DRAFT DRAFT DRAFT DRAFT September 4, 2015 Andy Slavitt Acting Administrator Centers for Medicare & Medicaid Services U.S. Department of Health and Human Services Attn: CMS-1625-P Mail Stop C4-26-05 7500 Security Boulevard Baltimore, MD 21244-1850 Submitted electronically via www.regulation.gov RE: CMS–1625-P Medicare and Medicaid Programs; CY 2016 Home Health Prospective Payment System Rate Update; Home Health Value-Based Purchasing Model; and Home Health Quality Reporting Requirements Dear Administrator Slavitt: The National Association for Home Care & Hospice (NAHC) is the largest trade association in the country representing home health care agencies. NAHC members represent the entire spectrum of home care agencies, including Visiting Nurse Associations, government-based agencies, multi-state corporate organizations, health system affiliated providers, and freestanding, proprietary home health agencies. NAHC members serve over several million Medicare home health care beneficiaries each year. NAHC members are significant stakeholders in Medicare and Medicaid, particularly Medicare home 1 health services and Medicaid Long Term Services and Supports (MLTSS). We appreciate the opportunity to submit comments regarding the above-entitled proposed rule. HHPPS Proposed Case Mix Weight Change Adjustment The Centers for Medicare and Medicaid Services (CMS) has proposed to adjust downward the Medicare home health services prospective payment system (HHPPS) rates in 2016 and 2017 to reflect its finding that case mix weights have increased during 2012-2014 at a level greater than any “real” change in case mix. CMS proposes to reduce the base episode payment rate by 1.72% in 2016 and an additional 1.72% in 2017 to capture 3.41% “nominal” growth in the average case mix weight. The CMS calculation is based on the following: CMS estimates that the total case mix increased by 2.76 percent between CY 2012 and CY 2013 CMS estimates that the total case mix increased by 1.41 percent between CY 2013 and CY 2014 CMS estimates that 15.97% of the change in average case mix relates to “real” changes in patients served, the remaining 84.03% is considered “nominal” case-mix growth that relates to up-coding or improvement in coding accuracy The combined “nominal” change in case mix weight is estimated at 3.41 percent (2.76(2.76 X 0.1597) + (1.41- 1.41 X 0.1597) The CMS estimate that 0.1597 of the case mix weight change is “real” is based upon an analysis of data from CY2000 through CY2010. CMS previously evaluated case mix change in three distinct periods: 1999-2005; 2005-2007; and 2007-2009. CMS estimated the “real” case mix change during those periods as 15.12%; 23.23%; and 13.77% respectively. The CMS estimate of “real” case mix change in its proposed rule is based upon its earlier evaluation of data from 2000-2009. CMS did not update its analysis with specific consideration of any of the data from 2012-2014. A July 2011 report, “Analysis of 2000-2009 Home Health Care-Mix Change,” issued by CMS contractor Abt Associates, sets out the factors that it concludes are “drivers” of real case mix change in contrast to nominal changes. https://www.cms.gov/Medicare/Medicare-Fee-for-ServicePayment/HomeHealthPPS/Home-Health-Prospective-Payment-System-Regulations-and-NoticesItems/CMS1249242.html?DLPage=1&DLEntries=10&DLSort=2&DLSortDir=descending. Abt evaluated as many as 921 variables in each model and concluded that only 4 variables predict a case-mix increase or decrease of more than 0.01. The “drivers” include: 2 1. The proportion of voluntary, non-profit HHAS to freestanding for profit HHAs. Real case mix increased with nonprofit HHAs. 2. The extent that patients had an inpatient hospitalization in the 14 days prior to the beginning of the home health episode. A decrease in the level of the mean/average of this variable is associated with higher real case mix change. 3. The number of hospital days preceding the episode. Less days is associated with lower real case mix change. 4. Changes in the proportion of patients with CVAs and knee joint replacements. Increases in such inpatients leads to high real case mix change. Several factors demonstrate that CMS’s estimate of “real” and “nominal” case mix change to support its proposal for rate reductions of 1.72% in each of 2016 and 2017 is unreliable. First, CMS did not evaluate any of the 921 variables previously considered in earlier adjustments for years contemporaneous with the case mix change years in issue, 2012-2014. Instead, CMS used its outdated earlier analysis that had applied an evaluation that synchronized the case mix change years under review with the consideration of the level of “real” case mix changes during that same time period. This is a significant problem with the proposed adjustment as the historical analyses conducted by CMS demonstrates that the level of “nominal” case mix weight change is not consistent from year to year and had instead, varied greatly. Accordingly, reliance on the findings from an unrelated time period is improper to support the proposed two-year adjustment and it is essential that CMS conduct an evaluation of the level of “nominal” change for the specific time period in issue rather than applying an unreliable surrogate. There is strong evidence that the level of “real” case mix change in 2012-2014 is different than the surrogate percentage applied by CMS in its proposal, including some of the “drivers” that the CMS contractor relied upon in its evaluation of case mix weight changes in 2000-2009. These changes include: 1. The proportion of home health services episodes NOT preceded by an inpatient hospitalization or post-acute care institutional stay declined during the period in issue. In 2013, 66 percent of home health episodes were not preceded by an inpatient hospital or PAC stay. MedPAC further found that the 2013 home health population includes more individuals that were dual Medicare-Medicaid eligibles, older, more likely afflicted with dementia, and experiencing longer lengths of service. These patients are generally associated with higher home health resource needs and higher case mix weight. http://www.medpac.gov/-documents-/data-book at 122. 2. The percentage of inpatient discharges to home health services grew from 15.2% in 2009 to 16.7% in 2012. http://www.medpac.gov/-documents-/data-book at 71. This is an indication that the acuity level patients admitted from hospital to home health has risen, a factor associated with a higher case mix weight. 3 3. The number of HHAs dropped between 2013 and 2014 for the first time in years. http://www.medpac.gov/-documents-/data-book at 113. For profit HHAs were more likely to close operations. A reduction in proprietary HHAs is associated with higher average case mix. 4. The average number of episodes per patient in home health services declined to 1.9 episodes even though the proportion of community admissions, which are usually associated with a higher incidence of episodes per patient, grew. That indicates a significant drop in the average length of service with a corresponding increase in “real” case mix. http://www.medpac.gov/documents-/data-book at 121. 5. Total hip and knee replacement surgeries are steadily increasing with over 400,000 such procedures in 2013. https://www.cms.gov/Newsroom/MediaReleaseDatabase/Factsheets/2015-Fact-sheets-items/2015-07-09.html. A higher incidence of such patients in home health services has been found to increase “real” case mix. 6. The proportion of Medicare enrollees in traditional Fee for Service Medicare remained relatively flat while the level of enrollment in Medicare Advantage Plans grew. There were 36.7 million FFS enrollees in 2005 with a slight increase to 37.2 million in 2012. At the same time, MA Plan enrollment grew from 7.5 million to 13.6 million over the same period. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-andReports/MedicareMedicaidStatSupp/2013.html at Table 2.1. MA Plan enrollment is associated with younger and healthier Medicare-eligibles leading to a disproportionate enrollment population of older, more infirm individuals in traditional Medicare FFS. That disproportionate enrollment trend indicates a likely higher “real” case mix change in home health services. 7. Inpatient hospital discharges have declined from 12.9 million (2005) to 12.3 million (2010) to 11.2 million (2012). Concurrently, the average hospital length of stay has decreased from 5.7 days (2005) to 5.4 days (2010) to 5.3 days (2012). https://www.cms.gov/Research-StatisticsData-and-Systems/Statistics-Trends-and-Reports/MedicareMedicaidStatSupp/2013.html at Table 5.1. As a result, it can be expected that the average acuity level of discharges has increased and triggered a higher “real” case mix change in home health patients admitted from an inpatient setting. There is no indication that CMS made any attempt to update its analysis of “real” and “nominal” changes in case mix to support its proposed payment rate adjustments to accurately reflect he time period in issue. At the same time, it is clear that data for such an analysis to be conducted is at least, in part, available. Further, it appears that the data is likely to show that the level of “real” case mix change is greater than the out-of-date estimate presented by CMS. In addition, it appears that previous analyses relied upon by CMS indicate that the variables evaluated to distinguish between “real” and “nominal” case mix changes fall short of what would be should be considered as reliable indicators. Specifically, the CMS contractor concluded that of the 921 variables it evaluated, only 4 predicted a case mix increase or decrease of more than 0.01. That means that a reliable analysis likely lies in other variables that were not considered or a reliable analysis is not actually possible. 4 However, even if the variables considered meet an acceptable standard of reliability, CMS has the tools and data needed to provide an updated calculation of “real” and “nominal” case mix change. With the changes that have occurred throughout Medicare in terms of policy reforms and health care practices, there are significant reasons to hold off on any case mix change adjustment until a properly time-synchronized evaluation is completed. Second, during the years in issue, 2012-2014, Medicare spending actually decreased while “nominal” increases in case mix weights should lead to increased spending. Publicly reported data indicates that spending decreased in the aggregate from $19.6 billion in 2010 to $18.3 in 2013. (2014 data is not yet available). http://www.medpac.gov/-documents-/data-book , Chart 8-2 at page 114. In addition, per episode revenue changes during the period in issue do not reflect the impact that would occur with true “nominal” case mix weight changes. If HHAs increased coding weights that did not reflect patient changes, the per-episode revenue would show a comparable increase. Instead, the change in per-episode revenue is less than the 4.17% overall case mix weight change as well as the 3.41% nominal case mix weight change estimated by CMS. TABLE 1. TABLE 1 Expected RPE with cuts and coding change** RPE Difference RPE Difference RPE ALL Actual* 2011 2012 2013 2931.89 2955.00 2707.30 2886.29 2906.11 2702.42 2913.91 2933.36 -19.45 RPE FS Actual* Expected RPE with cuts and coding change** YEAR 2938.35 2956.48 -18.13 RPE HB Actual* 2685.39 Expected RPE with cuts and coding change** RPE Difference 2708.65 -23.26 RPE= Revenue Per Episode (non-LUPA, non-outlier) *based on 8,803 filed cost reports from FYE2011-2013 **Expected RPE= 2011 RPE Actual X Net Annual Rate Adjustment*** X CMS Estimated Case Mix Weight Change ***Net Annual Rate Adjustment= Market Basket Index (2.4+2.3=1.047) X ACA MBI Reduction (1.0+1.0= .98) X Coding Weight Adjustment (3.79+1.32=.9489) The available Medicare spending data also raises a serious question as to whether any case mix weight adjustment should be applied at this time. Under the authorizing statute, 42 USC 1395fff(b)(3)(B)(iv), Medicare may adjust payment rates if there is “a change in aggregate payments…that are a result of changes in the coding or classification of different units of services that do not reflect real changes in case mix.” That provision establishes a discretionary power within Medicare to institute such rate adjustments, but ties exercise of that power to a condition precedent that spending ahs improperly increased. Here, Medicare has not increased during the years in issue. 5 That alone is a strong indication that there were no changes in coding that had an impact on spending justifying the use of the discretionary authority set out in section 1395fff(b)(3)(B)(iv). Further, even if spending had increased, CMS has not established a reasonable estimate of the “real” and “nominal” case mix changes given its application of an analysis from an unrelated time period as required by the statute. Until CMS performs a reliable and time-synchronized evaluation, it would be an unreasonable application of its discretionary power. Third, CMS has recalibrated the case mix adjustment model several times along with rate rebasing, thereby eliminating the effect of any impact from or relevance of previous “nominal” changes in case mix weights. In 2012, CMS adjusted case mix weights by modifying the impact of therapy utilization and eliminating certain hypertension diagnostic codes from the HHRG scoring system. In 2014, CMS imposed ICD-9 coding adjustments that had the effect of reducing home health spending that year by $100 million. In addition, CMS reset the average case mix weight to 1.000. Finally, in 2015, CMS completely restructured the case mix adjustment model, recalibrating all 153 case mix categories and significantly modifying the variable considered in the case mix weighting. While CMS claims that the 2014 wholesale recalibration was done in a budget neutral manner which might justify the imposition of the proposed case mix creep adjustment in 2016, several question arise in that regard. The 2015 recalibration used 2013 patient data in determining the budget neutrality adjustment of 1.0366. However, the proposed case mix creep adjustment in 2015 concerns 2012, 2013, and 2014 changes in case mix weights. In doing so, does the 2015 budget neutrality adjustment affect any “nominal” change in case mix weights in 2014? It would appear that the 2015 recalibration eliminated any spending impact of the 2014 “nominal” changes in case mix. The 2012 recalibration in eliminating two hypertension codes and modifying the impact of therapy utilization in the case mix model di not include any budget neutrality adjustment. Did the 2012 recalibration and the absence of any budget neutrality adjustment affect spending on home health services in a manner that requires modification of the proposed adjustments in the 2016 NPRM? In particular, the absence of a budget neutrality adjustment on 2012 led to a reduction in spending that would otherwise not occur if the case mix weights had not be recalibrated. That spending reduction existed during the entire period at issue in the NPRM. Should CMS take that spending reduction into account when determining whether to impose a case mix creep adjustment? A third recalibration of sorts occurred in the 2014 rate setting when CMS instituted ICD-9 coding adjustments along with the resetting of the average case mix weight to 1.000. These changes had an impact on spending that should be accounted for in evaluating whether any “nominal” change in case mix weights in 2012-2014 had an impact on Medicare home health spending. Overall, CMS must provided detailed explanation as to what impact any of the case mix weight recalibrations had on the spending that the proposed case mix creep adjustments are intended to address. It is highly notable that CMS did not even mention these actions in its NPRM. 6 Fourth, CMS should also consider whether to exercise its discretionary authority at this time. In its CY 2014 HHPPS rulemaking, CMS estimated that 43% of all HHAs would face negative Medicare margins, i.e. paid less than the cost of care, by 2017 with the impact of rate rebasing starting in 2014 and the application of the annual Productivity Adjustment starting in 2015. A recent analysis by NAHC indicates that the percentage of such-impacted HHAs is now forecast at 53.71% by 2017 with the addition of the case mix weight adjustment proposed by CMS. That forecast further indicates that some states are impacted to a much higher degree. For example, 76.2% of New York HHAs are forecast with negative margins. Other states significantly affected include: Montana (83.3%); Hawaii (80%); North Dakota (76.9%); Alaska (75%); and Oregon (70.6%). Overall, the forecast average Medicare margin is bleak (All HHAS: -0.37%; Freestanding: +2.69%; and Facility-Based: -28.66%). TABLE 2. It is noteworthy that invariably, an HHA’s overall margin is far lower than its Medicare margin. That translates to a very unstable financial picture for HHAs and a high risk of care access problems for Medicare FFS, Medicare Advantage, and Medicaid patients alike. TABLE 2 2017 Forecast Percentage of agencies with PM <=0% Free-standing Agencies All Agencies % of Agencies at or Below 0 Medicare Margin 36.8% Average Medicare Margin 3.8% % of Agencies at or Below 0 Medicare Margin 28.9% Average Medicare Margin 4.5% Alaska Arizona 75.0% -52.5% 66.7% 46.9% 1.5% 40.9% Arkansas 59.8% -5.5% California 58.7% Colorado Connecticut Hospital Based Agencies % of Agencies at or Below 0 Medicare Margin 66.7% Average Medicare Margin -3.6% -30.7% 83.3% -117.9% 4.4% 100.0% -37.9% 46.6% -1.7% 82.4% -14.9% -2.3% 55.9% 2.1% 100.0% -46.0% 44.9% 5.5% 36.3% 8.4% 93.8% -40.0% 29.4% 9.2% 25.4% 11.2% 80.0% -24.6% Delaware 37.5% 7.5% 28.6% 9.6% 100.0% -19.4% District of Columbia 40.0% 2.6% 40.0% 2.6% Florida 48.5% 2.2% 46.8% 3.6% 100.0% -40.0% Georgia 42.6% 3.5% 38.3% 5.2% 69.2% -17.8% Guam 0.0% 33.3% 0.0% 33.3% Hawaii 80.0% -13.6% 66.7% -5.0% 100.0% -37.0% Idaho 58.1% -8.2% 50.0% -2.2% 100.0% -63.7% Illinois 57.8% -2.8% 55.5% -0.1% 91.3% -22.6% Indiana 60.1% -6.5% 51.4% -0.8% 97.1% -30.7% Iowa 57.7% -1.6% 44.3% 6.4% 81.6% -21.0% Kansas 61.6% -4.6% 47.8% -0.3% 90.6% -22.8% State Alabama 7 Kentucky 48.0% 3.0% 41.0% 7.5% 70.8% -16.8% Louisiana 42.2% 4.5% 39.3% 5.4% 75.0% -17.9% Maine 62.5% -1.5% 59.1% -1.5% 100.0% -1.0% Maryland 48.9% 2.8% 42.9% 5.2% 100.0% -31.3% Massachusetts 40.8% 3.1% 38.9% 4.7% 71.4% -16.3% Michigan 59.7% -4.2% 57.3% -1.4% 91.7% -20.8% Minnesota 56.9% -3.0% 40.7% 5.7% 79.1% -33.3% Mississippi 45.5% -0.9% 37.1% 2.7% 77.8% -29.5% Missouri 59.7% -3.3% 51.4% 1.7% 82.5% -23.0% Montana 83.3% -15.0% 70.0% -11.6% 92.9% -23.0% Nebraska 66.1% -0.5% 33.3% 12.6% 88.6% -29.5% Nevada 64.2% -10.2% 64.8% -9.9% 50.0% -17.9% New Hampshire 42.9% 2.4% 36.0% 3.6% 100.0% -25.0% New Jersey 30.8% 7.0% 25.0% 9.4% 100.0% -29.4% New Mexico 46.7% -11.7% 40.7% -7.6% 100.0% -51.8% New York 76.2% -10.8% 70.4% -8.5% 100.0% -25.0% North Carolina 51.8% 3.6% 44.4% 7.2% 92.3% -16.8% North Dakota 76.9% -29.7% 66.7% -7.5% 85.7% -43.8% Ohio 37.7% 6.6% 32.7% 10.4% 79.5% -17.5% Oklahoma 57.7% -0.6% 52.6% 1.3% 90.0% -26.5% Oregon 70.6% -19.2% 50.0% -5.8% 95.7% -45.7% Pennsylvania 47.3% 1.7% 39.3% 9.6% 86.4% -30.4% Puerto Rico 77.8% -7.3% 75.8% -6.7% 100.0% -16.1% Rhode Island 14.8% 10.6% 8.0% 13.7% 100.0% -14.7% South Carolina 44.1% 3.9% 29.6% 9.8% 86.7% -19.1% South Dakota 50.0% -10.4% 36.4% 11.1% 58.8% -34.3% Tennessee 49.6% 4.1% 44.5% 5.3% 92.9% -37.2% Texas 56.0% -1.0% 54.8% 0.8% 89.0% -61.0% Utah 60.9% -3.8% 59.3% -3.6% 83.3% -8.5% Vermont 50.0% -1.3% 45.5% 1.1% 100.0% -58.6% Virgin Islands 100.0% -27.9% 100.0% -27.9% Virginia 49.7% -0.2% 42.6% 3.0% 92.6% -16.6% Washington 62.5% -6.3% 44.7% 2.9% 100.0% -37.2% West Virginia 45.1% 0.9% 30.8% 5.1% 91.7% -28.7% Wisconsin 63.9% -13.8% 51.0% -2.6% 95.2% -65.8% Wyoming 52.0% -13.9% 29.4% 0.9% 100.0% -80.2% National 53.71% -0.37% 49.47% 2.69% 88.02% -28.66% 8 Adding the 1.72% case mix weight adjustments in 2016 and 2017 weakens the prospects for continued access to care. As such, a reasonable exercise of discretionary power would caution against a premature application of any case mix weight adjustment. That caution is particularly well warranted where, as here, CMS has not analyzed the level of “real” and “nominal” coding changes during the time period involved. RECOMMENDATIONS: 1. CMS should withdraw the proposed case mix weight adjustments proposed for 2016 and 2017. No adjustments should be considered until CMS conducts a thorough analysis of real and nominal changes in case mix through evaluation of changes that occurred during the actual years of concern (2012-2014) with respect to the proposed adjustment and any adjustments that might be considered in future years. Such evaluation should analyze any variable that may reasonably explain changes in average case mix weights in addition to those variables considered in earlier analyses. 2. CMS should provide public notice and an opportunity for comment on any proposed payment rate adjustments with full disclosure of any technical analysis performed by CMS or its contractors prior to implementation. 3. No case mix weight change adjustment shall be imposed unless it can be demonstrated that Medicare spending on home health services exceeded forecasted spending. 4. CMS should fully evaluate the impact of case mix weight recalibration on case mix weight change and publicly disclose such evaluation. 5. CMS should develop program integrity measures to address provider-specific up-coding as an alternative to across-the-board case mix creep adjustments. 6. In the event that CMS does not withdraw the proposed adjustments, CMS should hold off on imposing the adjustments until the completion of rate rebasing in 2017. Alternatively, CMS should phase-in the adjustments over a five (5) year period. Home Health Value-Based Purchasing Pilot Program In the NPRM, CMS proposes to institute a Value-Based Purchasing (VBP) program that would be piloted for 5 years in nine states beginning in 2016 (2018 payment year). That program would be based upon 15 outcome measures, 10 process measures, and 4 additional process/reporting measures. The amount of payments at risk for providers would begin at 5% in 2018 and phase up to 8% in 2021 and 2022. The distribution of the payment impact would be based on performance attainment and performance improvement in comparison to a 2015 baseline year. NAHC supports exploration of innovative payment models such as VBP. However, the nature of an acceptable VBP program is one that incentivizes and rewards performance and performance improvement while permitting providers to have a fair opportunity to achieve such. In other words, the 9 VBP design itself cannot jeopardize or impede the ability of a provider to reach performance levels or performance improvements that are the desired end product of VBP innovations. In addition, a VBP must operate based on appropriate measures of performance that are consistent with the objectives of the Medicare program and the population of beneficiaries it is designed to serve. For example, if patient rehabilitation is a desired result of a particular benefit within Medicare, providers that achieve such rehabilitation should be considered high performers. At the same time, if the benefit is also intended to achieve a clinical outcome of stabilizing a patient’s condition or preventing rapid deterioration of that condition, the VBP design should reflect that performance outcome in its performance measures. The VBP proposal presented by CMS does meet appropriate design standards in several respects. First, it incorporates both performance attainment and performance improvement into its measures for rewards and penalties. Second, it uses some outcome and process measures that have been tested and are uniformly applied to Medicare participating home health agencies. Third, the payment impact minimizes cash flow problems in that monies are not withheld from all providers with a return to some at a later point in time. Instead, the proposed VBP integrates a payment adjustment two years after the performance year thereby permitting the provider to more reasonably budget for a know rate impact. However, the proposed VBP comes up short in other aspects of its design. Primary in that regard is the significant amount of payment that is put at risk (5-8%) over the five year term of the pilot. This level is far in excess of any Medicare VBP demonstration, pilot, or full program to date. The two provider sectors with VBP programs have a capped 2% at risk. HHAs already have 2% of payments at risk with the quality reporting programs where a failure to meet reporting requirements results in a 12 month payment reduction of 2%. While NAHC recognizes that it is unlikely that any HHA will face a full 5-8% payment reduction under VBP, the reductions forecast by CMS in the NPRM of as much as 2.98% in the first year of VBP (capped at 5% risk) is an unsustainable risk for the majority of HHAs. Medicare cost report for HHAs with fiscal years ending in 2013 indicate that if providers faced a risk of losing 5% of payments in that year, 39.5% of HHAs would have been at risk of receiving Medicare revenues less than their cost of care without consideration of costs such as telehealth, taxes, and marketing. NAHC further devised a forecast of “Medicare margins” for 2018 with consideration of the impact trend of rate reductions since 2010, the scheduled rate reductions in 2016-2018, Medicare payment sequestration, and the risks posed by the proposed VBP. That analysis forecasts that 59.64% of HHAs would be at risk of being paid less than the cost of care in 2018 based on the CMS estimate that the bottom 10% of HHAs would be penalized 2.98% in that year under its proposal. If the actual outcome is a 5% penalty, the forecast rises to 63.68%. TABLE 3. By 2022 when the VBP pilot reaches an 8% risk, NAHC forecasts that 69.31% of HHAs will be paid less than the cost of care. TABLE 3. 10 TABLE 3 2.98% VBP Penalty % of Agencies at or Below 0 Medicare Margin 40.4% Alaska 5% VBP Penalty Average Medicare Margin 0.8% % of Agencies at or Below 0 Medicare Margin 49.1% 75.0% -55.5% Arizona 56.1% Arkansas 8% VBP Penalty Average Medicare Margin -1.2% % of Agencies at or Below 0 Medicare Margin 58.8% Average Medicare Margin -4.2% 83.3% -57.5% 83.3% -60.5% -1.5% 58.2% -3.5% 62.2% -6.5% 67.4% -8.5% 72.8% -10.5% 78.3% -13.5% California 65.3% -5.2% 69.5% -7.2% 74.7% -10.2% Colorado 50.5% 2.5% 52.3% 0.5% 55.1% -2.5% Connecticut 35.3% 6.2% 39.7% 4.2% 50.0% 1.2% Delaware District of Columbia Florida 50.0% 4.6% 56.3% 2.6% 68.8% -0.5% 53.3% -0.3% 53.3% -2.3% 60.0% -5.3% 54.5% -0.8% 57.5% -2.8% 63.3% -5.8% Georgia 50.0% 0.5% 56.4% -1.5% 66.0% -4.5% Guam 0.0% 30.3% 0.0% 28.3% 33.3% 25.3% Hawaii 80.0% -16.6% 80.0% -18.6% 80.0% -21.6% Idaho 60.5% -11.2% 60.5% -13.2% 76.7% -16.2% Illinois 65.5% -5.8% 70.2% -7.8% 75.8% -10.8% Indiana 64.0% -9.5% 66.9% -11.5% 75.8% -14.5% Iowa 63.5% -4.6% 66.4% -6.6% 75.2% -9.6% Kansas 64.7% -7.6% 69.7% -9.6% 71.7% -12.6% Kentucky 53.9% 0.1% 62.8% -1.9% 69.6% -4.9% Louisiana 48.7% 1.5% 53.3% -0.5% 60.3% -3.5% Maine 66.7% -4.4% 75.0% -6.4% 79.2% -9.4% Maryland 51.1% -0.2% 59.6% -2.2% 70.2% -5.2% Massachusetts 45.0% 0.1% 47.5% -1.9% 50.8% -4.9% Michigan 66.5% -7.2% 71.1% -9.2% 77.3% -12.2% Minnesota 62.8% -6.0% 68.6% -8.0% 73.5% -11.0% Mississippi 54.6% -3.9% 61.4% -5.9% 65.9% -8.9% Missouri 67.1% -6.3% 69.8% -8.3% 77.9% -11.3% Montana 83.3% -18.0% 87.5% -20.0% 87.5% -23.0% Nebraska 69.5% -3.5% 72.9% -5.5% 76.3% -8.5% Nevada 69.5% -13.2% 74.7% -15.2% 79.0% -18.2% New Hampshire 46.4% -0.6% 57.1% -2.6% 60.7% -5.6% State Alabama 11 New Jersey 46.2% 4.0% 48.7% 2.0% 56.4% -1.0% New Mexico 48.3% -14.7% 50.0% -16.7% 61.7% -19.7% New York 78.2% -13.8% 83.2% -15.8% 86.1% -18.8% North Carolina 53.6% 0.6% 57.1% -1.4% 63.7% -4.4% North Dakota 76.9% -32.6% 84.6% -34.6% 92.3% -37.6% Ohio 41.1% 3.7% 43.8% 1.7% 49.9% -1.4% Oklahoma 64.0% -3.6% 71.2% -5.6% 77.5% -8.6% Oregon 78.4% -22.1% 80.4% -24.1% 82.4% -27.1% Pennsylvania 52.7% -1.3% 56.6% -3.3% 60.1% -6.3% Puerto Rico 77.8% -10.3% 77.8% -12.3% 80.6% -15.3% Rhode Island 25.9% 7.7% 33.3% 5.7% 37.0% 2.7% South Carolina 49.2% 1.0% 52.5% -1.1% 62.7% -4.1% South Dakota 53.6% -13.4% 53.6% -15.4% 53.6% -18.4% Tennessee 57.9% 1.2% 63.2% -0.9% 69.9% -3.9% Texas 62.1% -3.9% 65.6% -5.9% 70.5% -8.9% Utah 64.4% -6.8% 69.3% -9.0% 75.0% -12.0% Vermont 66.7% -4.3% 75.0% -6.3% 83.3% -9.3% Virgin Islands 100.0% -30.8% 100.0% -32.8% 100.0% -35.8% Virginia 57.1% -3.1% 60.3% -5.1% 64.6% -8.1% Washington 69.6% -9.3% 71.4% -11.3% 73.2% -14.3% West Virginia 51.0% -2.1% 60.8% -4.1% 64.7% -7.1% Wisconsin 65.3% -16.8% 69.4% -18.8% 75.0% -21.8% Wyoming 56.0% -16.9% 60.0% -18.9% 68.0% -21.9% National 59.64% -3.35% 63.68% -5.35% 69.31% -8.35% If it is assumed that some HHAs can continue to operate with less than cost Medicare reimbursement, it cannot be safely assumed that those providers will have sufficient financial resources to avoid a death spiral caused by a likely permanent reduction in payment. Performance improvement does not happen without adequate financial support. Once an HHA suffers the extreme penalty at risk in the proposed VBP (5-8%), it is unlikely that it can afford the necessary changes in its operation to improve its performance and avoid future penalties. Even if some of the nearly 40% of providers at risk are able to somehow find new resources to improve performance, it is most likely that the vast majority of those at risk will not. The above analysis fits regardless of which states are included in the pilot. In fact, certain states currently proposed for inclusion have a higher proportion of HHAs at risk than the national average referenced above. (Iowa, Nebraska, and Washington State). The selection of the range of penalty/reward in a VBP should be related to a determination as to what level of risk is need to affect provider behavior to change is care delivery in ways that achieve the desired outcomes of the VBP. CMS has not provided any analysis that supports a need to use a 5-8% 12 level of risk to change HHA behavior. The above data analysis indicates that HHAs are likely to react favorably to the risks at much lower levels than 5-8% given that a small percentage reduction in payments would spell financial ruin and the incapacity to meet costs let along increased costs that performance improvement would demand. Further evidence that a small amount at risk should be sufficient to trigger behavioral change is found in the overall margins of freestanding HHAs. Data from 7020 FYE2013 cost reports show an overall margin of 3.98% for these HHAs. That is the overall margin prior to the scheduled rate cuts between 2014 and 2018 that exceed 18.2 percentage points before any VBP penalty is imposed. Given that Medicare Fee-for-Service represents approximately 40% of HHA revenues and that Medicare Advantage and Medicaid makeup nearly all the rest and pay less than cost currently, it is reasonable to assume that overall home health agency margins will be less than 0% by 2018. Those numbers translate to a conclusion that even a very small amount at risk will change behaviors and achieve performance improvements. A second weakness in the proposed VBP is its structure that requires that each year, some HHAs must be penalized, regardless of their performance or performance improvement. The only way that would not occur would be if all HHAs ended up with the same score in the VBP scoring system. The odds of that are likely astronomical. As such, each year, some HHAs will end up in the bottom decile and others in the remaining nine. That would occur even if the difference between the highest and lowest scoring is only a few points. That weakness is further exposed when the pilot is carried through to its fifth year. If it is assumed that the bottom decile of HHAs do not have the resources to rise from the bottom and they close operation, by the fifth year the bottom decile will include HHAs whose performance equals or exceeds the average baseline performance, many of which had received bonus payments in previous years. There is some limit to performance improvement. The VBP at that point penalizes some HHAs that reached that stage. A third concern as to whether the VBP design meets acceptable standards is the performance measures proposed by CMS. The initial set of measures proposed for the VBP model utilizes data collected from the OASIS, Medicare claims, HHCAHPS survey data, and data reported directly from the HHAs to CMS. In total there are 29 proposed measures: 10 process measures, 15 outcome measures, and four new measures. Included in the outcome measures are the improvement measures for functional and clinical status currently reported on home health compare. CMS also proposes to measure Types and Sources of Assistance using the OASIS item M02102, and Prior Functioning in ADLs and IADLs using OASIS item M01900. Further, CMS proposes to include 4 new measures: Influenza vaccination for agency personnel, Herpes zoster vaccination surveillance of patients, Advanced Care planning, and Adverse Events for improper medication administration. Agencies would be awarded for reporting the new measures that are sent through a CMS portal set up specifically for reporting the new measures. 13 Of greatest concern is that these measures do not reflect the patient population serves under the Medicare home health benefit as the outcome measures focus on a patient’s clinical improvement and do not address patient’s with chronic illnesses, deteriorating neurological, pulmonary, cardiac, and other conditions, and some with terminal illness. This is a shortcoming of the recently unveiled Star Rating system as well and NAHC does not support extending that weakness into a VBP model that will have any even greater impact on home health care delivery. The Medicare home health benefit is not limited to individuals who can show improvement in their clinical condition. If it were, it would disserve the millions of aging beneficiaries afflicted with incurable chronic conditions or terminal illnesses. The expected outcome for many patients admitted to home health care is to stabilize or prevent decline of a condition or functional limitation. The recent settlement in the lawsuit in Jimmo v. Sebelius further confirms that the improvement standard does not apply to all Medicare home health patients. The longstanding and proper coverage standard is whether the individual need skilled care regardless as to whether care is intended to improvement function or clinical status, prevent accelerated deterioration, or care for a patient with a chronic, incurable condition. Recent data indicates that the highest per patient Medicare spending on home health services is for patients aged 85 and above. In that patient category, many are afflicted with multiple co-morbidities involving COPD, CHF, Diabetes, and other chronic illnesses. If this is the primary patient population using home health services, performance measures based on improvement are not very helpful NAHC believes CMS should balance the improvement measures with stabilization measures. Agencies should be judged and awarded on the number of patients who remain stable and avoid decline in their functional and clinical status. NAHC is also concerned that the proposed measures create an overly complex program. The measures should be limited to a select set that are most meaningful for patients and best reflect quality home health care. In addition, fewer measures would allow agencies to better focus on areas that impact quality of care within their agency. The absence of measures that focus on a significant segment of the home health patient population along with the complexities of the volume of measures proposed warrants a restructuring of the VBP pilot prior to its implementation. To do so simply through this current notice and comment rulemaking process will not work. Instead, NAHC recommends that CMS employ a consensus process that brings technical experts and stakeholders together for a robust and focused analysis of the best, reliable measures available that reflect the entire population receiving home health services. The consensus should be built around guiding principles that include: The measures should provide information relevant to the broad population served in home health services with consideration of the clinical outcomes that fit that population The measures employed should be tested and validated The integrity of the inputs for the measures should be evaluated, limiting the risk of manipulation 14 The measures should be limited to the minimum necessary to distinguish between positive and negative performing providers and permit guidance to providers on areas for meaningful performance improvement The measures should not be static, but instead subject to modification consistent with these principles and subject to a consensus process. It must also be recognized that providers need sufficient time to work with a new performance measure to understand how it affects quality of care and how the agency can implement an effective quality improvement program around the measure. Agencies should not be subjected to financial penalties for not being able to show achievement or improvement in a new, untested, performance measure. NAHC intentionally does not provide for a list of measures that is an alternative to those in the proposed rule. To do so would be contrary to the recommendation that measures be selected through a consensus process. Instead, NAHC intends to convey through this rulemaking comment process that CMS needs to establish a process that is conducive to achieving consensus on the measures to be employed. If CMS decides to move forward with VBP at this time, it will be essential that a detailed and comprehensive plan be developed that sets out a process for development and implementation of improved performance measures, the issuance of standards for VBP expansion, closure, and/or modification of published timetables, and monitoring of changes in quality of and access to care. That plan should be devised in coordination with VBP stakeholders as VBP will succeed only if done in true partnership with beneficiaries and providers of home health services. Comment on Specific Measures CMS’ choice to use the OASIS items M2102 and M01900 as measures is perplexing since these OASIS items, in of themselves, are not quality measures. Item M2102 is used only as a risk adjustment factor and as part of a measure for the Potentially Avoidable Event – Discharge to community needing medication or wound assistance. In addition, CMS proposes to use in the numerator and denominator “Multiple data elements”; however, these data elements are not defined. OASIS item M01900 also is used only as a risk adjustment factor. It is unclear how CMS intends to use OASIS items M02102 and M01900 as measures in the VBP program. CMS appears to propose to develop new measures using these OASIS items along with other undefined data elements. In the event that CMS maintains the use of the proposed measures, NAHC supports applying the proposed four new measures based on reporting performance. However, NAHC requests that agencies be permitted to demonstrate that the proposed measure for advanced planning is met by complying with the Standard at §484.10(c) (ii) The HHA complies with the requirements of subpart I of part 489 of this chapter relating to maintaining written policies and procedures regarding advance directives. The HHA must inform and distribute written information to the patient, in advance, concerning its policies on advance directives, including a description of applicable State law. The HHA may furnish advance directives information to a patient at the time of the first home visit, as long as the information is furnished before care is provided. 15 In addition, the measure for Adverse Event for Improper Medication Administration and/or Side Effects is currently being reported on the agency’s Potentially Avoidable Events Report and should be readily available for CMS to access. NAHC requests that CMS pull the data from the agency’s report for the measure, rather than require agencies send it through a CMS portal. These requests will help to alleviate some burden associated with collecting and reporting the new measures. RECOMMENDATIONS: 1. CMS should modify its proposal to establish a home health VBP with 5-8% of payment at stake. Instead, the VBP should include no more than 1-2% at risk for any provider as that level is sufficient to trigger desired behavioral change without created too high a risk of impacting care access and depriving HHAs of sufficient resources to support performance improvement. 2. CMS should convene a VBP Measure Consensus Panel made up of home health clinical experts and other stakeholders to develop VBP measures that have broad-based support, relevance, and efficacy consistent with the principle outlined above prior to the implementation of any home health VBP program. 3. CMS should establish access to real time data and information on the impact of VBP in the pilot states with primary focus on changes in utilization, OASIS and HHCAPS data reporting, access to care, admission practices, compliance with Conditions of Participation, inpatient discharge planning practices, and quality of care outcomes. The data and information should be made available to the public as soon as practicable. 4. CMS should establish a VBP Management and Monitoring Team with representatives from home health stakeholders to act as a sounding board for any potential VBP changes and to monitor VBP’s impact on care access and quality. CONCLUSION NAHC wishes to extend its sincere thanks to CMS for its open dialogue with the home health care community at this very important time in health care delivery and financing reform. NAHC and its members are available and willing to provide constructive comment on any reforms under consideration by CMS at any time. 16
© Copyright 2026 Paperzz