAC be released and subject to review since CMS is supporting its case-mix reduction using that report.
Response: We would like to assure the commenter that the analysis and associated methodology CMS used to differentiate between real and nominal case-mix growth involved extensive analysis, which is fully documented in the Abt report, publicly available via the HH PPS website at www.cms.hhs.gov/Reports/downloads/Coleman_final_April_2008.pdf.
We understand that the commenters are concerned about whether we are taking into consideration the financial conditions of hospital-based home health agencies. As MedPAC
hospital-based providers are impacted by the allocation of overhead costs from the hospital. We agree with this assessment and believe that using this information would not provide an accurate view of the overall industry margin or the impact of the proposed change to the payment system.