Nearly Half Of All Medicare Hospice Enrollees Received Care From Agencies Owned By Regional Or National Chains

Abstract

To date, analyses of ownership in the US hospice sector have focused on the growth of for-profit hospice and on aggregate differences in patient populations and service use patterns between for-profit and not-for-profit agencies. Although such comparisons are useful, they do not offer insights about the types of organizations that comprise the hospice sector, including the emergence of multi-agency chains. Using Medicare Cost Reports from 2000 to 2011, we track the evolution of the US hospice industry, not only to describe the market's composition by profit status but also to provide new information about the roles of regional and national chains. Almost half of all Medicare hospice enrollees received these services from a multi-agency chain in 2011. Although a handful of companies play a prominent role, the presence of smaller for-profit and not-for-profit hospice chains also has grown in recent years. By focusing on the role of the diverse organizations that provide hospice care, our analyses can help inform efforts to monitor and assure quality of care, to assess payment adequacy and options for reform, and to facilitate greater transparency and accountability within the hospice marketplace.

The Medicare hospice benefit has changed considerably over its 30-year history, evolving from a relatively small program centered on providing end-of-life care to cancer patients at home to one serving a much broader population living in community-based and institutional settings. The pace of change in Medicare's hospice program has been particularly fast over the last decade.(1) Between 2000 and 2012, the percent of Medicare decedents using hospice increased from 23% to 47%, with hospice spending growing from $2.9 to $15.1 billion.(2) Over this same time period, hospice became a predominantly for-profit (FP) enterprise, with the majority of agencies now operating on a FP basis (59% in 2012, compared to 30% in 2000).(2)

As hospice has grown, policymakers and others have paid particular attention to the emergence of the FP hospice sector, raising questions about the extent to which agencies (i.e., the individual units serving particular geographic areas) and the companies that own them aggressively target more profitable patients. The Medicare Payment Advisory Commission (MedPAC) and others have chronicled the increased role of FP agencies within the hospice provider market, noting that hospice expansion has been fueled almost entirely by the growth in FP agencies. Between 2000 and 2012, the number of FP agencies grew three-fold (from 672 to 2196 agencies), while the number of not-for-profit (NFP) and government agencies declined (from 1324 to 1313 and from 257 to 210 for NFP and government agencies, respectively).(2)

An emerging research literature has explored the role of ownership status in hospice service provision, primarily focusing on aggregate differences between FP and NFP agencies. For several years, MedPAC has documented differences between FP and NFP hospices, finding that FP agencies generally have longer mean lengths of stay, lower mean hospice costs per day, higher Medicare margins, and a higher proportion of institutionalized beneficiaries.(2-4) Other research has found that FPs typically have different referral networks (e.g., a greater emphasis on nursing home referrals) and patient populations (e.g., a higher proportion of non-cancer enrollees),(5-9) less skilled staffing and a narrower range of services,(6, 7, 10) more restrictive enrollment practices,(11) and fewer community benefits such as charity care and training sites.(8) To understand the mechanisms driving these differences, studies have examined traits such as agency size and tenure in the marketplace, finding that smaller and newer FP agencies have higher rates of live discharge and longer lengths of stay than other agencies.(12-14)

Although it is a useful starting point to examine differences by hospice agencies' profit status, the dichotomous distinction of FP/NFP is overly broad and offers few insights about the types of organizations that comprise the hospice sector, including the emergence of multi-agency chains. Even analyses that more closely examine agency size and tenure in the marketplace fail to consider the possible influence of parent companies that own many of these agencies. It is useful, for example, to explore the extent to which regional or national chains have entered the hospice provider market, structural characteristics of organizations delivering hospice, and the concentration of the hospice provider market among larger organizations. Entities such as MedPAC and the Office of the Inspector General have expressed interest in pursuing more detailed explorations of hospice ownership to inform payment and oversight policy.(3, 15) More generally, consumers, policymakers, and providers could utilize such information to consider organizational and company-level factors that might affect quality of care, to develop regulatory strategies and ensure sufficient oversight, and to guide quality improvement efforts.

Outside of media reports scrutinizing particular companies for aggressive enrollment practices and inadequate care,(16, 17) little research has been done on the role of hospice chains specifically. A more substantial literature exists in other areas of health care, most prominently the nursing home sector where the majority of facilities are owned by multi-facility chains.(18) Many studies have compared FP to NFP nursing homes, generally finding that FP facilities provide lower quality of care on average,(19) and a separate line of research has explored the role and performance of chain organizations. Although chains have the potential to achieve greater efficiency and quality through mechanisms like increased standardization, knowledge transfer, and enhanced information systems,(20-23) strong concerns have been raised historically about poor track records at particular companies and possible negative implications for nursing home quality, especially at the intersection of chain ownership and FP status where instances of inadequate care practices have been well documented.(24-26) In response to recent allegations of poor quality care and insufficient staffing, a newer point of emphasis in nursing home ownership studies has investigated the role and impact of private investment and corporate structure on quality of care.(27-29)

In the analyses below, we track the evolution of the hospice industry, not only in the extent to which FP and NFP agencies characterize the sector but also concerning the roles of regional and national chains and how these entities have grown in prominence. The goal of these descriptive analyses is to deepen our understanding of the hospice provider market and how it has changed over time. Identifying agencies within specific organizations across the hospice market is a necessary step to investigate whether particular FP and NFP chains have distinct structural traits and care practices that might inform policymakers, especially as they consider policies to monitor and assure quality of care and to spur greater transparency and accountability in the hospice sector more generally.

Data and Methods

The primary data for our analyses are Medicare Cost Reports. These reports must be submitted annually by all Medicare-certified hospice agencies and are publicly available from the Centers for Medicare and Medicaid Services (CMS). In addition to freestanding hospice agencies, we include hospice agencies that are listed as sub-providers in cost reports for hospitals, home health agencies, and skilled nursing facilities.

We use Medicare Cost Reports data from 2000-2011 to determine chain status (yes/no) and affiliation, geographic location, and enrollment for each hospice agency in the United States. Medicare cost reports do not include reliable variables identifying the parent organization or chain to which an agency belongs. Consequently, linking individual hospice agencies to their parent companies was done iteratively using information from various data fields such as the hospice name and address. Following previous work in the nursing home sector, (29-31) these activities involved extensive data cleaning and, where possible, referenced other available information as a check (e.g., provider websites). More detailed information about this coding effort is available in the Technical Appendix.(32)

First, we present information about the number of Medicare hospice agencies and enrollees from 2000 to 2011 by the following ownership categories: NFP non-chain, NFP chain, FP non-chain, FP chain, and government (Exhibits 1 and 2). Second, we present information about FP and NFP hospice chains for 2000 and 2011, including the total number of chains in the market, their size in agencies and enrollees, their geographic reach, the role of large chain organizations, and agency type (i.e., freestanding and hospital-, home health agency-, and skilled nursing facility-based) (Exhibit 3). As a frame of reference, we also include information about non-chain agencies, where the information was applicable to individual agencies. Third, we describe information about the five largest FP and NFP hospice chains in terms of Medicare enrollees served, including information about each organization's number of agencies, enrollees, and states of operation, and about the percent of enrollees served by different agency types (Appendix Exhibit 3 (32)). Finally, we present maps displaying the percent of Medicare hospice enrollees receiving care from chain providers in 2000 and 2011, by state (Exhibits 4 and 5).

Exhibit 1. Number of Hospice Agencies by Ownership, 2000-2011.

Source: Medicare Cost Reports

Exhibit 2. Number of Medicare Hospice Enrollees by Ownership, 2000-2011.

Source: Medicare Cost Reports

Exhibit 3. Not-for-Profit (NFP) and For-Profit (FP) Chain Hospice Providers, 2000 and 2011.

	2000		2011
	NFP	FP	NFP	FP
Size
Number of Chains	37	42	92	154
Avg. Chain Size (agencies)	2.6	5.6	3.3	6.9
Avg. Chain Size (Medicare hospice enrollees)	839	2,262	1,699	2,714
Median Chain Size (Medicare hospice enrollees)	432	910	994	722
Avg. Chain Agency Size (Medicare hospice enrollees/agency)	320	403	517	391
Geographic reach and market concentration
Avg. Number of States in Chain	1.2	2.7	1.6	3.0
Chains Operating in 3 or More States (%)	2.7%	23.8%	10.9%	26.0%
Agencies in Top 5 Chains	18	90	22	283
Beneficiaries in Top 5 Chains	14,802	60,760	41,381	189,664
Agencies in Top 10 Chains	36	147	57	465
Beneficiaries in Top 10 Chains	21,662	72,425	62,005	247,869
Agency type
Freestanding Agencies (%)	47.4%	95.8%	40.4%	91.0%
Hospital-Based Agencies (%)	42.3%	0.0%	39.1%	0.5%
Home Health Agency-Based Agencies (%)	10.3%	4.2%	20.5%	8.5%
Skilled Nursing Facility-Based Agencies (%)	0.0%	0.0%	1.3%	0.0%
NON-CHAIN AGENCIES (for reference)
Avg. Stand Alone Agency Size (Medicare hospice enrollees/agency)	289	212	542	222
Agency type
Freestanding Agencies (%)	37.2%	66.4%	48.8%	84.1%
Hospital-Based Agencies (%)	14.2%	23.7%	17.6%	13.9%
Home Health Agency-Based Agencies (%)	47.8%	8.4%	31.8%	1.3%
Skilled Nursing Facility-Based Agencies (%)	0.8%	1.5%	1.8%	0.8%

Source: Medicare Cost Reports

Exhibit 4. Percent of Medicare Hospice Enrollees in Chain Agencies - 2000.

Source: Medicare Cost Reports

Exhibit 5. Percent of Medicare Hospice Enrollees in Chain Agencies - 2011.

Source: Medicare Cost Reports

Limitations

Our results are primarily descriptive, but it is important to note key limitations. The identification and coding of hospice chains using available data is imprecise. We rely primarily on text fields and commonalities in provider names in the Medicare Cost Reports for our initial chain coding, supplementing with other information from provider websites and news articles where possible. Because of this approach, our count of hospice chains is likely an underestimate. For example, if a private investment company acquires ownership in multiple agencies and continues to operate them under their previous brands, we could misidentify these providers as non-chain agencies. Although our findings describe the shift in the hospice provider market in greater detail than has been available previously, the implications of these changes for service use and quality of care are unclear. Beyond refining the characterization of hospice providers and their structural differences, future research should integrate information about hospice ownership with other data on service use, quality of care, and broader market trends.

Results

The hospice provider market

Between 2000 and 2011, the ownership composition of US hospice agencies changed considerably (Exhibit 1). Over this time frame, the number of non-chain NFP hospice agencies was relatively stable (893 and 865 agencies in 2000 and 2011, respectively), but their share of the hospice provider market fell from 50% to 26% of all agencies (see Appendix Exhibit 1 for further detail(32)). In contrast, the number and share of FP hospice agencies increased considerably, especially among chain providers. The number of non-chain FP agencies grew from 379 to 917 agencies between 2000 and 2011 (21% and 28% of agencies in the hospice market, respectively), while the number of FP chain providers grew from 236 to 1070 agencies (13% and 33% of the market respectively). Although they represent a relatively modest portion of the market, the number of NFP chain agencies grew from 97 to 302 agencies over the study period (5% and 9% of the market, respectively). The number of government-owned agencies remained relatively stable (192 and 185 agencies in 2000 and 2011, respectively), although their share of the provider market diminished from 11% to 6% of all hospice agencies.

The number of Medicare beneficiaries receiving hospice services across the different ownership categories between 2000 and 2011 tells a similar story to the agency trends (Exhibit 2). Except for government-owned hospice agencies, all ownership categories were buoyed by a substantial increase in the number of enrollees using hospice over this timeframe. FP and NFP chain providers demonstrated the largest proportional increases in the relative share of enrollees served. FP chain providers went from serving 20% to 33% of Medicare hospice enrollees between 2000 and 2011 (95,023 and 418,022 enrollees, respectively), while NFP chain providers increased their share of hospice enrollees from 6% to 12% (31,038 to 156,265 enrollees, respectively). Non-chain NFP hospice agencies still served the most enrollees in 2011 (469,232 enrollees or 37%), but the share these agencies served was down from 53% (258,265 enrollees) in 2000 (see Appendix Exhibit 2 for further detail(32)).

Reflecting the growth of FP and NFP chain hospices, the role of multi-agency chains in the US hospice market increased substantially over the study period. From 2000 to 2011, the share of chain agencies in the market increased from 19% to 41%, and the share of enrollees these agencies served increased from 26% to 45%. Around three-quarters of the hospice chain market in agencies and enrollees is FP, a share that was relatively stable over the study period.

Characteristics of FP and NFP chains

As shown in Exhibit 3, the number of FP chain organizations in the hospice market increased from 42 to 154 between 2000 and 2011; for NFP chains, these numbers were 37 and 92 companies, respectively. FP and NFP hospice chains differed considerably in their features, and these differences were relatively consistent across study years. Focusing on 2011, FP chains were larger on average, with 6.9 agencies and 2714 enrollees per chain compared to 3.3 agencies and 1699 enrollees for NFP chains. These aggregate differences were driven, in part, by the role of larger FP chains in the market. While the largest 5 FP chains served 189,664 enrollees in 2011 (15% of the overall hospice market and 31% of the FP hospice market), the 5 largest NFP chains served 41,381 enrollees (3% of the overall hospice market and 7% of the NFP market). FP hospice chains operated in almost twice as many states, on average, as NFP chains (3.0 states for FP chains and 1.6 for NFP chains, respectively). Further, 26% of FP chains operated in 3 or more states, compared to 11% of NFP chains. NFP chain agencies were similar in size to NFP non-chain agencies (517 enrollees per NFP chain agency versus 542 enrollees per NFP non-chain agency), whereas FP chain agencies were substantially larger in size compared to FP non-chain agencies (391 enrollees per FP chain agency versus 222 enrollees per FP non-chain agency). Almost all FP chain hospice providers (91%) operated as freestanding agencies in 2011, with only small percentages affiliated with home health agencies (8.5%) and hospitals (0.5%). Most FP non-chain agencies also were freestanding (84.1%) but the proportion of agencies that were hospital-based (13.9%) and home health-based (1.3%) differed. In contrast, 40% of NFP chain hospices were freestanding, with 21% and 39% affiliated with home health agencies and hospitals, respectively. A similar proportion of NFP non-chain agencies were freestanding (48.8%), with somewhat different proportions in the hospital-based (17.6%) and home health agency-based (31.8%) categories.

As when comparing FP and NFP chains generally, several differences emerge when comparing the largest FP and NFP chains (Appendix Exhibit 3 (32)). The FP chains are relatively large compared to the NFP chains, with the two largest FP chains – Vitas and Gentiva – each serving more enrollees than do the largest 5 NFP chains combined. In fact, only one NFP chain – Hospice of the Valley – is among the 10 largest hospice chains overall. Relative to the largest NFP chains, the large FPs had a much larger number of agencies (57 vs. 4, on average) and operated in a much larger number of states (21 vs. 2, on average). With a few exceptions, the largest FP and NFP chains were predominantly comprised of freestanding agencies.

Geographic distribution of hospice chains

Exhibits 4 and 5 show the state-by-state percentages of hospice enrollees in all chain agencies, in 2000 and 2011. The maps convey substantial growth in these percentages over time and across states. In 2000, 11 states and the District of Columbia did not have any enrollees being served by chain hospices (FP or NFP), compared with only 1 state by 2011. At the other end of the distribution, 13 states had more than 30% of all hospice enrollees being served by chain providers in 2000, compared with 36 such states by 2011. The 5 states with the highest percent of hospice enrollees receiving services from chain providers in 2011 were Delaware (98%), Alaska (92%), Virginia (75%), South Dakota (72%), and Louisiana (64%).

If one narrows the focus to be on FP chain hospices only, a similar story emerges, albeit a more geographically concentrated one (see Appendix Exhibits 4 and 5 for more detail(32)). In 2000, 18 states and the District of Columbia did not have any beneficiaries receiving hospice services from FP chain agencies, compared with only 3 states by 2011. Only 5 states had more than 30% of hospice enrollees receiving services from FP chains in 2000, compared with 25 states by 2011. Although the share of hospice enrollees receiving services from FP chain providers generally increased nationwide, these providers played an especially prominent role in the South. The 5 states with the highest percent of hospice enrollees receiving services from FP chain providers in 2011 were Louisiana (61%), Delaware (56%), Alabama (56%), Georgia (55%), and Texas (55%).

Discussion

Our findings confirm and extend previous analyses describing the increased role of FP companies in the US hospice sector. Not only has hospice provision become increasingly dominated by FP companies over the last decade, this growth has been driven mostly by the increased presence of FP chains. Although a handful of large hospice companies account for almost half of the FP chain market, the presence of smaller FP and NFP hospice chains also has become more substantial in recent years. In total, almost half of all Medicare hospice enrollees received these services from either a FP or NFP organization that owns multiple hospice agencies in 2011.

Closer examination reveals insights into key differences between the FP and NFP chain markets. FP hospice chains tend to be bigger, although this reflects the presence of a relatively small number of large FP companies, such as Vitas, Gentiva, and Heartland. Interestingly, FP chains (and FP agencies generally) are comprised mostly of freestanding agencies, whereas hospital-and home health-based hospice agencies play a larger role in the NFP sector. Not only do these types of providers have different organizational and cost structures, they also likely differ in their patient referral networks, something suggested by prior studies.(6, 9) We also find substantial geographic variation in the distribution of hospice providers, with FP chain agencies having a relatively greater concentration in the South, a trait that is shared with the nursing home sector.(18)

As detailed above, analyses of ownership in the hospice sector have focused primarily on the growth of FP hospice and on aggregate differences between FP and NFP agencies. These findings have highlighted substantial differences in practice and raised a number of concerns. Despite documentation of these differences, the policy implications of these findings have been unclear. Absent restrictions on FP entry into the hospice market (which seems unlikely) or greater regulatory scrutiny across all FP providers (which seems poorly targeted), aggregate comparisons of hospice based on profit status alone offer a provocative, but somewhat diffuse window into how organizations differ in their approach to delivering hospice care. By providing a detailed examination of the types of organizations that comprise the hospice provider community, our analyses offer a foundation from which to examine their role going forward. Although there might be other uses of such data, we envision their primary usefulness to be in the areas of regulatory oversight, quality measurement and reporting, and payment reform.

Regulatory oversight

Although it has garnered periodic attention over the last two decades, oversight of the US hospice industry has attracted increased scrutiny in recent years as a result of federal investigations and media reports identifying concerns about aggressive recruitment practices, enrollment of patients with uncertain eligibility, allegations of inadequate care, and questionable billing practices.(15, 17, 33-37) In fact, President Obama recently signed into law bi-partisan legislation (the IMPACT Act of 2014) that included provisions to bolster hospice oversight.(38) Considerations of ownership often have featured prominently in discussions of oversight, whether concerning the practices of particular companies or the financial relationships that hospices have with other providers, such as nursing homes. For example, two of the largest chains we identify (Heartland and Aseracare) are owned by entities that also operate two of the largest nursing home chains in the country (HCR ManorCare and Golden Living). Without reliable information that describes and connects the practices of agencies within particular companies, such endeavors are difficult to pursue on more than an ad hoc, or agency-by-agency basis. Similarly, tracking ownership and related changes across the provider market can inform even routine oversight functions such as provider certification, compliance monitoring, and the investigation of complaints. For instance, regulatory entities might consider past performance of a particular company's agencies as they evaluate licensure renewals, certificate of need applications, and the approval of mergers and acquisitions. Although these efforts typically have focused on the individual provider (i.e., the agency), mechanisms such as the corporate integrity agreements used by the HHS Office of the Inspector General offer an example where oversight has centered at the ownership-level to work with particular companies to monitor and correct identified deficiencies.(39)

Quality measurement and reporting

As required by Section 3004 of the Affordable Care Act, all Medicare-certified hospice agencies must report a set of hospice quality measures starting in fiscal year 2014 or face payment reductions (a 2 percentage point reduction in the market basket update for that year).(40) These measures include information about pain screening and assessment; dyspnea screening and assessment; the percent of opioid users who are offered or prescribed a bowel regimen; and documented discussions about treatment preferences and patient's beliefs and values. The ACA also stipulates that hospice quality measures ultimately will be publicly reported (the timetable has yet to be announced), as CMS already does with quality measures for other types of Medicare providers. Such measures presumably will be reported at the level of the agency, but purchasers, referring providers, and even consumers could find information at the level of the parent organization useful. Even though it is the local agency that provides the care received by enrollees, organizational systems, strategy, and culture presumably play a role in shaping it. Developing a company-level measurement focus seems especially relevant as integrated provider networks seek to build contractual relationships across geographic markets and as policymakers seek to evaluate their impact.

Payment reform

In recent years, Medpac and others have proposed hospice payment reforms, such as paying higher amounts at the beginning and end of hospice stays (and lower amounts in the middle), including hospice services in Medicare Advantage benefits, and modifying the way hospice is reimbursed for nursing home residents.(2) The Affordable Care Act also calls on the Secretary of Health and Human Services to implement revisions to hospice payment methodology no earlier than October 1, 2013. Although the development of reform proposals does not depend on such information, analyses that consider more detailed ownership information can offer insights into the potential impact of policy changes and offer an early window into practice changes that might result. For instance, MedPAC assesses profit margins in evaluating payment adequacy across different segments of the hospice industry. Although these analyses include comparisons of FP to NFP providers, they have not focused on the financial performance of hospice chains. Similarly, examining service patterns and financial performance of companies with different strategic emphases (e.g., providing hospice to nursing home residents) could offer insights that help assess the impact of reforms that might affect segments of the hospice sector differentially. Finally, company-level analyses could offer a useful perspective on hospice providers' role in integrated payment reforms such as the Accountable Care Organization demonstrations, bundled payment initiatives, and state demonstrations to integrate care for dually eligible beneficiaries.

Conclusions

In sum, major changes have occurred in the hospice provider market over the last decade. Although analyses to date have described part of this shift in their exploration of differences between FP and NFP hospices, researchers, policymakers, and regulators should pursue a deeper understanding of the types of organizations providing hospice care in the US. In particular, having greater clarity about the changing hospice marketplace and the role of particular companies within it can help inform efforts to monitor and assure quality of care, to assess the adequacy of payment rates and the potential impact of related reforms, and – more broadly – to facilitate greater transparency and accountability of the diverse set of companies providing end-of-life care to Medicare beneficiaries.

Technical Appendix: Hospice Chain Matching Protocol

1. From the Cost Report data, the following variables are used for matching purposes:

   Variable name	Explanation
   Chain	Chain ID from cost report data
   ID	OSCAR ID number
   Bgn Dt	Cost report begin date
   End Dt	Cost report end date
   Owner	Non-profit/For-profit status
   Medicare	Medicare Enrollees (unduplicated Census Count)
   Hospice Name
   Address
   City
   State
   Zip
   County

2. From the Hospice Name field, we matched phrases containing hospice names to construct chains of hospices that may not have been captured by the “Chain” field

3. First we pre-processed some of the names by deleting punctuation (periods, commas, dashes, slashes) and the following words: “INC” and “LLC”

   1. Any double spaces in the names (caused by these deletions or errors) were then converted to single spaces.
4. Some common variations on larger hospice chains were corrected:

   Variation	Parent Name
   VISTA CARE	VISTACARE
   VISTA HOSPICE	VISTACARE
   SOUTHERN CARE	SOUTHERNCARE
   COMMUNITY HOSPICES OF AMERICA	COMMUNITY HOSPICE OF AMERICA
   FIRST HEALTH HOSPICE	FIRSTHEALTH HOSPICE
   HPC	HOSPICE PREFERRED CHOICE

5. Using the name of each hospice's parent organization from the previous year, we checked for any changes in ownership status by comparing the two names.

   1. If a given hospice was part of a chain in the previous year, we searched in the current year's name for that parent name.
   2. If the previous parent name was in the current hospice name, we considered this hospice still in the previous parent company.
   3. Otherwise, the hospice name was checked manually to see if the name change was a typo, or actual ownership changes occurred.
   4. Parent organizations that were left with only one agency in the next year were now no longer considered chains.

6. If duplicate cost reports were present for any year, the most recent one was used to determine assignment for that year.

7. To check if any chains from the previous year added new hospices, we searched for each parent name in the unmatched hospices from the current year.

   1. These potential matches were manually checked to confirm that were part of the chain to which they matched, based on location and non-profit status.

8. Each unique hospice name is now parsed into single words and phrases. Each of these is searched for in the names of all other hospices for that year.

9. For every match to one or more hospices, a potential chain is formed with the matching phrase

   1. The OSCAR IDs, hospice name, state, address and profit-status are compiled for each potential match so they can be compared manually.
   2. I then verified the potential chains by checking if their addresses matched or if they operated in the same state/region and shared the same profit status.

      1. Potential chains with implausibly large numbers of matches (i.e. “Hospice of”) were ignored for being too large and unlikely. More specific matches (i.e. “Hospice of Northern Oklahoma”) were investigated under the more specific potential chain name.
      2. Questionable matches were furthered verified by searching the hospices' names and state(s) in Google. For instance if two hospices with the same name operated in different states, I Googled the hospice name along with the states. If a company website confirmed that one company operated in both states then I considered this a match, otherwise these hospices were not considered chains.

10. For subproviders, we initially implemented steps 8 and 9 using the overarching Hospital, Home Health Agency or SNF information, if available in the data.

11. The name of the parent to which each hospice belongs is now merged back into the yearly cost report files.

Appendix Exhibit 1. Share of Hospice Agencies by Ownership, 2000-2011.

Source: Medicare Cost Reports

Appendix Exhibit 2. Share of Medicare Hospice Enrollees by Ownership, 2000-2011.

Source: Medicare Cost Reports

Appendix Exhibit 4. Percent of Medicare Hospice Enrollees in For-Profit Chain Agencies – 2000.

Source: Medicare Cost Reports

Appendix Exhibit 5. Percent of Medicare Hospice Enrollees in For-Proflt Chain Agencies – 2011.

Source: Medicare Cost Reports

Appendix Exhibit 3. Five Largest For-Profit and Not-for-Profit Hospice Chains, 2011.

For-Profit Chains
Parent Company	Agencies	Medicare Hospice Enrollees	FP Chain Rank	Overall Rank (size)	States in Operation	Freestanding Agencies	HHA-Based Agencies	Hospital-Based Agencies
Vitas	37	63,479	1	1	17	100%	0%	0%
Gentiva	103	58,911	2	2	30	100%	0%	0%
Heartland	74	33,490	3	3	25	88%	12%	0%
Amedisys	53	20,916	4	4	19	85%	15%	0%
Seasons Hospice	16	12,868	5	6	14	100%	0%	0%
Not-for-Profit Chains
Parent Company	Agencies	Medicare Hospice Enrollees	NFP Chain Rank	Overall Rank (size)	States in Operation	Freestanding Agencies	HHA-Based Agencies	Hospital-Based Agencies
Hospice of the Valley	4	13,795	1	4	1	75%	25%	0%
Chapters	2	9,231	2	12	1	100%	0%	0%
Providence Hospice	7	8,267	3	15	3	29%	43%	29%
Covenant Hospice	2	5,304	4	19	2	100%	0%	0%
Sutter	7	4,784	5	22	1	71%	14%	14%

^*Source: Medicare Cost Reports.

None of the listed chains have SNF-based agencies