STATE OF FLORIDA
DIVISION OF ADMINISTRATIVE HEARINGS
SELECT SPECIALTY HOSPITAL- MARION, INC.,
Petitioner,
vs.
AGENCY FOR HEALTH CARE ADMINISTRATION,
Respondent.
)
)
)
)
)
) Case No. 03-2483CON
)
)
)
)
)
)
KINDRED HOSPITALS EAST, LLC,
Petitioner,
vs.
AGENCY FOR HEALTH CARE ADMINISTRATION and SELECT SPECIALTY HOSPITAL-MARION, INC.,
Respondents.
)
)
)
)
) Case No. 03-2810CON
)
)
)
)
)
)
)
)
RECOMMENDED ORDER
Pursuant to notice, a final hearing was held in these cases on April 20-22, 2004, in Tallahassee, Florida, before T. Kent Wetherell, II, the designated Administrative Law Judge of the Division of Administrative Hearings.
APPEARANCES
For Select Specialty Hospital-Marion, Inc. (Select-Marion): Mark A. Emanuele, Esquire
Deborah S. Platz, Esquire
Panza, Maurer, & Maynard, P.A.
Bank of America Building, Third Floor 3600 North Federal Highway
Fort Lauderdale, Florida 33308-6225 For Kindred Hospital East, LLC (Kindred):
Patricia A. Renovich, Esquire
M. Christopher Bryant, Esquire Oertel, Fernandez & Cole, P.A.
301 South Bronough Street, Fifth Floor Post Office Box 1110
Tallahassee, Florida 32302-1110
For Agency for Health Care Administration (Agency): Richard J. Saliba, Esquire
Agency for Health Care Administration Fort Knox Building III, Mail Station 3 2727 Mahan Drive
Tallahassee, Florida 32308 STATEMENT OF THE ISSUE
The issue is whether the Agency should approve the Certificate of Need applications filed by Select-Marion and/or Kindred, each of which proposes to establish a new long-term care hospital in Marion County.
PRELIMINARY STATEMENT
Select-Marion and Kindred filed competing Certificate of Need (CON) applications in the first “hospital beds and facilities” batching cycle of 2003. The application filed by Select-Marion, CON 9647, proposed to establish a freestanding
bed long-term care hospital (LTCH) in Marion County. The application filed by Kindred, CON 9648, proposed to establish a 31-bed LTCH within an existing hospital in Marion County.
After a comparative review of the applications, the Agency published notice of its intent to deny both applications in the Florida Administrative Weekly. Thereafter, Select-Marion and Kindred each timely filed petitions with the Agency in which they raised disputed issues of material fact and requested hearings on the Agency’s decision to deny their respective applications.
The Agency referred the petitions to the Division of Administrative Hearings (Division) for the assignment of an administrative law judge to conduct the hearings requested by Select-Marion and Kindred. Select-Marion's petition was designated DOAH Case No. 03-2483CON, and Kindred's petition was designated DOAH Case No. 03-2810CON. The cases were consolidated by Order dated August 13, 2003.
The final hearing was initially scheduled to begin on November 17, 2003, but it was continued on three occasions at the parties’ request. The hearing ultimately commenced on April 20, 2004.
A Notice of Change of Agency Position was filed on
March 15, 2004. The Notice advised the parties that the Agency “now supports [the] issuance of CON No. 9648 to Kindred,” and it
reaffirmed the Agency’s position against the approval of Select- Marion’s application.
At the hearing, Select-Marion presented the testimony of Marsha Webb-Medlin, who was accepted as an expert in nursing; Patricia Greenberg, who was accepted as an expert in health care planning and health care finance; and the deposition testimony of Theresa Hunkins (Exhibit S-8), John Grant (Exhibit S-9), Jeffrey Gregg (Exhibit S-10), and John Caron1 (Exhibit S-11).
Select-Marion’s Exhibits S-1 through S-12 were received into evidence.
Kindred presented the testimony of Carol Cregan, who was accepted as an expert in health care business development;
Mr. Grant; Clarence Wurdock, who was accepted as an expert in health care planning; James Novak, who was accepted as an expert in health care administration and LTCH administration;
Mr. Gregg, who was accepted as an expert in health care planning and CON review; and the deposition testimony of Dr. Sean Muldoon2 (Exhibit K-2) and Paul Clark (Exhibit K-3). Kindred’s Exhibits K-1 through K-9 were received into evidence.
An aerial photograph of Munroe Regional Medical Center and the surrounding area was attached to Mr. Clark’s deposition.
There was no reference to that photograph in Mr. Clark’s deposition, so there was no basis for the undersigned to receive the photograph as part of Mr. Clark’s deposition; however,
Ms. Cregan’s hearing testimony provided an adequate foundation for the admission of the photograph. See Transcript, at 265-67. Therefore, the photograph has been separated from Mr. Clark’s deposition, and it is hereby received into evidence as Exhibit
K-10.
The Agency adopted Mr. Gregg’s testimony as its case-in- chief, and did not present the testimony of any other witnesses. The Agency’s Exhibit AHCA-1 was received into evidence.
The five-volume Transcript of the hearing was filed on
May 17, 2004. The parties requested and were given 20 days from that date to file their proposed recommended orders (PROs). The deadline was subsequently extended to June 14, 2004, upon Select-Marion’s unopposed motion.
Kindred and the Agency timely filed a joint PRO on June 14, 2004. Select-Marion filed its PRO on June 15, 2004. Due consideration was given to the parties’ PROs in preparing this Recommended Order.
Through an Amended Request for Judicial Notice filed June 18, 2004, Select-Marion requested that the undersigned “take judicial notice of” the rules proposed by the United States Department of Health and Human Services, Centers for Medicare and Medicaid Services (CMS) and published in the
Federal Register on May 18, 2004, at pages 28196 through 28389. A copy of the proposed rules was filed by Select-Marion on
June 21, 2004. Kindred filed a response in opposition to the Amended Request for Judicial Notice on June 25, 2004, in which the Agency joined. Select-Marion filed a reply to Kindred's response on July 8, 2004. The Amended Request for Judicial Notice, which is treated as a motion for official recognition, is hereby denied.3 See Collier Medical Center, Inc. v. Dept. of
Health & Rehabilitative Servs., 462 So. 2d 83, 84 (Fla. 1st DCA 1985) (affirming hearing officer’s denial of a motion for official recognition of events occurring after the close of the evidentiary hearing).
FINDINGS OF FACT
Based upon the testimony and evidence received at the hearing and the parties' stipulations, the following findings are made:
Parties
Select-Marion is a recently-created subsidiary of Select Medical Corporation (Select). Select has its corporate headquarters in Pennsylvania.
Select operates 79 LTCHs in 24 states, including two in Florida. Select’s Florida LTCHs are located in Jacksonville and Miami. Another Select LTCH has been approved in Orlando, but that facility has not yet opened.
Kindred is a subsidiary of Kindred Healthcare, Inc., which has its corporate headquarters in Kentucky.
Kindred Healthcare operates 72 LTCHs in 26 states, including seven in Florida. Kindred operates 17 of those LTCHs, including six of the seven Florida LTCHs.
Kindred’s Florida LTCHs are located in Green Cove Springs, St. Petersburg, Hollywood, Ft. Lauderdale, Miami, and Tampa. Another subsidiary of Kindred Healthcare operates a second LTCH in Tampa.
Select and Kindred Healthcare have each been in the business of operating LTCHs since the 1980’s.
The Agency is the state agency responsible for administering the CON program and licensing LTCHs and other health care facilities.
Relevant Demographics of Marion County
The LTCHs proposed by Select-Marion and Kindred are to be located in Ocala, which is in Marion County.
Marion County is in District 3 for health planning purposes.
District 3 is a geographically large district. It includes 16 counties: Marion, Hamilton, Suwannee, Columbia, Lafayette, Dixie, Gilchrist, Union, Bradford, Alachua, Putnam, Levy, Citrus, Sumter, Lake, and Hernando.
Marion County is the most populous county in District
The 2003 population of Marion County was approximately
275,000, which represented approximately 20 percent of the total population of District 3.
Approximately 25 percent of Marion County’s 2003 population was in the 65 and older (65+) age cohort, and approximately 23 percent of the population of the respective primary service areas defined by Select-Marion and Kindred (see
Part D(3) below) was in that age cohort. Those percentages are higher than the statewide average of 17.5 percent in the 65+ age cohort.
The population of Marion County is projected to increase approximately 8.9 percent by January 2007, and during that same period, the population of the 65+ age cohort in the county is projected to increase approximately 11.3 percent. Similar relative growth rates are projected in the primary service area defined by Kindred through July 2005, which was to be the beginning of the facility's second year of operation.4
Marion County has a higher percentage of persons in the 65+ age cohort than do any of the counties that currently have an LTCH except for Sarasota County, which has 31.2 percent of its population in the 65+ age cohort. There are 40 approved but not yet operational LTCH beds in Sarasota County.
Hospitals in Marion County and the Surrounding Areas
There are three acute care hospitals in Marion County: Munroe Regional Medical Center (Munroe), Ocala Regional Medical
Center (Ocala Regional), and West Marion Community Hospital (West Marion). Each of those hospitals is in Ocala.
Munroe is a not-for-profit hospital. It is the oldest and largest acute care hospital in Marion County. Munroe currently has 323 beds, and it is in the midst of a $78 million expansion that will increase its capacity to 421 beds.
Ocala Regional and West Marion are for-profit hospitals owned by HCA, Inc. Ocala Regional, which has more than 200 beds, is located directly across the street from Munroe. West Marion, which opened in 2002 and has approximately
70 beds, is a satellite facility of Ocala Regional.
There is very little in-migration to the Marion County hospitals. For fiscal year 2002, approximately 83 percent of Munroe’s discharges were residents of Marion County. A similar percentage of the discharges from Ocala Regional were Marion County residents.
In addition to the three hospitals in Marion County, there are nine more acute care hospitals in District 3 within a 40-mile radius of the parties’ proposed Ocala LTCHs. Those facilities include Shands Hospital (Shands) at the University of Florida in Gainesville, Leesburg Regional Medical Center (Leesburg Regional), and The Villages Regional Hospital (Villages Regional).
Shands is located to the north of Marion County in Alachua County. Leesburg Regional and Villages Regional are located to the south of Marion County in Lake County and Sumter County, respectively.
The record does not reflect how many total acute care beds are in the hospitals outside of Marion County within the 40-mile radius, nor does it reflect whether any of those hospitals (other than Shands) are trauma centers or have other specialty programs that might impact, either positively or negatively, the potential patient pool for the parties’ proposed LTCHs.
The record does not reflect the number, type, quality of care, or utilization at the nursing homes and other traditional post-acute care facilities in Marion County or
District 3.
The Parties’ Proposed LTCHs
Select-Marion and Kindred each submitted timely letters of intent and CON applications for their proposed LTCHs in the first “hospital beds and facilities” batching cycle of 2003.
Kindred
Kindred’s proposed LTCH is a 31-bed “hospital-within- a-hospital” (HIH).
Kindred's proposed LTCH will be located on a wing of the fifth floor of Munroe’s existing hospital. The wing includes 11,606 square feet of contiguous space that Kindred is leasing from Munroe. Kindred is renovating the leased space to include 15 semi-private (i.e., double-occupancy) rooms and one “isolation room.”
The total project cost of Kindred’s proposed LTCH is approximately $1.4 million, or approximately $45,300 per bed. That cost will be funded by Kindred Healthcare’s “cash on hand.”
The lease agreement between Kindred and Monroe has been executed. The terms of the executed lease are slightly different than the terms of the draft lease included in Kindred’s CON application.
The lease is for an initial term of seven years with two five-year renewal periods upon the agreement of the parties. The draft lease included in the CON application contained a 10- year initial term with the two five-year renewals at Kindred’s option.
Because renewal of the lease will require the mutual agreement of Kindred and Munroe, there is no guarantee that the lease will be renewed at the end of the seven-year initial term. However, Kindred’s expert in LTCH business development testified that HIH leases typically are renewed.
There is also no guarantee that space will be made available within Munroe for future expansions of Kindred’s LTCH. However, Munroe’s Chief Operating Officer testified that Munroe would “entertain whatever discussions that would facilitate the expansion.”
Kindred's application states that its proposed LTCH will include the same types of services that are provided at Kindred’s other Florida LTCHs, including respiratory/life support, wound care, and neurological.
Kindred’s proposed LTCH will not include an intensive care unit (ICU) or a pediatric program.
Kindred will contract with Munroe for services such as laboratory, radiology, surgery/operating room, physical and speech therapy, and cardiology. An “ancillary services agreement” for these services has been executed by Kindred and Munroe.
A patient transfer agreement between Munroe and Kindred has been negotiated. The agreement cannot be formally executed until Kindred’s LTCH is approved.
Kindred projected that its proposed LTCH would begin operating in June 2004, which is one year after the “initial decision deadline date” for the applicable batching cycle.
The need projections in Kindred’s application were based upon July 2005 population figures, and the utilization and
financial projections in Kindred’s application were for the first two years of operation ending May 31, 2005 and 2006.
Select-Marion
Select-Marion’s proposed LTCH is a 60-bed freestanding facility to be located in Ocala.
The precise site of the facility has not yet been determined by Select-Marion. The application identifies a “top priority location” east of Interstate 75 on State Road 200 in the general vicinity of Central Florida Community College.
Select-Marion’s proposed LTCH will consist of 44,434 square feet of new construction. The bed complement at the facility will be 47 private (i.e., single-occupancy) rooms, five semi-private rooms, and three “isolation rooms.”
The total project cost of Select-Marion’s proposed LTCH is approximately $12.2 million, or approximately $204,100 per bed. That cost will be funded by Select from its cash flow from operations and/or through borrowings from Select’s line of credit with its bank.
The services at Select-Marion's proposed LTCH will include the same "core" services found at other Select LTCHs. Those services include the treatment of pulmonary and ventilator patients, neuro-trauma and stroke patients, medically complex patients, and wound care.
Select-Marion’s proposed LTCH will include a “step down” unit where ICU-level care will be provided. The facility will not include a pediatric program.
Select-Marion has not negotiated patient transfer agreements with any of the area hospitals.
Select-Marion’s application includes a letter from West Marion’s Administrator, which states that West Marion “anticipates implementing a transfer agreement with [Select- Marion’s] facility” in the event that the LTCH “achieves appropriate licensure and certification.” No weight is given to that letter because, as discussed below, West Marion did not have any potential LTCH patients and it is a satellite facility of Ocala Regional, which provided a letter of support to Kindred.
Select-Marion projected that its proposed LTCH would begin operating in June 2005, which is one year later than Kindred's proposed LTCH was projected to open. The delay is a result of Select-Marion's proposed LTCH being a new freestanding facility, which has a longer construction period than does the renovation of the existing hospital space for Kindred's HIH LTCH.
The need projections in Select-Marion’s application were focused on the facility’s third year of operation, which is the 12-month period ending June 2008. The utilization and
financial projections in Select-Marion’s application were for the first two years of operation ending May 31, 2006 and 2007.
Proposed Service Areas
Kindred’s application defines the primary service area (PSA) for its proposed LTCH as Marion, Alachua, Levy, Citrus, Sumter, and Lake Counties. The secondary service area is defined as the remaining 10 counties in District 3.
Select-Marion’s application defines the PSA for its proposed LTCH as Marion County and a 40-mile radius around the proposed facility. A secondary service area is not specifically defined, but the application states that Select-Marion’s proposed LTCH will serve the entire district “to the degree possible.”
The geographic scope of the PSA proposed by Select- Marion is similar to the geographic scope of the PSA proposed by Kindred; the 40-mile radius defined by Select-Marion includes all of Marion County and substantial portions of the remaining five counties included in Kindred’s PSA.
It is not reasonable to expect that a material number of the admissions to either of the proposed LTCHs would come from outside of the PSAs defined in the applications. On this issue, the undersigned accepts the expert testimony of Select- Marion’s health planner that a realistic PSA for an LTCH in Ocala would be Marion County, and based upon the limited
in-migration to Marion County hospitals and the testimony of Kindred's expert health planner that patients generally prefer to receive their long-term care close to their homes, the undersigned draws the inference and finds that realistic secondary service area for an LTCH in Ocala would be the remainder of the PSAs defined by the parties.
It is also not reasonable to expect that either of the proposed LTCHs will serve District 3 in its entirety despite the parties’ expressed intentions and desire to do so. Indeed, Kindred’s application states that it expects that residents of District 3 who are geographically closer to the existing LTCHs in Green Cove Springs and Tampa -– which would include a number of the counties in Kindred’s proposed secondary service area -- will continue to leave the district for LTCH services. Similar statements are included in Select-Marion’s application as part of its explanation of the proposed 40-mile PSA.
Letters of Support
Select-Marion’s application included only one letter of support, which was from West Marion. Although the letter represented that “[m]any of our long-stay acute patients are candidates for care in [an LTCH,]” the evidence establishes that West Marion did not have any long-stay patients that would be potential LTCH patients. Moreover, West Marion is a satellite facility of Ocala Regional, which is located across the street
from Munroe and provided a letter of support for Kindred’s proposed LTCH.
Kindred’s application included 13 letters of support, including letters from five hospitals in District 3 (i.e., Munroe, Ocala Regional, Leesburg Regional, Villages Regional, and Citrus Memorial Hospital), a nursing home in Ocala, a home health care organization affiliated with Munroe, a local physician, and several local and state politicians.
The support expressed in the letters included in Kindred’s application was general in nature. Indeed, many of the letters, including the letters from Ocala Regional and several of the other hospitals, appeared to be form letters.
Other than the letter from Monroe’s director of case management, the letters included in Kindred’s application did not attempt to quantify the number of LTCH patients that would likely be generated for Kindred’s proposed LTCH. Nor did the letters detail access or other problems for potential LTCH patients that would help to demonstrate need for an LTCH in Marion County.
General letters of support typically carry little weight with the Agency in its evaluation of need for a new health care facility. Indeed, the Bureau Chief of the Agency’s CON program, Jeffrey Gregg, testified that general letters of support are only significant to the Agency where one applicant
has received a number of such letters and the other applicant has not received any letters of support.5
Because of their general nature, the letters of support included in Kindred’s application do not in and of themselves demonstrate need for a proposed LTCH in Marion County, nor do they validate the numeric bed need projected in either of the applications. However, the volume of the letters included in Kindred’s application from hospitals and health care providers as compared to the single letter included in Select- Marion’s application demonstrates that there is greater support from the relevant community for Kindred’s proposed LTCH than there is for Select-Marion’s proposed LTCH.
Application Review and Preliminary Agency Action
The applications filed by Select-Marion (CON 9647) and Kindred (CON 9648) complied with all of the application content and submittal requirements in the governing statutes and the Agency’s rules. The required filing fees were paid, and the local health council submittal requirements were met.
The applications were comparatively reviewed by the Agency in accordance with the Agency’s rules and standard procedures.
On June 11, 2003, the Agency issued its State Agency Action Report (SAAR) based upon its comparative review of the applications. The SAAR recommended denial of both applications
based primarily on the Agency’s determination that neither applicant had adequately demonstrated need for its proposed LTCH.
The Agency’s decision to deny the applications was published in the Florida Administrative Weekly as required by statute and Agency rule, and Select-Marion and Kindred each timely requested a hearing on the Agency’s decision. This consolidated proceeding followed.
In March 2004, the Agency changed its position with respect to Kindred’s application and provided timely notice of that change in position to the parties. The Agency now supports approval of Kindred’s application, even though it continued to maintain at the hearing that need has not been adequately demonstrated by either applicant.
LTCHs
Generally
LTCHs are health care facilities that provide extended medical and rehabilitative care to patients with multiple, chronic, and/or clinically complex acute medical conditions. To qualify as an LTCH, the facility must serve a patient population whose average length of stay (ALOS) exceeds 25 days.
LTCHs fit into the continuum of care between traditional, “short-term” acute care hospitals and traditional post-acute care facilities such as nursing homes, skilled
nursing facilities (SNFs), skilled nursing units (SNUs), or comprehensive medical rehabilitation (CMR) facilities.
LTCHs are designed to serve patients that would otherwise have to be maintained in a traditional acute care hospital where the reimbursement rates may be insufficient to cover the costs associated with a lengthy stay, or be moved to a traditional post-acute care facility where the patient may not receive the level of curative care needed.
LTCH services are most highly utilized by persons in the 65+ age cohort because those persons are more likely to have complex and/or co-morbid medical conditions that require long- term acute care.
The typical LTCH patient is still in need of considerable acute care, but a traditional acute care hospital may no longer be the most appropriate or lowest cost setting for that care.
Most LTCH admissions are patients transferred directly from a traditional acute care hospital. It is not uncommon for an LTCH patient to be transferred on life support from a critical care unit or ICU after the patient has been diagnosed and stabilized.
Nursing homes, SNFs, SNUs, CMR facilities, and home health care are not appropriate for the typical LTCH patient because the patient's acuity level and medical/therapeutic needs
are higher than those generally treated in those settings. Indeed, unlike traditional post-acute care settings, which typically do not admit patients who still require acute care, the core patient-group served by LTCHs are patients who require considerable acute care through daily physician visits and intensive nursing care in excess of eight hours of direct patient care per day.
LTCH patients are often discharged to a traditional post-acute care facility such as a nursing home, SNF, SNU, CMR facility, or home health care. Thus, those facilities cannot be considered as "substitutes" for LTCHs, even though there is some overlap between the services provided to lower acuity LTCH patients and higher acuity patients in those traditional post- acute care facilities.
The family of a patient in an LTCH is generally encouraged to be more involved in the patient’s care than it would be if the patient was in the ICU of a traditional acute care hospital. For example, the visiting hours at LTCHs are typically more liberal than the visiting hours at a traditional acute care hospital for the ICU and, in some cases, the general medical/surgery floor.
Medicare reimbursements are the primary source of revenue for LTCHs because on average 75 to 85 percent of LTCH patients are covered by Medicare. In this case, Kindred
projected that approximately 91 percent of the patient days and
86 percent of the net patient revenues for its proposed LTCH would be generated by Medicare patients, and Select-Marion projected that approximately 76 percent of its patient days and approximately 71 percent of its net patient revenue at its proposed LTCH would be generated by Medicare patients.
In 2002, CMS adopted a Medicare prospective payment system (PPS) specifically for LTCHs. That system recognizes the LTCH patient population as being distinct from the patient populations treated by traditional acute care hospitals and post-acute care facilities such as nursing homes, SNFs, SNUs, and CMR facilities, even though there may be some overlap between the patient populations served by LTCHs and those other types of facilities.
Under the LTCH PPS, services are reimbursed by Medicare at a predetermined rate that is weighted based upon the patient's diagnosis and acuity, regardless of the cost of care. This reimbursement system is similar to, but uses Diagnosis Related Groups (DRGs) that are different than the DRGs used in the PPS for traditional acute care hospitals.
The Medicare reimbursement rates for services to long- stay patients in an LTCH are generally higher than the reimbursement rates for the same services to long-stay patients at a traditional acute care hospital. As a result, there is a
financial incentive for hospitals to transfer their long-stay patients to an LTCH.
It takes approximately six months for a new LTCH to receive its Medicare certification from CMS. During that period, the LTCH is reimbursed by Medicare at the lower rates applicable to traditional acute care hospitals, and, there is a financial incentive for the LTCH to keep its patient census as low as possible while still meeting the 25-day minimum ALOS. Accordingly, the first year of operation of an LTCH is not necessarily representative of the facility’s utilization, patient mix, or payor-source mix over time.
The Agency has concerns about the long-term viability of the LTCH industry because it is largely dependent upon CMS's continuing the LTCH PPS with its higher Medicare reimbursement rates.
The Agency’s concerns have caused it to take a very conservative approach in evaluating CON applications for new LTCHs so as not to allow the LTCH market to over-develop in the same way that freestanding psychiatric hospitals were over- developed at a time when the reimbursement rate for those facilities was favorable, which is no longer the case.
HIH Verses Freestanding LTCHs
There are two distinct types of LTCHs: HIHs and freestanding facilities. Both types are accepted in the industry, and both types are found nationwide.
Almost all of Select’s LTCHs (i.e., 76 out of 79) are HIHs.
Approximately 80 percent of Kindred Healthcare’s LTCHs, including all of its Florida LTCHs, are freestanding facilities.
Initially, all of the LTCHs approved in Florida were freestanding facilities because of the Agency’s concern that the applicable building codes precluded HIHs and/or because the Agency’s interpretation of the building codes made the development of HIH LTCHs cost-prohibitive.
The Agency no longer has concerns about the building code issue and, as reflected by its change in position in this case, it is willing to approve HIH LTCHs. Indeed, the two recently-approved Orlando LTCHs are HIHs.
The Agency does not have a formal preference for HIHs over freestanding LTCHs, or vice versa.
There are no material operational differences between HIH LTCHs and freestanding LTCHs, except that HIHs typically contract with the “host hospital” (i.e., the hospital in which
the LTCH is located) for ancillary services such as laboratory, radiology, and operating room.
The acuity level and mix of patients is materially the same at HIHs and freestanding LTCHs.
The ALOS at HIH LTCHs is slightly less than the ALOS at freestanding LTCHs, but the difference is not material.
HIH LTCHs draw the largest percentage of their admissions from the host hospital. That percentage is typically at least 25 percent, and it is not uncommon for the percentage to range from 50 to 90 percent.
Typically, HIH LTCHs are less expensive to operate than freestanding LTCHs because of lower overhead and the ability to obtain ancillary services from the host hospital; however, both types of LTCHs can be financially viable.
Currently, the Medicare reimbursement rates under the LTCH PPS are the same at HIH LTCHs and freestanding LTCHs.
LTCHs in Florida
At the time Select-Marion’s and Kindred’s applications were filed, there were only nine LTCHs operating in Florida with a total of 683 licensed beds. Those facilities were concentrated in six counties – Dade, Broward, Hillsborough, Pinellas, Duval, and Clay – and five of the State’s 11 heath planning districts – Districts 4, 5, 6, 10 and 11.
There were an additional 182 LTCH beds that had been approved by the Agency but which were not operational at the time that Kindred’s and Select-Marion’s applications were filed, including new facilities in Panama City (District 2), Orlando (District 7), and Sarasota (District 8). Those 182 beds do not include Select’s recently-approved 40-bed LTCH in Orlando.
There are no licensed or approved LTCHs in District 3.
The closest operational LTCH to Ocala is Kindred’s 60-bed facility in Green Cove Springs, which is 50 to 75 miles away. The approved but not yet operational Orlando LTCHs will be approximately 75 miles from Ocala.
The Florida LTCHs are generally well utilized; the occupancy rates at the facilities range from 54.6 percent to
99.2 percent. Four of the nine facilities, including Kindred’s LTCH in Green Cove Springs, have occupancy rates higher than
80 percent, and the average occupancy rate for all of the Florida LTCHs is 76.6 percent.
Data reported to the Agency reflects that counties and districts where LTCHs are located have considerably more LTCH admissions (both in number and as a percentage of total patients whose length of stay was greater than 25 days) than do counties and districts where there is no LTCH. See Exhibit S-4, at 5-6, 8-9. This data indicates that patients generally remain closer
to home for their long-term care, which is consistent with the expert testimony of the health planners in this case.
During calendar year 2002, a total of 202 District 3 residents were admitted to a Florida LTCH. Over half of those admissions were residents of the District 3 counties that are closer to an existing LTCH, which indicates that residents of those counties have better access to LTCH services than do residents of the other counties in the district.6 By contrast, only 20 Marion County residents were admitted to a Florida LTCH in calendar year 2002.
The admission experience at Kindred’s Green Cove Springs LTCH is consistent with that data. Between 20 and 25 percent of the admissions at that facility are District 3 residents, most of whom come from Putnam, Bradford, and other northern counties in District 3.
The ALOS for all patients discharged from Florida LTCHs between July 2001 and June 2002 was 42.2 days. The 65+ age cohort accounted for approximately 77 percent of the LTCH discharges in Florida for that period.
The ALOS for Florida LTCHs is considerably higher than the ALOS experienced nationwide by either Select or Kindred Healthcare. The ALOS at Select’s LTCHs is between 25 and 30 days; the ALOS at Kindred Healthcare’s LTCHs is in the low 30's.
The higher ALOS at the Florida LTCHs is due, at least in part, to the limited number of LTCHs in Florida and their tendency to treat only the “sickest of the sick,” which are the patients that generally have longest lengths of stay. The ALOS at the Florida LTCHs is expected to become lower as the number of LTCHs in Florida increases and as medical technology improves such that the more complex cases can be better managed.
Determination of Need for LTCH Beds in District 3 and Marion County
Generally
Select-Marion and Kindred have the burden to demonstrate need for their proposed LTCHs.
The Agency does not publish a fixed-need pool for LTCHs, nor is there an Agency rule which provides a specific formula or methodology for determining numeric need for LTCH beds.7 Indeed, it was clear from Mr. Gregg’s testimony at the hearing and in his deposition that the Agency’s policy toward LTCHs and, more specifically, need projections for LTCHs is still evolving.
The criteria that must be addressed by the applicant to demonstrate need are set forth in Florida Administrative Code Rule 59C-1.008(2)(e)2.a. through d. The criteria are population demographics and dynamics; availability, utilization and quality of "like services" in the district; medical treatment trends; and market conditions.
In addition to or as part of the demonstration of need based upon the rule criteria, the applicant must demonstrate that there is numeric need for at least the number of beds that it is proposing.
Kindred and Select-Marion stipulated that there is a numeric need for at least 31 LTCH beds in District 3. The Agency did not join that stipulation, and even though Mr. Gregg testified at the hearing and in his deposition that the Agency “presumes” that there is some level of need for LTCH beds in District 3, the Agency continued to take the position that neither applicant demonstrated need for its proposed LTCH.8
Methodologies Used by Kindred and Select-Marion to Project Numeric Need
Kindred and Select-Marion each used a “use rate” methodology and a “length of stay” methodology to project numeric need for their proposed LTCHs.
“Use Rate” Methodologies
Use rate methodologies are commonly used by health planners to project need for acute care hospital beds and other types of services. However, use rate methodologies do not produce reliable projections of bed need in the LTCH context because the existing LTCHs are not evenly distributed statewide and the utilization rates for the existing LTCHs vary significantly.9
The use rate methodologies used by Kindred and Select-Marion each projected the number of LTCH patient days that will likely be generated for their proposed LTCHs based upon the utilization rates at the existing LTCHs in Florida. Those patient days were then converted into an average daily census (ADC) and, in the case of Select-Marion, a bed need based upon a presumed occupancy rate.
The use rate methodology used by Kindred calculated the average utilization rate of the existing Florida LTCHs for both the population as a whole and for the 65+ age cohort. Those use rates were then applied to the respective 2005 populations of the primary and secondary service areas for Kindred’s proposed LTCH.
Kindred’s use rate methodology projected an ADC of 79 applying the use rate for the total population and an ADC of 109 applying the use rate for the 65+ age cohort. The ADCs projected for the PSA, which as noted above is a more reasonable geographic scope from which Kindred’s proposed LTCH will likely draw its patients, were 56 applying the use rate for the total population and 80 applying the use rate for the 65+ age cohort.
Select-Marion’s use rate methodology calculated separate “statewide” utilization rates for the 0-44, 45-64, 65- 74, 75-84, and 85+ age cohorts. Those rates were then applied to the projected 2008 population of Marion County in the
respective age cohorts in order to calculate a projected number of patient days that will be generated by Marion County residents in the third year of operation at Select-Marion’s proposed LTCH. Those patient days were then "grossed up” by an occupancy standard of 80 percent and then "grossed up" by an additional 44.5 percent to account for patients coming to the facility from outside of Marion County.
Select-Marion’s use rate methodology projected that Marion County residents would generate an ADC of 55 in 2008, which resulted in a projected need for 123 LTCH beds when the ADC was “grossed up” to account for the 44.5 percent of out-of- county patients and to reflect the 80 percent occupancy standard.
The use of age cohort-specific utilization rates is generally more reliable than the use of a utilization rate for the total population or only the 65+ age cohort. Kindred’s failure to use age cohort-specific utilization rates causes the bed need projections based upon its use rate methodology to be unreliable.
The age cohort-specific utilization rates used by Select-Marion in this case are unreliable because the LTCHs in Miami-Dade and Pinellas Counties were excluded from the calculations because their utilization rates were, according to Select-Marion’s application, “misleadingly conservative.” The
effect of excluding those counties is that the utilization rates applied by Select-Marion were inflated.
The 44.5 percent out-of-county rate was purportedly derived from the experience of the other LTCHs operating in Florida, but there is no evidence in the record to corroborate that rate. Moreover, Select-Marion’s own health planning expert testified that the 44.5 percent rate was too high; she testified that a more appropriate out-of-county rate would have been 20 to
25 percent, which is more consistent with the in-migration rate experienced by existing hospitals in Marion County.
Based upon the general unreliability of use rate methodologies in the LTCH context and the specific flaws in the use rate methodologies applied by Kindred and Select-Marion, the ADC and bed need projected under those methodologies are not reliable or reasonable.
“Length of Stay” Methodologies
The length of stay methodologies used by Kindred and Select-Marion each attempted to quantify the need for LTCH beds in their respective service areas by analyzing discharges from the hospitals in those areas. The methodologies each focused on “long-stay patients” discharged from those hospitals with “LTCH- appropriate DRGs” as the potential patient population for the proposed LTCHs. This general approach is reasonable from a health planning perspective.
The lists of LTCH-appropriate DRGs used by Kindred and Select-Marion in their respective methodologies were generally consistent. Although Select-Marion’s list included a greater number of DRGs than did Kindred’s, both lists are reasonable.
The parties’ respective methodologies each used the geometric mean length of stay (GMLOS) as the starting point for defining long-stay patients, which is reasonable and appropriate.10
The GMLOS is calculated by CMS. It is a statistically-adjusted value for all cases within a DRG that takes into account certain types of cases that could skew an arithmetic average length of stay.
Patients who have lengths of stay substantially longer than the GMLOS due to co-morbidities, complex medical conditions, or frailties due to age are typically appropriate LTCH patients, particularly if the patient would otherwise remain in the ICU of a traditional acute care hospital. In such circumstances, an LTCH would likely be the more appropriate setting for the patient from both a financial and patient-care standpoint.
Kindred defined long-stay patients as adult patients who were residents of District 3 and whose length of stay was at least 17 days longer than the GMLOS. Select-Marion’s definition
of long-stay patients included adult patients whose length of stay was at least 15 days longer than the GMLOS, and was not limited to District 3 residents.
Kindred’s definition of long-stay patients produced a more conservative (and more reasonable) estimate of potential LTCH patient days than did Select-Marion’s definition, primarily because Kindred’s definition included only District 3 residents. In this regard, Kindred’s methodology appropriately recognizes and takes into account the fact that patients discharged from District 3 hospitals who are not residents of the district are less likely to stay in the district for long-term acute care.
Kindred’s methodology was also more conservative (and more reasonable) in the manner that it calculated the number of potential LTCH patient days generated by the identified long- stay patients. Kindred calculated the potential LTCH patient days starting seven days after the GMLOS, whereas Select-Marion calculated the potential LTCH patient days starting at the GMLOS. In this regard, Kindred’s methodology appropriately recognizes and takes into account the fact that hospitals typically do not consider the transfer of patients to an LTCH until after the GMLOS and that it typically takes several days for the transfer to be coordinated once the patient has been identified as a potential LTCH patient.
Both parties’ methodologies assumed a 100 percent capture rate of the potential LTCH patients identified through their respective definitions of long-stay patients with LTCH- appropriate discharges.
The assumed 100 percent capture rate is not realistic, particularly when applied to the potential LTCH patients outside of the PSAs defined by the parties in the northernmost and southernmost counties in District 3. The evidence establishes that those patients will likely continue to leave the district for LTCH services, and as a result, a considerably lower capture rate would be expected from those counties.
The assumed 100 percent capture rate within the PSAs is more reasonable for Kindred’s methodology than it is for Select-Marion’s methodology because Kindred’s methodology includes only District 3 residents in the pool of potential LTCH patients whereas Select-Marion’s methodology also includes non- District 3 residents.
The result of Kindred’s length of stay methodology was an estimate of 14,269 potential LTCH patient days generated by residents of its PSA and 4,762 potential LTCH patient days generated by residents of its secondary service area. Those patient days translated into an ADC of 39 from the PSA and an ADC of 13 from the secondary service area.
The result of Select-Marion’s length of stay methodology was an estimate of 51,221 potential LTCH patient days generated by discharges from hospitals within the 40-mile radius around Ocala. Those patient days translated into an ADC of 140 and a bed need of 175 based upon an 80 percent occupancy standard.
Kindred’s length of stay methodology is, on balance, more reasonable than Select-Marion’s, and as a result, its projection of need is more reliable than the projection based upon Select-Marion’s methodology. Specifically, the preponderance of the evidence establishes that Select-Marion’s methodology overstates need by including patient days generated by non-residents of District 3 (who would not be as likely to use an LTCH in Marion County, and would certainly not use the facility at the assumed 100 percent capture rate) and by calculating the potential LTCH patient days starting at the GMLOS, rather than the seven days after the GMLOS which reasonably reflects the period of time necessary to identify, assess, and discharge a patient to the LTCH.
Ultimate Findings Related to Numeric Need
There was no credible evidence that there is a need for more than 91 LTCH beds so as to allow for the approval of both applications. As discussed above, the parties' use rate methodologies and Select-Marion’s length-of-stay methodology,
which all projected need for more than 91 LTCH beds, were not reliable.
The most reliable projection of need for a proposed LTCH in Marion County is the ADC of 39 projected by Kindred’s length of stay methodology for its PSA.11 The assumptions underlying that methodology are more reasonable than the assumptions underlying the length-of-stay methodology used by Select-Marion, and Kindred’s PSA, which is essentially the same as Select-Marion’s proposed 40-mile service area, is the most reasonable projection of the area from which an LTCH in Ocala could expect to draw virtually all of its patients. See Part of D(3) above.
An 80 percent occupancy standard is reasonable and appropriate because it is conservative figure and better reflects the lower bed turn-over by LTCH patients than does the
75 percent occupancy standard typically applied to traditional, “short-term” acute care hospitals.
Applying the 80 percent occupancy standard to an ADC of 39 results in a numeric need for 49 LTCH beds in the service area from which an LTCH in Ocala could reasonably be expected to draw its patients.
Ultimate Findings Related to Need Based Upon the Criteria in Florida Administrative Code
Rule 59C-1.008(2)(e)2.
The population demographics and dynamics of Marion County and the PSAs support the establishment of an LTCH in Ocala. The 65+ age cohort, which is the group that most highly utilizes LTCH services, represents approximately 25 percent of Marion County’s population and approximately 23 percent of the population of the parties' proposed PSAs. That age cohort is projected to grow at a higher rate than the overall population of the county and the PSAs over the planning horizons of the proposed LTCHs.
There are no LTCHs in District 3, and there are no LTCHs reasonably accessible to the residents of Marion County. The closest LTCH is more than 50 miles away in Green Cove Springs, and the occupancy/utilization rate at that facility is higher than 80 percent.
The absence of an LTCH in District 3 does not mean that there are no “like services” available at the existing hospitals in District 3 and, potentially, at some of the traditional post-acute care facilities in the district. The availability of an LTCH in District 3 would, however, provide an alternative and, in some cases, more appropriate and cost- effective setting for those services to be provided.
There is a general trend towards the increased utilization of LTCHs due to higher reimbursement rates available under the relatively new LTCH PPS, but the evidence failed to establish that the trend will, in fact, impact the delivery of health care services in Marion County or District 3. Instead, the evidence establishes that the utilization of a LTCH in Ocala will depend in large part on how well the local physicians are “educated” about the availability of LTCH services and whether the physicians have positive experiences with their patients at the facility.
Market conditions in Marion County support the establishment of an LTCH. The favorable market conditions include the size and projected growth of the 65+ age cohort and the letters of support provided by the two largest hospitals in Marion County as well as several of the hospitals in the surrounding counties.
The letters demonstrate a general level of support in the community for the establishment of an LTCH in Marion County, even though they do not specifically quantify the number of LTCH patients that would be generated from those hospitals. Given the stage in the process at which the letters were obtained, the level of specificity in the letters is reasonable, at least for purposes of demonstrating market support for the proposed LTCHs.
These factors, on balance, demonstrate the need for an LTCH in District 3 and Marion County, and they also support the projection of numeric need set forth above.
Comparative Analysis of the
CON Applications Based Upon the Criteria in Section 408.035, Florida Statutes (2003)12
Assuming the Agency accepts the foregoing findings that there is a demonstrated need for only 49 LTCH beds, it is not necessary to comparatively evaluate Kindred’s application against Select-Marion’s application, which proposes a 60-bed LTCH; it is only necessary to evaluate Kindred’s application against the applicable statutory criteria. However, in an abundance of caution, a comparative evaluation of the applications is set forth below in the event that the Agency (or an appellate court) determines that the evidence demonstrates a need for more than 60, but less than 91 LTCH beds.
Criteria Upon Which All of the Parties Agree
The parties stipulated that Kindred’s and Select- Marion’s applications each satisfied the criteria in Subsections 408.035(3), (5), (6), (10), and (12), Florida Statutes, or that the criteria in those subsections are not applicable. See Joint Pre-hearing Statement, at 7.
Notably, the parties stipulated that both Kindred and Select have a history of providing high quality of care to their patients. The evidence fails to establish that either of the
proposed LTCHs would provide a materially higher quality of care to its patients than would the other facility.13
Criteria That Are in Dispute
Section 408.035(1), Florida Statutes (Need in Relation to District Health Plan)
There are no preferences related to the development of LTCHs in the District 3 health plan; the plan does not even address LTCH services. Accordingly, this criterion is not implicated in this case.
Section 408.035(2), Florida Statutes (Availability, Quality of Care, etc. of Existing Services)
There are no existing LTCHs in District 3. As a result, either of the proposed LTCHs will enhance the availability and accessibility of LTCH services in the district to some degree. That impact will be most significant in Marion County and the adjacent counties because the evidence establishes that residents of the northernmost and southernmost counties in District 3, which are closer to the existing LTCHs in the adjacent districts, already have access to LTCH services.
Kindred’s proposed HIH LTCH will primarily serve its host hospital, Munroe. The most persuasive evidence on this issue was that Kindred's LTCH will likely receive 80 to 90 percent of its admissions from Munroe and Ocala Regional, which is across the street from Munroe. That will leave only 10 to 20 percent of its admissions coming from the remainder of District
3, which equates to 3 to 6 of the facility’s beds being available to serve residents of District 3 outside of Marion County.
Because Select-Marion’s proposed LTCH is nearly twice the size of Kindred’s proposed LTCH, it would have more beds available to serve residents of District 3 outside of Marion County. However, because it is unreasonable to expect that there will be significant in-migration to Marion County for LTCH services, the evidence does not support the inference that Select-Marion’s larger facility will, in fact, better enhance access to LTCH services for residents of District 3 than will Kindred’s facility.
Accordingly, each application satisfies the criterion in Section 408.035(2), Florida Statutes, and that criterion does not materially weigh in favor of the approval of one application over the other.
Section 408.035(4), Florida Statutes (Special Health Care Services)
The phrase “special health care services” in Section 408.035(4), Florida Statutes, is not defined in statute or Agency rule. Kindred and Select-Marion contend that LTCH services are “special health care services”; the Agency contends that they are not. See Joint Pre-hearing Statement, at 10.
The health care services that will be provided in the proposed LTCHs are services that are currently being provided in
District 3, either at a traditional acute care hospital (with respect to acute-level care) or at a post-acute care facility (with respect to rehabilitative care). The approval of the CON applications at issue in this proceeding would simply provide an alternative, and potentially more cost-effective setting in which those services could be provided.
Accordingly, LTCH services are not “special health care services” for purposes of Section 408.035(4), Florida Statutes, and the criterion in that subsection is not implicated in this case.
Section 408.035(7), Florida Statutes (Enhancing Access)
The approval of an LTCH in Ocala will enhance access to LTCH services in Marion County and the immediate surrounding areas; however, access to such services will not be significantly enhanced for residents of District 3 who are closer to and currently have reasonable access to LTCH services in an adjacent district.
Moreover, as discussed above in relation to Section 408.035(2), Florida Statutes, there is no credible evidence that access would be improved to a greater extent by the approval of Select-Marion’s proposed LTCH over Kindred’s, or vice versa.
Accordingly, each application satisfies the criterion in Section 408.035(7), Florida Statutes, and that criterion does
not materially weigh in favor of the approval of one application over the other.
Section 408.035(8), Florida Statutes (Financial Feasibility)
Issues Upon Which All of the Parties Agree
The parties stipulated that both of the proposed LTCHs are financially feasible in the short-term.
The parties also stipulated that Select-Marion’s proposed LTCH, which is projected to generate a net profit of approximately $939,000 in its second year of operation is financially feasible in the long-term.
Disputed Issues Related to the Long-term Financial Feasibility of Kindred's Proposed LTCH
The Agency concluded in the SAAR that Kindred’s proposed LTCH is financially feasible in the long-term, and that position was reaffirmed through Mr. Gregg’s testimony at the hearing. However, Select-Marion disputes the long-term financial feasibility of Kindred’s proposed LTCH.
The general rule for assessing the long-term financial feasibility of a CON project is if the project will at least break even by the end of the second year of operation, then the project is financially feasible in the long-term; if, however, the project continues to show a loss in the second year of operations and it is not demonstrated that the project will
reach a break-even point within a reasonable period of time, then the project is not financially feasible in the long-term.
Schedule 8A in Kindred’s application projects a net profit from operations of $263,134 for Kindred’s proposed LTCH in its second year of the operation.
Omission of Public Medical Assistance Trust Fund Assessment
Kindred concedes that the net profit from operations is overstated by $117,734 because Schedule 8A did not include a deduction for the Public Medical Assistance Trust Fund assessment in that amount. That assessment is imposed on all hospitals at a rate of 1.5 percent of the facility’s gross revenues.
Understatement of Property Taxes
The preponderance of the evidence establishes that the net profit from operations is overstated by an additional
$92,918 because the taxes shown on Schedule 8A are understated by that amount. The $92,918 represents the difference between the amount of taxes shown on Schedule 8A and the total estimated property taxes on the proposed facility, which according to John Grant, the Kindred witness responsible for calculating the taxes for Schedule 8A, were based upon the average property taxes paid on Kindred’s other Florida LTCHs.
There is no competent evidence, such as expert testimony from a property appraiser, to support Kindred’s
position that the $32,136 of taxes shown on Schedule 8A is a reasonable estimate of the taxes for its proposed LTCH.14
Alleged Understatements in the Nursing Costs Projected for the Second Year of Operation
Select-Marion also contends that Kindred’s net profit from operations is overstated by an additional $191,856, as a result of various understatements in the nursing costs on Schedule 8A, which are based upon the staffing projections on Schedule 6 of Kindred’s application. Specifically, Select- Marion contends that the nursing costs are understated because of an alleged error in the calculation of the nursing salary expense for the second year of the facility’s operation based upon the number of nursing full-time equivalents (FTEs) identified for that period; because of the additional nursing staff (and related costs) that would be necessary to achieve the nursing-hours-per-patient-day figure and the licensed-to- unlicensed-nurse ratio that were referenced in the text of Kindred’s CON application; and because a benefit rate of 17 percent was used instead of 20 percent.
Number of FTEs Used in Projecting Staffing Costs
Kindred’s CON application does not include staffing projections on Schedule 6 for the second year of operation; however, Mr. Grant, the Kindred employee who developed the assumptions underlying the financial schedules in Kindred’s application testified in his deposition that at the end of the
second year of operation, Kindred’s facility would have 41.7 nursing FTEs and 11.8 ancillary FTEs. Those projections were not disputed, and are reasonable.
Mr. Grant and other Kindred witnesses testified that the staffing projections shown on Schedule 6 were the number of FTEs at the end of the year, rather than the average number of FTEs during the year. Those witnesses further testified that the nursing costs shown on Schedule 8A are based upon the average number of FTEs during the year, not the year-end FTEs. That testimony is found to be credible.
The opinion of Select-Marion’s expert financial witness, Patricia Greenberg, that there is an error in the calculation of the nursing costs shown on Schedule 8A was based upon the incorrect assumption that the nursing costs on Schedule 8A and the staffing projections on Schedule 6 were both based upon the year-end figures. Accordingly, Ms. Greenberg’s opinion on that issue is rejected.
Nursing-Hours-Per-Patient-Day
Kindred’s application states that Kindred Healthcare's LTCHs currently provide an “average [of] 10.5 nursing hours per patient day.” The financial projections in Kindred’s application used the 10.5-hour figure for the proposed Ocala LTCHs' first year of operation only. The figure used for
the second year of operation was approximately 9.6 nursing hours per patient day.
Kindred’s application did not represent that 10.5 hours of nursing care per day would actually be provided at the proposed Ocala LTCH. There are no regulations that require that amount (or any amount) of nursing care per day at LTCHs.
Kindred’s application refers to the 10.5 hours as an “average.” Kindred’s medical director testified in his deposition that 10.5 hours is “a little on the high side” based upon his experience and another Kindred witness testified that the nursing hours would only be “around that number.” Select- Marion did not offer any credible evidence to the contrary.
Accordingly, the net profit from operations for Kindred’s proposed LTCH is not overstated in any amount because of the failure to project nursing costs for the second year of operation based upon the 10.5 hours of nursing care per day, and Ms. Greenberg’s opinion on that issue is rejected.
Licensed-to-Unlicensed-Nurse Ratio
Kindred’s application states that the ratio of licensed to unlicensed nursing personnel in Kindred Healthcare's LTCHs “averages 3.5 to 1.” The ratio actually used in the financial projections for Kindred’s proposed Ocala LTCH was approximately 2.46-to-one, which can be calculated by dividing
the 23.2 FTEs shown on Schedule 6 for “R.N.’s” and “L.P.N.’s” by the 9.4 FTEs shown on that schedule for “Nurses Aides.”
Kindred’s application did not represent that the 3.5- to-one ratio would actually be provided at the proposed Ocala LTCH. There are no regulations that require that ratio (or any ratio) of licensed-to-unlicensed nursing personnel at LTCHs.
Accordingly, the net profit for Kindred’s proposed facility is not overstated in any amount because of the failure to project nursing costs for the second year of operation based upon the 3.5-to-one ratio, and Ms. Greenberg’s opinion on that
issue is rejected.
Benefit Rate
The preponderance of the evidence establishes that Kindred’s benefit rate is 17 percent (not 20 percent), and that the 17 percent rate was used for the salary projections included in Kindred’s application.15 Accordingly, the net profit from operations projected for Kindred’s proposed LTCH is not understated in any amount based upon its use of a 17 percent benefit rate, and Ms. Greenberg’s opinion on that issue is rejected.
Summary
In sum, the evidence establishes the net profit from operations of Kindred’s proposed LTCH in the second year of operation is overstated by a total of $210,652. Even after that
amount is deducted from the net profit from operations shown on Schedule 8A, Kindred’s facility will still have a net profit from operations in the amount of $52,482. Therefore, Kindred’s proposed LTCH is financially feasible in the long-term.
Kindred and Select-Marion are both for-profit entities whose parent corporations are headquartered in other states, and there is no credible evidence that the local community would directly benefit from the profitability of either facility. As a result, the fact that Select-Marion’s proposed facility is projected to be “more financially feasible" in the long-term than Kindred’s proposed facility is immaterial to the outcome of this case.
Moreover, on the latter point, the net profit projected for Select-Marion’s facility is likely overstated to some degree because it is based on the patient revenues generated by a 60-bed LTCH operating at an average utilization rate of 68 percent in the second year of operation, which translates into an ADC of approximately 41. That ADC is higher than the ADC projected through the need methodology that was found to be the most reasonable and reliable. See Part G(3) above.
Section 408.035(9), Florida Statutes (Fostering Competition that Promotes Cost-effectiveness)
Competition for LTCH services in District 3 will not be fostered by the approval of only one of the applications, and
need has not been established for two LTCHs. The approved applicant would be the sole provider of such services in District 3, and there would likely not be any significant overlap in and competition for the patients served by the LTCH in Ocala and the existing LTCHs in Green Cove Springs, Jacksonville, and Tampa.
Accordingly, the criterion in Section 408.035(9), Florida Statutes, does not materially weigh in favor of the approval of either application.
Section 408.035(11), Florida Statutes (Medicaid and Indigent Care)
Kindred and Select-Marion stipulated that their applications were conditioned on the provision of Medicaid and charity care at levels consistent with the statewide average. However, the Agency took the position in the SAAR and in the Joint Pre-hearing Statement that the conditions offered by each applicant were lower than the statewide average.
Kindred conditioned its application on a combined two percent of its total patient days being provided to Medicaid and charity patients.
Schedule 7A of Kindred’s application reflects only
0.4 percent of the patient days being Medicaid patients and 0.3 percent of the patient days being charity patients for the first year of operation. In the second year of operation, Schedule 7A reflects 1.6 percent of the patient days being Medicaid patients
and 1.2 percent of the patient days being charity patients, for a total of 2.8 percent.
Select-Marion conditioned its application on a combined 2.8 percent of its patient days being provided to Medicaid and charity patients.
Schedule 7A of Select-Marion’s application projects two percent of its patient days being charity patients and 0.8 percent of the patient days being Medicaid patients in each of the first two years of operation.
The SAAR states that the statewide averages for Medicaid and charity patient days are two percent and 1.7 percent, respectively, for a total of 3.7 percent. There is no credible evidence in the record to corroborate those figures.
Even if those percentages were assumed to accurately reflect the data reported to the Agency, the evidence establishes that the Medicaid percentage is skewed based upon the pediatric program in Kindred’s St. Petersburg LTCH, which attracts an inordinately high level of Medicaid patients. The St. Petersburg facility is the only LTCH in Florida that has such a program.
Neither of the proposed LTCHs at issue in this proceeding will have a pediatric program. As a result, it is not reasonable to compare the skewed statewide average to the
Medicaid and charity care commitments proposed by the applicants in this case.
If the Medicaid patient days from the St. Petersburg LTCH are removed from the calculation of the statewide average, the “adjusted” statewide average for Medicaid patient days would between 1.2 and 1.5 percent. That range is comparable to the levels projected by the applicants in this case by the second year of operation of their respective LTCHs.
Select-Marion projects that it will provide a higher percentage of patient days for Medicaid and charity care in the first year of operation than does Kindred; however, by the second year of operation, the total percentage of patient days for Medicaid and charity care projected for each facility is the same.
The Agency recognizes that an LTCH may not meet its Medicaid and charity care conditions in the first year of the facility’s operation because the facility must control its admissions during a portion of that year in order to ensure that its ALOS is long enough to obtain certification from CMS.
In sum, Select-Marion and Kindred both satisfy the criterion in Section 408.035(11), Florida Statutes, when their projected Medicaid and charity care commitments are compared to the statewide average that is adjusted to exclude pediatric patient days; and, because the Medicaid and charity care
commitments by the Select-Marion and Kindred are materially the same by the second year of the facilities’ operation, this criterion does not favor either applicant over the other.
Ultimate Findings Based upon Comparative Analysis
On balance, Kindred’s application satisfies the applicable criteria in Section 408.035, Florida Statutes.
On balance, Select-Marion’s application satisfies the applicable criteria in Section 408.035, Florida Statutes.
Neither application is materially superior to the other in relation to the statutory criteria.
Faced with two comparatively equal applications, it is not unreasonable for the Agency to view the more conservative application more favorably. In this case, the more conservative application is Kindred’s because it is proposing to establish only 31 beds in existing, unused hospital space with a cost of only $1.4 million (or $45,300 per bed), as compared to Select- Marion’s 60 beds in a newly-constructed facility with a cost of
$12.2 million (or $204,100 per bed).
The higher costs associated with Select-Marion’s proposed LTCH will be borne primarily by Select, not the facility’s patients or the general public because approximately
75 percent of the facility’s revenues will come from Medicare reimbursements, which are made under a PPS rather than a cost- based system. The same is true for the costs of Kindred’s LTCH;
it will just take Kindred Healthcare less time to recoup its investment than it will for Select to recoup its investment.
The applicants also bear the primary financial risk if the Medicare reimbursement system for LTCHs becomes less favorable. However, as Mr. Gregg pointed out in his deposition, there is also a risk to the health care system in those circumstances because the approved LTCH would have an incentive to serve patients who might also be equally well-served in an another, less expensive post-acute care setting in order to recoup its capital investment. This consideration weighs in favor of approval of Kindred’s LTCH, which has less of a capital investment to be recouped.
Other considerations also weigh in favor of the approval of Kindred’s application over Select-Marion’s application. Specifically, Kindred garnered more community support for its proposed LTCH, including support from the two largest acute care hospitals in Ocala from which most of the admissions to a proposed LTCH in Ocala would come, and Kindred has also already negotiated a transfer agreement with one of the hospitals, making it more likely that Kindred’s LTCH would actually receive the admissions that it has projected. Additionally, Kindred’s LTCH will be operational approximately a year sooner than Select-Marion’s LTCH.
CONCLUSIONS OF LAW
The Division has jurisdiction over the parties to and subject matter of this proceeding pursuant to Sections 120.569, 120.57(1), and 408.039(5), Florida Statutes.
Select-Marion and Kindred have the burden to prove based upon a preponderance of the evidence that their respective CON applications should be approved. See Boca Raton Artificial Kidney Center, Inc. v. Dept. of Health & Rehabilitative Servs.,
475 So. 2d 260, 263 (Fla. 1st DCA 1985); Select Specialty Hospital – Sarasota, Inc. v. Agency for Health Care Admin., Case No. 03-2484CON, at 21 (DOAH Mar. 15, 2004; AHCA May 20, 2004);
§ 120.57(1)(j), Fla. Stat.
When comparatively evaluating competing CON applications, a balanced consideration of the applicable statutory and rule criteria must be made; the appropriate weight to be given to each criterion is not fixed, but rather varies based upon the facts of the case. See, e.g., Morton F. Plant Hospital Ass’n, Inc. v. Dept. of Health & Rehabilitative Servs.,
491 So. 2d 586, 589 (Fla. 1st DCA 1986) (quoting North Ridge General Hospital, Inc. v. NME Hospitals, Inc., 478 So. 2d 1138, 1139 (Fla. 1st DCA 1985)).
An LTCH is defined as “a hospital licensed under chapter 395 which meets the requirements of 42 C.F.R. s. 412.23(e) and seeks exclusion from the Medicare prospective
payment system for inpatient hospital services.” § 408.032(13), Fla. Stat. Accord Fla. Admin. Code R. 59C-1.002(28).
Because the Agency does not publish a fixed need pool for LTCHs or a formula or methodology for projecting need for LTCH beds, the determination of need for new LTCH beds is governed by Florida Administrative Code Rule 59C-1.008(2)(e)2. That rule provides:
(e) . . . . If an agency need methodology does not exist for the proposed project:
* * *
2. . . . the applicant will be responsible for demonstrating need through a needs assessment methodology which must include, at a minimum, consideration of the following topics, except when they are inconsistent with the applicable statutory and rule criteria:
Population demographics and dynamics;
Availability, utilization and quality of like services in the district, subdistrict or both;
Medical treatment trends; and
Market conditions.
The preponderance of the evidence establishes a need for 49 LTCH beds in the service area from which an LTCH in Ocala could reasonably be expected to draw its patients. That projection of need is supported by and takes into account each of the factors listed in Florida Administrative Code Rule 59C- 1.008(2)(e)2.a. through d.
Kindred’s application satisfies, on balance, the applicable criteria in Section 408.035, Florida Statutes, and its application should be approved.
Even though Select-Marion’s application also satisfies, on balance, the applicable criteria in Section 408.035, Florida Statutes, its application must be denied because it proposes a 60-bed LTCH and the evidence establishes a need for only 49 LTCH beds in the service area from which an LTCH in Ocala could reasonably be expected to draw its patients.
Even if the evidence had demonstrated a need for more than 60, but less than 91 LTCH beds,16 Kindred’s application should be approved over Select-Marion’s application because it is the more conservative of the two otherwise equal applications. Cf. Morton F. Plant Hospital Ass’n, supra
(affirming denial of a CON based upon system-wide cost considerations even though evidence established that there was a need for the proposed beds and that the applicant had the ability to fund the project).
Based upon the foregoing findings of fact and conclusions of law, it is
RECOMMENDED that the Agency issue a final order approving Kindred’s application (CON 9648) and denying Select-Marion’s application (CON 9647).
DONE AND ENTERED this 14th day of July, 2004, in Tallahassee, Leon County, Florida.
S
T. KENT WETHERELL, II Administrative Law Judge
Division of Administrative Hearings The DeSoto Building
1230 Apalachee Parkway
Tallahassee, Florida 32399-3060
(850) 488-9675 SUNCOM 278-9675
Fax Filing (850) 921-6847 www.doah.state.fl.us
Filed with the Clerk of the Division of Administrative Hearings this 14th day of July, 2004.
ENDNOTES
1/ The parties stipulated at the hearing that Mr. Caron is an expert in financial feasibility analysis for LTCHs.
2/ There was no objection at Dr. Muldoon’s deposition to Kindred’s tender of him as an expert in pulmonary disease, internal medicine, preventative medicine, and critical care medicine. Accordingly, Dr. Muldoon is accepted as an expert in those fields.
3/ Even if the Amended Request for Judicial Notice had been granted, no weight would have been given to the proposed rules in the evaluation of the parties’ competing CON applications because the comment period for the proposed rules did not close until July 12, 2004 (69 Fed. Reg. 28196), and several options are set forth in the proposed rules regarding the level at which CMS will pay for services rendered at hospital-within-hospital LTCHs that do not meet the 75 percent admission threshold in the proposed rules (id. at 28326-27). The proposed rules are not expected to be finalized until September 2004, and until they are, the potential impact of the proposed rules on the LTCHs at issue in this proceeding (and the financial viability of parties’ parent corporations) is entirely speculative.
4/ The projected operational dates for both facilities have been delayed as a result of this proceeding.
5/ Mr. Gregg testified as follows regarding the types of support letters that the Agency finds most useful in its evaluation and “validation” of an applicant’s need methodology:
[T]he most useful thing would be a statement which would say I’ve had patients and I couldn’t place them. The second most useful thing would be this is about how many patients we think we’ve had that would be in the ballpark for this [new service].
Transcript, at 588. He later added:
I think I’ve described what I would think of sort of the hierarchy of letters, and then I would add volume of letters, comparatively, is significant . . . .
Id. at 596. On this latter point, Mr. Gregg explained in response to a question regarding the value of a general letter of support:
Well, I would consider it of value depending upon what else I’m comparing it to. If I’m comparing this type of letter to an applicant that has submitted nothing, then I would give it greater value than I would if I was comparing it to an applicant that gave me that type of information that I said would be most valuable, which would be, we have tried to place people and can’t.
Id. at 589. See also Exhibit S-10, at 22 (Mr. Gregg’s testimony that if the applicant is able to get letters of commitment and/or support from discharge planners to corroborate the need projections than the applicant has “done the best possible job as far as we’re [the Agency] concerned”).
6/ Specifically, 58 of the 202 admissions were residents of Bradford and Putnam Counties, which are the counties adjacent to Clay County where Kindred’s Green Cove Springs LTCH is located, and 48 of the 202 admissions came from Lake, Hernando, and Sumter Counties, which are the southernmost counties in District
3 and are closer to Kindred’s Tampa LTCHs.
7/ At some point in the past, the Agency published a proposed rule that included a methodology for projecting need for LTCH beds. That proposed rule was subsequently withdrawn by the
Agency, and the parties’ stipulated that its terms are not relevant to this proceeding.
8/ The Agency’s position on this issue cannot be squared with its formal change of position in this case and its joinder in Kindred’s PRO, which advocates the approval of Kindred’s CON application. Unlike other Agency rules, Florida Administrative Code Rule 59C-1.008(2)(e)2. does not allow the Agency to approve a CON application in the absence of numeric need if “special” or “not normal” circumstances are established. Instead, the rule requires the applicant to demonstrate need and if it fails to do so, then its CON application must be denied. Ultimately, it is not necessary to resolve the apparent conflict between the Agency’s position on this issue and the rule because, as discussed below, the evidence in this case establishes a need for 49 LTCH beds.
9/ In making this finding, the undersigned did not overlook the fact that use rate methodologies have been used by applicants and accepted by the Agency in prior cases to demonstrate need for new LTCH beds. See Kindred Hospitals East, LLC d/b/a Kindred Hospital South Florida v. Mercy Medical Development, Inc. and Agency for Health Care Admin., Case No. 01-2712CON, at 19-21, 52 (DOAH July 23, 2002; AHCA November 1, 2002); Vencor
Hospitals South v. Agency for Health Care Admin., Case No. 97- 1181CON, 1998 WL 870012, at *3 (DOAH March 3, 1998). More
recently, however, use rate methodologies have been found to be inherently unreliable in the LTCH context. See Select Specialty Hospital – Sarasota, Inc. v. Agency for Health Care Admin., Case No. 03-2484CON, at 12-14, 18, 21 (DOAH Mar. 15, 2004; AHCA May
20, 2004). See also Transcript, at 163 (testimony of Select- Marion’s health planner acknowledging that the Agency “doesn’t rely upon [the use rate methodology] anymore as a demonstration of need [for LTCHs],” but disagreeing with the Agency’s decision not to do so in this case).
10/ Select-Marion’s application included an alternative length- of-stay methodology that focused on the potential LTCH patient days generated by patients whose length of stay was longer than
15 days notwithstanding the GMLOS for the patient’s diagnosis. That methodology is unreasonable because it did not use the GMLOs as the starting point for defining long-stay patients, and as a result, grossly overstated the potential LTCH patient pool. Accordingly, the higher bed need projections based upon this alternative length-of-stay methodology are inherently unreliable. Accord Select Specialty Hospital – Sarasota, supra.
11/ In making this finding, the undersigned has not overlooked the bed need projected by Select-Marion’s health planner in her “sensitivity analyses.” See Exhibit S-4, at 3-4. Those projections are rejected for several reasons. First, the potential LTCH patient days were calculated by multiplying the number of identified long-stay discharges by the inflated statewide ALOS of 42.7 days, rather than by using the actual number of days beyond the GMLOS that patient was in the hospital (which would more accurately reflect the number of potential LTCH patient days generated by each discharge) or the lower ALOS experienced by Select nationwide. Second, the projections include discharges of non-District 3 residents who, as discussed above, would not be as likely to use an LTCH in Ocala. Third, the projections use a 75 percent occupancy rate, rather than the more conservative and more appropriate 80 percent occupancy rate. Fourth, the inclusion of patients whose length of stay was seven days longer than the GMLOS as potential LTCH discharges (page 3 of Exhibit S-4) is not appropriate because the evidence establishes that it often takes at least seven days after the GMLOS to identify and coordinate the transfer of a potential LTCH patient. Each of these points caused the bed need projected in the "sensitivity analyses" to be inflated and unreasonable.
12/ All statutory references in this Recommended Order are to the 2003 version of the Florida Statutes.
13/ In making this finding, the undersigned has not overlooked the fact that Select-Marion’s proposed LTCH, which has a much larger compliment of private rooms than does Kindred’s facility, would be better able to accommodate patients of different genders or patients needing isolation due to their medical conditions. However, the evidence also establishes that the semi-private rooms at Kindred’s proposed LTCH create operational efficiencies by effectively doubling the amount of time that a nurse is in the patient’s room, and they also may help to reduce the loneliness that LTCH patients may have during their long lengths of stay.
14/ The undersigned expressly rejects Kindred’s argument that Exhibit K-6, which is a print-out from the Marion County Property Appraiser’s office, supports the inference that the deficiency in the property taxes calculated by Ms. Greenberg is overstated. See Kindred/Agency Joint PRO, at 15-16. Neither the print-out nor Mr. Grant’s testimony regarding his understanding of what was shown on the print-out provides a sufficient basis for the undersigned to infer an appraised value of Kindred’s leasehold interest from the appraised value of
Munroe’s facility. Indeed, it was clear from Mr. Grant’s testimony that he had no expertise in the appraisal of real property, and his testimony regarding the millage rate in effect in Munroe County (which would be applied against the assessed value of the leasehold to calculate the taxes) was imprecise and was uncorroborated hearsay, as was the related testimony elicited from Select-Marion's financial expert, Patricia Greenberg, on cross-examination. Simply put, the undersigned found Ms. Greenberg’s calculation of the estimated property tax for Kindred’s facility, which used the same methodology that Mr. Grant purportedly used in preparing the application, to be more reasonable than the inferences that Kindred seeks to draw from Exhibit K-6.
15/ Kindred employees Theresa Hunkins and John Caron each testified that Kindred’s benefit rate was “approximately 20 percent.” Their testimony is not inconsistent with Mr. Grant’s unequivocal and more precise testimony that the actual benefit rate is 17 percent and that he used that figure when he prepared the financial pro formas for Kindred’s CON application.
16/ For example, if the Agency concludes contrary to the undersigned’s findings that it is reasonable to expect that a proposed LTCH in Ocala will serve the entire district and not just the six-county PSA defined in Kindred’s application or the 40-mile radius defined in Select-Marion’s application, then the preponderance of the evidence would establish a need for 65 LTCH beds. That figure is based upon the ADC of 52 projected through Kindred’s length of stay methodology for its primary and secondary service areas and an 80 percent occupancy standard.
COPIES FURNISHED:
Alan Levine, Secretary
Agency for Health Care Administration Fort Knox Building, Suite 3116
2727 Mahan Drive
Tallahassee, Florida 32308
Valda Clark Christian, General Counsel Agency for Health Care Administration Fort Knox Building, Suite 3431
2727 Mahan Drive
Tallahassee, Florida 32308
Lealand McCharen, Agency Clerk Agency for Health Care Administration 2727 Mahan Drive, Mail Station 3
Tallahassee, Florida 32308
Mark A. Emmanuele, Esquire Panza, Maurer, & Maynard, P.A.
Bank of America Building, Third Floor 3600 North Federal Highway
Fort Lauderdale, Florida 33308-6225
Patricia A. Renovitch, Esquire
Oertel, Hoffman, Fernandez & Cole, P.A.
301 South Bronough Street, Fifth Floor Post Office Box 1110
Tallahassee, Florida 32302-1110
NOTICE OF RIGHT TO SUBMIT EXCEPTIONS
All parties have the right to submit written exceptions within
15 days from the date of this Recommended Order. Any exceptions to this Recommended Order should be filed with the agency that will issue the Final Order in this case.
| Issue Date | Document | Summary |
|---|---|---|
| Sep. 14, 2004 | Agency Final Order | |
| Jul. 14, 2004 | Recommended Order | Respondent should approve Kindred`s Certificate of Need (CON) application for a 31-bed, Long-Term Care Hospital (LTCH) and deny Select`s CON application for a 60-bed LTCH because evidence showed 49 beds needed and Kindred`s proposal was more conservative. |
