Medicare Program; Hospice Wage Index for Fiscal Year 2009

SUMMARY: This final rule sets forth the hospice wage index for fiscal year 2009. In addition, this final rule finalizes the policy to phase out the Medicare hospice budget neutrality adjustment factor, and clarifies two wage index issues pertaining to the definition of rural and urban areas and multi- campus hospital facilities.

EFFECTIVE DATE: Effective Dates: These regulations are effective on October 1, 2008.

FOR FURTHER INFORMATION CONTACT: Katie Lucas (410) 786-7723 or Randy Throndset (410) 786-0131.

SUPPLEMENTARY INFORMATION:

I. Background

A. General

1. Hospice Care

Hospice care is an approach to treatment that recognizes that the impending death of an individual warrants a change in the focus from curative care to palliative care for relief of pain and for symptom management. The goal of hospice care is to help terminally ill individuals continue life with minimal disruption to normal activities while remaining primarily in the home environment. A hospice uses an interdisciplinary approach to deliver medical, nursing, social, psychological, emotional, and spiritual services through use of a broad spectrum of professional and other caregivers, with the goal of making the individual as physically and emotionally comfortable as possible. Counseling services and inpatient respite services are available to the family of the hospice patient. Hospice programs consider both the patient and the family as a unit of care.

Section 1861(dd) of the Social Security Act (the Act) provides for coverage of hospice care for terminally ill Medicare beneficiaries who elect to receive care from a participating hospice. Section 1814(i) of the Act provides payment for Medicare participating hospices.

2. Medicare Payment for Hospice Care

Our regulations at 42 CFR part 418 establish eligibility requirements, payment standards and procedures, define covered services, and delineate the conditions a hospice must meet to be approved for participation in the Medicare program. Part 418 subpart G provides for payment in one of four prospectively-determined rate categories (routine home care, continuous home care, inpatient respite care, and general inpatient care) to hospices based on each day a qualified Medicare beneficiary is under a hospice election.

B. Hospice Wage Index

Our regulations at [Section] 418.306(c) require each hospice's labor market to be established using the most current hospital wage data available, including any changes by the Office of Management and Budget (OMB) to the Metropolitan Statistical Areas (MSAs) definitions, which have been superseded by the Core Based Statistical Areas (CBSAs).

The hospice wage index is used to adjust payment rates for hospice agencies under the Medicare program to reflect local differences in area wage levels. The original hospice wage index was based on the 1981 Bureau of Labor Statistics hospital data and had not been updated since 1983. In 1994, because of disparity in wages from one geographical location to another, the Hospice Wage Index Negotiated Rulemaking Committee was formulated to negotiate a wage index methodology to be used for updating the hospice wage index. This Committee, functioning under a process established by the Negotiated Rulemaking Act of 1990, signed an agreement for the methodology to be used for updating the hospice wage index on April 13, 1995.

On August 8, 1997, we published in the Federal Register a final rule (62 FR 42860) implementing a new methodology for calculating the hospice wage index based on the recommendations of the negotiated rulemaking committee. The committee statement was included in the appendix of that final rule (62 FR 42883).

The hospice wage index is updated annually. Our most recent annual update final rule (72 FR 50214) published in the Federal Register on August 31, 2007, set forth updates to the hospice wage index for fiscal year (FY) 2008.

1. Raw Wage Index Values (Raw Pre-Floor, Pre-Reclassified Hospital Wage Index)

As described in the August 8, 1997 hospice wage index final rule (62 FR 42860), the pre-floor and pre-reclassified hospital wage index is used as the raw wage index for the hospice benefit. These raw wage index values are then subject to either a budget neutrality adjustment or application of the hospice floor to compute the hospice wage index used to determine payments to hospices.

Raw pre-floor, pre-reclassified hospital wage index values of 0.8 or greater are adjusted by the Budget Neutrality Adjustment Factor (BNAF). Raw pre-floor, pre-reclassified hospital wage index values below 0.8 are adjusted by the greater of: (1) The hospice BNAF; or (2) the hospice floor (which is a 15 percent increase) subject to a maximum wage index value of 0.8.

The BNAF has been computed and applied annually to the labor portion of the hospice payment. Currently, the labor portion of the payment rates is as follows: for routine home care, 68.71 percent; for continuous home care, 68.71 percent; for general inpatient care, 64.01 percent; and for respite care, 54.13 percent. The non-labor portion is equal to 100 percent minus the labor portion for each level of care.

2. Changes to Core-Based Statistical Area (CBSA) Designations

The annual update to the hospice wage index is published in the Federal Register and is based on the most current available hospital wage data, as well as any changes by the OMB to the definitions of MSAs, which now include CBSA designations.

3. Definition of Rural and Urban Areas

Each hospice's labor market is determined based on definitions of MSAs issued by OMB. In general, an urban area is defined as an MSA or New England County Metropolitan Area (NECMA) as defined by OMB. Under 42 CFR 412.64(b)(1)(ii)(C), a rural area is defined as any area outside of the urban area. The urban and rural area geographic classifications are defined in [Section] 412.64(b)(1)(ii)(A) through (C), and have been used for the Medicare hospice benefit since implementation.

4. Areas Without Hospital Wage Data

When adopting OMB's new labor market designations in FY 2006, we identified some geographic areas where there were no hospitals, and no hospital wage index data on which to base the calculation of the hospice wage index. Beginning in FY 2006, we adopted a

[Page Number 46465]

policy to use the FY 2005 raw pre-floor, pre-reclassified hospital wage index value for rural areas when no hospital wage data were available. Under the CBSA labor market areas, there are no hospitals in rural locations in Massachusetts and Puerto Rico. We also adopted the policy that for urban labor markets without a hospital from which hospital wage index data could be derived, all of the CBSAs within the State would be used to calculate a statewide urban average raw pre-floor, pre-reclassified hospital wage index value to use as a reasonable proxy for these areas. The only affected CBSA is 25980, Hinesville-Fort Stewart, Georgia.

In the FY 2008 final rule (72 FR 50214, 50217), in cases where there was a rural area without rural hospital wage data, we used the average raw pre- floor, pre-reclassified hospital wage index data from all contiguous CBSAs to represent a reasonable proxy for the rural area. This approach does not use rural data; however, the approach uses raw pre-floor, pre-reclassified hospital wage data, and is easy to evaluate, easy to update from year-to-year, and uses the most local data available. In the FY 2008 rule (72 FR at 50217), we noted that in determining an imputed rural raw pre-floor, pre-reclassified hospital wage index, we interpret the term "contiguous" to mean sharing a border. For example, in the case of Massachusetts, the entire rural area consists of Dukes and Nantucket counties. We determined that the borders of Dukes and Nantucket counties are contiguous with Barnstable and Bristol counties. Under the adopted methodology, the raw pre-floor, pre-reclassified hospital wage index values for the counties of Barnstable (CBSA 12700, Barnstable Town, MA) and Bristol (CBSA 39300, Providence-New Bedford-Fall River, RI-MA) were averaged, resulting in an imputed raw pre-floor, pre- reclassified rural hospital wage index for FY 2008.

We also noted that we do not believe that this policy would be appropriate for Puerto Rico, as there are sufficient economic differences between hospitals in the United States and those in Puerto Rico, including the payment of hospitals in Puerto Rico using blended Federal/Commonwealth-specific rates. Therefore, we believe that a separate and distinct policy for Puerto Rico is necessary. Any alternative methodology for imputing a raw pre-floor, pre- reclassified hospital wage index for rural Puerto Rico would need to take into account the economic differences between hospitals in the United States and those in Puerto Rico. While we have not yet identified an alternative methodology for imputing a raw pre-floor, pre-reclassified hospital wage index for rural Puerto Rico, we will continue to evaluate the feasibility of using existing hospital wage data and, possibly, wage data from other sources. For FY 2008, we used the most recent raw pre-floor, pre-reclassified hospital wage index available for Puerto Rico, which is 0.4047.

5. CBSA Nomenclature Changes

The OMB regularly publishes a bulletin that updates the titles of certain CBSAs. In the FY 2008 final rule (72 FR 50218), we noted that the FY 2008 rule and all subsequent hospice wage index rules and notices would incorporate CBSA changes from the most recent OMB bulletins. The OMB bulletins may be accessed at http://www.whitehouse.gov/omb/bulletins/index.html.

6. Hospice Payment Rates

Payment rates have been updated according to section 1814(i)(1)(C)(ii)(VII) of the Act, which states that the update to the payment rates for FYs since 2002 will be the market basket percentage for the fiscal year. According to section 1814(i)(1)(C) of the Act, hospices are to use the inpatient hospital market basket as a proxy for a hospice market basket.

Historically, the rate update has been published through a separate administrative instruction issued annually in the summer to provide adequate time to implement system change requirements. Providers determine their payments by applying the hospice wage index in this final rule to the labor portion of the published hospice rates.

Requirements for Issuance of Regulations

Section 902 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) amended section 1871(a) of the Act and requires the Secretary, in consultation with the Director of the OMB, to establish and publish timelines for the publication of Medicare final regulations based on the previous publication of a Medicare proposed or interim final regulation. Section 902 of the MMA also states that the timelines for these regulations may vary but shall not exceed 3 years after publication of the preceding proposed or interim final regulation except under exceptional circumstances.

This final rule finalizes provisions proposed in the May 1, 2008 proposed rule. In addition, this final rule has been published within the 3-year time limit imposed by section 902 of the MMA. Therefore, we believe that the final rule is in accordance with the Congress' intent to ensure timely publication of final regulations.

II. Provisions of the Proposed Rule and Analysis of and Responses to Public Comments

On May 1, 2008, we published a proposed rule in the Federal Register (73 FR 24000) that set forth the proposed hospice wage index for FY 2009. We received 540 timely items of correspondence. The following is a summary of each of the proposals followed by our responses to these public comments.

A. Clarification of New England Deemed Counties

In the May 1, 2008 proposed rule, we proposed to amend [Section] 418.306(c) to cross-reference to the definitions of urban and rural in the Inpatient Prospective Payment System (IPPS) regulations in 42 CFR Part 412 subpart D. In that proposed rule, we addressed the IPPS change in the designation of "New England deemed counties," which are listed in [Section] 412.64(b)(1)(ii)(B). These counties were deemed to be part of urban areas under section 601(g) of the Social Security Amendments of 1983. However, under the OMB geographic definitions, these counties were considered rural. In the FY 2008 IPPS final rule, CMS adopted a change that resulted in these counties no longer being "deemed" urban. The counties include Litchfield County, Connecticut; York County, Maine; Sagadahoc County, Maine; Merrimack County, New Hampshire; and Newport County, Rhode Island. Of these five "New England deemed counties," three (York County, Sagadahoc County, and Newport County) are included in metropolitan statistical areas defined by OMB and are therefore urban under the current IPPS labor market area definitions in [Section] 412.64(b)(1)(ii)(A). The remaining two counties, Litchfield County and Merrimack County, are geographically located in areas that are rural under the current IPPS labor market area definitions.

In the August 22, 2007 FY 2008 IPPS final rule with comment period (72 FR 47130), [Section] 412.64(b)(1)(ii)(B) was revised such that the two "New England deemed counties" that are still considered rural under the OMB definitions (Litchfield County, CT and Merrimack County, NH) are no longer considered urban effective for discharges occurring on or after October 1, 2007. Therefore, these two counties are considered rural in accordance with [Section] 412.64(b)(1)(ii)(C). However, for

[Page Number 46466]

purposes of payment under the IPPS, acute care hospitals located within those areas are treated as being reclassified to their deemed urban area effective for discharges occurring on or after October 1, 2007 (see 72 FR 47337 through 47338). We also noted that this policy change was limited to the "New England deemed counties" IPPS hospitals only, and that any change to non-IPPS provider wage indexes would be addressed in the respective payment system rules. The hospice program does not provide for such geographic reclassification as the IPPS does.

The recommendations to adjust payments to reflect local differences in wages are codified in [Section] 418.306(c) of our regulations; however there is no explicit reference to [Section] 412.64 in [Section] 418.306(c). Although [Section] 412.64 is not explicitly referred to, the hospice program has used the definition of urban in [Section] 412.64(b)(1)(ii)(A) and (B), and the definition of rural as any area outside of an urban area in [Section] 412.64(b)(1)(ii)(C). We proposed to explicitly refer to those provisions in [Section] 412.64 to make it absolutely clear how we define urban and rural for purposes of the hospice wage index. We received no comments on this proposal and will implement it as proposed.

Litchfield county, CT and Merrimack county, NH are considered rural areas for hospital IPPS purposes in accordance with [Section] 412.64. Effective October 1, 2008, Litchfield county, CT will no longer be considered part of urban CBSA 25540 (Hartford-West Hartford-East Hartford, CT), and Merrimack county, NH will no longer be considered part of urban CBSA 31700 (Manchester- Nashua, NH). Rather, these counties will be considered to be rural areas within their respective States under the hospice payment system. When the raw pre-floor, pre-reclassified hospital wage index was adopted for use in deriving the hospice wage, it was decided not to take into account IPPS geographic reclassifications. This proposed policy to follow OMB designations of rural or urban, rather than considering some counties to be "deemed" urban, is consistent with our policy of not taking into account IPPS geographic reclassifications in determining payments under the hospice wage index.

We received no comments on this proposal, and will implement it as proposed without change.

B. Wage Data for Multi-Campus Hospitals

Historically, under the Medicare hospice benefit, we have established hospice wage index values calculated from the raw pre-floor, pre-reclassified hospital wage data (also called the IPPS wage index) without taking into account geographic reclassification under sections 1886(d)(8) and (d)(10) of the Act. The wage adjustment established under the Medicare hospice benefit is based on the location where services are furnished without any reclassification.

For FY 2009, the data collected from cost reports submitted by hospitals for cost reporting periods beginning during FY 2004 were used to compute the 2008 raw pre-floor, pre-reclassified hospital wage index data without taking into account geographic reclassification under sections 1886(d)(8) and (d)(10) of the Act. This 2008 raw pre-floor, pre-reclassified hospital wage index was used to derive the applicable wage index values for the hospice wage index because these data (FY 2004) are the most recent complete data (for information on the data used to compute the FY 2008 IPPS wage index, refer to the FY 2008 IPPS final rule with comment period (72 FR 47308 through 47309, 47315)).

Beginning in FY 2008, the IPPS apportioned the wage data for multi-campus hospitals located in different labor market areas (CBSAs) to each CBSA where the campuses are located (see the FY 2008 IPPS final rule with comment period (72 FR 47317 through 47320)). We are continuing to use the raw pre-floor, pre- reclassified hospital wage data as a basis to determine the hospice wage index values for FY 2009 because hospitals and hospices both compete in the same labor markets, and therefore, experience similar wage-related costs. We note that the use of raw pre-floor, pre-reclassified hospital (IPPS) wage data, used to derive the FY 2009 hospice wage index values, reflects the application of our policy to use that data to establish the hospice wage index. The FY 2009 hospice wage index values presented in this final rule were computed consistent with our raw pre-floor, pre-reclassified hospital (IPPS) wage index policy (that is, our historical policy of not taking into account IPPS geographic reclassifications in determining payments for hospice). For the FY 2009 Medicare hospice benefit, the wage index was computed from IPPS wage data (submitted by hospitals for cost reporting periods beginning in FY 2004 (just like the FY 2008 IPPS wage index)), which allocated salaries and hours to the campuses of two multi-campus hospitals with campuses that are located in different labor areas, one in Massachusetts and another in Illinois. Thus, the FY 2009 hospice wage index values for the following CBSAs are affected by this policy: Boston-Quincy, MA (CBSA 14484), Providence-New Bedford-Falls River, RI-MA (CBSA 39300), Chicago-Naperville-Joliet, IL (CBSA 16974), and Lake County-Kenosha County, IL-WI (CBSA 29404).

We received no comments on this proposal, and will implement it as proposed without change.

C. FY 2009 Hospice Wage Index With Phase Out of the Budget Neutrality Adjustment Factor (BNAF)

1. Background

The hospice final rule published in the Federal Register on December 16, 1983 (48 FR 56008) provided for adjustment to hospice payment rates to reflect differences in area wage levels. We apply the appropriate hospice wage index value to the labor portion of the hospice payment rates based on the geographic area where hospice care was furnished. As noted earlier, each hospice's labor market area is based on definitions of Metropolitan Statistical Areas (MSAs) issued by the OMB. For FY 2009, we proposed to use a raw pre-floor, pre-reclassified hospital wage index based solely on the CBSA designations.

As noted above, our hospice payment rules utilize the wage adjustment factors used by the Secretary for purposes of section 1886(d)(3)(E) of the Act for hospital wage adjustments. Again, we proposed to use the raw pre-floor and pre-reclassified hospital wage index data to adjust the labor portion of the hospice payment rates based on the geographic area where the beneficiary receives hospice care. We believe the use of the raw pre-floor, pre- reclassified hospital wage index data results in the appropriate adjustment to the labor portion of the costs. For the FY 2009 update to hospice payment rates, we proposed using the most recent raw pre-floor, pre-reclassified hospital wage index available at the time of publication.

Comment: A few commenters were unhappy with CMS' use of the raw pre-floor, pre-reclassified hospital wage index as the input for the hospice wage index, and suggested it is flawed. Some commenters noted that the hospital-based wage index has undergone multiple changes over the past 10 years and that providers were not invited to provide comment for CMS to consider when formalizing these changes. One commenter added that the existence of exceptions to the hospital wage index system in the form of reclassifications demonstrates the unfairness and inadequacy of the hospital-based wage index system.

[Page Number 46467]

Several commenters mentioned that a 2007 MedPAC report on the hospital wage index suggested that CMS repeal the existing hospital wage index, and develop a new one. The commenter stated that MedPAC recommended that CMS evaluate the use of the revised wage index in other Medicare payment systems, which includes hospice. A commenter asked CMS to devise a hospice-specific reimbursement system, rather than using the hospital-based wage index. Several of these commenters offered to work with CMS in reforming the wage index, and recommended use of the collaborative negotiated rulemaking process. They suggested that CMS use the established wage index methodology, including the BNAF, until a viable alternative is found.

In addition, a commenter wrote that hospices compete in the same labor market as hospitals for staff but hospitals do not use the same wage index, and that the wage index does not reflect the reality of wages in a healthcare community.

Response: The raw pre-floor, pre-reclassified hospital wage index was adopted in 1998 as the wage index from which the hospice wage index is derived. The Negotiated Rulemaking Committee considered several wage index options: (1) Continuing with Bureau of Labor Statistics data; (2) using updated hospital wage data; (3) using hospice-specific data; and (4) using data from the physician payment system. The Committee determined that the raw pre-floor, pre-reclassified hospital wage index was the best option for hospice. The raw pre-floor, pre-reclassified hospital wage index is updated annually, and reflects the wages of highly skilled hospital workers.

We agree that the hospital-based wage index has undergone some changes in the past 10 years. Those changes were put forward through rulemaking, which provided the public an opportunity to provide comments. Therefore, we disagree that hospice providers have not had an opportunity to comment on hospital wage index changes.

The reclassification provision provided at section 1886(d)(10) of the Act is specific to hospitals. We believe the use of the most recent available raw pre-floor and pre-reclassified hospital wage index results in the most appropriate adjustment to the labor portion of hospice costs as required in 42 CFR 418.306(c). Additionally, use of the unadjusted hospital wage data avoids further reductions in certain rural statewide wage index values that result from reclassification. We also note that the wage index adjustment is based on the geographic area where the beneficiary is located, and not where the hospice is located.

We continue to believe that the unadjusted hospital wage index, which is updated yearly and is used by many other CMS payments systems including home health, appropriately accounts for geographic variances in labor costs for hospices. Home health agencies and hospices are Medicare's only home-based benefits, and home health agencies and hospices share labor pools. In the future, when looking into reforming the hospice payment system, we will consider wage index alternatives, to include those recommended by MedPAC.

We are implementing as final the proposal to continue to use the raw pre- floor, pre-reclassified hospital wage index.

2. Areas Without Hospital Wage Data

In adopting the CBSA designations, we identified some geographic areas where there are no hospitals, and thus no hospital wage data on which to base the calculation of the hospice wage index. These areas were described in section I.B.4 of the proposed rule (73 FR 24004). Beginning in FY 2006, we adopted a policy that, for urban labor markets without an urban hospital from which a raw pre-floor, pre-reclassified hospital wage index can be derived, all of the urban CBSA raw pre-floor, pre-reclassified hospital wage index values within the State would be used to calculate a statewide urban average raw pre-floor, pre-reclassified hospital wage index to use as a reasonable proxy for these areas. Currently, the only CBSA that would be affected by this policy is CBSA 25980, Hinesville-Fort Stewart, Georgia. We proposed to continue this policy for FY 2009.

Currently, the only rural areas where there are no hospitals from which to calculate a raw pre-floor, pre-reclassified hospital wage index are in Massachusetts and Puerto Rico. In August 2007 (72 FR 50217) we adopted the following methodology for imputing rural raw pre-floor, pre-reclassified hospital wage index values for areas where no hospital wage data are available as an acceptable proxy. We imputed an average raw pre-floor, pre-reclassified hospital wage index value by averaging the raw pre-floor, pre-reclassified hospital wage index values from contiguous CBSAs as a reasonable proxy for rural areas with no hospital wage data from which to calculate a raw pre- floor, pre-reclassified hospital wage index. In determining an imputed rural raw pre-floor, pre-reclassified hospital wage index, we define "contiguous" as sharing a border. In the proposed rule, we proposed to apply this methodology for imputing a rural raw pre-floor, pre-reclassified hospital wage index for those rural areas without rural hospital wage data in FY 2009. For Massachusetts, rural Massachusetts currently consists of Dukes and Nantucket Counties. We determined that the borders of Dukes and Nantucket counties are "contiguous" with Barnstable and Bristol counties. We did not receive any comments on this proposal, and are implementing it as proposed.

As we noted in our proposed rule, we do not believe that this methodology for imputing a rural raw pre-floor, pre-reclassified hospital wage index value is appropriate for Puerto Rico. We noted that there are sufficient economic differences between the hospitals in the United States and those in Puerto Rico, including the fact that hospitals in Puerto Rico are paid on blended Federal/Commonwealth-specific rates, to make a separate distinct policy for Puerto Rico necessary.

We did not receive any comments on this proposal, and are implementing it as proposed without change. Therefore, in this final rule, for FY 2009, we are continuing to use the most recent raw pre-floor, pre-reclassified hospital wage index value available for Puerto Rico, which is 0.4047. This raw pre- floor, pre-reclassified hospital wage index value is then adjusted upward by the hospice floor in the computing of the final FY 2009 hospice wage index.

3. Phase Out of the Budget Neutrality Adjustment Factor (BNAF)

As previously stated, the current hospice wage index methodology was developed through a negotiated rulemaking process and implemented in 1997. The rulemaking committee sought to address the inaccuracies in the original Bureau of Labor Statistics (BLS)-based hospice wage index, account better for disparities from one geographic location to another, and develop a wage index that would be as accurate, reliable and equitable as possible. The resulting hospice wage index reflects a special adjustment (a BNAF) to ensure payments in the aggregate are budget neutral to payments using the original 1983 hospice wage index. The adjustment, which is still in place today, results in providers currently receiving about 4 percent more in payments than they would have received if the adjustment factor were not applied. We believe the rationale for maintaining this adjustment is outdated, as explained in detail below, particularly given the

[Page Number 46468]

amount of time that has elapsed since it was first put into place and the continuing growth that is occurring in the hospice benefit. In the proposed rule, we proposed to phase out this adjustment over 3 years, reducing it by 25 percent in FY 2009, by an additional 50 percent for a total of 75 percent in FY 2010, and eliminating it completely in FY 2011. Additionally, from a parity perspective, because hospices and home health agencies have a similar labor mix, we believe that adjusting for geographic variances in both of these Medicare home-based benefits with the raw pre-floor and pre-reclassified hospital wage index is appropriate.

The original hospice wage index that was used when the benefit was first implemented was based on the 1981 BLS hospital data and had not been updated from 1983 until the current raw pre-floor and pre-reclassified hospital wage index was adopted. During earlier attempts to update the original hospice wage index, the hospice industry raised concerns over the adverse financial impact of a new wage index on individual hospices and a possible overall reduction in Medicare payments. Thus, the result was that in the absence of agreement on a new wage index, we continued to use the original wage index that was clearly obsolete for geographically adjusting Medicare hospice payments (see "Medicare Program; Notice Containing the Statement Drafted by the Committee Established to Negotiate the Wage Index to Be Used to Adjust Hospice Payment Rates Under Medicare", November 29, 1995, 60 FR 61264).

Changing to a new, more accurate wage index would result in some areas gaining as their wage index value would increase, but other areas would see declines in payments as their wage index value dropped. In 1994, we noted that a majority of hospices would have their wage index reduced with the new wage index that is based on using the raw pre-floor, pre-reclassified hospital wage index. These reductions would have occurred for two key reasons: (1) Hospices were located in areas where the original hospice wage index was artificially high due to flaws in the 1981 BLS data; and (2) hospices were located in areas where wages had gone down relative to other geographic areas (see "Hospice Services Under Medicare Program: Intent to Form Negotiated Rulemaking Committee", October 14, 1994, 59 FR 52130).

Because of the negative impact to certain areas that was expected with the change to a new wage index, a committee (the Committee) was formulated in 1994, under the process established by the Negotiated Rulemaking Act of 1990 (Pub. L. 101-648). The Committee was established to negotiate the hospice wage index methodology rather than to go through the usual rulemaking process. On September 4, 1996, we published a proposed rule (61 FR 46579) in which we proposed a methodology to update the hospice wage index used to adjust Medicare hospice payment rates. This proposed methodology contained the negotiated rule making committee's recommendations.

In formulating the provisions of that proposed rule, the Committee considered criteria in evaluating the available data sources. These criteria included the need for fundamental equity of the wage index, data that reflected actual work performed by hospice personnel, compatibility with wage indexes used by CMS for other Medicare providers, and availability of the data for timely implementation.

The Committee agreed that the hospice wage index be derived from the 1993 hospital cost report data and that these data, prior to reclassification, would form the basis for the FY 1998 hospice wage index. That is the raw pre- floor, pre-reclassified hospital wage index would not be adjusted to take into account the geographic reclassification of hospitals in accordance with sections 1886(d)(8)(B) and 1886(d)(10) of the Act. The methodology is codified in [Section] 418.306(c). The hospice wage index for subsequent years would be based on raw pre-floor, pre-reclassified hospital wage index data.

The Committee was also concerned that while some hospices would see increases in their payments, use of the raw pre-floor, pre-reclassified hospital wage index as the wage index for hospices would result in a net reduction in aggregate Medicare payments for hospices. As noted above, a majority of hospices would have had their wage index lowered by using the new wage index because the prior hospice wage indices were based on outdated data which were artificially high due to flaws in the 1981 BLS data, and because some hospices were located in areas where wages had gone down relative to other geographic areas. The reduction in overall Medicare payments if a new wage index were adopted was noted in the November 29, 1995 final rule (60 FR 61264). Therefore, the Committee also decided that for each year in updating the hospice wage index, aggregate Medicare payments to hospices would remain budget neutral to payments as if the 1983 wage index had been used.

As decided upon by the Committee, budget neutrality means that, in a given year, estimated aggregate payments for Medicare hospice services using the updated hospice values will equal estimated payments that would have been made for these services if the 1983 hospice wage index values had remained in effect, after adjusting the payment rates for inflation. Therefore, although payments to individual hospice programs may change each year, the total payments each year to hospices would not be affected by using the updated hospice wage index because total payments would be budget neutral as if the 1983 wage index had been used. To implement this policy, a BNAF would be computed and applied annually.

The BNAF is calculated by computing estimated payments using the most recent completed year of hospice claims data. The units (days or hours) from those claims are multiplied by the updated hospice payment rates to calculate estimated payments. The updated hospice wage index values are then applied to the labor portion of the payments. For this final rule, that means estimating payments for FY 2009 using FY 2007 hospice claims data as of March 2008, and applying the estimated updated FY 2009 hospice payment rates (updating the FY 2008 rates by the FY 2009 market basket update). The final FY 2009 hospice wage index values are then applied to the labor portion only. The procedure is repeated using the same claims data and payment rates, but using the 1983 BLS- based wage index instead of the updated raw pre-floor, pre-reclassified hospital wage index. The total payments are then compared, and the adjustment required to make total payments equal is computed; that adjustment factor is the BNAF.

All raw pre-floor, pre-reclassified hospital wage index values of 0.8 or greater would be adjusted by the BNAF, which would be calculated and applied annually. Also, all raw pre-floor, pre-reclassified hospital wage index values below 0.8 would receive the greater of the following: (1) A 15-percent increase subject to a maximum hospice wage index value of 0.8; or (2) an adjustment by the BNAF.

While the Committee sought to adopt a wage index methodology that would be as accurate, reliable, and equitable as possible, the Committee also decided to incorporate a BNAF into the calculation of the hospice wage index that would otherwise apply in order to mitigate adverse financial impacts some hospices would experience through a decrease in their wage index value by transitioning to a raw pre-floor, pre-reclassified hospital wage index.

[Page Number 46469]

In the August 8, 1997, final rule (62 FR 42860), we indicated that the annual updates of the hospice wage index values would be made in accordance with the methodology agreed to by the Committee. We also noted that in the event that we decide to change this methodology by which the hospice wage index is computed, we would propose to do so in the Federal Register . In the May 2008 proposed rule, we proposed to change this methodology.

In FY 1998, the BNAF was 1.020768 and in FY 2008, the BNAF was 1.066671. Any raw pre-floor, pre-reclassified hospital wage index value greater than 0.8 was increased by over 2 percent in FY 1998 and increased by almost 7 percent in FY 2008. In FY 2008, this adjustment resulted in hospice providers receiving about 4 percent more in payments than they would have received if the BNAF had not been applied.

The Committee also recommended that the transition to the new hospice wage index occur over 3-years, from FY 1998 to FY 2001. The intent of both the 3- year transition and the budget neutrality adjustment was to mitigate the negative financial impact to many hospices resulting from the wage index change. Additionally, the committee sought to ensure that access to hospice care was not jeopardized as a result of the wage index change.

We believe that the rationale for maintaining the BNAF is outdated for several reasons.

First, the original purpose of the BNAF was to prevent reductions in payments to the majority of hospices whose wage index was based on the original hospice wage index which was artificially high due to flaws in the 1981 BLS data. Additionally, the BNAF was adopted to ensure that aggregate payments made to the hospice industry would not be decreased or increased as a result of the wage index change. While incorporating a BNAF into hospice wage indices could be rationalized in 1997 as a way to smooth the transition from an old wage index to a new one, since hospices have had plenty of time to adjust to the then new wage index, it is difficult to justify maintaining in perpetuity a BNAF which was in part compensating for artificially high data to begin with.

Second, the new wage index adopted in 1997 resulted in increases in wage index values for hospices in certain areas. The BNAF applies to hospices in all areas. Thus, hospices in areas that would have had increases without the BNAF received an artificial boost in the wage index for the past 11 years. We believe that continuation of this excess payment can no longer be justified.

Third, an adjustment factor that is based on 24-year-old wage index values is not in keeping with our goal of using a hospice wage index that is as accurate, reliable, and equitable as possible in accounting for geographic variation in wages. We believe that those goals can be better achieved by using the raw pre-floor, pre-reclassified hospital wage index, without the outdated BNAF, consistent with other providers. For instance, Medicare payments to home health agencies, that utilize a similar labor mix, are adjusted by the raw pre-floor, pre-reclassified hospital wage index without any budget neutrality adjustment. We believe that using the raw pre-floor, pre-reclassified hospital wage index provides a good measure of area wage differences for both these home-based reimbursement systems.

Fourth, in the 13 years since concerns about the impact of switching from an old to a new wage index were voiced, the hospice industry and hospice payments have grown substantially. Hospice expenditures in 2006 were $9.2 billion, compared to about $2.2 billion in 1998. Aggregate hospice expenditures are increasing at a rate of about $1 billion per year. MedPAC reports that expenditures are expected to grow at a rate of 9 percent per year through 2015, outpacing the growth rate of projected expenditures for hospitals, skilled nursing facilities, and physician and home health services. We believe that this growth in Medicare spending for hospice indicates that the original rationale of the BNAF, to cushion the impact of using the new wage index, is no longer justified. These spending growth figures also indicate that any negative financial impact to the hospice industry as a result of eliminating the BNAF is no longer present, and thus the need for a transitional adjustment has passed.

Fifth, 13 years ago the industry also voiced concerns about the negative financial impact on individual hospices that could occur by adopting a new wage index. In August 1994 there were 1,602 hospices; currently there are 3,111 hospices. Clearly any negative financial impact from adopting a new wage index in 1997 is no longer present, or we would not have seen this growth in the industry. The number of Medicare-certified hospices has continued to increase, with a 26 percent increase in the number of hospice providers from 2001 to 2005. This ongoing growth in the industry also suggests that phasing out the BNAF would not have a negative impact on access to care. Therefore, for these reasons, we believe that continuing to apply a BNAF for the purpose of mitigating any adverse financial impact on hospices or negative impact on access to care is no longer necessary.

Finally, we proposed to phase out the BNAF over a 3-year period, reducing the BNAF by 25 percent in FY 2009, by 75 percent in FY 2010, and eliminating it in FY 2011. We believe that the proposed 3-year phase-out period will reduce any adverse financial impact that the industry might experience if we eliminated the BNAF in a single year. We also proposed to maintain the hospice floor, which offers protection to hospices with raw pre-floor, pre- reclassified hospital wage index values less than 0.8, noting that the steps in the calculation which involve the BNAF will become unnecessary. We are implementing the BNAF phase-out as proposed, and maintaining the hospice floor as proposed.

We received several comments on the phase-out of the BNAF. Specific comments and our responses to these comments are as follows:

Comment: Several commenters disputed CMS' description of the purpose of the BNAF in the proposed rule. The commenters stated CMS asserted that the purpose was to smooth the transition from an outdated BLS-based wage index to the hospital-based wage index in 1998, the language in several payment rules suggested that the BNAF was not a time-limited adjustment and was to be applied annually, during and after the transition to the hospital-based wage index. One comment supported keeping the BNAF, stating that a payment reduction for FY 2009 to FY 2011 is no less disruptive than any payment reduction which occurred through the wage index transition in 1997. Another commenter stated that the hospice negotiated wage index rule that was finalized by CMS in 1997 recognized the need to include a budget neutrality adjustment to offset the flaws in the hospital wage index, and therefore protect the viability of hospices. The commenter also stated that reason remains as valid today as eleven years ago. Another commenter said CMS' rationale for phasing out the BNAF suggested that eliminating the BNAF would restore fairness to the hospice wage index, when in reality no wage index methodology is perfect. Other commenters stated that CMS has previously recognized that BNAF protects hospices from inadequacies in the hospital wage index, and inadequacies in the hospice payment rates. Another commenter stated that the BNAF was put into place because of the dramatic changes triggered by

[Page Number 46470]

implementation of the new wage index, so that access to care was protected.

In addition, a commenter asserted that the fundamental reason for the BNAF was that no component of the current reimbursement system accurately replicates hospice costs. A commenter also indicated that CMS stated that hospice payments and providers had increased over the past 10 years, and that the hospice wage index methodology is dated. The commenter further stated that by those standards, the wage index model used by every Medicare provider type would need revision. Furthermore, a commenter asked why, other than time passing, is the BNAF outdated. Commenters indicated that the rationale for applying the BNAF originally is still valid.

Response: We continue to believe that the hospice wage index negotiating committee intended the BNAF to mitigate the negative financial impact of the 1998 hospice wage index change. We continue to believe that because of the growth in the industry and the amount of time that has passed since the transition, the rationale for maintaining the BNAF is no longer justified. In addition, from a parity perspective, we believe that an raw pre-floor, pre- reclassified hospital wage index is appropriate for use in adjusting rates for geographic variances in both of our home-based benefits, hospice and home health. Nothing in our data analysis has shown us that hospice labor costs differ substantially from home health labor costs. Therefore, we believe we can no longer justify the 6 percent increase in the hospice wage index, which results from the BNAF. We agree with the commenter that BNAF was put into place so that beneficiary access to hospice care would be protected. We believe the Committee was primarily concerned about those areas of the country that would see their payments reduced as a result of the wage index change. The Committee was concerned that the payment reductions might affect the viability of hospices in these areas, thus ultimately risking access to care. The Committee intended that aggregate payments to hospices not be reduced as a result of the wage index change. We do not believe that the Committee foresaw the amount of growth in the number of new hospices that would occur over the following decade. While we agree with the commenter that our regulations describe that the BNAF be applied during and after the transition to the new wage index, we continue to believe that those decisions were made as part of the negotiations to address transitional issues, and do not negate our ability to make future policy changes. We believe that our regulations, the negotiating committee statement, and the negotiating committee workgroup notes support these beliefs. We also believe that given the current industry climate, it is appropriate that a policy change now occur.

The decision to transition from the BLS-based wage index to the hospital- based wage index was a long process. In the October 14, 1994, proposed rule (59 FR 52130), we noted that both CMS (formally HCFA) and industry projections indicated that most hospices would have their wage indices lowered if a new wage index were based on unadjusted hospital data. The preamble of the final rule stated that, "During the discussions preliminary to developing a new wage index, the industry voiced concerns over the adverse financial impact of a new wage index on individual hospices and a possible reduction in overall Medicare hospice care payments" (59 FR 52130). There were also concerns that access to hospice care could be affected. We noted that as a result of the impact of the lower payments to hospices in the aggregate, the new wage index would have to be at least budget neutral (59 FR 52131). The Committee Statement of April 13, 1995, which was published in a notice on November 29, 1995 (60 FR 61265), said that we would apply a factor to achieve budget neutrality, and noted that budget neutrality meant that aggregate Medicare hospice payments using the new hospital-based wage index would have to equal estimated payments that would have been made under the original hospice wage index.

We do not believe that the Committee foresaw the tremendous growth in the industry. As a result of this growth, the surge of new entrants into the industry over the past 10 years has benefited from this adjustment. We continue to believe that the committee adopted the BNAF to help existing hospices transition to the 1998 wage index change, and did not expect that the BNAF would result in these payment increases to new providers in perpetuity. Impact analysis performed by participants in the negotiating process showed pockets of the country where the migration to the new hospital wage index would result in wage index values decreasing nearly 30 percent. The committee was clearly concerned about hospice viability in those areas of the country, with a corresponding concern about access to care. We continue to believe that the unique BNAF methodology, coupled with the 3-year transition period, served to address those transitional concerns. It also continues to be our belief that because of the growth in the number of hospices, and the growth in the beneficiaries served that has occurred during the last decade, the committee's goal to ensure that access to hospice care not be reduced as a result of the wage index change has been achieved. Therefore, we believe that this unique methodology for achieving budget neutrality has served its purpose and is no longer necessary to be continued.

We disagree with the commenters who wrote that the BNAF was intended to offset flaws in the hospital wage index or address inadequacy of the hospice payment rates. None of our hospice regulations or notices from 1994 to 1998 which deal with the transition to a new wage index indicated that the BNAF was put into place because of flaws in the hospital-based wage index, rate inadequacies, or because of any inaccurate replication of hospice costs under the current reimbursement system. We continue to believe, as the Committee did, that the raw pre-floor, pre-reclassified hospital wage index is currently the best choice for use in deriving the hospice wage index.

We agree with the commenter that the language in the August 8, 1997 final rule indicated that the BNAF would be applied during and after the transition period (62 FR 42862), however this language did not imply that the BNAF could not be changed or eliminated. That same final rule also included a provision for us to change the wage index methodology, through notice and comment rulemaking (62 FR 42863).

In our rationale for the BNAF phase-out, we noted the increase in payments and in the number of providers to show that the hospice industry was growing. Growth such as this, rather than industry contraction, typically occurs in a favorable business climate. The presence of a favorable business climate suggests that concerns about the financial impact of changing to a new wage index had passed. Finally, we did not state that all hospice wage index methodology was outdated, but only that the BNAF was outdated, and we continue to believe that is the case.

Given that the impact of the BNAF phase-out is relatively small (1.1 percent payment reduction for FY 2009), and is being offset by a 3.6 percent market basket update, we do not feel that the phase-out will be disruptive to the hospice industry. However, we will monitor the impact as the phase-out occurs.

Comment: A commenter wrote that CMS justified phasing out the BNAF in part because the combination of increases in the wage index in certain

[Page Number 46471]

areas with the BNAF led to an artificial boost in the wage index for the past 11 years, which CMS concluded was an excess payment. The commenter also stated that CMS said that if there had been no wage index change in 1997, the total payments to hospices would be greater than the payments that will be made if the proposal is implemented. The commenter concludes that there is no excess spending triggered by the BNAF, but instead there is an unauthorized reduction under the CMS proposal.

Another commenter felt that CMS is singling out the BNAF because some hospices benefited more from it than others. The commenter also suggested that CMS change the methodology for the limited number of hospices that benefited unduly from the "artificial boost" given by the BNAF.

In addition, a commenter stated that CMS had indicated one reason for the BNAF phase-out was because the growth in hospice expenditures indicates that any negative financial impact from the transition to the hospital-based wage index in 1998 was no longer present. The commenter indicated that CMS assumed this growth in spending was excess spending, and that CMS had put forward no evidence that there was excess spending in hospice versus appropriate increases in spending.

Response: We continue to believe that applying the BNAF to the raw hospital-based wage index does not, as accurately as possible, account for geographic variances in hospice labor costs. When the hospice industry changed from the BLS-based wage index to the raw pre-floor, pre-reclassified hospital wage index, it began using more accurate, more current data which are updated annually. When that transition occurred, there were hospices whose wage index value increased, but many hospices saw their wage index value decrease. This is because the BLS-based wage index values, which were applied to hospice payments, were artificially high in some areas of the country. The Committee itself acknowledged that the BLS data were "inaccurate and outdated" in its Committee Statement (62 FR 42883). The hospital-based wage index was considered more accurate, even though its wage index values were lower for many hospices. Therefore before the transition to the hospital-based wage index, many hospices were receiving payments that were inflated due to the artificially high BLS-based wage index.

In addition, the BNAF was put into place to mitigate the adverse financial impact to hospice providers of changing wage indices, since the change would lead to a reduction in payments, which could threaten access to care. However, as we previously described in the comment above, the BNAF has been applied not only to those hospices that were in existence at the time of the wage index change, but also to those new hospices that were established after 1998. We continue to believe that these new entrants have received an artificial boost to their payments as a result of the BNAF, which was not the intent of the negotiating committee.

The commenter is correct that if the hospice industry had not adopted the hospital-based wage index, but had remained with the BLS-based data, each year's total Medicare hospice payments would be higher than they will be when the BNAF is phased out. However, as noted above, because of the inaccuracy and outdatedness of the BLS-based wage data, those payments would also be inaccurate, and CMS must do its best to ensure the accuracy of Medicare payments.

The commenter correctly noted that we feel that the growth in hospice expenditures indicates that the need to mitigate any adverse financial impact from the change to a hospital-based wage index has passed. However we did not assume that this growth was due to excess spending associated with the BNAF. We recognize that many factors contribute to expected and appropriate growth in spending, including increased numbers of Medicare beneficiaries eligible for hospice care; increased awareness of the benefit by beneficiaries, their families, and physicians; some longer lengths of stay; etc.

We believe that the growth in Medicare hospice expenditures indicates the overall good financial health of the hospice industry and that this further demonstrates that the BNAF has outlived its usefulness and is no longer appropriate. As stated previously, we believe that given the current industry climate, it is appropriate that a policy change to phase out the BNAF be implemented.

Comment: A commenter wrote that CMS had justified the BNAF phase-out by noting that there had been an 86 percent increase in growth in the number of hospices. The commenter maintained that growth in the number of hospice providers does not demonstrate that hospices can absorb the payment reduction triggered by the BNAF phase-out. The commenter also stated that CMS does not know the financial status of those hospices or the level of demand for their services.

Several commenters stated that CMS has concluded that the growth in the hospice benefit was due to the BNAF, thereby justifying its elimination. The commenters noted a number of factors that have contributed to the hospice industry's growth, including an increased number of beneficiaries using the benefit, longer lengths of stay, increased acceptance of hospices for end-of- life care by the physician and patient/family communities, changes in the mix of patients using hospice, and educational efforts by providers and by CMS to beneficiaries and health care providers.

Several commenters felt that the proposed BNAF reduction is a reaction to increasing hospice reimbursements overall. Another commenter stated that hospice is a small portion of all Medicare spending.

Response: We appreciate these comments. As we indicated in our responses, the FY 2009 financial impact of the BNAF phase-out is no more than a 1.1 percent reduction in payments. Therefore, with a 3.6 percent market basket update factor for FY 2009, we do not believe that there will be a significant adverse effect on the new providers, or on long-standing providers. We agree that demand for hospice services is growing as the U.S. population ages, and as the baby boomer generation begins to be eligible for Medicare.

We disagree with the commenter's suggestion that CMS does not know the financial status of hospices. In fact, the Medicare Payment Advisory Commission (MedPAC) has performed extensive analysis on various aspects of hospice financial performance and utilization trends over the last few years, including an assessment of growth trends in the hospice industry. We believe that both the growth in hospice expenditures and the growth in the number of hospices are indicators of financial stability in the industry, especially given the growth surge in the number of for-profit providers. MedPAC noted that hospice care has changed considerably since the benefit's implementation. In 1983 most providers were nonprofits affiliated with religious or community organizations, but now for-profit hospices constitute the majority of providers and the vast majority of new entrants into the program since 2000 (MedPAC, p. 206). In 1998, for-profit providers comprised 26.9 percent of the industry (63 FR 53456), while in 2007, for-profits comprised 51 percent (MedPAC, p. 216). The growth in not-for-profit hospices since 1998 has remained relatively flat.

[Page Number 46472]

MedPAC has also provided some information about the financial health of hospices, particular those who are new entrants into the market. MedPAC noted that hospices that began participating in the market in 2000 or after had consistently and substantially higher margins than those participating in Medicare before 2000. In addition, these higher margins are consistent with the growth in the number of for-profit providers (MedPAC, p. 223-224). Therefore, we do not believe that the newer entrants will be more affected by the BNAF reduction than older hospices.

We disagree with the comment that we asserted that the growth in the hospice industry was due to the BNAF or that the BNAF reduction is a reaction to the growth in hospice reimbursements. However, the commenters correctly noted several factors that have contributed to industry growth. We indicated that the BNAF phase-out was not a reaction to that growth--in the proposed rule, rather we stated that the BNAF was put in place to mitigate any adverse financial impact that individual hospices might experience as a result of transitioning to the new hospital-based wage index in 1998. We note that industries do not typically expand and grow during times of financial adversity; often there is industry contraction instead. We stated that the growth in the industry is an indication that any adverse financial effects of transitioning to a new wage index had ended.

We agree that relative to all Medicare spending, hospice spending is a small portion accounting for an expected 2.3 percent of spending overall in FY 2009. However, we estimate that hospice spending will more than double in the next 10 years. The growth in hospice spending has outpaced the rate of growth for other Medicare provider types. Furthermore, CMS has a responsibility to safeguard trust fund dollars by paying accurately and appropriately for all Medicare services. Finally, we disagree with the commenter that the proposed reduction in the BNAF is simply a reaction to increasing hospice reimbursements. Rather, as we have stated in the previous responses, we believe that the purpose of the BNAF was to mitigate the negative financial impact of a 1998 wage index change. We believe this mitigation for the transition to a "new" wage index is no longer necessary. We also believe that phasing out the BNAF places both Medicare home-based benefits on a more equal footing in terms of recruiting staff.

Comment: A commenter stated that cutting hospice payments disregards the significant, collaborative progress made in the Medicare hospice program over the last decade. A few commenters stated that CMS circumvented Congress by going through rulemaking to propose and possibly finalize a BNAF phase-out. Several commenters suggested CMS should use negotiated rulemaking to refine payment policy such as the BNAF. Another commenter stated that the wage index calculation is not and never has been intended to be used as a method to form payment policy. This commenter stated that role historically has been reserved for Congress. Another commenter stated that the BNAF phase-out was an administrative proposal put forward in the President's budget, and therefore should be enacted by Congress rather than effectuated through CMS rulemaking. Another commenter stated that the Secretary of the Department of Health and Human Services is required to propose reforms to the wage index calculations. One commenter stated that the proposed rulemaking process administratively circumvented the legislative intent to maintain and ensure adequate hospice funding levels.

Response: We appreciate these comments, but respectfully disagree with the commenters. The BNAF was put into place through use of a Negotiated Rulemaking Committee. We recognized that the wage index methodology, including the BNAF, could be changed when we included the following statement in Section IV (B) of the August 8, 1997 Final Rule entitled "Medicare Program; Hospice Wage Index" (62 FR 42863):

The annual updates will update the hospice wage index values according to the methodology agreed to by the rulemaking committee and implemented by this final rule. In the event that we decide to change the methodology by which the wage index is computed, this will be reflected in a proposed rule published in the Federal Register .

The "we" in this paragraph refers to CMS (formally HCFA), which published the final rule in 1997. It is clear from this statement that the wage index methodology, including the BNAF, is subject to changes by CMS, and that any such changes do not have to go through negotiated rulemaking, but rather through our rulemaking process of publishing proposed and final rules in the Federal Register . There is no statutory requirement that requires wage adjustment methodology changes to go through Congress. While legislative proposals in the President's Budget require Congressional action, administrative proposals in the budget simply indicate intended administrative action, and do not require any Congressional action.

We believe that the intent of the BNAF was to protect hospice payments during the transition to the hospital-based wage index. The growth in the industry and in payments suggests that the industry has adequate funding levels, and is one reason for our proposal to phase out the adjustment.

We value the collaborative process, but do not feel that it is limited to Negotiated Rulemaking. The notice and comment rulemaking process, which we are following, allows for industry input and comment and is the general process by which changes to Medicare payment policy most often occur. We look forward to continuing to work with the industry in the future.

Comment: Many commenters stated that achieving budget neutrality was always a goal of the negotiated rulemaking process. A commenter disputed CMS' assertion that the BNAF was intended to prevent reductions in payments to the majority of hospices whose wage index was based on the original hospice wage index, which was artificially high due to flaws in the BLS data, and stated that the BNAF was never a point of contention during the Negotiated Rulemaking Process. The commenter stated that before the Negotiated Rulemaking Committee began its negotiations, CMS (formally HCFA) indicated that the wage index could not be used as a tool to increase payments to hospices, nor would it be used as a tool to lower aggregate payments to hospices. The commenter quoted our regulation (59 FR 52131), which stated that "We [HCFA] consider it a given of negotiation that any revised wage index would have to be at least budget neutral; that is, total aggregate payments for the same services could not be more using the revised wage index than if such payments were made using the current index."

Response: We stand by our assertion that the BNAF was intended to prevent reductions in payment, and point to the quote from the 1994 proposed rule (59 FR 52131), which the commenter included that the BNAF was designed to protect aggregate payments. We did not state or imply that the adoption of the BNAF was a point of contention during Negotiated Rulemaking. Rather, we said that the BNAF was now outdated, and is no longer needed for the reasons given in the proposed rule (73 FR 24006).

It is clear that we have the authority to make changes to the hospice wage index methodology, as noted in the August 8, 1997 final rule (62 FR 42862):

[Page Number 46473]

"In the event we decide to change the methodology by which the wage index is computed, this will be reflected in a proposed rule published in the Federal Register .

a. Effects of Phasing Out the BNAF Using the Published FY 2008 Hospice Wage Index

In the proposed rule, we used the August 31, 2007 FY 2008 hospice wage index (72 FR 50214) to illustrate the effects of phasing out the BNAF over 3 years. This analysis and discussion is for illustrative purposes only and does not affect any of the hospice wage index values for FY 2008.

The BNAF that was calculated and applied to the 2007 raw pre-floor, pre- reclassified hospital wage index values was 6.6671 percent. We will reduce the BNAF by 25 percent for FY 2009, by 75 percent for FY 2010, and eliminate it altogether for FY 2011 and beyond. A 25 percent reduction in the BNAF can be accomplished by blending 75 percent of the FY 2008 hospice wage index that applied the full 6.6671 percent BNAF with 25 percent of the FY 2008 hospice wage index that used no BNAF. This is mathematically equivalent to taking 75 percent of the full BNAF value, or multiplying 0.066671 by 0.75, which equals 0.050003, or 5.0003 percent. The BNAF of 5.0003 percent reflects a 25 percent reduction in the full BNAF. The 25 percent reduction in the BNAF of 5.0003 percent would be applied to the raw pre-floor, pre-reclassified hospital wage index values of 0.8 or greater used in the published FY 2008 hospice wage index.

The hospice floor calculation will still apply to any raw pre-floor, pre- reclassified hospital wage index values less than 0.8. Currently, the floor calculation has 4 steps--(1) Raw pre-floor, pre-reclassified hospital wage index values that are less than 0.8 are first multiplied by 1.15; (2) the minimum of 0.8 or the raw pre-floor, pre-reclassified hospital wage index value times 1.15 is chosen as the preliminary hospice wage index value; (3) the raw pre-floor, pre-reclassified hospital wage index value is multiplied by BNAF; and (4) the greater result of either step 2 or step 3 is chosen as the final hospice wage index value. We left the hospice floor calculation unchanged, noting that steps 3 and 4 will become unnecessary once the BNAF is eliminated.

For the simulations of the BNAF phase-out for FY 2010 and FY 2011, we used the same raw pre-floor, pre-reclassified hospital wage index values and claims data as the example above, and simply changed the value of the BNAF to reflect either a 75 percent reduction for FY 2010 or a 100 percent reduction for FY 2011. In both cases we started with the full BNAF of 6.6671 percent. We changed the calculation to take 25 percent of the full BNAF to reflect a 75 percent reduction for FY 2010, or eliminated the BNAF altogether to reflect a 100 percent reduction for FY 2011. For FY 2010, the reduced BNAF or the hospice floor was then applied to the 2007 raw pre-floor, pre-reclassified hospital wage index as described previously. For FY 2011 and subsequent years, the raw pre-floor, pre-reclassified hospital wage index values would be unadjusted unless they are less than 0.8, in which case the hospice floor calculation would be applied. Again, we note that the steps in the calculation that involve the BNAF will become unnecessary once the BNAF is phased out.

For our simulations, the calculations of the BNAF are as follows:

. A 75 percent reduction to the BNAF in FY 2010 would be 0.066671 x 0.25 = 0.016668 or 1.6668 percent.

. A 100 percent reduction or elimination of the BNAF in FY 2011 would be 0.066671 x 0.0 = 0.0 or 0 percent.

We examined the effects of phasing-out the BNAF versus using the full BNAF of 6.6671 percent on the FY 2008 hospice wage index. The FY 2009 BNAF reduction of 25 percent resulted in approximately a 1.55 to 1.57 percent reduction in the hospice wage index values. The FY 2010 BNAF reduction of 75 percent would result in an estimated additional 3.12 to 3.13 percent reduction from the FY 2009 hospice wage index values. The elimination of the BNAF in FY 2011 would result in an estimated final reduction of the FY 2011 hospice wage index values of approximately 1.55 to 1.57 percent compared to FY 2010 hospice wage index values.

Those CBSAs whose raw pre-floor, pre-reclassified hospital wage index values had the hospice floor calculation applied before any change to the BNAF would not be affected by the phase-out of the BNAF. These CBSAs, which typically include rural areas, are protected by the hospice floor calculation. Additionally, those CBSAs, which were eligible for the hospice floor calculation, but whose hospice wage index values were previously 0.8 or greater after the calculation was applied, but which would have values less than 0.8 after the calculation using a reduced BNAF was applied, would see a smaller reduction in their hospice wage index values. We have estimated the number of CBSAs that would have their raw pre-floor, pre-reclassified hospital wage index value eligible for the floor calculation after applying the 25, 75, and 100 percent reductions in the BNAF. Three CBSAs would be affected by the 25 percent reduction, 12 would be affected by the 75 percent reduction, and 22 would be affected by the 100 percent reduction. Because of the protection given by the hospice floor calculation, these CBSAs would see smaller percentage decreases in their hospice wage index values than those CBSAs that are not eligible for the floor calculation. This will benefit those hospices with lower hospice wage index values, which are typically in rural areas.

Finally, the hospice wage index values only apply to the labor portion of the payment rates; the labor portion was described in section I.B.1 of the proposed rule (73 FR 24002). Therefore, the estimated reduction in payments due to the phase-out of the BNAF would be less than the percentage reductions to the hospice wage index values that would result from reducing or eliminating the BNAF. In addition, the effects of the phase-out of the BNAF will also be mitigated by a hospital market basket update in payments, which in FY 2008 was a 3.3 percent increase in payment rates. The hospital market basket update for FY 2009 will be 3.6 percent. This update and the FY 2009 payment rates will be officially communicated through an administrative instruction and not through rulemaking. The estimated effects on payment described in column 5 of Table 2 in section V of this final rule include the projected effect of a FY 2009 3.6 percent hospital market basket update.

b. Effects of Phasing Out the BNAF Using the Updated Raw Pre-Floor, Pre- Reclassified Hospital Wage Index Data (FY 2009 Proposal)

In this final rule, for FY 2009, we are updating the hospice wage index using the 2008 raw pre-floor, pre-reclassified hospital wage index and the most complete claims data available (FY 2007 claims as of March 2008). Using these data, we computed a full BNAF of 6.6255 percent. For the first year of the BNAF phase-out (FY 2009), the BNAF will be reduced by 25 percent, or 0.066255 x 0.75 = 0.049691, to 4.9691 percent. This will decrease hospice wage index values by approximately 1.55 to 1.56 percent from wage index values with the full BNAF applied. As noted in the previous discussion on the effects of the BNAF reduction in the published FY 2008 hospice wage index, those CBSAs which already have raw

[Page Number 46474]

pre-floor, pre-reclassified hospital wage index values that have the hospice floor applied before implementing a BNAF reduction will be completely unaffected by this BNAF reduction (for example, rural West Virginia, and CBSA 13900, Bismarck, ND). Those CBSAs which are eligible for the hospice floor, and which previously had hospice wage index values above 0.8 after applying the full BNAF (as part of the floor calculation), but which now are below 0.8 with the 25 percent reduction in the BNAF, will be less affected by the BNAF reduction than those CBSAs which are 0.8 or above after applying the BNAF. They are protected by the hospice floor calculation (for example, rural Alabama would realize a decrease in its wage index value of only 0.40 percent, and CBSA 27780, Johnstown, PA, would realize a decrease in its wage index value of only 0.53 percent). Additionally, the final hospice wage index is only applied to the labor portion of the payment rates, so the actual effect on estimated payment will be less than the anticipated percent reduction in the hospice wage index value. Furthermore, that effect will be mitigated by a market basket update. The final market basket update for FY 2009 will be 3.6 percent rather than the 3.0 percent estimated in the proposed rule.

Column 3 of Table 2 (section V of this final rule) shows the impact of using the most recent wage index data (the 2008 raw pre-floor, pre- reclassified hospital wage index not including any reclassification under section 1886(d)(8)(B) of the Act) compared to the 2007 raw pre-floor, pre- reclassified hospital wage index data which was used to derive the FY 2008 hospice wage index. Column 4 of Table 2 in Section V of this final rule shows the impact of incorporating the 25 percent reduction in the BNAF in the FY 2009 hospice wage index along with using the most recent wage index data (2008 raw pre-floor, pre-reclassified hospital wage index). Finally, column 5 of Table 2 shows the combined effects of using the updated raw pre-floor, pre- reclassified hospital wage index, the 25 percent reduced BNAF, and a FY 2009 market basket update of 3.6 percent. The FY 2009 rural and urban hospice wage indexes can be found in Addenda A and B of this final rule. The raw pre-floor, pre-reclassified hospital wage index values were adjusted by the 25 percent reduced BNAF or by the hospice floor.

Comment: Many commenters stated that the impact of the phased elimination of the budget neutrality provision is much greater than the 1.1 percent reduction in payment that was described in the proposed rule. These commenters stated that some providers will experience reductions ranging from 5 percent to in excess of 14 percent over the 3 years as a result of the BNAF phase-out, and stated that the cuts would create hardship for hospices. Several commenters gave specific examples of CBSAs where wage index values decreased more than 1.1 percent, or of wage index values in contiguous CBSAs which decreased, but by differing amounts. The commenters stated that they cannot match pay scales with such a disparity in wage index values, and that there are no differences in medical costs between adjoining CBSAs.

Response: We appreciate the comments about the financial impact of the proposed rule, but are very concerned that incorrect percentage impacts on payments from the BNAF reduction are being cited in many of the comments. Some commenters may have confused the effect of the BNAF reduction with the effect of fluctuations in the wage index values from the raw pre-floor, pre- reclassified hospital wage index. This hospital wage index is used to derive the hospice wage index. We emphasize that the BNAF reduction will result in a payment reduction to hospices of no more than 1.1 percent in FY 2009. The large payment cuts which hospices repeatedly cited are not due to the BNAF reduction. The impact table in the proposed rule shows the effects of using the updated wage index values, the combined effects of using updated wage index values and the 25 percent reduction in the BNAF, and the combined effects of the updated wage index values, the 25 percent BNAF reduction, and the market basket update. Given the apparent confusion, we will clarify our methodology for calculating the FY 2009 hospice wage index to include the BNAF reduction, and we will expand our explanation of the associated impacts.

In the proposed rule, using the most current data available, we first calculated the unreduced BNAF for FY 2009, which was 6.5357 percent. We reduced that number by 25 percent, to arrive at 4.9018 percent. The raw wage index values from the raw pre-floor, pre-reclassified hospital wage index were increased by 4.9018 percent instead of by 6.5357 percent for every CBSA or rural area with a wage index value of 0.8 or greater (if the raw wage index value was less than 0.8, the hospice floor applied). The difference in the wage index value was 6.5357 percent - 4.9018 percent = 1.6339 percent. However, the wage index value only applies to the labor portion of payments, so the effect on payments is less. The labor percentages do not vary by hospice or by CBSA; they are the same for every provider. Therefore the impact of the BNAF reduction on payments does not and cannot vary from one location to another.

In the proposed rule, our impacts showed the effect of the 25 percent BNAF reduction (not including the effect of using the updated wage index) on total payments to be a 1.0 percent reduction (1.1 percent in this final rule), compared to what hospices would have received if the full BNAF had been used. We noted in the proposed rule that this reduction would be offset by a market basket update that was estimated at 3.0 percent. Because the BNAF reduction is applied across the board to the raw pre-floor, pre-reclassified hospital wage index values as a percentage reduction, all hospices (except those subject to the floor) are affected the same in that this final rule's estimated reduction in payments to all hospices is approximately 1.1 percent. Over the course of the 3-year phase-out, the elimination of the BNAF will reduce payments by about 4 percent: We estimated a 1.1 percent reduction in FY 2009, an additional 2 percent reduction in FY 2010, and an additional 1 percent reduction in FY 2011. However those reductions do not include 3 years of market basket updates for FY 2009, FY 2010, and FY 2011. Therefore, assuming market basket updates' inclusion in FY 2010 and FY 2011, hospices will still have a net gain in payments over the 3 years. While we do not know what the market basket updates will be for FY 2010 and FY 2011, hospices received market basket updates ranging from 3.3 percent to 3.7 percent from FY 2005 to FY 2008. The market basket update for FY 2009 is 3.6 percent. Because we do not know how the commenters calculated the percentage reductions they cited, it is unclear whether they accounted for market basket updates in their analyses. Therefore, not knowing the details of the analysis, we are unable to comment further on, or substantiate, the commenters' analysis.

As with the estimated reduction for FY 2009, the reductions in FY 2010 and FY 2011 payments will apply uniformly to all hospices with wage index values >/=0.8. Therefore, for hospices with raw pre-floor, pre-reclassified hospital wage index values >/=0.8, the reductions over the 3 years will not be larger for one hospice versus another. As noted in the proposed rule, those with raw pre-floor, pre-reclassified wage index values < 0.8 will be less affected or unaffected by the BNAF phase-out.

[Page Number 46475]

In this final rule, we calculated the BNAF using updated claims data (2007 claims as of March 2008). The full BNAF was slightly higher at 6.6255 percent; the 25 percent reduced BNAF was 4.9691 percent, which is slightly higher than the BNAF in the proposed rule. Therefore, the raw pre-floor, pre-reclassified hospital wage index values used to derive the hospice wage index were increased by 4.9691. Because of the increase in the BNAF itself from the proposed rule to the final rule, the hospice wage index values in this final rule are slightly higher than those that were in the proposed rule.

Additionally, the final market basket update for FY 2009 is 3.6 percent rather than the 3.0 percent estimated in the proposed rule. That means the total impact of using an updated wage index, of reducing the BNAF by 25 percent, and of the market basket update is estimated to be a 2.5 percent increase in payments to hospices in FY 2009.

The impact of the BNAF reduction does not vary from hospice to hospice (except for those subject to the floor) as the same adjustment was applied across the board to the raw pre-floor, pre-reclassified hospital wage index values. Likewise, the impact of the market basket update does not vary from hospice to hospice, as the 3.6 percent increase is applied to the same base rates across the board. Therefore our impacts do not and cannot mask the effects of the BNAF reduction by presenting aggregate data.

The only place for variation in payment at the individual hospice level is within the raw pre-floor, pre-reclassified hospital wage index values themselves. These raw wage index values are the input values which are adjusted by either the BNAF or the hospice floor calculation to derive the hospice wage index. To show the changes from FY 2008 to FY 2009 in the raw pre-floor, pre-reclassified hospital wage index, from which the hospice wage index is derived, see Addendum C in this final rule.

Addendum C shows that large fluctuations in some wage index values exist from year to year, some positive and some negative. These fluctuations are the source of negative and positive effects on payment to hospices beyond the 1.1 percent reduction due to the BNAF and the 3.6 percent increase due to the market basket update. Between FY 2008 and FY 2009, there were 21 CBSAs or rural areas with raw pre-floor, pre-reclassified hospital wage index values which decreased 5 percent or more, and 16 CBSAs or rural areas with raw pre- floor, pre-reclassified hospital wage index values which increased 5 percent or more. We have also included Addendum D, comparing FY 2008 raw pre-floor, pre-reclassified hospital wage index values with those from FY 2007 to demonstrate that fluctuations in raw wage index values occur every year. Addendum D shows that there were actually more fluctuations between FY 2007 and FY 2008 than between FY 2008 and FY2009; Addendum D also shows that between FY 2007 and FY 2008, 23 CBSAs or rural areas had raw pre-floor, pre- reclassified hospital wage index values that decreased by 5 percent or more, and 17 CBSAs or rural areas that increased by 5 percent or more. We remind commenters that these raw pre-floor, pre-reclassified hospital wage index values are adjusted upward by the hospice floor if the value is below 0.8. These fluctuations do not translate into an equivalent increase or decrease in payments, as the wage index value only applies to the labor portion of payments. Additionally, in considering the total impact on payments, commenters would need to account for the market basket increase that applies to hospice payment rates.

The raw pre-floor, pre-reclassified hospital wage index originates from data provided on each hospital's cost reports. Hospitals must report their wages paid; Medicare takes those data for the hospitals in each CBSA and computes a CBSA average hourly rate. It also takes the data for all hospitals and computes a national average hourly rate, which becomes the standard. The raw pre-floor, pre-reclassified wage index values for each CBSA are computed by dividing the CBSA's average hourly rate by the national average hourly rate. Therefore, if a wage index value is increasing or decreasing, it is because hospital wages within that CBSA are increasing or decreasing relative to the national average.

CMS performs an intensive review of the hospital wage data, mostly through use of edits to identify aberrant data. The Fiscal Intermediary/MAC then revises or verifies the data elements that resulted in specific edit failures.

Table 1 below shows calculation of the hospice wage index for both FY 2008 and FY 2009, beginning with the raw pre-floor, pre-reclassified wage index value (the input), and applying the BNAF or hospice floor for 3 CBSAs. For the first CBSA (31020), the raw pre-floor/pre-reclassified hospital wage index for FY 2009 is greater than it was in FY 2008. Conversely, the raw pre-floor/pre- reclassified hospital wage index values for CBSAs 41780 and 48540 are less in FY 2009 than in FY 2008. Table 1 shows the computation of the hospice wage index values for these CBSAs for FY 2008 and for FY 2009 (using proposed rule BNAF values). The table also demonstrates that the hospice floor protects values < 0.8 from the effects of the BNAF reduction. In this case (CBSA 48540), the FY 2009 proposed wage index value is unchanged from the final wage index value for FY 2008. Therefore, as we noted in the proposed rule, the BNAF reduction had no effect in this circumstance.

The cities and counties which make up CBSAs are not determined by CMS, but instead are set by the OMB. Information about CBSA designations is available at the following Web site: http://www.whitehouse.gov/omb/bulletins/fy2008/b08- 01.pdf. We continue to believe that OMB's CBSA designations reflect the most recent available geographic classifications and are a reasonable and appropriate way to define geographic areas for the purposes of determining wage index values.

Currently there are limited data available for analysis of the impact of the phase-out of the BNAF on quality. The new claims data (with visit reporting beginning July 1st, 2008) and the new Conditions of Participation will provide data related to quality of care. We will monitor these data for any unanticipated effects of the BNAF phase-out.

[Page Number 46476]

Table 1--Examples of How the Proposed FY 2009 Wage Index Values Were Derived

CBSA FY08 FY 08 BNAF-- Output-- FY09 FY 09 Output--

input-- Full BNAF = FY 2008 input-- proposed FY 2009

raw pre- 0.066671 final raw pre- BNAF-- NPRM

floor pre- increases hospice floor pre- 25% reduced hospice

reclassi- input value wage index reclassi- BNAF = wage index

fied for all value fied 0.049018 value

hospital providers hospital increases

wage index w/WI wage index input value

for FY08 values >/= for FY09 for all

0.8; or apply providers

hospice floor w/WI

values >/=

0.8; or apply

hospice floor

31020 1.0011 x 1.066671 1.0678 1.0827 x 1.049018 1.1358

41780 0.9302 x 1.066671 0.9922 0.8822 x 1.049018 0.9254

FY08 input < FY09 input <

0.8; hospice 0.8; hospice

floor applies floor applies

0.7010 x 1.15 0.6961 x 1.15

= 0.8062; = 0.8005;

(Subject to (Subject to

0.8000 max) 0.8000 max)

0.7010 x 0.6961 x

1.066671 = 1.049018 =

0.7477 0.7302

48540 0.7010 Take greater 0.8000 0.6961 Take greater 0.8000

of 15% of 15%

increase increase

(subject to (subject to

0.8 maximum) 0.8 maximum)

or BNAF or BNAF

increase increase

Comment: A number of commenters referred to the BNAF phase-out as a "rate reduction" or stated that CMS was cutting rates. In addition, a commenter asked CMS to publish the rate updates as part of the rule in the Federal Register , rather than in an administrative notice such as a Change Request.

Response: The BNAF is an adjustment which increases the raw pre-floor, pre- reclassified hospital wage index values that are 0.8 or greater, with the result being the hospice wage index. Raw pre-floor, pre-reclassified hospital wage index values < 0.8 have the hospice floor calculation applied instead.

The hospice payment rates are per diems for routine home care, continuous home care, respite care, and general inpatient care. They were put into place by Congress, and are updated annually by the market basket update. We have not proposed any cut to the payment rates. In the proposed rule, we estimated that the per diems would increase due to a 3.0 percent market basket update; for the final FY 2009 rule the market basket update to be applied to the per diems increased to 3.6 percent. Therefore, we are not cutting hospice payment rates. Conversely, hospice payment rates for FY 2009 will be increased by the hospital market basket update of 3.6 percent, and will be communicated through a separate administrative instruction/issuance this summer.

We appreciate the comment about where payment rate updates are published. Historically, the payment rate updates have been issued through a separate administrative instruction or administrative issuance in the summer of each year to provide adequate time to implement the necessary system changes. In previous years, the hospice wage regulation was often published after August 1st, which does not allow sufficient time for system changes to be made to accommodate the October 1st implementation of payment updates. We will look into including the updated payment rates in the Federal Register in the future.

Comment: CMS received a number of comments suggesting that a BNAF phase-out would limit access to hospice care. Multiple commenters noted that costs were rising, including gasoline, wages, pharmacy costs, medical supplies, insurance, utilities, and food, and that hospices cannot absorb these costs in addition to the BNAF reduction without adversely affecting Medicare beneficiary hospice care. One commenter mentioned the high cost of converting to electronic health records. Many commenters stated that the BNAF reduction exacerbates the financial strain that CMS has already imposed on hospices this year, noting new hospice costs resulting from CMS's requirements to implement new Hospice Conditions of Participation (CoPs), and CMS's new visit data collection requirements (CR 5567). A few commenters stated that agencies are having difficulty soliciting donations in this economic downturn. Some rural providers commented that their staff may drive 100 miles each way to visit patients, and that they cannot afford the rising cost of gasoline. Rural and urban commenters stated that they could not survive a reduction in payments in the face of rising costs. These commenters stated that as a result of the BNAF reduction, they would have to limit the geographic areas they service, thus limiting hospice access to beneficiaries, especially in rural areas.

In addition, few commenters stated that there was anecdotal evidence that an increasing number of hospices had ceased operations or were in danger of closing, due to rising gas prices and to cap overpayments. Other commenters stated that they may have to delay expansion. Some commenters stated that they would have to discontinue programs, including bereavement programs, outreach programs, programs to specific underserved groups (for example, to inner city beneficiaries), complementary treatments (that is, acupuncture and art therapy) or comfort items such as overlay airflow mattresses, and charity care. Several commenters mentioned that hospices would be forced to limit access by restricting admissions, limiting the number of admissions for costly, medically complex patients such as cancer patients needing expensive palliative treatments. There was concern among many commenters that the BNAF phase-out would lead to cost-cutting within hospices, including staff reductions that would jeopardize quality patient care. Commenters stated that quality care takes time, but if staffing is reduced, the nurse-to-patient ratio will be unfavorable, and hospice workers will spend less time with patients. One commenter stated that hospices may cut the number of visits they make, and use phone contact instead. Another stated that the reduction disproportionately punishes best practices which hold true to the hospice concept. A commenter felt that the BNAF reduction punishes high quality, high quantity providers.

Response: We appreciate the commenters' concerns about rising costs and about access to hospice care. We agree that costs are rising and that it is vital to preserve access to hospice care for Medicare beneficiaries. As noted in our response to a previous comment, it appears that the changes in payments to individual hospices have been

[Page Number 46477]

misunderstood as being due to the BNAF reduction. For FY 2009, the BNAF reduction cannot affect any hospice by more than 1.1 percent. As stated in an earlier response, it is worth noting that while it is true that, in a given year, some areas will see what could be considered a significant decrease in their raw pre-floor, pre-reclassified hospital wage index value, other areas will see significant increases in their raw pre-floor, pre-reclassified hospital wage index value. These fluctuations in the raw pre-floor, pre- reclassified hospital wage index occur every year (Addenda C and D of this final rule show fluctuations from FY 2007 to FY 2008 and from FY 2008 to FY 2009). These fluctuations are not related to the BNAF reduction, and we believe the 1.1 percent impact of the BNAF, which is offset by the 3.6 percent FY 2009 market basket update, will not force hospices to close their doors or otherwise affect access to the quality, compassionate care which beneficiaries expect and deserve.

We agree that rising gas prices are a concern for hospices, and note that the hospital market basket update which is used by hospices includes an energy component that is sensitive to petroleum costs. It is reasonable to expect that future market basket updates will continue to account for any continuation of rising fuel costs. The FY 2009 market basket update increased from 3.0 percent in the proposed rule to 3.6 percent in this final rule, partly due to rising energy costs. We refer the reader to the comment about the market basket update later in this section for more details on the market basket update.

In addition, we believe that the requirements associated with the CoPs and CR5567 are part of the cost of doing business, and that the industry has had ample time to plan and budget for these changes. We do not believe that these requirements will have adverse affects on admissions or services, but instead expect that the emphasis on quality and the increased awareness of visits provided could enhance services.

We believe that in a time of inflationary pressure, all businesses, including hospices, will seek to operate more efficiently. We do not believe that the BNAF reduction will lead to the type of cost-cutting that would jeopardize quality care. However, we plan to monitor the effect of the BNAF reduction to assess whether unanticipated effects occur.

Comment: Several commenters mentioned the Medicare Payment Advisory Commission's (MedPAC's) June 2008 report which includes a discussion of hospice margins. Commenters stated that MedPAC reports hospice margins for the period 2001 through 2005 averaged 3.4 percent, and margins for some categories of providers were even lower. For example, commenters stated that MedPAC reported margins for not-for-profit hospices that were -2.8 percent during this period. Commenters stated that hospices cannot withstand the payment reductions resulting from the BNAF phase-out, stating that complete elimination of the BNAF would result in negative margins for all hospice provider types. Commenters were especially concerned about the effect any BNAF reduction would have on not-for-profit hospices, stating that the reduction will further reduce margins which are already in the negative range. Commenters were concerned that the BNAF reduction will reduce profitability and increases losses such that hospices will close.

Response: In June 2008, MedPAC published a report entitled "Evaluating Medicare's Hospice Benefit" (MedPAC, pp. 203-237). This is the first time margins have been analyzed for hospices. MedPAC estimated Medicare hospice margins using Medicare claims and cost report data for the period from 2001 to 2005. Their report stated, "These margins may not provide a full picture of hospices' financial status. Nonprofit hospices derive revenues from philanthropic donations, which are an integral part of their operations and mission; these revenues are not consistently reported on Medicare cost reports. These revenues may help offset the generally negative margins we observe for nonprofit hospice providers. Additionally * * * hospitals may find it desirable to operate hospices, even in light of negative hospice margins. Harrison and colleagues (2005) found that hospitals that operated hospice programs had higher return on assets and higher hospital occupancy rates, as well as shorter lengths of stay, than hospitals without hospices" (MedPAC, p. 224).

As noted above, the margins that MedPAC showed in its report may not tell the full story of hospice profitability. MedPAC noted that financial analysts have estimated margins of 6 to 15 percent for 2006 for the 3 largest publicly traded hospice chains. Further, MedPAC noted that Security and Exchange Commission (SEC) filings on publicly traded hospices estimated margins based on all revenues and costs (not just Medicare, which accounts for more than 90 percent hospice revenues). Two major for-profit chains had 7 percent and 7.8 percent margins in CY 2006. A third chain showed a loss of 3 percent for FY 2007, but attributed it partly to corporate restructuring costs and their cap liability (they have since reduced their cap exposure). MedPAC noted that the chain with the 7.8 percent margin was seeking to acquire the chain with the loss. In addition, analysts project that the prospective buyer's margins will be 11 to 12 percent over the next several years.

We will continue to work with MedPAC to assess the appropriateness of future enhancements to the Hospice payment system. We will work closely with them as they review and refine their margin analysis and assess the impact of the BNAF reduction on margins. Phasing out the BNAF rather than eliminating it all at once enables us to iteratively assess the impact of the reduction.

Comment: Many commenters wrote about the financial hardship to hospices as a result of rising fuel costs. Two commenters said they do not reimburse volunteers for mileage, and that they were losing volunteers because of high gasoline prices, with rural areas being particularly hard hit. Several providers, particularly rural ones, noted that their staff may drive 100 miles each way to see a patient or that their service area may cover over 1,000 square miles. A commenter also asked that we develop a hospice-specific reimbursement system that would entail developing a hospice-specific market basket. A commenter noted that the per diems are updated by an index which is not hospice specific, and that the assumptions in the per diem components had changed. Pharmacy costs, travel costs, and salaries have risen at rates that are not provided for in the subsequent market basket updates. Several noted that using the hospital market basket was disadvantageous to hospices because of differences in the way hospices and hospitals operate; some mentioned that hospice providers have been adversely impacted by the tremendous rise in gasoline prices far more than hospitals, because hospice services are provided where the patient resides rather than in a single facility. A commenter added that the market basket updates are not driven by an objective analysis of hospice costs.

Response: Section 1814(i)(1)(C)(ii)(I) of the Act requires that the hospice PPS market basket update factor be based on the IPPS hospital market basket (as defined in Section 1886(b)(3)(B)(iii) of the Act). The IPPS market basket reflects the operating cost structures of IPPS hospitals and the inflationary pressures

[Page Number 46478]

facing these providers. The market basket update factor includes any associated price changes with labor, energy, insurance, food, pharmacy, and other IPPS hospital operating costs.

Hospitals and hospice facilities both hire staff from the same healthcare worker labor pool; however, hospitals tend to have a higher skill occupation mix compared to hospice facilities. The IPPS market basket update factor reflects the inflationary pressures on these highly skilled healthcare occupations.

The hospice per diems are updated using the hospital market basket, which we agree is not hospice-specific. While the hospital market basket does not have a specific transportation factor, it does include energy costs. It also includes pharmacy costs and wage costs. To see the components in the hospital market basket, we refer the reader to: http://www.cms.hhs.gov/MedicareProgramRatesStats/downloads/mktbskt-pps- hospital-2002.pdf. To better understand the market baskets and how they are constructed, we also refer the reader to: http://www.cms.hhs.gov/MedicareProgramRatesStats/downloads/info.pdf.

Regarding the development of a hospice-specific market basket, CMS will investigate the cost structures specific to hospice facilities and how they compare to IPPS hospitals. However, we believe that Congress intended us to use the hospital market basket to adjust for hospice inflationary pressures, and therefore, the authority to create and use a hospice-specific market basket is determined by Congress.

Comment: Several commenters stated that there were likely to be major revisions to the hospital wage index in the future and that CMS should not remove the BNAF when the wage index is fluctuating. A few commenters indicated that CMS has not studied the impact on the hospice payment system of hospital wage index changes which occurred over the past 10 years. These commenters also stated that CMS justified the application of hospital-specific changes to the hospice payment system by application of the BNAF, which serves as a "cushion" for hospice. They also indicated that hospice providers were never given an opportunity to study the impact of these changes. Finally, these commenters stated that eliminating the BNAF now would subject hospices to multiple, significant changes over a short period of time. One commenter stated that CMS proposal to phase-out the BNAF is not reform. Another commenter indicated that wage index is as important to the stability of hospice care as are base payment rates, the annual inflation rate update, and the aggregate annual hospice cap, and suggested that stability would be lost if the BNAF were phased-out.

Furthermore, a commenter indicated that the current hospital-based wage index was not accurate, reliable, or equitable. The commenter stated that some hospices are part of health systems that have wage indices lower than a hospital in the same system and the same geographic area. Therefore, the commenter asserts, CMS cannot reason that the BNAF phase-out is needed for accuracy, reliability, and equity.

Response: In the May 1, 2008 proposed rule, we proposed phasing out the BNAF over 3 years because this adjustment has served its purpose, which was to mitigate the adverse financial impact of transitioning to a new wage index in 1998. The need for the BNAF has passed for a variety of reasons, as stated in section II.C.3 of this final rule. We did not propose to reform the raw pre- floor, pre-reclassified hospital wage index or the hospice payment system, but only to phase out the BNAF. We did not propose to phase out the hospice floor calculation, which continues to apply when raw pre-floor, pre-reclassified hospital wage index values are < 0.8, though we noted that the steps involving the BNAF would become unnecessary once the phase-out is complete.

The raw pre-floor, pre-reclassified hospital wage index is the same wage index used by the other home-based Medicare benefit, home health, which draws from the same labor pool as hospices. Home health agencies experience the same wage index fluctuations, but do not receive an adjustment such as the BNAF; therefore, we do not believe that phasing out the hospice BNAF in the presence of normal wage index fluctuation will be detrimental to the industry or threaten stability.

Our purpose in phasing out the BNAF is not to reform the wage index, but rather to phase out an adjustment which served its purpose. The BNAF helped hospices adjust to the negative financial impact of changing from the Bureau of Labor Statistics-based wage index to the raw pre-floor, pre-reclassified hospital wage index. There is no longer a need for this adjustment given that the transition occurred over 10 years ago and the growth that has occurred in the industry.

In our proposed rule, we stated that one reason for phasing out the BNAF is that it is not as accurate, reliable, and equitable as possible in accounting for geographic variation. The BNAF ties payments back to an outdated 1981 BLS- based wage index that the Negotiated Rulemaking Committee called "inaccurate" in its Committee Statement (62 FR 42883). We are unable to respond to the comment about one health system's wage index being different from that of a hospital in the same area as it is outside the scope of this rule.

Comment: One commenter suggested that CMS conduct a study to determine the appropriate per diems for rural hospices, asserting that rural costs are not adequately addressed in the current payment system. This commenter suggested CMS include an additional time and distance factor for rural hospices to align the per diem with rural costs.

Response: Medicare pays one of four daily rates to hospice providers, based on the intensity level of care the patient requires. These per diem payment rates are the same, regardless of whether the services are provided in an urban area or a rural area. The hospice wage index, which includes a floor calculation which benefits many rural providers, is the vehicle we use to adjust for geographic variances in labor costs. In a time of high gasoline costs, we are sensitive to concerns from rural hospices that the additional time and distance required to visit a rural patient adds significantly to their costs, and their assertion that payments are not adequate. However, we believe that an additional payment for rural providers, which is sometimes called a rural add-on payment, would have to be legislated.

Comment: Several commenters indicated that hospices sometimes care for high cost patients such as those who require more expensive palliative radiation or chemotherapy, and that hospice reimbursements are not sufficient to provide quality care to these complex patients. Another commenter stated that the original hospice per diems were arbitrarily set by CMS (formally HCFA) and had to be raised by Congress. A commenter stated that increased case mix and new treatment options have also affected hospice costs. A commenter recommended that CMS study the real costs of hospice care and devise a 21st century, hospice-specific reimbursement system. One commenter recommended that CMS raise the wage-related per diem.

Several commenters asked CMS not to decrease payments because they are already providing services above the level of reimbursement that is presently provided in the daily rate. Another

[Page Number 46479]

commenter stated that Medicare does not pay for staff consultations, pre- admission consults, travel time to patients' homes, or the cost of fuel. A few commenters stated that hospices now provide an ever-expanding array of costly palliative treatments.

Response: While these comments addressed issues that are beyond the scope of this final rule because they concern issues about which we did not make any proposals, we will address them briefly since they pertain to overall hospice reimbursement. In its June 2008 report, MedPAC wrote that hospice payments are generally adequate in the aggregate, but noted that individual hospices' performance varies (MedPAC, p. 205).

The hospice per diems were designed to cover hospice costs and were developed based on cost information gathered from a demonstration project that began in October, 1980 and which included 26 hospices. The per diems include costs of the services hospices provide, plus overhead costs such as maintenance, depreciation, general accounting, capital, and other administrative costs in the calculation of the individual service components (that is, nursing services and aide services). A cost component is included for hospital outpatient charges for palliative radiation and chemotherapy, based on a sample of Medicare patients who died from cancer in 1980, and adjusted for inflation. In the future, we plan to perform analyses on hospice resource utilization, in the hopes of refining our hospice payment system. We will consider the high cost of certain palliative treatments in future payment refinements analyses.

As some commenters stated there have been changes in hospice case mix, with proportionally fewer cancer patients and more patients with other diagnoses such as dementia and congestive heart failure. At this time, we do not have the data on services provided to patients with specific diagnoses, and therefore cannot easily determine the adequacy of Medicare payments relative to the cost of care. However, hospices were required to begin reporting visits for nurses, physicians, social workers, and aides as of July 1, 2008 (This is just a beginning in data collection efforts that should provide the information needed to refine payments in the future). We note that because hospice rates are currently described in statute, any hospice payment refinements which affect those rates would need to be enacted by Congress.

Comment: Many commenters praised the benefits of hospice, which provides care to the most vulnerable Medicare beneficiaries for example, the dying and does so in a cost-effective fashion that saves Medicare money. They suggested that if hospice utilization declines, beneficiaries would be forced to use more costly medical care, driving up Medicare costs in the long run. Others commented on the quality, compassionate care and comfort that hospice provides to patients and their families. Another commenter stated that hospice provides tremendous aid at end of life, serving the dying equally, regardless of color and economic status. These commenters and others asked that CMS not reduce payments. Several commenters also questioned the timing of CMS' proposal to eliminate the BNAF. They felt CMS should not be cutting hospice benefits at a time when the demand for services is growing. One commenter stated that the industry growth is a testament to the service gap that hospice fills, the growing awareness of hospice, and preference for a hospice death. A few mentioned that CMS has encouraged hospice usage. One commenter indicated that studies show that patients live longer on hospice, but there is still a need for earlier referrals to the benefit. The commenters stated that the population is aging and the baby boomer generation is getting older.

Response: We agree that the Medicare hospice benefit has been of tremendous benefit to those at end-of-life and to their families, and applaud those who serve the dying as hospice staff and volunteers. We also agree that the hospice benefit often saves Medicare money, and appreciate the studies which have highlighted the areas where it provides costs savings to the Medicare program. However, hospice care does not save money in every instance. MedPAC has noted that "hospice's net reduction in Medicare spending decreases the longer the patient is enrolled and beneficiaries with very long hospice stays may incur higher Medicare spending than those who do not elect hospice." (MedPAC, p. 209). We do not believe that hospice utilization will decline due to the BNAF phase-out, and therefore, do not believe that Medicare costs will be shifted from hospice to more expensive forms of care. As noted in our response to a previous comment, the FY 2009 impact of the BNAF phase-out is, at most, 1.1 percent for every provider, and is being offset by a 3.6 percent FY 2009 market basket update, resulting in a 2.5 percent increase in payments to hospice providers in FY 2009.

We agree that the hospice industry is growing and that the demand for hospice services is likely to grow in the future, particularly with an aging population. We also agree that CMS has encouraged hospice usage. We have not cut Medicare hospice benefits in any way--terminally ill Medicare patients are still eligible to receive the same quality, compassionate end-of-life care that has been the hallmark of hospice. We expect the hospice benefit to continue to grow: CMS' Office of the Actuary projects that Medicare hospice spending will more than double in the next 10 years. However, we will monitor the impact of the BNAF phase-out for any unintended impact.

Comment: Several commenters recommended that CMS delay any BNAF reduction or phase-out. They suggested waiting for better and more reliable data, for time to evaluate the impact of the requirements from the new Conditions of Participation and from CR 5567. A commenter stated that CMS lacks the data to conduct a reasoned analysis, citing MedPAC's June 2008 report which stated that CMS needs substantially more data. One commenter recommended that CMS consider a 1-year freeze on any reductions and wait for more available data. Another commenter recommended that CMS delay until we had gathered data on the potential impact on hospice operations and quality of care of eliminating the BNAF. A commenter stated that there was not enough time to prepare for a change due to a BNAF phase-out. Other commenters asked CMS to wait until MedPAC completes its hospice analysis and CMS updates its hospice payment system, before eliminating the BNAF. Several other commenters recommended phasing out the BNAF more gradually. Some recommended a gradual phase-out that would help minimize the economic impact of such a change, particularly at a time of rising gas prices. One commenter recommended a 5-year phase-out, and another recommended a 7-year phase-out, with a 25 percent reduction in FY 2009, a 15 percent reduction in FY 2010, and a 10 percent reduction through FY 2016. Another commenter recommended that implementing the BNAF reduction on a more gradual schedule would allow hospices time to evaluate the impact of the reduction of the new CoPs, hospital versus hospice costs, and of the June 2008 MedPAC report. Even with the proposed schedule for reducing, and ultimately eliminating the BNAF, commenters suggested that by CMS not providing impacts for the 2nd and 3rd year of reducing the BNAF, the industry was not afforded information as to the impact on hospices beyond the

[Page Number 46480]

proposed first year's reduction of the BNAF.

Response: We appreciate the commenters' input. However, we continue to believe that our phase-out approach as described in the proposed rule is appropriate. One reason we decided to implement a phased-out approach to reducing the BNAF was to ensure we would have the ability to assess the impact of the BNAF reduction iteratively. We plan to monitor the impact of the BNAF reduction for unintended effects. The financial impact of the phase-out is very clear; in FY 2009 it is estimated to affect hospices by no more than negative 1.1 percent. We continue to believe that because the FY 2009 market basket update is 3.6 percent, and hospices will receive 2.5 percent more in payments in FY 2009 than they received in FY 2008, the FY 2009 reduction should not have an adverse impact on hospice operations or quality.

The hospice CoPs were developed with input from the industry. The CoPs have been in development for 9 years and have widespread support from the industry. Hospices have been expecting these since the Conditions of Participation proposed rule was published in May of 2005, providing time to plan and budget for these changes accordingly. Hospices were required to report certain visit data as of July 1, 2008. This requirement was implemented through CR 5567 after MedPAC, the General Accounting Office, and the Office of the Inspector General recommended that CMS gather data on Medicare hospice service utilization. While the most recent revision of CR 5567 occurred on April 29, 2008, the initial issuance of the CR occurred on July 20, 2007, providing ample time for hospices to plan and budget for these requirements. CR 5567 requires reporting of visit data that many hospice software programs already track. We do not feel it will cause undue burden, especially since we recently revised the CR 5567 requirements to suspend a subset of the reporting requirements to address an industry concern. Additionally, we believe the industry has ample time to prepare for the BNAF phase-out. The FY 2009 President's budget was published in February, 2008 and contained a provision to phase out the BNAF. In the interest of transparency, we made public our intent to propose the BNAF reduction via FY 2009 rulemaking shortly after publication of the budget, over 2 months in advance of our proposed rule publication.

We agree that MedPAC recommended that CMS collect additional data to better understand what services we are paying for, and for use in future payment refinements. We did not propose to reform the payment system, but simply to remove an outdated adjustment.

We acknowledge that the impacts reflected in this rule are for FY 2009 only. The purpose of this final rule, so far as impacts, is to show the estimated impacts on hospice providers for FY 2009. We have, in the proposed rule (72 FR 24005, 24006) and in a response to a previous comment, communicated that over the course of the three-year phase-out, the elimination of the BNAF will reduce payments by about 4 percent: We estimated a 1.1 percent reduction in FY 2009, an additional 2 percent reduction for a cumulative reduction of 3.1 percent in FY 2010, and an additional 1 percent reduction for a cumulative reduction of 4.1 percent in FY 2011. However those reductions do not include 3 years of market basket updates for FY 2009, FY 2010, and FY 2011. Therefore, assuming market basket updates' inclusion in FY 2010 and FY 2011, hospices will still have a net gain in payments over the 3 years. While we do not know what the market basket updates will be for FY 2010 and FY 2011, hospices received market basket updates ranging from 3.3 percent to 3.7 percent from FY 2005 to FY 2008. The market basket update for FY 2009 is 3.6 percent. We will provide similar impacts for FY 2010 and FY 2011 in future rulemaking.

For all of these reasons, we do not feel that there is a reason to delay the phase-out of an outdated adjustment. While we believe that a 3-year phase- out is fair and appropriate, we will monitor the effects of the phase-out as it occurs.

Comment: A commenter wrote that hospitals and home health agencies receive the BNAF. The commenter said that since hospices compete with home health agencies and hospitals for staff, phasing out the BNAF for hospice creates an uneven playing field for recruiting and retaining staff. A few commenters stated that CMS should not justify removing the BNAF so that the hospice and home health wage indices were consistent. One commenter stated that the new CoPs recognized the differences between hospice and home care patients, noting that