The Issue The ultimate issue is whether the application of Petitioner, University Medical Park, for a certificate of need to construct a 130-bed acute care hospital in northern Hillsborough County, Florida should be approved. The factual issues are whether a need exists for the proposed facility under the Department's need rule and, if not, are there any special circumstances which would demonstrate the reasonableness and appropriateness of the application notwithstanding lack of need. The petitioner, while not agreeing with the methodology, conceded that under the DHRS rule as applied there is no need because there is an excess of acute care beds projected for 1989, the applicable planning horizon. The only real factual issue is whether there are any special circumstances which warrant issuance of a CON. The parties filed post-hearing findings of fact and conclusions of law by March 18, 1985, which were read and considered. Many of those proposals are incorporated in the following findings. As indicated some were irrelevant, however, those not included on pertinent issues were rejected because the more credible evidence precluded the proposed finding. Having heard the testimony and carefully considered the Proposed Findings of Fact, there is no evidence which would demonstrate the reasonableness and appropriateness of the application. It is recommended that the application be denied.
Findings Of Fact General Petitioner is a limited partnership composed almost entirely of physicians, including obstetricians/gynecologists (OB/GYN) and specialists providing ancillary care, who practice in the metropolitan Tampa area. (Tr. Vol. 1, pp. 103-104). Petitioner's managing general partner is Dr. Robert Withers, a doctor specializing in OB/GYN who has practiced in Hillsborough County for over thirty years. (Tr. Vol. 1, pp. 24- 26, 28-29.) Dr. Withers was a prime moving force in the founding, planning and development of University Community Hospital and Women's Hospital. (Tr. Vo1. 1, pp. 26-28, 73; Vol. 4, pp. 547-548.) Petitioner seeks to construct in DHRS District VI a specialty "women's" hospital providing obstetrical and gynecological services at the corner of 30th Street and Fletcher Avenue in northern Hillsborough County and having 130 acute care beds. 1/ (Tr. Vol. 1, pp. 34, 74-75, Vol. 5, pp. 678-679, Northside Ex.-1, pp. 1-2, Ex.-4A.) The proposed hospital is to have 60 obstetrical, 66 gynecological and 4 intensive care beds. (Tr. Vol. 8, P. 1297, Northside Ex.-1 Table 17, Ex.-B.) DHRS District VI is composed of Hardy, Highlands, Hillsborough, Manatee and Polk counties. Each county is designated a subdistrict by the Local Health Council of District VI. Pasco County, immediately north of Hillsborough, is located in DHRS District V and is divided into two subdistricts, east Pasco and west Pasco. If built, Northside would be located in the immediate vicinity of University Community Hospital (UCH) in Tampa, Hillsborough County, Florida. Less than 5 percent of the total surgical procedures at UCH are gynecologically related, and little or no nonsurgical gynecological procedures arc performed there. (Tr. Vol. 4, p. 550.) There is no obstetrical practice at UCH, although it has the capacity to handle obstetric emergencies. The primary existing providers of obstetrical services to the metropolitan Tampa area are Tampa General Hospital (TGH) and Women's Hospital (Women's). (Tr. Vol. 1, p. 79, Northside Ex.-4, Tr. Vol. 7, pp. 1074-1075.) TGH is a large public hospital located on Davis Islands near downtown Tampa. (Tr. Vol. 1, pp. 47-48, Vol. 8, pp. 1356, 1358.) TGH currently has a 35 bed obstetrical unit, but is currently expanding to 70 beds as part of a major renovation and expansion program scheduled for completion in late 1985. (Tr. Vol. 7, pp. 1049, 1095, Vol. 8, pp. 1367-1368, Vol. 10, P. 1674, Northside Ex.- 2, P. 3.) In recent years, the overwhelming majority of Tampa General's admissions in obstetrics at TGH have been indigent patients. (Tr. Vol. 1, P. 61, Vol. 8, pp. 1375- 1379; Vol. 9, P. 1451; TGH Ex.-3.) Tampa General's internal records reflect that it had approximately 2,100 patient days of gynecological care compared with over 38,000 patient days in combined obstetrical care during a recent eleven month period. (TGH Ex.-3..) Women's is a 192 bed "specialty" hospital located in the west central portion of the City of Tampa near Tampa Stadium. (Tr. Vol. 1, pp. 63-64, 66-67; Vol. 10 P. 1564; Northside Ex.-4.) Women's Hospital serves primarily private-pay female patients. (Vol. 1, pp. 79, 88-89; Vol. 6, pp. 892-893.) Humana Brandon Hospital, which has a 16 bed obstetrics unit, and South Florida Baptist Hospital in Plant City, which has 12 obstetric beds, served eastern Hillsborough County. (Tr. Vol. 7, P. 1075; Northside Ex.-2, P. 3; Northside Ex.-4 and Tr. Vol. 1, P. 79; Northside Ex.-4.) There are two hospitals in eastern Pasco County, which is in DHRS District V. Humana Hospital, Pasco and East Pasco Medical Center, each of which has a six bed obstetric unit. Both hospitals are currently located in Dade City, but the East Pasco Medical Center will soon move to Zephyrhills and expand its obstetrics unit to nine beds. (Tr. Vol. 1, pp. 108- 109; Tr. Vol. 7, P. 1075; Vol. 8, pp. 1278-1281; Northside Ex.-4.) There are no hospitals in central Pasco County, DHRS District V. Residents of that area currently travel south to greater Tampa, or, to a lesser extent, go to Dade City for their medical services. (Tr. Vol. 2, pp. 266-267, 271-272; Vol. 7, p. 1038.) Bed Need There are currently 6,564 existing and CON approved acute care beds in DHRS District VI, compared with an overall bed need of 5,718 acute care beds. An excess of 846 beds exist in District VI in 1989, the year which is the planning horizon use by DHRS in determining bed need applicable to this application. (Tr. Vol. 7, pp. 1046-1047, 1163, 1165-66; DHRS Ex.-1.) There is a net need for five acute care beds in DHRS District V according to the Department's methodology. (Tr. Yolk. 7, pp. 1066, 1165; DHRS Ex.-1.) The figures for District VI include Carrollwood Community Hospital which is an osteopathic facility which does not provide obstetrical services. (Tr. Vol. 1, P. 158; Vol. 7, p. 1138; Vol. 8, P. 1291.) However, these osteopathic beds are considered as meeting the total bed need when computing a11 opathic bed need. DHRS has not formally adopted the subdistrict designations of allocations as part of its rules. (Tr. Vol. 7, pp. 1017-1017, 1019; Vol. 8, pp. 1176, 1187.) Consideration of the adoption of subdistricts by the Local Health Council is irrelevant to this application. 2/ Areas of Consideration in Addition to Bed Need Availability Availability is deemed the number of beds available. As set forth above, there is an excess of beds. (Nelson, Tr. Vol. VII, P. 1192.) Tampa General Hospital and Humana Women's Hospital offer all of the OB related services which UMP proposes to offer in its application. These and a number of other hospitals to include UCH, offer all of the GYN related services proposed by Northside. University Community Hospital is located 300 yards away from the proposed site of Northside. UCH is fully equipped to perform virtually any kind of GYN/OB procedure. Humana and UCH take indigent patients only on an emergency basis, as would the proposed facility. GYN/OB services are accessible to all residents of Hillsborough County regardless of their ability to pay for such services at TGH. (Williams, Tr. Vol. IX, P. 1469; Baehr, Tr. Vol. X, P. 1596; Splitstone, Tr. Vol. IV, P. 582; Hyatt, TGH Exhibit 19, P. 21.) Utilization Utilization is impacted by the number of available beds and the number of days patients stay in the hospital. According to the most recent Local Health Council hospital utilization statistics, the acute care occupancy rate for 14 acute care hospitals in Hillsborough County for the most recent six months was 65 percent. This occupancy rate is based on licensed beds and does not include CON approved beds which are not yet on line. This occupancy rate is substantially below the optimal occupancies determined by DHRS in the Rule. (DHRS Exhibit 4; Contis, Tr. Vol. VII, P. 1069.) Utilization of obstetric beds is higher than general acute care beds; however, the rules do not differentiate between general and obstetric beds. 3/ Five Hillsborough County hospitals, Humana Women's, St. Joseph's, Tampa General, Humana Brandon, and South Florida Baptist, offer obstetric services. The most recent Local Health Council utilization reports indicate that overall OB occupancy for these facilities was 82 percent for the past 6 months. However, these computations do not include the 35 C0N-approved beds which will soon be available at Tampa General Hospital. (DHRS Exhibit 4). There will be a substantial excess of acute care beds to include OB beds in Hillsborough County for the foreseeable future. (Baehr, Tr.w Vol. X, pp. 1568, 1594, 1597.) The substantial excess of beds projected will result in lower utilization. In addition to excess beds, utilization is lowered by shorter hospital stays by patients. The nationwide average length of stay has been reduced by almost two days for Medicare patients and one day for all other patients due to a variety of contributing circumstances. (Nelson, Tr. Vol. VII, P. 1192; Contis, Tr. Vol. VII, P. 1102; Baehr, Tr. Vol. X, pp. 1583-84; etc.) This dramatic decline in length of hospital stay is the result of many influences, the most prominent among which are: (1) a change in Medicare reimbursement to a system which rewards prompt discharges of patients and penalizes overutilization ("DGRs"), (2) the adaptation by private payers (insurance companies, etc.) of Medicare type reimbursement, (3) the growing availability and acceptance of alternatives to hospitalization such as ambulatory surgical centers, labor/delivery/recovery suites, etc. and (4) the growing popularity of health care insurance/delivery mechanisms such as health maintenance organizations ("HMOs"), preferred provider organizations ("PPOs"), and similar entities which offer direct or indirect financial incentives for avoiding or reducing hospital utilization. The trend toward declining hospital utilization will continue. (Nelson, Tr. Vol. VII, pp. 1192-98; Baehr, Tr. Vol. X, pp. 1584-86; etc.) There has been a significant and progressive decrease in hospital stays for obstetrics over the last five years. During this time, a typical average length of stay has been reduced from three days to two and, in some instances, one day. In addition, there is a growing trend towards facilities (such as LDRs) which provide obstetrics on virtually an outpatient basis. (Williams, Tr. Vol. IX, P. 1456; Hyatt, Tr. Vol. IV, P. 644.) The average length of stay for GYN procedures is also decreasing. In addition, high percentage of GYN procedures are now being performed on an outpatient, as opposed to inpatient, basis. (Hyatt, Tr. Vol. IV, P. 644, etc.) The reduction in hospital stays and excess of acute care beds will lower utilization of acute care hospitals, including their OB components, enough to offset the projected population growth in Hillsborough County. The hospitals in District VI will not achieve the optimal occupancy rates for acute care beds or OB beds in particular by 1989. The 130 additional beds proposed by UMP would lower utilization further. (Paragraphs 7, 14, and 18 above; DHRS Exhibit 1, Humana Exhibit 1.) Geographic Accessibility Ninety percent of the population of Hillsborough County is within 30 minutes of an acute care hospital offering, at least, OB emergency services. TGH 20, overlay 6, shows that essentially all persons living in Hillsborough County are within 30 minutes normal driving time not only to an existing, acute care hospital, but a hospital offering OB services. Petitioner's service area is alleged to include central Pasco County. Although Pasco County is in District V, to the extent the proposed facility might serve central Pasco County, from a planning standpoint it is preferable to have that population in central Paso served by expansion of facilities closer to them. Hospitals in Tampa will become increasingly less accessible with increases in traffic volume over the years. The proposed location of the UMP hospital is across the street from an existing acute care hospital, University Community Hospital ("UCH"). (Splitstone, Tr. Vol. IV, P. 542.) Geographic accessibility is the same to the proposed UMP hospital and UCH. (Smith, Tr. Vol. III, P. 350; Wentzel, Tr. Vol. IV, p. 486; Peters, Tr. Vol. IX, P. 1532.) UCH provides gynecological services but does not provide obstetrical services. However, UCH is capable of delivering babies in emergencies. (Splitstone, Tr. Vol. IV, p. 563.) The gynecological services and OB capabilities at UCH are located at essentially the same location as Northside's proposed site. Geographic accessibility of OB/GYN services is not enhanced by UMP's proposed 66 medical-surgical beds. The accessibility of acute care beds, which under the rule are all that is considered, is essentially the same for UCH as for the proposed facility. As to geographic accessibility, the residents of Hillsborough and Pasco Counties now have reasonable access to acute care services, including OB services. The UMP project would not increase accessibility to these services by any significant decrease. C. Economic Accessibility Petitioner offered no competent, credible evidence that it would expand services to underserved portions of the community. Demographer Smith did not study income levels or socioeconomic data for the UMP service area. (Smith, TR. Vol. III, pp. 388, 389.) However, Mr. Margolis testified that 24 percent of Tampa General's OB patients, at least 90 percent of who are indigents, came from the UMP service area. (Margolis, Tr. Vol. X, P. 1695.) The patients proposed to be served at the Northside Hospital are not different than those already served in the community. (Withers, Tr. Vol. II, P. 344.) As a result, Northside Hospital would not increase the number of underserved patients. Availability of Health Care Alternative An increasing number of GYN procedures are being performed by hospitals on an outpatient basis and in freestanding ambulatory-surgical centers. An ambulatory-surgical center is already in operation at a location which is near the proposed UMP site. In fact, Dr. Hyatt, a UMP general partner, currently performs GYN procedures at that surgical center. (Withers, Tr. Vol. I, P. 150; Hyatt, Tr. Vol. IV, pp. 644, 646. Ambulatory surgical centers, birthing centers and similar alternative delivery systems offer alternatives to the proposed facility. Existing hospitals are moving to supply such alternatives which, with the excess beds and lower utilization, arc more than adequate to preclude the need for the UMP proposal. (Nelson, Tr. Vol. VII, P. 1204, 1205, 1206; Williams, Tr. Vol. IX, pp. 1453, 1469; Contis, Tr. Vol. VII, pp. 1154; Contis, Tr. Vol. VII, pp. 1151, 1154.) Need for Special Equipment & Services DHRS does not consider obstetrics or gynecology to be "special services" for purposes of Section 381.494(6)(c)6, Florida Statutes. In addition, the services proposed by UMP are already available in Hillsborough and Pasco Counties. (Nelson, Tr. Vol. VII, pp. 1162, 1210.) Need for Research & Educational Facilities USF currently uses Tampa General as a training facility for its OB residents. TCH offered evidence that the new OB facilities being constructed at Tampa General were designed with assistance from USF and were funded by the Florida Legislature, in part, as an educational facility. (Powers, Tr. Vol. IX, P. 1391; Williams, Tr. Vol. IX, pp. 1453-1455.) The educational objectives of USF for OB residents at Tampa General are undermined by a disproportionately high indigent load. Residents need a cross section of patients. The UMP project will further detract from a well rounded OB residency program at Tampa General by causing Tampa General's OB Patient mix to remain unbalanced. (Williams, Tr. Vol. IX, P. 1458; Margolis, Tr. Vol. X, P. 1695.) UMP offered no evidence of arrangements to further medical research or educational needs in the community. (Nelson, Tr. Vol. VII, P. 1213. UMP's proposed facility will not contribute to research and education in District VI. Availability of Resources Management UMP will not manage its hospital. It has not secured a management contract nor entered into any type of arrangement to insure that its proposed facility will be managed by knowledgeable and competent personnel. (Withers, Tr. Vol. I, p. 142.) However, there is no alleged or demonstrated shortage of management personnel available. Availability of Funds For Capital and Operating Expenditures The matter of capital funding was a "de novo issue," i.e., evidence was presented which was in addition to different from its application. In its application, Northside stated that its project will be funded through 100 percent debt. Its principal general partner, Dr. Withers, states that this "figure is not correct." However, neither Dr. Withers nor any other Northside witness ever identified the percentage of the project, if any, which is to be funded through equity contributions except the property upon which it would be located. (UMP Exhibit 1, p. 26; Withers, Tr. Vol. I, P. 134.) The UMP application contained a letter from Landmark Bank of Tampa which indicates an interest on the part of that institution in providing funding to Northside in the event that its application is approved. This one and one half year year old letter falls short of a binding commitment on the part of Landmark Bank to lend UMP the necessary funds to complete and operate its project and is stale. Dr. Withers admitted that Northside had no firm commitment as of the date of the hearing to finance its facility, or any commitment to provide 1196 financing as stated in its application. (UMP Exhibit I/Exhibit Dr. Withers, Tr. Vol. I, P. 138.) Contribution to Education No evidence was introduced to support the assertion in the application of teaching research interaction between UMP and USF. USF presented evidence that no such interaction would occur. (Tr. Vol. IX, P. 1329.) The duplication of services and competition for patients and staff created by UMP's facility would adversely impact the health professional training programs of USF, the state's primary representative of health professional training programs in District VI. (Tr. Vol. IX, pp. 1314-19; 1322-24; 1331-1336.) Financial Feasibility The pro forma statement of income and expenses for the first two years of operation (1987 and 1988) contained in the UMP application projects a small operating loss during the first year and a substantial profit by the end of the second year. These pro formas are predicated on the assumption that the facility will achieve a utilization rate of 61 percent in Year 1 and 78 percent in its second year. To achieve these projected utilization levels, Northside would have to capture a market share of 75-80 percent of all OB patient days and over 75% of all GYN patient days generated by females in its service area. (UMP, Exhibit 1; Withers, Tr. Vol. I, P. 145, Dacus; Tr. Vol. V, P. 750-755.) These projected market shares and resulting utilization levels are very optimistic. It is unlikely that Northside could achieve these market shares simply by making its services available to the public. More reasonable utilization assumptions for purposes of projecting financial feasibility would be 40-50 percent during the first year and 65 percent in the second year. (Margolis, Tr. Vol. X, P. 1700; Baehr, Tr. Vol. X, pp. 1578, 1579, 1601.) UMP omitted the cost of the land on which its facility is to be constructed from its total project cost and thus understates the income necessary to sustain its project. Dr. Withers stated the purchase price of this land was approximately $1.5 million and it has a current market value in excess of $5 million. (Withers, Tr. Vol. I, pp. 139, 140.) Dr. Withers admitted that the purchase price of the land would be included in formulating patient charges. As a matter of DHRS interpretation, the cost of land should be included as part of the capital cost of the project even if donated or leased and, as such, should be added into the pro formas. UMP's financial expert, Barbara Turner, testified that she would normally include land costs in determining financial feasibility of a project, otherwise total project costs would be understated (Withers, Tr. Vol. I, P. 141; Nelson, Tr. Vol. VII, pp. 1215, 1216; Turner, Tr. Vol. X, P. 1714.) In addition, the pro formas failed to include any amount for management expenses associated with the new facility. Dr. Withers admitted UMP does not intend to manage Northside and he anticipates that the management fee would be considerably higher than the $75,000 in administrator salaries included in the application. (Withers, Tr. Vol. I, pp. 143, 144.) Barbara Turner, UMP's financial expert, conceded that the reasonableness of the percent UMP pro formas is predicated on the reasonableness of its projected market share and concomitant utilization assumptions. These projections are rejected as being inconsistent with evidence presented by more credible witnesses. The UMP project, as stated in its application or as presented at hearing, is not financially feasible on the assumption Petitioner projected. VIII. Impact on Existing Facilities Approval of the UMP application would result in a harmful impact on the costs of providing OB/GYN services at existing facilities. The new facility would be utilized by patients who would otherwise utilize existing facilities, hospitals would be serving fewer patients than they are now. This would necessarily increase capital and operating costs on a per patient basis which, in turn, would necessitate increases in patient charges. (Nelson, Tr. Vol. VII, pp. 1217-1219; Baehr, Tr. Vol. X, P. 1587.) Existing facilities are operating below optimal occupancy levels. See DHRS Exhibit 4. The Northside project would have an adverse financial impact on Humana, Tampa General Hospital, and other facilities regardless of whether Northside actually makes a profit. See next subheading below. The Northside project would draw away a substantial number of potential private-pay patients from TGH. Residents of the proposed Northside service area constitute approximately 24 percent of the total number of OB patients served by TGH. The Northside project poses a threat to TGH's plans to increase its non- indigent OB patient mix which is the key to its plans to provide a quality, competitive OB service to the residents of Hillsborough County. (Nelson, Tr. Vol. VIII, P. 1225; Margolis, Tr. Vol. X, P. 1695.) Impact Upon Costs and Competition Competition via a new entrant in a health care market can be good or bad in terms of both the costs and the quality of care rendered, depending on the existing availability of competition in that market at the time. Competition has a positive effect when the market is not being adequately or efficiently served. In a situation where adequate and efficient service exists, competition can have an adverse impact on costs and on quality because a new facility is simply adding expense to the system without a concomitant benefit. (Baehr, Tr. Vol. X, p. 1650.) Competition among hospitals in Hillsborough County is now "intense and accelerating." (Splitstone, Tr. Vol. IV, p. 558.) Tampa General is at a competitive disadvantage because of its indigent case load and its inability to offer equity interests to physicians in its hospital. (Blair, Tr. Vol. VI, pp. 945, 947-948); Powers, Tr. Vol. IX, P. 1405.) Tampa General Hospital is intensifying its marketing effort, a physician office building under construction now at Tampa General is an illustration of Tampa General's effort to compete for private physicians and patients. (Powers, Tr. Vol. IX, pp. 1405-1406.) The whole thrust of Tampa General's construction program is to increase its ability to compete for physicians. (Nelson, Tr. Vol. VII, P. 1224; Powers, Tr. Vol. IX, p. 1442.) The Tampa General construction will create new competition for physicians and patients. (Contis, Tr. Vol. VII, p. 1099.) Patients go to hospitals where their doctors practice, therefore, hospitals generally compete for physicians. (Splitstone, Tr. Vol. IV, P. 563; Blair, Tr. Vol. VI, pp. 898, 928.) Because many of the UMP partners are obstetricians who plan to use Northside exclusively, approval of the Northside project would lessen competition. (Popp, TGH Exhibit 18, P. 11.) It is feasible for Tampa General to attract more private pay OB patients. (Williams, Tr. Vol. IX, pp. 1460- 1461.) At its recently opened rehabilitation center, Tampa General has attracted more private pay patients. (Powers, Tr. Vol. IX, pp. 1393-1396.) USF OB residents at Tampa General are planning to practice at Tampa General. (Williams, Tr. Vol. IX, pp. 1460-1461.) The state-of-the-art labor, delivery, recovery room to be used at Tampa General will be an attractive alternative to OB patients. (Williams, Tr. Vol. IX, pp. 1460- 1461); Popp, TGH Exhibit 18, p.26) IX. Capital Expenditure Proposals The proposed Northside hospital will not offer any service not now available in Tampa. (Hyatt, TGH Exhibit 19, p. 21).
Recommendation Petitioner having failed to prove the need for additional acute care beds to include OB beds or some special circumstance which would warrant approval of the proposed project, it is recommended that its application for a CON be DENIED. DONE and ORDERED this 25th day of June, 1985, in Tallahassee, Florida STEPHEN F. DEAN Hearing Officer Division of Administrative Hearings The Oakland Building 2009 Apalachee Parkway Tallahassee, Florida 32399-1550 (904) 488-9675 Filed with the Clerk of the Division of Administrative Hearings this 25th day of June, 1985.
The Issue The issue in this case is whether the Respondent, the Department of Health and Rehabilitative Services (HRS), should grant the application of the Petitioner, RHPC, Inc., d/b/a Riverside Hospital (Riverside), for a certificate of need, CON Action No. 6582, for the addition of 31 acute care beds.
Findings Of Fact The Applicant and the Application. The applicant, the Petitioner, RHPC, Inc., d/b/a Riverside Hospital (Riverside), is a 102 bed acute care hospital 1/ located at 6600 Madison Street, New Port Richey, Florida, in the West Pasco County Subdistrict of HRS Service District 5, which also includes Pinellas County and East Pasco County. Included among its complement of beds are 14 obstetrical (OB) beds. There are no existing pediatric beds. Riverside's application is for a certificate of need to spend approximately $2,000,000 to renovate its existing OB unit, add 14 beds to the OB unit, add 11 medical/surgical beds and add six pediatric beds. The addition of the pediatric unit will be accomplished by relatively minor alterations to existing space and existing beds, and the cost attributable to this phase of the application is negligible. Similarly, the 11 additional med/surg beds will be accomplished by adding beds to existing private rooms, to create semi-private rooms, at a cost of only approximately $44,000. (Gas and electric lines for the additional beds already have been run to the headwall of these rooms and can be connected without difficulty or much expense.) Most of the $2 million total capital expenditure proposed in the application is attributable to the cost of modernizing the OB unit, with the addition of 14 beds in the process. The addition of 14 beds to the unit does not add significantly to what the modernization effort would cost without the addition of the 14 beds. The proposed new OB unit would include private rooms, to go along with the semi-private rooms that make up the existing 14-bed unit. In addition, the proposed modernized 28-bed OB unit would consist of the combined labor/delivery/recovery/post-partum (LDRP) rooms now preferred by most patients. Pertinent State Health Plan Provision. The 1989 State of Florida Health Plan states at the outset of a list of preferences to be utilized in comparing applications for additional acute care beds: No additional acute care beds should generally be approved unless the subdistrict occupancy rate is at or exceeds 75 percent, or, in the event of an existing facility, an applicant shall demonstrate that the occupancy rate for the most recent 12 months is at or exceeds 80 percent. The Need Methodology. Using the F.A.C. Rule 10-5.038 methodology, the district and subdistrict would show numeric need of approximately 201 and 230, respectively. See F.A.C. Rule 10-5.038(5). Regardless of the calculated bed need, HRS does not normally approve additional beds in a subdistrict unless the annual average acute care bed occupancy rate is 75 percent or higher during the 12-month base period of July, 1989, through June, 1990. See F.A.C. Rule 10-5.038(7)(d). The 670 licensed beds in the West Pasco Subdistrict reported only 68.92% occupancy during the 12- month base period, resulting in no projected need for additional acute care beds in the subdistrict for the applicable 1996 planning horizon. Even when a subdistricts's need for additional acute care beds projected by the methodology is zero, an application by an existing hospital still may be approved where that hospital's annual average occupancy rate exceeds 75 percent for the 12-month base period (again, in this case, from July, 1989, through June, 1990.) See F.A.C. Rule 10-5.038(7)(e). During the 12-month base period from July, 1989, through June, 1990, Riverside's occupancy averaged 72.40%, not high enough to be approved under F.A.C. Rule 10-5.038(7)(e). Observation Bed Days. Three types of beds days are included in a category of so-called "outpatient observation bed days." First, "twenty-three hour patients" are patients who are not eligible for inpatient services under the Health Care Finance Administration (HCFA) criteria for the Medicare program. Second, "observation patients" are similar non-Medicare patients. Third, some outpatients (or ambulatory surgery patients) also use beds for part of a day. With new cost containment and review/regulation developments in hospital care, more patients are spending up to 23 hours in the hospital before a decision is made that further hospitalization in not needed. As a result, "observation" bed use has increased. Outpatient observation services have been recognized and defined by HCFA. Blue Cross and Blue Shield of Florida (the Medicare intermediary) and the Health Care Cost Containment Board (HCCCB) have addressed issues such as reimbursement, billing and reporting of observation beds. Services are provided to "observation bed" patients under doctor's orders, including diagnostic services, observation and monitoring by nursing personnel and/or medical intervention or treatment. Calculation of occupancy rates under the HRS need methodology does not take into account the so-called "observation bed days." 2/ There was no evidence that any part of District V or the West Pasco Subdistrict are inaccessible geographically. Other Need Factors. The evidence showed that there is a seasonal peak utilization and occupancy of acute care beds in District V and in the West Pasco Subdistrict during approximately October or November through March or April each year. This seasonal peak is reflected by the statistics. As previously stated, Riverside's occupancy averaged 72.40% during the period from July, 1989, through June, 1990. During the first quarter of 1990, occupancy was 86.83%. Riverside's average occupancy for calendar year 1990 was 73.87%. For the period from March, 1990, through February, 1991, average occupancy for Riverside's acute care beds was 71.2%. 3/ For the period from March, 1990, through February, 1991, occupancy for Riverside's obstetrics beds was 92.9%. There is no acute care pediatric unit in the West Pasco subdistrict. Subdistrict residents (as well as others in Riverside's general service area) needing level II pediatric services generally go to a Pinellas County or East Pasco County hospital for them. Given the choice, some but not all of these patients likely would prefer to get these services at Riverside, depending primarily on the severity of the particular medical needs. But the evidence did not quantify the number predicted to switch to Riverside. Also, occupancy of pediatric beds in Pasco county was less than 15% during 1987 and 1988. Medical Care for the Poor. The State Health Plan also notes that the uncompensated care burden on hospitals has grown during the 1980s because of a growing number of low-income persons; simultaneously, the proportion of persons covered by Medicaid has dropped. Numerous statewide studies, moreover, have shown that hospitals' uncompensated care is increasing at the same time that their ability to absorb the cost of care is decreasing. Riverside's predecessor bought the hospital from Pasco County in 1982. As a condition to the purchase, Riverside's predecessor agreed to provide Medicaid and indigent care for Pasco County in perpetuity. When Riverside purchased the hospital on December 29, 1983, it assumed the contractual obligation to provide Medicaid and indigent care in perpetuity. Riverside is a disproportionate share provider within the meaning of the State and local health plans. Approximately, 13% of Riverside's total annual patient days are for Medicaid patients. In 1990, 2,647 of Riverside's obstetrical, and 4,272 of its non-obstetrical patient days, were Medicaid. Riverside's charity care deduction from gross patient revenue for fiscal year 1990 was 1.07% of gross patient revenue. Riverside's Medicaid deduction from gross patient revenue for fiscal year 1990 was 5.96% of gross patient revenue. Approximately, 14.8% of Riversides's services go to Medicaid and indigent patients. Although Riverside has only 14% of the beds in the West Pasco subdistrict, it does more than 90% of the non-emergency, non-OB Medicaid care. Approval of the Riverside application would enable Riverside to spread its administrative and overhead costs over a larger base, thereby reducing average charges. Approval of the Riverside application also would make Riverside more profitable and thereby better able to absorb the cost of the Medicaid and indigent care it provides. If Riverside converts existing acute care beds to pediatric or OB beds, it probably would have to squeeze out paying patients during seasonal occupancy peaks, thereby losing more revenue and profits. Competition. If the Riverside application is approved, Riverside's share of the market represented by the West Pasco subdistrict will rise from approximately 14% to approximately 18%. HCA controls the rest of the market. There are no existing OB beds in the West Pasco subdistrict other than at Riverside. The HCA hospital in New Port Richey had an OB unit which it recently abandoned. As a result of the grant of Bayonet Point's application, CON Action No. 6583, with which Riverside had been in direct competition in this application review cycle, Bayonet Point now is approved for a seven-bed OB unit as part of its bed complement. Upgrading its existing OB unit and adding 14 more OB beds will enable Riverside to capture more private paying patients, which will better enable it to compete with the HCA hospitals. At present, Riverside's OB unit is utilized almost exclusively by indigent and Medicaid patients because of the hospital's contract with Pasco County. This unit now is operating at close to absolute capacity. With the upgrades and additional beds, Riverside can work to capture some private pay patients; without them, Bayonet Point will capture the private pay patients. Financial Feasibility. Riverside operated at a deficit from 1983 essentially to the present. By the end of 1990, Riverside had accumulated a deficit of $8.8 million. Riverside's corporate parent, American Healthcare Management, Inc. (AHM), was funding the deficit. From 1985 through December, 1989, AHM was in Chapter 11 bankruptcy proceedings. During that time period, there was legitimate concern whether AHM would be able to continue to fund Riverside deficits. AHM emerged from bankruptcy in December, 1989, stronger financially. It has since become stronger still. AHM reduced its debt by approximately $88 million. Part of the debt reduction was achieved by the sale of $43 million of underperforming assets. In addition, $45 million of bond debt was exchanged for common stock on September 30, 1991. The interest savings on the bond-for-stock exchange is $6 million a year. As a result, AHM's current debt-to-equity ratio is approximately $160 million to $130 million. AHM's corporate staff has been reduced from about 102 to 65. Its corporate office were transferred from expensive quarters in Dallas, Texas, to less expensive quarters in King of Prussia, Pennsylvania. Corporate expenses have been greatly reduced as a result. Accounts receivable have been reduced by better collection methods, and the $43 million of assets sold to reduce corporate debt had been underperforming. AHM had $21 million cash and short-term investments as of December 31, 1989. As of the date of the final hearing, it had $18 million cash and short- term investments. Riverside's gross margin (profit) for the first nine months of 1991 was $4 million. After depreciation, amortization, and interest and home office costs, Riverside generated approximately $1.2 million for the first nine months of 1991. Internal cash flow generated by AHM and Riverside would be sufficient to finance Riverside's application project. Since the capital costs of Riverside's proposed project are relatively small, financial feasibility is relatively easy to achieve. Besides costing relatively little, the 31 new beds will not increase intercompany interest or management fees significantly. In addition, the 31 new beds would enable Riverside to better compete for private pay patients. Given the expected utilization of the new beds, the proposed project will be to the financial benefit of the applicant. The pro forma bears this out. It projects 75.11% occupancy for the 31 new beds in the second year of operation (July, 1994, to June, 1995). (This projection does not include expected "observation bed days.") A profit of $2,477,199 for the 31 beds is projected for the second year of operation (not counting any portion of the preexisting intercompany interest or management fees).
Recommendation Based on the foregoing Findings of Fact and Conclusions of Law, it is recommended that HRS enter a final order denying the Riverside application for a certificate of need, CON Action No. 6582, for the addition of 31 acute care beds. RECOMMENDED this 28th day of January, 1992, in Tallahassee, Florida. J. LAWRENCE JOHNSTON Hearing Officer Division of Administrative Hearings The DeSoto Building 1230 Apalachee Parkway Tallahassee, Florida 32399-1550 (904) 488-9675 Filed with the Clerk of the Division of Administrative Hearings this 28th day of January, 1992.
The Issue Whether proposed Rule 10-5.011(1)(v), Florida Administrative Code, is an invalid exercise of delegated legislative authority?
Findings Of Fact General Background; Neonatologists in Contemporary Hospital Settings. The rule at issue regulates neonatal intensive care services. Neonates are newborn infants, less than one month old.The care of neonates is provided in hospital nurseries. Before the birth, the mother and fetus are ordinarily under the care of an OB-GYN physician, although in cases of problem pregnancies, a subspecialist known as a maternal-fetal specialist may also participate in the care of the mother and fetus. After birth, normal newborns are cared for by physicians who specialize as pediatricians. They usually have office practices which cannot be interrupted continuously by travel to a hospital nursery to care for a neonate in crisis. Ill neonates, especially those of low birth weight or born before full term, often need non-routine or intensive care, including intravenous administration of medication, alimentation using stomach tube feeding (gavage), assistance with breathing through an oxygen hood or by mechanical ventilation and intensive medical and nursing care. Intensive care of neonates commonly is managed by neonatologists. Neonatologists usually have a hospital-based practice, not an office practice. They are physicians who specialized as pediatricians and received three years pediatrics training after licensure and successfully completed a board examination, and then received three additional years training in a neonatology fellowship and successfully completed a further examination, leading to board eligibility or certification in the sub-specialty of neonatology. The first board examination in neonatology was given in 1975. There are now about 1,900 board-certified neonatologists in the United States; about 105 practice in Florida. As subspecialists, neonatologists command substantial salaries due to their experience and lengthy training. They are the second highest paid subspecialists in pediatrics. They often do not find practice at community hospitals attractive because the low incidence of ill neonates at a community hospital does not allow them to keep their skills current; they tend to work in larger medical centers. Neonatologists can maximize their income by working at proprietary hospitals on a fee-for-service basis if the hospital has a large obstetric volume of healthy mothers (i.e., few low income mothers who lack prenatal care or present complicating problems such as drug addiction which may affect the child). Infants such as premature babies of crack addict mothers can be very sick, but as the cost of their care rises, the percentage of reimbursement to the neonatologist and the hospital from any insurers or public sources diminishes. A very few physicians take additional training and board certification examinations as sub-subspecialists in extremely narrow areas of care, such as neonatal cardiovascular surgery, or in areas such as pediatric surgery, pediatric neurology, pediatric neurosurgery, or pediatric cardiac catheterization. They almost always work at regional, tertiary care hospitals. The Guidelines for Perinatal Care (second edition, 1988), (Perinatal Guidelines) of the American Academy of Pediatrics and the American College of Obstetricians and Gynecologists point out that: [a]lthough outcome is not necessarily influenced by facility or service size, the availability of appropriate equipment and personnel, such as obstetric and pediatric specialists, subspecialists, and experienced nurses, is very important to the effective management of the problems presented by the small proportion of the perinatal population with the highest risks. In the recent past, the outcome for very low birth weight infants (750-1,500 g) has steadily improved. Neonatal intensive care and interhospital transport of high-risk and very ill infants have positively influenced the survival and well-being of small and premature babies. The outcome is better for high-risk infants born in regional perinatal centers than for those transferred after birth. Perinatal Guidelines, Chapter 1 at pages 1-2. As might be expected in such matters, OB-GYN physicians and neonatologists in private practice do not always agree with neonatologists on the faculty of teaching hospitals on the best way to design a system to provide for the care of very ill neonates. Practicing OB-GYNs and neonatologists want to have more hospitals designated as intensive care centers. The Perinatal Guidelines themselves recognize that "competition [among hospitals] is eroding regionalized care," and contain the caveat that "local circumstances must dictate the way in which these guidelines are best interpreted to meet the needs of a particular hospital, community, or system." Perinatal Guidelines, Preface, at pages xv-xvi. As the field of neonatology has grown, and new neonatologists entered the employment market, hospitals have competed for obstetric admissions by escalating their level of neonatal care. Traditional levels of care under the Perinatal Guidelines are described in Findings 28 to 30, post. The range of services required for severely ill neonates in large part is directly related to their prematurity, and/or their birth weight. The more premature, and/or the lower the birth weight, the more likely the neonate will present multiple problems, not just cardio-pulmonary ones, requiring substantial interdisciplinary coordination in care, which is best accomplished in regional intensive care centers. Gestational age and birth weight problems are exacerbated by poor prenatal care, birth by very young mothers or by mothers addicted to drugs such as cocaine. These mothers and their babies require extensive, expensive care but often are indigent. The Department intends to develop a rule for regulation of neonatal intensive care services which will achieve these goals: Ensure an adequate supply of beds to satisfy need. Develop units of an efficient size, taking into account quality of care and economies of scale. Minimize disruption to existing patient flow patterns which had been developed in the existing system of neonatal intensive care. Provide higher quality neonatal care. The rule designed by the Department divides neonatal intensive care into two categories, Level II, and Level III. Rule 10-5.011(1)(v)2.e. (II) and (III), Florida Administrative Code. Traditional Level I care is the ordinary care, sometimes called "normal newborn" or "well baby care" provided by regular pediatricians in hospital nurseries. Rule 10-5.011(1)(v)2.e.(I), Florida Administrative Code. As a practical matter, all care not swept into the proposed definition of Level II or Level III care becomes Level I care. This expands the care which may be provided in Florida Level I nurseries over that which could be provided under the hierarchy of care embodied in the Perinatal Guidelines. The proposed rule defines the minimum requirements for personnel, equipment, and support services for the two levels of neonatal intensive care services, methodologies to be used to determine the need for additional neonatal intensive care beds for each level of care, and a means to establish a beginning inventory of Level II and Level III neonatal intensive care beds in Florida. This implicitly involves an issue of grandfathering. The beginning inventory of the neonatal intensive care units in hospitals in Florida will include some units which were initiated without prior certificate of need approval through accretion of services over time. Existing units which met certain standards contained in the rule, which the Department believes are indicators of quality neonatal intensive care services, as of October 1, 1987, will be placed on the beginning inventory without being required to submit a CON application for those services, along with units which had obtained prior CON approval and units which are now licensed because they predated the CON program. A study by the Robert Wood Johnson Foundation has found that nationally: In urban areas, Level Is are upgrading to Level II status, and many Level IIs are upgrading to near- tertiary status. (Mease Exhibit 7 at 4) and Upgraded Level IIs and proliferating Level IIIs in a single area compete for a limited supply of high risk patients. These trends raise worrisome questions regarding system-wide cost efficiencies and comparative patient outcomes by hospital level. (Id. at 5) These competitive pressures account to a large extent for the blurring of roles among providers in the traditional three level hierarchy of care embodied in the Perinatal Guidelines construct. Neonatologists employed at Level I units tend to provide Level II services, which can be inappropriate. Although the neonatologist may be well trained, the provision of intensive care services is not an issue dominated by considerations of physician training but by issues of the experience and availability of nurses and other staff such as respiratory technicians, and allocation of resources by the hospital administration. The level of service a neonatologist has been trained to provide and wants to provide may not be cost-effective or represent wise public policy. Some hospitals have decried the Department's attempt to redefine in the proposed rule Level I, II, and III care in a manner which diverges from the Perinatal Guidelines. The Department is not required to accept the Guidelines developed by private orzanizations, but may craft a rule suited to Florida's needs. Those Guidelines definitions do not reflect current Florida practices; they are standards from the past. They were based on an implicit assumption that the health care system should and would marshal scarce resources to provide higher levels of manpower and technology in regional neonatal intensive care centers to serve as large a proportion of the population as possible. Hospitals today see a competitive advantage in offering higher level services and consequently do so. Competition has eroded the former altruistic sharing or transfer of patients which was designed to provide the greatest medical benefits to high risk neonates. (Mease Ex. 7 at 7). The percentage of neonates who require prolonged assisted ventilation, and thereby also require related nursing and other services which are doubtlessly intensive care services under any definition, are by birth weight: Birth Weight in Grams Percentage 2,500 and above---------------------1/2 of 1% 2,499 to 2,000 2% 1,999 to 1,500 4-5% 1,499 to 1,000 20% 999 and below-----------------------at least 75% The State of Florida, through the Department of Health and Rehabilitative Services, already has established a statewide system of Regional Perinatal Intensive Care Centers, known as the RPICC program. Chapter 10J-7, Florida Administrative Code. The hospitals participating in the program are large hospitals which have the facilities, equipment, and specialized medical, nursing, and technical staff to coordinate the transportation of and manage care for preterm and other severely ill neonates who need Level III care. RPICC step-down hospitals provide intensive nursing to neonates who do not require prolonged ventilation, or who cannot be placed because all RPICC beds are full. The Legislature has recognized that intensive care of neonates is a tertiary health care service which cannot be accomplished in a large number of hospitals offering obstetric services. See, Finding 16. Smaller community hospitals do not have the patient population to justify the significant expenditures necessary to recruit and retain subspecialists and sub-subspecialists, experienced nurses, and the equipment necessary to provide a high quality intensive care program for these neonates. A few community hospitals now offer neonatal intensive care services by engaging a firm of neonatologists to provide coverage at a hospital when the need for the services of a neonatologist arises, thus avoiding the need to have a neonatologist in the employ of the hospital. This practice can be acceptable if the neonatologist(s) are nearby and able to provide and direct care around the clock. It is also necessary for quality care that the neonatologist(s) regularly practice at the hospital and be on the active staff of the hospital, i.e., have full privileges at the hospital, without a cap on the number of admissions or on the duration of care, which is often the case with "courtesy" hospital staff. The point is that the neonatologist(s) should actually provide care at the hospital on a regular basis, and not merely lend their name(s) and credentials to a hospital through listing an itinerant neonatologist or members of a group of neonatologists on a form as a means for the hospital to meet regulatory staffing requirements. The proposed rule's staffing requirement that Level II units be directed by a neonatologist or group of neonatologists on the active staff with unlimited privileges is reasonable and appropriate. Rule 10- 5.011(1)(v)8.a.(I), Florida Administrative Code. See, also, Finding 32, post. Community hospitals have financial incentives to seek to be inventoried as Level II or even Level III neonatal intensive care hospitals. OB-GYNs and some pediatricians feel most comfortable in concentrating their medical practices at hospitals where experienced neonatologists, nurses, and specialized equipment are near at hand in emergencies. The risk of malpractice claims causes OB-GYNs to feel this pressure more acutely than many other physicians. They will even relocate their medical practices away from hospitals which cannot provide neonatal intensive care, in part because they view the potential for bad outcomes and malpractice claims as directly related to the consultants and resources available at their hospital. OB-GYN physicians who practice at hospitals which currently provide some neonatal intensive care services want those hospitals to retain the ability to provide those services to avoid disruption of their practices. Such a disruption would result if their sensitivity to malpractice claims impels them to move their practices to a larger hospital, because their current hospital fails to qualify as a Level II or Level III neonatal intensive care provider. Community hospitals in more affluent areas with a base of patients able to pay for their care through insurance or otherwise will sometimes incur the expense of employing one or more staff neonatologists. This encourages OB- GYNs to use the hospital for their obstetric patients, and is reassuring to more highly educated and demanding expectant mothers, who know a neonatologist and a supporting cluster of services are immediately available if needed. Hospitals understand that OB-GYN physicians tend to use the same hospital for both their obstetrics and gynecology practices. If the absence of a staff neonatologist causes OB-GYN physicians to avoid delivering babies at a hospital, that hospital will lose those doctors' gynecological surgery patients as well. Gynecology patients are desirable because hospitals find those services profitable. Background of the Regulatory Controversy. Before the certificate of need program was enacted, hospitals were free to establish services, including neonatal intensive care services. After enactment of the program, a hospital would be licensed to provide neonatal intensive care services or other services only if it obtained a CON to add new institutional health services. Section 381.494(1)(i), Florida Statutes (1981), Section 395.003(4), Florida Statutes (1983). After October 1, 1987, certificate of need approval has been required specifically for "[t]he establishment of tertiary health care services," Section 381.706(1)(m), Florida Statutes (1989). The statute defines a tertiary health service as a health service which, due to its high level of intensity, complexity, specialized or limited applicability, and cost, should be limited to, and concentrated in, a limited number of hospitals to insure quality, availability, and cost-effectiveness of such service. Examples of such service include, but are not limited to, . . . neonatal intensive care units . . . . Section 381.702(20), Florida Statutes (1989). The Department has proposed to adopt several versions of Rule 10-5.011(1)(v), regulating neonatal intensive care services. The Department published its first neonatal intensive care services rule on March 10, 1989. 15 F.A.W. 1056. That proposed rule would have (a) regulated the establishment of new Level II and Level III neonatal intensive care units, as defined in the rule, (b) established minimum requirements for personnel, equipment, and support services for these levels of neonatal intensive care, (c) established methodologies for determining the need for additional Level II and Level III beds, (d) established criteria which would treat certain current Level II and Level III services as eligible for placement on the beginning inventory if those services already had been licensed or received certificate of need approval, or were being offered in the manner described in the rule on October 1, 1987, and continuously thereafter. The rule initially proposed on March 10, 1989, would have qualified few hospitals for the Level III designation (the most intense level of care) due to the broad range of services required, which included availability of sub- subspecialists such as pediatric cardiac catheterization, neonatal cardiovascular surgery, and pediatric neurosurgery. HRS notified the parties of tentative amendments to the proposed rule on July 25, 1989, which were ultimately published in the Administrative Weekly on October 6, 1989. See, 15 F.A.W. 4574. These amendments would have enabled more hospitals to qualify as Level III units. The March 10, 1989, edition of the rule sparked a number of rule challenge proceedings, which were consolidated at the Division of Administrative Hearings as cases 89-1883R through 89-1892R. After the Department published its October 6, 1989, amendments to the proposed rule, it completely withdrew that rule on November 3, 1989. See, 15 F.A.W. 5140. The administrative challenges to the March version of the rule (cases 89-1883R through 89-1892R) therefore became moot, and were closed by an Order entered on November 16, 1989. The current version of the proposed rule was published on November 3, 1989, see, 15 F.A.W. 5052. Timely challenges to that edition of the rule were filed, and consolidated in this proceeding. All parties agree that each has a substantial interest in the outcome of this proceeding. All parties have standing. Brief Description of the Parties. Baptist Hospital is a 520 bed not-for-profit tertiary care facility located in Pensacola, Florida. It provides comprehensive obstetrical and newborn baby services, but does not serve newborns with birth weights of less than 1,000 grams (Mease Ex. 5). If the proposed rule is adopted, Baptist hopes to obtain grandfather status for a five-bed Level II unit, thereby avoiding the necessity of obtaining a certificate of need to continue to provide what the rule describes as Level II services. The rule in its current form may preclude this, because neonatal services at Baptist were not provided under the direction of a neonatologist on and after October 1, 1987. Florida Hospital is a 1,100 bed tertiary care hospital in the Orlando area which has provided services which would be Level II and Level III neonatal intensive care services under the proposed rule. Its neonatal intensive care unit began in 1983. It holds CON #3062, issued July 24, 1984 for Level II and III beds (Mease Ex. 5). In 1987 it opened a 19 bed unit. More than 40% of its neonatal intensive care services are rendered to indigents and are reimbursed by Medicaid. In the fiscal year from October 1, 1986, to September 30, 1987, (the last year before Section 381.706(1)(m) became effective), most neonates with birth weights less than 1,000 grams were transferred elsewhere (Mease Ex. 5). More recently, it has served a significant number of low birth weight infants; some were transferred there from Orlando Regional Medical Center. Holmes Regional Medical Center is a current provider of neonatal intensive care services with 15 Level II and 14 Level III beds. It has not received CON approval for these beds (Mease Ex. 5). Holmes regularly receives transfers of infants from Indian River Memorial Hospital, Lawnwood Medical Center, Cape Canaveral Hospital, Jess Parrish Hospital, and West Volusia Memorial Hospital. Roughly one-half of these transfers are Medicaid patients or indigents. In fiscal year 1986-87, its only neonate of less than 1,000 grams birth weight was transferred to another hospital (Mease Ex. 5). Recently it has served some very low birth weight neonates who were born at or transferred to its Level III unit. Mease Health Care, Inc., operates Mease Hospital, a not-for-profit community hospital in Dunedin, Florida. Under the proposed definition of Level II and Level III, Mease currently operates six Level II and four Level III neonatal intensive care beds. It has not obtained a CON for its beds (Mease Ex. 5). It serves few infants of 1,000 grams birth weight, 2 in fiscal year 1986- 87, 3 in fiscal year 1987-88, but this jumped to 10 in the first half of fiscal year 1989. (Id.) Its services are directed by a neonatologist and Mease has, on its staff, a second full-time neonatologist. Mease has provided neonatal intensive care services since at least 1981. Before Mease hired a staff neonatologist, pediatricians were uncomfortable in caring for high-risk mothers or neonates there, but the hospital also found that patients were upset if they had to be transferred to unfamiliar facilities due to their health care needs. If the intensive care services at Mease are not grandfathered, some OB-GYN physicians will likely relocate their practices to other hospitals where Level II or Level III neonatal intensive care services are available. Local physicians who currently hold admitting privileges at Mease regard this as a reversion to an earlier, unsatisfactory state of affairs. Winter Park Memorial Hospital Association operates Winter Park Hospital in a suburb of Orlando, Florida. It is located not far from Florida Hospital, and is also near Orlando Regional Medical Center. Since 1981 it has provided mechanical ventilation services to neonates who required it. Since 1983 Winter Park has had approximately 1,800 deliveries of newborns annually. It operates five Level II beds, which it has reported to the local health council, but it has not obtained a CON for these beds (Mease Ex. 5). It serves no infants of 1,000 grams birth weight or less. Fifty-three percent of all Level II care (by admissions) at Winter Park is provided to Medicaid patients. Since 1983 Winter Park has provided Level II care through the services of three consulting neonatologists. While these neonatologists are on the active staff of Orlando Regional Medical Center, they have admitting privileges at Winter Park, and can attend babies at Winter Park when necessary. A neonatal intensive care network in the Orlando area is developing among Winter Park, Orlando Regional Medical Center and Florida Hospital. Through this arrangement, Winter Park has provided Level II neonatal intensive care services without having a full-time neonatologist at the hospital. The past volume of neonatal intensive care services at Winter Park has not been such that a neonatologist would be willing to work exclusively at Winter Park. In fiscal year 1986-87, the hospital served 58 Level II neonates, who cumulatively used 136 patient days. When recruiting obstetrical staff, Winter Park administrators find that the first matter of concern to an obstetrician is the level of intensity of services which the hospital's nursery offers. If Winter Park fails to qualify for Level II status, OB-GYN physicians probably would move their practices from Winter Park to hospitals which would offer Level II services. 24 a. North Broward Hospital District is an independent special taxing district located in Broward County, Florida. It owns and operates Broward General Medical Center and Coral Springs Medical Center. Broward General is a 744 bed acute care hospital in Fort Lauderdale which presently operates 36 neonatal intensive care beds providing Level II and 27 beds providing Level III services. It holds a certificate of need for neonatal intensive care services issued on April 7, 1982. It serves a substantial number of neonates, including 67 in fiscal year 1986-87, of less than 1,000 grams birth weight (Mease Ex. 5). Coral Springs Medical Center is a 220 bed acute care hospital located in Coral Springs, in Broward County. It operates a six bed Level II neonatal intensive care facility under a certificate of need issued on August 13, 1984. It does not serve neonates of less than 1,000 grams birth weight (Mease Ex. 5). b. Similarly, the South Broward Hospital District is also an independent special taxing district. It operates Memorial Hospital in Hollywood, Florida, a large public hospital which received certificate of need approval on February 18, 1982, for 8 Level II and 15 Level III beds. In fiscal year 1986-87, it served 40 neonates of less than 1,000 grams birth weight. It is a major indigent care provider. Sacred Heart Hospital is a 398 bed acute care hospital in Pensacola, Florida. It provides high risk obstetrical care, and had about 3,400 deliveries in 1989. Its neonatal intensive care unit has 22 Level II and 18 Level III beds staffed by five board certified neonatologists. Some data from this hospital is skewed by its practice of admitting all neonates who need intensive care to its Level III unit for evaluation. Those with lesser needs are then stepped-down into its Level II beds. Sacred Heart began neonatal intensive care in the early 1970's, before the CON program began, and became a Regional Perinatal Intensive Care Center in 1974. It serves a 10-county area in Northwestern Florida. In 1987, it served 65 neonates of less than 1,000 grams birth weight, which was 11% of total admissions. Medicare reimburses the hospital for half of its neonatal intensive care patients, which indicates that it serves many indigents. Community Health Systems, Inc. d/b/a Women's Medical Center is a licensed acute care hospital which will be substantially affected by the proposed rule, as is Alachua General Hospital. The Service Level Definitions of the Proposed Rule. Introduction. The Department conducted two surveys which were sent to hospitals providing obstetrical services, one in 1987 and one in 1989, to assess the types of services that were being offered, the levels of staffing (including such things as the number of nurses per shift and whether the hospital had a full- time neonatologist), the number of neonatal intensive care patient days, whether the unit had obtained CON approval and other information. The purpose of these surveys was to obtain a preliminary inventory of the types of services that were actually being offered in Florida hospitals and give a general overview of the existing service delivery system which had developed in Florida. During the rule promulgation process, the Department also received comments from a number of neonatologists practicing in Florida hospitals and from hospitals. According to the surveys, only 2 of 49 hospitals which claimed to operate Level II units lacked ventilators. Traditional Perinatal Guidelines Levels of Care. The terms Level I, Level II, and Level III neonatal intensive care originate from the Perinatal Guidelines. Under the traditional Guidelines definition, a Level I nursery provides routine care to normal full term or pre- term neonates who weigh more than 2,000 grams at birth. According to a summary table in the Perinatal Guidelines, the babies in a traditional Level I unit present uncomplicated, emergency and remedial problems such as the lack of progress in labor, immediate resuscitation of asphyxiated neonates, uterine antony, nursing care of large premature neonates [those weighing greater than 2,000 grams] without risk factors, or physiologic jaundice. Guidelines, Table A-1, at 288. They are usually cared for in bassinets. Level II units under the traditional Guidelines definition provide services to sick neonates who do not require what the Guidelines define as intensive care, but who need six to twelve hours of nursing care each day. Hours of nursing care are determined by dividing the number of nurses by the number of babies in the nursery. Thus, a hospital which maintains three nurses on the nursery floor for six babies is said to provide twelve hours of nursing care to each baby. This is not a measure of the time actually devoted to each newborn, but a measure of the intensity of staffing in the nursery. These are related however, because hospitals try to minimize personnel costs. Hospital administrators staff a nursery to provide more than 6 hours of care per newborn only if the condition of the newborns require it. Under the Perinatal Guidelines, Level II units provide Level I services, plus selected problems such as preeclampsia, premature labor at 32 weeks and later, mild to moderate respiratory distress syndrome, suspected neonatal sepsis, hypoglycemia, neonates of diabetic mothers, post- asphyxia without life-threatening sequelae. Guidelines, Table A-1 at 288. Level II services described by the Guidelines are sometimes also characterized as "intermediate care". These neonates may be cared for in radiant warmers or incubators, which can accommodate IV lines, pumps and monitors. The radiant warmers are expensive, costing about $25,000 each after installation. Care involving IV lines or monitoring cannot be done in a bassinet. Newborns who require complex surgical care, or cardio-pulmonary care such as prolonged assisted mechanical ventilation (ventilation for more than resuscitation and stabilization purposes) are ordinarily moved to an intensive care area within the Level II unit, which the Perinatal Guidelines describe as a Level III unit. Infants who need prolonged ventilation require the care of a neonatologist and a minimum of 12 hours nursing care (i.e., a ratio of at least 1 nurse for 2 neonates). Based on the Department's survey, most Florida Level II units had ventilation services, and met the Perinatal Guidelines definition of a Level III unit. It is essential to understand that the Perinatal Guidelines are primarily standards of care, and they do not contain operational definitions that are suited to regulatory enforcement. The definitions of Level I, II, and III neonatal intensive care services found in the Perinatal Guidelines do not describe the actual system of health care delivery to neonates in Florida. For example, the Guidelines recommend 3-4 Level II and one bed per 1,000 births for Level III services. Florida had approximately 185,000 births in 1988, which under the guidelines would show a need for only 185 Level III neonatal beds. In reality, the Department's preliminary survey data indicated that Florida had 391 Level III beds. In part because the Perinatal Guidelines definitions would indicate a significantly lower number of Level III beds than are in use, the Department determined that the Guidelines need methodology should not be adopted without modification. Division Among Levels of Care in the Proposed Rule. The new definitions the Department has proposed to adopt for Level II and Level III use functional distinctions that are significant indicators of the level or intensity of services provided in a nursery. These distinctions reflect the reality that there is a need for a hierarchical clustering of services, in terms of the availability of medical specialists, subspecialists and sub-subspecialists, the intensity of nursing care, and the availability of respiratory therapy technicians and equipment, as hospitals deal with smaller and smaller neonates, who require prolonged ventilation and other interdisciplinary care. Level I versus Level II Neonatal Intensive Care Services. The definition of Level I services in the proposed rule encompasses intravenous medication and tube feedings (gavage) to neonates, and administration of oxygen by hood, which would be traditional Level II intermediate care services under the Perinatal Guidelines. The major distinction between the services authorized under the proposed rule for Level I versus Level II nurseries are that a Level I unit cannot offer prolonged assisted ventilation; it can provide mechanical ventilation to a newborn for resuscitation and stabilization only. A Level II unit may provide prolonged ventilation, and it must provide those services under the direction of a neonatologist or group of neonatologists. Rule 10-5.011(1)(v)2.e.(I)(II) and 8.a.(I), Florida Administrative Code. Under contemporary standards of care, prolonged mechanical ventilation is a complex procedure which requires extensive training due to its significant mortality rate. Providing artificial ventilation to an infant is not the only distinction in the proposed rule between Level I and Level II services, however. A unit routinely providing less than six hours of nursing care to its newborns is providing care which is appropriately rendered by a pediatrician. The proposed definition of Level II neonatal intensive care services also includes provision of no less than six hours of nursing care per day. Management of conditions which require six or more hours nursing care per day likely transcend the skills and training of an ordinary pediatrician and require care directed by a neonatologist. This is not a bright line standard, but relates to the clustering of services typically required by seriously ill neonates. If a unit commonly serves neonates who are so ill that the hospital routinely must staff the unit so that the neonates receive more than six hours of nursing care a day, the unit is likely operating an intensive care nursery. This is so even though the proposed rule's definition of Level I care includes some services traditionally regarded as intermediate care. Cost pressures will dissuade genuine Level I units from staffing to provide more than six hours of nursing care per day. There is, however, no restriction on Level I units forbidding them from providing six hours or more of nursing care to its newborns. Transport Between Units. The need to transport a neonate from a Level I to Level II or a Level II to a Level III unit can be minimized by appropriate pre-natal care. Women in pre-term labor with gestation which correlates with low birth weight can be directed to a Level II unit, or if gestation is compatible with very low birth weight (i.e. less than 1,000 grams), to a Level III unit which will have the services of a maternal-fetal specialist who can attempt to arrest the premature labor. See, Rule 10-5.011(1)(v) 8.a. (II), Florida Administrative Code. This transfer of the mother and fetus as a unit is preferable to transfer of a neonate. When that is not possible, however it is appropriate to arrange transfer from a Level I to Level II or Level II to Level III unit. It is true that there are some risks in transferring a neonate of less than 1,000 grams from a Level II to a Level III center. It can be difficult to maintain the infant's temperature, and it is possible that the ventilation equipment being used during transport can be dislodged from the infant's windpipe, and that there can be some risk of brain hemorrhaging. Fewer than 1% of neonates who are transported die during transport, however, and the increased likelihood of a good outcome by treating the infant in an enriched setting with respect to the experience of the medical, nursing and technical staff and the availability of equipment outweighs this risk. Similarly, the increased cost of transportation to the parents is counter-balanced by the higher level of care available in a Level III center for low birth weight neonates. For example, approximately 3/4 of infants between 500 and 1,000 grams require prolonged assisted ventilation, that is, assisted ventilation for several hours or even many days. They commonly require the care of a single nurse on a 24-hour basis because they are on a ventilator, have an IV, and multiple pumps and monitors. The practical effect of birth weight transfer rule should not be the transfer of most low birth weight infants, but the direction of mothers who are likely to have low birth weight babies to Level III centers in the first instance. On an institution-specific basis, there may be a change in transfer patterns at a small number of hospitals with low service volumes. Level II versus Level III Neonatal Intensive Care Services. The proposed rule also requires that Level II units restrict themselves to care to neonates of 1,000 grams birth weight and over, with the exception that a Level II facility may serve a neonate of less than 1,000 grams birth weight only while waiting to transport the baby to a Level III facility. Rule 10-5.011(1)(v)2.e., (II), (III), Florida Administrative Code. The 1,000 gram birth weight limit on Level II units is a reasonable one. Few neonates weigh less than 1,000 grams at birth; the population is small, and the technologic requirement to care for them are great. (Tr. 1179). Data demonstrate that morbidity and mortality for infants of less than 1,000 grams is not as good when there is a small number of infants cared for as compared to care in regional Level III centers which handle a moderate volume of such infants. (Tr. 1259-63; Sacred Heart Ex. 1; see, also Finding 5, above). Small neonates often need interdisciplinary medical care, including care by a cardiologist as well as by a neonatologist. Anecdotal data from community hospitals such as Mease hospital, which recently began caring for a number of very low birth weight neonates and experienced good results with that small sample of children does not alter this fact. Consequently, the rule provision which would restrict Level II centers from handling infants with less than 1,000 grams birth weight has a reasonable medical basis. It is also an operational standard suitable for regulatory enforcement. This provision impacts, to some extent, on the judgment of neonatologists about whether to care for a neonatal of less than 1,000 grams birth weight in a Level II center; the issue involved is not only the physician's assessment of his or her own skill and ability to care for the infant in a Level II center, however. It also raises questions about the training of nurses, their skills, and the resources available in a Level II hospital, such as access to subspeciality care. A survey of Florida neonatologists indicates that they disagree among themselves on the appropriate birth weight cut-off for transfer to a Level III unit. Opinions range from requiring transfer at 1,250 grams, 1,000 grams, or 800 grams to deletion of the transfer requirement. As congressional hearings found in another context (dealing with the efficacy of drugs), "impressions or beliefs of physicians, no matter how fervently held, are treacherous." Hearings on S. 1552 before the Subcommittee on Antitrust and Monopoly of the Senate Committee on the Judiciary 87th Cong., 1st Sess., pt. 1 at 195, 282, 411-12.1 Data should be preferable to physician plebiscite in establishing standards of care. Dr. Curran's testimony concerning studies in Florida and Texas and Dr. Alpin's testimony on Florida data on outcomes of treatment of low birth weight infants form an adequate basis for choosing the 1,000 gram cut off. (Tr. 1411, Sacred Heart Ex 1.) The argument that the cut point is the result of impermissible compromise is rejected. The choice was not an unprincipled one, such as a compromise jury verdict or the compromise invalidated in the Adam Smith case, discussed post at Dr. Curran merely acknowledged that colleagues could come to different conclusions on the weight standard, and he might choose a higher one were the matter left up to him. The actual practice of hospitals such as Baptist, Florida Hospital, Holmes Regional, Winter Park and Coral Springs in fiscal year 1986-87 of transferring almost all neonates of less than 1,000 grams show that the threshold is reasonable and consistent with contemporary medical practice. See, Findings 19, 20, 21 and 23, above. Level III units not only care for very low birth weight neonates, but for all neonates who need sub-subspecialist care, such as complex pediatric surgery, neonatal cardiovascular surgery, the interdisciplinary efforts of pediatric neurology and neurosurgery (usually neonates suffering from hydrocephalus) or pediatric cardiac catheterization. These are services appropriately limited to regional medical centers, because very few physicians provide such care. In recognition of this, the rule definition of Level III care permits Level III centers to make care available for major congenital anomalies that require the care of a pediatric surgeon, pediatric cardiovascular surgery and cardiac catheterization by means of a written agreement with another Level III unit in the same or adjacent service district. Rule 10- 5.011(1)(v)2.e. (III), Florida Administrative Code. Each Level III provider need not have these sub-subspecialists on active staff. The rule contains a conflict, however. Rule 10-5.011(1)(v)10.a., requires in the "Level III Neonatal Intensive Care Unit Standards" that Level III units "shall have a pediatric surgeon on active staff who is either board certified or board eligible in pediatric surgery." The definition of a Level III unit and this substantive provision are mutually exclusive. Because pediatric surgeons are so few, the testimony that the definition should prevail, and the substantive provision be dropped, is persuasive. (Tr. 1299-1301). These services can appropriately be provided through written agreement with a pediatric surgeon nearby. Grandfather Provisions. General. Proposed Rule 10-5.011(1)(v)14.a., Florida Administrative Code, would place on the beginning inventory of Level II and III units hospitals which obtained a certificate of need to provide neonatal intensive care services, but limits them to the number of neonatal intensive care beds approved in the certificate of need, unless the provisions of Rule sub-subparagraph 14d. authorizes additional beds. Hospitals which provided Level II or Level III neonatal intensive care services before October 1, 1987, and continuously thereafter, which were disclosed on the hospital's licensure application and which were provided under the direction of a neonatologist or a group of neonatologists will be inventoried under Rule 10-5.011(1)(v)14.b., Florida Administrative Code, for the number of neonatal intensive care beds which the Department has accepted on the hospital's most recent application for licensure, unless the provisions of Rule sub-subparagraph 14.d. authorizes additional beds. Finally, proposed Rule 10-5.011(1)(v)14.f., Florida Administrative Code, permits hospitals which claim to have provided Level II or Level III services before October 1, 1987, and continuously thereafter which had not been authorized by a certificate of need or licensed by the Department, to provide Level II and Level III services if certain conditions are met. For Level II units those hospitals must have instituted services before October 1, 1987, under the direction of a neonatologist or a group of neonatologists who are board certified or board eligible in neonatal medicine, who had been on the active staff of the hospital with unlimited privileges, and provided 24-hour coverage for the hospital. Rule 10-5.011(1)(v)14.f.(I), Florida Administrative Code. Hospitals which claimed to have had Level III units before October 1, 1987, must have provided services under the direction of a neonatologist or a group of neonatologists, and at least one of the following must also have been true: The average length of stay for all neonatal intensive care patients, regardless of whether they were Level II or Level III patients, must have been at least 10 days; At least 5% of all neonates admitted to the intensive care unit, regardless of Level II or Level III status, weighed less than 1,000 grams at birth; or At least 50% of all neonates admitted, regardless of their reported Level II or Level III status, were classified within diagnosis related groups (DRGs) 385, 386, 387 or 388. Rule 10-5.011(1)(v)14.f.(II)(B) I., II., and III, Florida Administrative Code. In all cases, however, the number of beds inventoried for Level II and III services will be at least five. Rule 10-5.011(1)(v)14.d., g., Florida Administrative Code. In other words, the number of beds authorized will be the greater of the number listed on a certificate of need, the hospital license application, or based upon the need calculation set forth in sub-subparagraph 14.c., but in no event fewer than five. One of the Department's goals is to avoid a disruption of existing service capacity when implementing its proposed rule. See, Finding 9, above. In the absence of a grandfather provision, facilities which instituted neonatal intensive care services after the certificate of need program became effective but which failed to obtain a certificate of need would be required to terminate their services. Given the incentives which have led many hospitals to escalate the level of neonatal intensive care services they offer, such as 1) pressures from OB/GYN physicians, 2) confusion over whether traditional intermediate care under the Perinatal Guidelines required a CON, and 3) competitive pressures, such a reduction in the existing number of beds would be contrary to the best interests of the public. The Department has therefore structured grandfather provisions to permit hospitals which a) obtained certificates of need, b) were licensed for neonatal intensive care beds because their service pre-dated the CON program, or c) actually provided neonatal intensive care services without a CON or licensure in a manner which meets certain standards, to continue to do so. This avoids a significant adverse impact on the availability of neonatal intensive care services during the time between the effective date of the rule and the replacement of those existing beds through the certificate of need process. Although a hospital may receive authorization for Level II or Level III beds, there is no requirement that a hospital actually equip, staff and license all those beds. Review of CON Applications for Neonatal Intensive Care Services Before October 1, 1987. Historial Background. The Department issued a policy memorandum concerning standards for the licensure designation of hospital bassinets and neonatal beds on January 2, 1981. The memorandum memorialized an agreement between the Department's Office of Health Planning and Development, which issued certificates of need, and its Office of Licensure and Certification, which issued hospital licenses. Under the agreement, a hospital was not entitled to a license for a neonatal intensive care bed unless the hospital had obtained a certificate of need for that bed, or unless the hospital had that bed in service before the certificate of need program came into effect. A bassinet was not separately licensed as a bed at the hospital because a bassinet was regarded as "the normally expected adjunct to the mother's occupancy of a post-partum hospital bed." HRS Exhibit 1b. On the other hand, a neonatal intensive care bed was "an infant unit within a neonatal continuing or a intensive care area which is designed to provide care to high risk neonates presenting a complexity of problems and requiring the coordinated activities of a highly skilled hospital personnel team using specialized equipment" (Id.). These definitions were derived from a manual entitled Standards and Recommendations for Hospital Care of Newborn Infants published by the American Academy of Pediatrics. As a result, normal newborn bassinets were considered an adjunct to the mother's bed and were not counted in the hospital's bed inventory. In conformity with this policy memorandum, the Department advised Broward General Medical Center in September, 1981 to obtain a certificate of need for its existing compliment of 30 unlicensed intermediate care beds in connection with an expansion project in which that hospital wished to add 33 new beds. As a result, Broward General applied for and received certificate of need Number 1971 for a total of 63 neonatal intensive care beds. Similarly, Coral Springs Medical Center obtained a certificate of need for neonatal intensive care services in 1984 to convert up to 12 of its previously approved 220 acute care beds to neonatal intensive care services. The Department has not consistently followed its memorandum since 1981. The Department knew that a significant but indeterminate number of hospitals throughout the State of Florida had initiated what the Perinatal Guidelines would have classified as Level II and even Level III intensive care services without obtaining certificates of need for those beds. In one instance, the Department filed an Administrative Complaint against HCA Health Services of Florida d/b/a Northwest Regional Hospital for establishing a neonatal intensive care service without obtaining a certificate of need, which is a prerequisite for licensure of neonatal intensive care beds. That matter has been placed in abeyance pending the outcome of this rule challenge proceeding. The practical effect of the inventory process proposed in the rule would be to authorize the continued operation of neonatal intensive care services by hospitals which initiated them before October 1, 1987, and operated them according to certain standards without requiring them now to obtain a certificate of need. To avoid a substantial disruption in the availability of neonatal intensive care services, some grandfather provision of a type roughly similar to that proposed by the Department is essential. 2. Department Action on CON Applications. When the Department acted upon applications to establish neonatal intensive care services submitted since 1981, it had no uniform definition promulgated in rule for what constituted Level II and Level III neonatal intensive care services. The Department relied instead on whatever definition of services the applicant used in its CON application. The Department has no statutory authority to eliminate certificates of need it has already issued for the establishment of Level II and Level III intensive care services. The lack of prior standards means that some hospitals which applied for and received CONs for traditional Pediatric Guidelines Level II neonatal intensive care units (which would not have provided prolonged assisted ventilation), would be treated as having Level II approval under the proposed rule, and be permitted to provide prolonged assisted ventilation. The surveys which the Department sent to hospitals currently offering neonatal intensive care demonstrate that hospitals which have Level II certificate of need approval are, in fact, usually providing prolonged assisted ventilation today. This is consistent with the general accretion of neonatal intensive care services which has been taking place in Florida hospitals. It is rational, therefore, for the Department to treat previously-issued certificates of need for Level II and Level III beds as beds which should be placed in the beginning inventory notwithstanding the variations in the definitions of Level II and Level III services among the Perinatal Guidelines, the CON applications, and the proposed rule. These providers have one year to bring their previously approved units into compliance with the quality of care standards contained in subparagraphs 8; 9; 10; 11; and 12. of the proposed rule. Rule 10-5.011(1)(v) 14.i., Florida Administrative Code. Other Challenges to Level II Grandfathering Provisions. Baptist Hospital has objected to testimony from Robert Griffin, the Assistant Secretary in charge of the Department Office promulgating this rule, that the Department intends to automatically deem regional perinatal care center "step-down" hospitals as meeting the Level II grandfather requirement. RPICC step-down units provide ventilator assistance to neonates for stabilization purposes, under contract with the Department's Office of the Children's Medical Services, and must immediately transport the neonate to a RPICC neonatal intensive care unit if prolonged ventilation is necessary. Rule 10J- 7.003(2)(b)4., Florida Administrative Code. No provision in the text of the proposed rule provides for the automatic grandfathering of all RPICC step-down units, notwithstanding Mr. Griffin's testimony. An RPICC step-down hospital which has no certificate of need for neonatal intensive care beds, nor licensure approval for Level II neonatal intensive care beds, must demonstrate that it has provided Level II services under the direction of a neonatologist prior and subsequent to October 1, 1987, just like any other hospital. RPICC step-down units do have the capability to provide prolonged assisted ventilation, and out of necessity have provided ventilation in the past. According to HRS survey data, all step-down units have a neonatologist and all are equipped with ventilators. Their directors should be neonatologists. Rule 10J- 7.003(2)(a)1.a., Florida Administrative Code. Whether they will qualify to be inventoried for grandfathering purposes is beyond the scope of this rule challenge proceeding. Such issues may be raised in the challenge to the inventory. Rule 10-5.01(1)(v)15., Florida Administrative Code. Baptist Hospital also is concerned that the rule might have the effect of terminating its Level II neonatal intensive care services. The question in this proceeding is whether the proposed rule is valid, not the effect that the rule may have on Baptist Hospital, or any other individual hospital. Under Rule 10-5.011(1)(v) 14.b., Florida Administrative Code, providers which filed license applications disclosing neonatal intensive care beds which pre-dated the CON program will be inventoried if they provided their Level II services under the direction of a neonatologist by October 1, 1987, and continuously thereafter. This is reasonable because it is consistent with contemporary standards of care, and because survey data collected by the Department in preparing the rule show that all hospitals claiming Level II beds which pre-dated the CON program had a neonatologist prior to October 1, 1987, and thereafter. For the rule to have provided otherwise would contract the availability of neonatal services in an unwarranted manner. For hospitals which claimed to have provided Level II services before October 1, 1987, without a CON or licensure, the survey data shows that 90% had a neonatologist directing their units. Of 27 providers who had a CON or license application disclosing a Level II unit, only two reported that they did not have a neonatologist as a director before October 1, 1987. The survey data also shows that of the facilities which did not have a neonatologist, the diagnostic related group distributions were different from facilities which had a neonatologist. The average length of stay was significantly different between these groups. Consensus developed during the proceedings leading to rule promulgation that prolonged assisted ventilation should be provided under the direction of a neonatologist. It is therefore medically reasonable for proposed Rule 10- 5.011(1)(v)14.f.(I), Florida Administrative Code, to require that neonatologists have been available to direct coverage on a 24-hour basis. Although the presence of a neonatologist is not conclusive evidence that a facility provided prolonged assisted ventilation, it is a very reliable indicator of a pre-existing Level II unit. It is quite unlikely that a facility with a neonatologist would be providing only normal newborn or well baby services under the Perinatal Guidelines. Such a neonatologist would be significantly under- employed in terms of her skills. Similarly, the mere listing of an itinerant neonatologist on the hospital roster who was not actually available for around-the-clock care would not be an indication of a genuine pre-existing Level II unit. Thus, the provisions of proposed Rule 10-5.011(1)(v)14.b and 14.f do not arbitrarily discriminate against unlicensed or uncertified traditional Level II providers which had no director of neonatology on staff by October 1, 1987. Baptist also objects because the text of the rule does not include a specific definition of what it means to have had a Level II unit before October 1, 1987 under the "direction" of a neonatologist. Rule 10-5.011(1)(v)14.f.(I), Florida Administrative Code. That portion of the rule is designed to screen out those hospitals which seek to be inventoried when they only had a neonatologist who occasionally came to the hospital on a itinerant basis, and did not provide services at the hospital as a member of the regular active staff. Such limitation is reasonable. Tripartite Test for Grandfathering Level III Providers Who Have no CON or License Designation. Proposed Rule 10-5.011(1)(v)14.f.(II)(B) describes grandfather criteria for hospitals claiming Level III services if the hospital had no certificate of need or license designation for Level III beds. Such hospitals must meet one prong of a three part or tripartite test to show that they actually operated a Level III unit, and therefore should be grandfathered. The rule requires that a) the average length of stay for all neonatal intensive care services must have been at least ten days, or b) at least 5% of the neonates admitted must have weighed less than 1,000 grams at birth, or c) 50% of all neonates admitted must have been classified into diagnosis related groups 385, 386, 387, or 388. In the RPICC Level III facilities, the average length of stay is approximately 20 days. These are clearly Level III facilities, and by requiring only one-half their average length of stay, the Department's choice is reasonable. Similarly, for RPICC centers, 10% of their neonates weighed less 1,000 grams at birth, so requiring a hospital seeking grandfathering to demonstrate that at least 5% of their patient census consisted of neonates with a birth weight less than 1,000 grams is reasonable. Lastly, the DRG classifications that would normally include Level III neonates are predominantly the ones identified by the Department, although some Level III babies may also fall within DRG 389. DRG 389 is "newborn with other conditions" and refers to a large group of patients. It is the single DRG for full term babies that is most susceptible to difficulty in interpretation. Babies in this DRG include those with jaundice and septic work-ups, many of whom have short lengths of stay and do not require intensive, Level III services. It was reasonable, therefore, for the Department to omit this DRG. The Department's survey indicates that almost every facility claiming to have Level III services would meet at least 2 of the 3 criteria or prongs of the tripartite test, with the exception of one hospital (Miami Children's Hospital, which is one of the few hospitals which is not a party to this proceeding). Of the remaining four facilities which assert that they have Level III services, but for which preliminary survey data indicate they would not qualify for grandfathering, none meet even one of the three criteria in the tripartite test. This consistency in the results across all three tests is an indication that the factors embodied in each prong of the test are reasonable. It is true that small neonatal intensive care units may have difficulty in reaching the 5% cut-off in the second prong of the tripartite test, but due to the tertiary nature of Level III units, it is to be expected small units will have difficulty qualifying either under the grandfathering provision, or under the rule itself. This is inherent in tertiary services. Need Methodology. Introduction. The opponents of the rule attack the need methodology, in large part, because it will restrict the hospitals which will qualify as Level II and Level III units. They believe it improperly stifles competition. Petitioners overlook the significant difference between this rule, and many of the Department's other need methodologies: this methodology relates to a tertiary service, which by its nature, and by Section 381.702(20), Florida Statutes, is designed to be limited to and concentrated in few hospitals in order to insure quality care, service availability, and cost effectiveness. The Department specifically seeks to discourage the proliferation of new tertiary care services, and to protect the financial viability of the neonatal intensive care units currently operated by large urban facilities which provide a substantial amount of Medicaid and charity care. Most of these hospitals are already Regional Perinatal Intensive Care Centers. Moreover, there is an established correlation between indigency and high risk births, so it is rational to attempt to concentrate neonatal intensive care services in hospitals which provide a great deal of charity care. The proposed rule contains separate need methodologies for Level II and Level III neonatal intensive care services. The methodology for Level II services is contained in Rule 10-5.011(1)(v)3.c. and d. An alternative methodology for certain existing providers is also provided in subparagraph 3.g., for existing Level II providers who find that their occupancy rate exceeds 90%. This provision effectively creates a 2-tier need assessment process. The Occupancy Standards. To obtain CON approval for a neonatal intensive care unit, hospitals which do not already have neonatal intensive care beds must normally demonstrate that all Level II neonatal intensive care beds in the district as a whole had an 80% occupancy rate, and the need formula must predict a net numerical need for additional beds. On the other hand, existing providers will be permitted to add beds to their units if their occupancy rate exceeds 90%. This means that the high-volume indigent care hospitals will have more access to additional beds than potential new providers. This 2-tier approach for new and existing providers does not violate the statutory requirement that there be uniform need methodology, because this same methodology applies throughout the state from district to district. It is therefore uniform. It is difficult to understand the argument advanced by Mease Hospital that "the proposed rule does not contain any extraordinary circumstances provision" (Supplemental Proposed Recommended Order at 17, paragraph 84). The proposed rule states that "a favorable need determination for Level II or Level III beds will not normally be granted unless a bed need exists according to the need methodology . . ." Rule 10-5.011(1)(v)3.a., Florida Administrative Code. This "not normally" language embraces extraordinary circumstances. The desired occupancy standard of 80% in order to approve beds for new providers is within the range of occupancy generally accepted in health care planning, viz., 80-85%. While health planners may disagree on which percentage should be adopted, HRS received and evaluated criticisms of the initial proposal for an 85% occupancy rate, and lowered it to 80% as a means of accommodating the criticism that it might be appropriate to include a "Poisson adjustment" to the desired occupancy standard. A Poisson adjustment is sometimes used in health care planning to expand the number of beds available where the service involved has seasonal peak demand. While reducing the occupancy rate from 85% to 80% is less elegant than incorporating a Poisson adjustment, the Department's approach is reasonable. The second tier of the need methodology, which permits providers with extremely high occupancy rates (over 90%) to obtain beds despite an average occupancy rate in the district of below 80% and despite the absence of numeric need on a district-wide basis according to a utilization formula is reasonable. In the rule promulgation process, the Department learned that many publicly funded units are operating at very high occupancy rates. For example, Jackson Memorial in Miami operates at 122.4% occupancy, Hollywood Memorial in South Broward at 138.1%, St. Mary's Hospital in Palm Beach at 114.8%, Orlando Regional Medical Center at 120.5%, and Shands Hospital in Gainesville at 164.7%. On the other hand, other hospitals have rather low occupancy rates in their neonatal units. In District X, Broward County, where Hollywood Memorial operates at 138.1% occupancy, Broward General operates at about 41.3%, Coral Springs at 30%, Holy Cross at 63.2%, Humana South Broward at 6.8%, and Plantation General at 19.3%. The high occupancy providers are largely publicly-funded hospitals, and they ought to be able to add beds to deal with the population which they serve (which is often indigent) even if there is no numerical need as a consequence of low utilization at other hospitals within the district. The 2-tier need methodology furthers the goal of making neonatal intensive care services economically accessible. Similarly, the proposed rule contains a preference when a number of hospitals are competing for a limited pool of neonatal intensive care beds for hospitals which have contracts with the Department's Children's Medical Services program. Rule 10-5.011(1)(v)3.i., Florida Administrative Code. A large percentage of neonatal intensive care unit patient days are provided at Regional Perinatal Intensive Care Centers and associated step-down unit hospitals. This preference insures that large high-volume indigent care providers will have beds to serve a population which is largely medically indigent and which is disproportionately the source of neonates requiring intensive care. Similarly, the comparative preference written into Rule 10-5.011(1)(v)3.k., Florida Administrative Code, for hospitals which provide care to non-Children's Medical Service patients who need charity care, according to the manner in which charity care is determined by the Health Care Cost Containment Board, is reasonable and appropriate. Minimum Unit Sizes and Birth Volumes. The provision for minimum unit sizes found in the rule is consistent with the legislative determination in a Section 381.702(20), Florida Statutes (1989), that neonatal intensive care services be concentrated at a limited number of hospitals. The National Health Planning Guidelines recommend minimum unit sizes of 15 beds for Level II and Level III services. 43 Fed. Reg. 13040, 13047 (March 28, 1978). The federal guidelines note that both standards are supported by the American Academy of Pediatrics. Id. These 10 and 15 bed minimum unit sizes were used as a basis for developing Florida's Perinatal Intensive Care Program, and are related to quality of care by insuring that a facility has a sufficient volume to support adequate staffing and ancillary support, as well as enough work for medical personnel to keep their skills current. Although there was some testimony that units as small as four beds could maintain staff proficiency, contrary testimony (see, e.g., Tr. 1153), was more persuasive. Such small unit sizes lead to diseconomies of scale. They are clearly inconsistent with the legislative mandate that these services be limited to and concentrated in a small number of hospitals. As a related matter, the minimum birth volume requirements, which provide that a hospital will "not normally be approved for" a Level II unit without a minimum volume of 1,000 live births or Level III without a minimum service volume of 1,500 live births is reasonable and appropriate. See, Rule 10-5.011(1)(v) 6., Florida Administrative Code. Approximately 10-11% of neonates require neonatal intensive care services. Hospitals with relatively low birth volumes cannot adequately support an intensive care service. A minimum number of patients are necessary to generate adequate physician and ancillary support. Economic Impact Statement. The Department engaged in a reasonable analysis of the possible economic impact of the proposed rule. Assessing the impact is, necessarily, a difficult process. The economic impact statement accompanying the proposed rule addresses the cost to the agency of implementing the proposed rule; an estimate of the cost to persons directly affected by the proposed rule; an estimate of the impact of the proposed action on competition; a statement of the data and method used in making those estimates; and an analysis of the impact on small business as defined in the Florida Small and Minority Business Assistance Act of 1985. (Nu-Med Ex. 7). The rule will allow Level I units to keep some babies that would have been transferred to Level II units under the traditional Perinatal Guidelines categorization of units. It will also allow Level II units to conduct prolonged ventilation, which was a Level III service under the traditional Perinatal Guidelines, thus reducing the need for transfers from Level I to Level II units and from Level II to Level III units. In terms of overall impact, there is no reason to believe that there will be a significant change in the total transportation patterns of neonates throughout the state. There may be some impact on neonatologists, if they have been employed by hospitals which did not have authority to conduct a Level II or Level III neonatal intensive care service. There is little likelihood that the rule will have an adverse economic impact on the overall employability of neonatologists in the general market place, but they maybe required to relocate their practices from small community hospitals to larger medical centers. The economic impact statement does not take into consideration the losses which might follow if a hospital is prohibited from advertising its nursery as a Level II or Level III unit, because it fails to qualify under the rule, and consequently OB-GYN physicians decide to move their practices elsewhere. That would be an extremely difficult calculation to make, and the failure of the economic impact statement to attempt to quantify that possible shift in OB-GYN practices does not impair the fairness of the rulemaking proceeding.
The Issue Whether the Agency for Health Care Administration has a policy regarding the determination of the need for long term care beds which constitutes a rule and, if so, whether rulemaking is feasible and practicable.
Findings Of Fact Vencor Hospitals South, Inc. (Vencor), applied for a certificate of need (CON No. 8614) to establish a 60-bed long term care hospital in Agency for Health Care Administration (AHCA) District 8, for Fort Myers, Lee County, Florida. AHCA is the state agency authorized to administer the CON program for health care services and facilities in Florida. AHCA reviewed and preliminarily denied Vencor's application for CON No. 8614. The reasons for AHCA's actions on this or any other CON application are memorialized in documents called State Agency Action Reports (SAARs). Vencor alleges that the following statement generally describes AHCA's policy in regard to the review of CON applications for long term care hospitals: Long term care is not a separate category of health service, but is instead merely an allowable form of reimbursement pursuant to Medicare regulations. The care provided in acute care hospitals, hospital based skilled nursing beds, "subacute" care in nursing homes, and care at rehabilitation facilities, are all equivalent to the care provided at long term care hospitals. Therefore, in evaluating the need for long term care hospital beds, AHCA will assess the availability of other categories of beds and services to meet the need for the services proposed by the applicant for long term care hospital beds. Need for long term care beds is determined on a regional basis. Prior to 1994, long term care hospitals were not regulated separately and were considered comparable to general acute care hospitals. In 1994, AHCA amended the CON rules to establish long term care beds and hospitals as separate categories of health care providers. In 1994, AHCA defined and continues to the present to define long term care hospital as follows: "Long term care hospital" means a hospital licensed under Chapter 395, Part I, F.S., which meets the requirements of Part 412, subpart B, paragraph 412.23(e), [C]ode of Federal Regulations (1994), and seeks exclusion from the Medicare prospective payment system for inpatient hospital services. Rule 59C-1.002(29), Florida Administrative Code. In the federal regulations referenced by the AHCA rule, long term care hospital is more specifically defined as a hospital with an independent governing structure, an average length of stay greater than 25 days, referral of at least 75 percent of total patients from separate hospitals, and which meets the requirements for Medicare participation. 42 CFR Ch. IV, Subch. B, Pt. 412, Subpt. B, s. 412.23. AHCA also distinguishes long term care in its rules governing the conversions from one type of health care provider to another. The applicable conversion rules provide: "Conversion from one type of health care facility to another" means the reclassification of one licensed facility type to another licensed facility type, including reclassification from a general acute care hospital to a long term care hospital or specialty hospital or from a long term care hospital or specialty hospital to a general acute care hospital. Rule 59C-1.002(14), Florida Administrative Code (emphasis added); and "Conversion of beds" means the reclassification of licensed beds from one category to another including, for facilities licensed under Chapter 395, F.S., conversion to or from acute care beds, neonatal intensive care beds, hospital inpatient psychiatric beds, comprehensive medical rehabilitation beds, hospital inpatient substance abuse beds, distinct part skilled nursing facility beds, or beds in a long term care hospital; and, for facilities licensed under Chapter 400, Part I, F.S., conversion to or from skilled beds and intermediate care beds in a facility that is not certified for both skilled and intermediate nursing care if such conversion effects a change in the level of care of 10 beds or 10 percent of the total bed capacity of the facility within a 2-year period, or conversion to or from sheltered beds and community beds. Rule 59C-1.002 (15), Florida Administrative Code (emphasis added). AHCA also defined "substantial change in health services" to include: The conversion of a general acute care or specialty hospital licensed under Chapter 395, Part I, F.S., to a long term care hospital. Rule 59C-1.002(41)(c), Florida Administrative Code. Taken together AHCA's rules recognize long term care hospitals or beds as a separate and distinct category. Elfie Stamm was responsible for the development of the rules and is currently the chief of the CON and Budget Review Office at AHCA. Ms. Stamm testified in a 1994 rule challenge case, when AHCA was drafting a rule with a numeric need methodology for long term care beds, that: long term care hospitals serve patients who cannot be cost effectively treated in an acute care hospital, who do not have the same needs for the same types of service; it would not be fair for an applicant for the new construction of a long term care hospital to be compared to an acute care hospital; comprehensive medical rehabilitation (CMR) services are different than services in a long term care hospital; a long term care hospital with an average length of stay of 25 days or more is different from an acute care hospital that generally has a length of stay of 5 to 6 days but provides a full range of services; the patient populations in long term care hospitals are different from those in an acute care hospital in terms of overall patient characteristics, including older than average age, higher percentage of patients with particular diagnoses, such as ventilator dependency, higher overall mortality rates than acute care hospitals, and a much higher percentage of admissions by referrals from acute care hospitals. [T. 262-283]. See also Tarpon Springs Hospital Foundation, etc. v. AHCA, et al., DOAH Case No. 94-0958RU (R.O. 8/2/94). On behalf of AHCA, Ms. Stamm testified in this proceeding that: AHCA has changed its mind on whether or not it is appropriate to leave a patient in an acute care setting rather than transfer to long term care, specifically with regard to cost-effectiveness. [T. 373]. AHCA has not changed its mind and still says acute care hospitals and long term care hospitals should be reviewed separately, because if they would be reviewed comparatively, . . . there would be no chance for any [long term] beds ever because we don't show any need for acute care beds anywhere in the state. [T. 376]. But in evaluating Vencor's application for long term care hospitals in District 8 that would be located in Lee County, the Agency viewed hospital-based skilled nursing units, community nursing home subacute beds and comprehensive medical rehab beds throughout the entire district as existing and like potential alternatives to the proposed project. [T. 389]. AHCA does not necessarily agree that CMR services are different from long term care hospital services. [T. 265]. AHCA does not have a clearly identified population group for whom long term care would be more cost-effective, or to determine a numeric need methodology. [TR. 324]. Although there is a population that does need services that exceed 25 days or prolonged ventilator service, AHCA is not sure what is the most appropriate setting for their care because of inadequate data on comparative costs and outcomes. [TR. 327-8]. AHCA attributes its change in position to the publication titled Subacute Care: Policy Synthesis And Market Area Analysis, submitted to the Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, on November 1, 1995, by Lewin-VHI, Inc. The document is commonly referred to as the Lewin Report. The Lewin Report concludes that long term care hospitals serve patients who are also served in other subacute settings, including CMR beds and hospitals, acute care hospital skilled nursing units, and skilled nursing units in freestanding nursing homes. As a result of the conclusions in the Lewin Report, AHCA maintains that it is unable to develop a numeric need methodology without an identifiable patient population. AHCA has not, however, repealed the rules establishing long term care as a separate type of health care service. Rather, the agency intends to wait for additional studies, including one being conducted for Vencor. The Medicare prospective payment system (PPS) for acute care hospitals created the market for subacute and long term care. Under the PPS, acute care hospitals receive a fixed payment based on the patient's diagnosis or diagnostic related group (DRG). Upon discharge to a subacute or long term setting, the patient's care is no longer reimbursed on a fixed basis, but at actual, reasonable costs. AHCA maintains that financial pressures created the current system, but without cost/benefit or outcomes analyses to demonstrate the appropriateness of using long term care hospitals. Therefore, AHCA considered the occupancy levels of acute care hospitals and available nursing home beds in determining the need for Vencor's project. AHCA has no rule defining subacute care, no inventory of subacute care units in nursing homes, and no reporting requirements from which it can determine the level of care or services provided in hospital based skilled nursing units. AHCA has no reports on specific levels or types of services provided in CMR beds. AHCA, nevertheless, presumed that the services are like those provided in long term care beds based on the Lewin Report. In rejecting Vencor's attempts to distinguish itself from other types of health care providers, AHCA relied, in part, on its finding that 1995 District 8 acute care hospital occupancy averaged 47.69 percent and peaked at 60.26 percent. By not adopting rules for determining the numeric need for long term care, AHCA also failed to establish the appropriate service area for determining need. AHCA considers the need for long term care services on a regional basis. In support of AHCA's decision to deny a long term care hospital application in District 9, Ms. Stamm's predecessor, Elizabeth Dudek, testified that long term care is a regional service. As further evidence of AHCA's position, the SAARs issued by AHCA on long term care hospital applications, have examined available services beyond the limits of the district. AHCA contends that long term care is regional, but determines its need by comparison to available hospital based skilled nursing units and subacute beds in community nursing homes, which are evaluated on a subdistrict basis, and CMR services which are tertiary but evaluated on a district-wide basis. See Finding of Fact 22. Since November 1995, AHCA has preliminarily denied all CON applications for long term care hospitals. Its policy of comparing the need for long term care to available beds in nursing homes and other types of hospitals is consistently repeated in the portions of the SAARs which address need. In analyzing the need for long term care hospitals in AHCA District 1, the SAAR dated January 10, 1997, includes the following statements: Vencor Hospitals South, Inc. defines its patient population as those currently being treated in ICUs and belonging to roughly 10 DRGs (which account for approximately 83% of Vencor patients. . . .) However these DRGs could also [be] appropriate for acute care, hospital based freestanding skilled nursing care, skilled nursing facility care and comprehensive medical rehabilitation care and the applicant does not demonstrate that these services are not available to residents of District 1. and The applicant [Baptist Health Affiliates Inc.] also discusses the differences between its proposed patient population and that of an acute care hospital, nursing home and those treated at home. However, there is no documentation provided which demonstrates the applicant's potential patients could not receive appropriate care in the District's existing rehabilitation facility, hospital based or nursing home skilled subacute nursing units. . . . Vencor Exhibit 12, pages 3-4 and 8. AHCA reviewed a CON application filed by Columbia of Pinellas County, Inc., to convert acute care beds to a long term care hospital in District 5, and concluded: The patient population represented by the DRGs listed above (by the applicant) are typical of freestanding nursing home with subacute units and hospital based SNUs in the state. There appear to be strong similarities between the subacute patient population of nursing homes/units and those of a long term care hospital. Vencor Exhibit 13, page 8. The SAAR issued on the Columbia of Pinellas County CON application continued with an extensive discussion of the Lewin Report. The SAAR reported AHCA's finding that CMR hospitals are alternatives since they admit patients who do not fit federal guidelines for CMR admissions (being able to tolerate three hours of therapy a day), and who might otherwise be in long term care hospitals. In the SAAR issued after the review of long term care applications for District 7, the same statement appears: The patient population represented by the DRGs listed above [by Orlando Regional Hospital] are typical of freestanding nursing home with subacute units and hospital based SNUs in the state. There appear to be strong similarities between the subacute patient population of nursing homes/units and those of a long term care hospital. Vencor Exhibit 14, page 11. Finally, in reviewing applications from Palm Beach County in District 9, AHCA concluded again: The applicant states that generally speaking the long term care hospital patients have respiratory complications, . . . tracheostomies, . . . chronic diseases, an infectious process requiring antibiotic therapy, . . . skin complications . . . need a combination of rehabilitation and complex medical treatment or are technology dependent individuals requiring high levels of nursing care. However, these patients could also [be] appropriate for acute care, hospital based skilled nursing care, skilled nursing facility care and comprehensive medical rehabilitation care and the applicant does not demonstrate that these services are not available to the residents of District IX. Vencor Exhibit 15, page 4. AHCA relies on the statutory review criteria in Subsection 408.035(1)(b), Florida Statutes, as authority for its consideration of all beds and facilities which may serve the same patients. That provision requires consideration of: (b) The availability, quality of care, efficiency, appropriateness, accessibility, extent of utilization, and adequacy of like and existing health care facilities and health services in the service district of the applicant. The expert witness for AHCA, however, distinguished between "like and existing" services for purposes of determining numeric need and the statutory criteria. She noted that once numeric need is established and published for nursing beds or CMR beds, for example, that same category of beds outside the appropriate health service planning subdistrict or district is not considered "like and existing." Similarly, within the district or subdistrict, there is a factual issue in each case but no presumption that beds of a different category are "like and existing." AHCA contends that it has no policy related to long term care and any comparable services. Since 1995, long term care CON applicants, according to AHCA, have failed to meet the requirements of Rule 59C-1.008(e), which provides in pertinent part: If no agency policy exists, 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 where they are inconsistent with the applicable statutory or 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. (Emphasis added). AHCA's argument ignores the fact that its expert witness provided competent, substantial evidence that it has redefined and expanded the meaning of "like services" for purposes of demonstrating need through a needs assessment methodology. It also ignores the fact that AHCA has expanded the comparison of need beyond the geographical limits of the district. AHCA's argument that it is waiting for additional data before adopting a need methodology, including data from a Vencor study, is to no avail since AHCA has already changed its policy. After reviewing a total of eighteen CON applications for long term care hospitals, AHCA has issued two CONs, one as part of a settlement agreement and the other approving an application filed by St. Petersburg Health Care Management, Inc. (St. Petersburg), for CON 8213. The St. Petersburg application demonstrated need using an identical methodology prepared by the same health planner as Vencor in this case. Referring to CON 8213, AHCA's expert witness candidly admitted . . . "I want to make clear that particular application was actually submitted and approved prior to the Lewin study." (T. 393). Subsequent to the Lewin study, AHCA has consistently denied applications for long term care beds or hospitals.
The Issue Whether Trustees of Mease Hospital, Inc., d/b/a Mease Hospital Dunedin ("Mease") is entitled to be included on the Neonatal Intensive Care Unit ("NICU") inventory, as authorized to provide Level III NICU services in five (5) Level III beds in Department of Health and Rehabilitative Services ("HRS") District 5. Whether All Children's established its standing to intervene in Case No. 90-6255, as an existing provider of Level III NICU services in HRS District 5. Whether Morton Plant established is standing to intervene in the consolidated cases, as an existing provider of Level II NICU services in HRS District 5.
Findings Of Fact On September 14, 1990, Mease a 278-bed acute care hospital, located in Dunnedin, Florida, timely challenged the inventory of neonatal intensive care beds published for District 5 (Pasco and Pinellas Counties) by HRS. The preliminary inventory, published on August 24, 1990, authorized five (5) Level II and no Level III beds at Mease. All Children's is a 168-bed specialty children's hospital, which has a 24-bed Level III NICU, located in St. Petersburg, Florida, in Pinellas County. Morton Plant is a 750-bed acute care hospital with Level II NICU beds, located in Clearwater, Florida, in Pinellas County. Of the 2,670 babies delivered at Morton Plant in 1990, 598 came from the Mease area; 569 of the 2,670 were classified as not normal or in need of some NICU services, and 148 of the 569 not-normal newborns came from the Mease service area. HRS is the department with responsibility for promulgating NICU rules pursuant to legislation passed in May 1987, effective in October 1987. See, Section 381.702(20) and 381.706(1)(a), (e), (h), and (m), Florida Statutes (1989). The NICU rule became final in August 1990, and included a "grandfather" provision for providers of NICU services prior to October 1, 1987, to avoid any disruption in the availability of NICU services. The grandfather provision includes, as one of three tests for determining if NICU services were offered prior to October 1, 1987, the requirement that 50% of the neonates admitted to Level II and III units from October 1, 1986 through September 30, 1987, were classified in Diagnostic Related Groups (DRGs) 385, 386, 387 or 388. See, Florida Administrative Code, Rule 10-5.042(14)(f) (2)(b) (III). In August 1987, HRS, in preparation for the promulgation of the NICU rule, mailed surveys to various hospitals, including Mease, requesting information about the provision of NICU services at those hospitals. Using the Guidelines for Perinatal Care to distinguish the level of services it was providing, Mease responded to the August 1987 survey by reporting that it had three (3) Level II beds and no Level III beds. In response to the survey questions, Mease also reported that its NICU services began on April 1, 1987, coinciding with the time that a neonatologist, Mary Newport, M.D., joined the staff at Mease. In March 1987, Board Certified Neonatologist Mary Newport began providing 24-hour coverage at Mease, receiving final approval for active staff privileges in early May 1987, so that Mease could treat rather than transfer sick neonates. Mease resubmitted the August 1987 survey in February 1989, reporting that it had increased from three to six Level II beds, after October 1, 1987, and from one to four Level III beds beginning on October 1, 1987, although such changes required certificate of need ("CON") approval effective October 1, 1987. HRS sent out a second survey in April 1989, to which Mease responded that it currently had six Level II beds and no Level III services, under the more stringent requirements included within the proposed rule as compared to the Guidelines for Perinatal Care standards used in the first survey. In August 1989, Mease resubmitted the second survey and reported a current total of six Level II beds and four Level III beds of which three Level II beds and one Level III bed were operating on September 30, 1987. Mease, in the August resubmittal of the second survey, also reported that 37 neonates were admitted to the Level III bed for 63 patient days and that a total of 188 neonates were admitted to both Level II and Level III beds, from October 1, 1986 through September 30, 1987. Of the 188 neonates admitted to Level II and Level III beds from October 1, 1986 through September 30, 1987, Mease claimed that 87 of those had DRG's 385, 386, 387 or 388. On September 10, 1990, Mease submitted documentation to HRS showing that from October 1, 1986 through September 30, 1987, NICU admissions totaled 122 patients, of which 77 were in DRGs 385, 386, 387 or 388. Subsequently, Mease claimed to have had 18 fewer admissions. On December 18, 1990, Mease claimed to have had 107 NICU patients, of which 69 were in DRGs 385, 386, 387 or 388. At the final hearing, Mease claimed that its review of the available records of 1512 of the 1520 deliveries from October 1, 1986 through September 30, 1987, showed that, when diagnoses and codes were changed retrospectively, there were 126 NICU admissions, of which 74 were in DRGs 385, 386, 387, and 388. The number of neonates in intensive care and the number of neonates in the specified DRGs asserted by Mease are both unreliable. There is no credible evidence to support Mease's claim that it started offering NICU services upon the arrival of Dr. Newport on the courtesy staff on March 28, 1987, or upon her becoming a member of the active staff in early May, 1987. Credible expert testimony was presented that a time lag occurs between the arrival on staff of a neonatologist and the initiation of NICU services. In fact, Dr. Newport testified that after she arrived at Mease, she tried out various rented equipment and evaluated it before making purchases, and altered the locations of the nurseries. The Mease Perinatal Committee Agenda dated May 26, 1987 included an item "Task Force for Development of Level II Facility." Even assuming arguendo that Mease did establish NICU services on April 1, 1987, as reported to HRS in response to the first HRS survey, a substantial number of the neonates Mease claimed to have served in its NICU were, in fact, discharged from Mease prior to April 1, 1987. 1/ Mease has failed to submit documentation that it had neonatal intensive care services from October 1, 1986 through September 30, 1987. Mease has failed to submit documentation that it admitted 126 neonates to intensive care services from October 1, 1986 through September 30, 1987. Mease has failed to submit documentation that 74 neonates were classified into DRGs 385, 386, 387 or 388. Credible expert testimony supports the conclusion that retrospective changes in diagnoses are not reliable, in view of the fact that observation of a patient, not just the patient's record, is significant in making a diagnosis. Mease reported to the Health Care Cost Containment Board that it had no revenue from the operation of a NICU from October 1, 1986 to September 30, 1987.
Recommendation Based upon the foregoing Findings of Fact and Conclusions of Law, it is RECOMMENDED that a Final Order be entered: Including Mease on the final inventory as an authorized Level II neonatal intensive care unit with five beds, based on the Summary Recommended Order of April 9, 1991, entered without objection; and Excluding Mease from the final inventory as an authorized provider of Level III neonatal intensive care services. DONE and ENTERED this 1st day of November, 1991, at Tallahassee, Florida. ELEANOR M. HUNTER Hearing Officer Division of Administrative Hearings The DeSoto Building 1230 Apalachee Parkway Tallahassee, Florida 32399-1550 (904) 488-9675 Filed with the Clerk of the Division of Administrative Hearings this 1st day of November, 1991.
The Issue Whether the certificate of need (CON) applications filed by New Port Richey Hospital, Inc., d/b/a Community Hospital of New Port Richey (Community Hospital) (CON No. 9539), and Morton Plant Hospital Association, Inc., d/b/a North Bay Hospital (North Bay) (CON No. 9538), each seeking to replace and relocate their respective general acute care hospital, satisfy, on balance, the applicable statutory and rule criteria.
Findings Of Fact The Parties AHCA AHCA is the single state agency responsible for the administration of the CON program in Florida pursuant to Chapter 408, Florida Statutes (2000). The agency separately reviewed and preliminarily approved both applications. Community Hospital Community Hospital is a 300,000 square feet, accredited hospital with 345 licensed acute care beds and 56 licensed adult psychiatric beds, located in southern New Port Richey, Florida, within Sub-District 5-1. Community Hospital is seeking to construct a replacement facility approximately five miles to the southeast within a rapidly developing suburb known as "Trinity." Community Hospital currently provides a wide array of comprehensive inpatient and outpatient services and is the only provider of obstetrical and adult psychiatric services in Sub-District 5-1. It is the largest provider of emergency services in Pasco County with approximately 35,000 visits annually. It is also the largest provider of Medicaid and indigent patient days in Sub-District 5-1. Community Hospital was originally built in 1969 and is an aging facility. Although it has been renovated over time, the hospital is in poor condition. Community Hospital's average daily census is below 50 percent. North Bay North Bay is a 122-bed facility containing 102 licensed acute care beds and 20 licensed comprehensive medical rehabilitation beds, located approximately one mile north of Community Hospital in Sub-District 5-1. It serves a large elderly population and does not provide pediatric or obstetrical care. North Bay is also an aging facility and proposes to construct a replacement facility in the Trinity area. Notably, however, North Bay has spent approximately 12 million dollars over the past three years for physical improvements and is in reasonable physical condition. Helen Ellis Helen Ellis is an accredited hospital with 150 licensed acute care beds and 18 licensed skilled nursing unit beds. It is located in northern Pinellas County, approximately eight miles south of Community Hospital and nine miles south of North Bay. Helen Ellis provides a full array of acute care services including obstetrics and cardiac catheterization. Its daily census average has fluctuated over the years but is approximately 45 percent. Mease Mease operates two acute care hospitals in Pinellas County including Mease Dunedin Hospital, located approximately 18 to 20 miles south of the applicants and Mease Countryside Hospital, located approximately 16 to 18 miles south of Community and North Bay. Each hospital operates 189 licensed beds. The Mease hospitals are located in the adjacent acute care sub-district but compete with the applicants. The Health Planning District AHCA's Health Planning District 5 consists of Pinellas and Pasco Counties. U.S. Highway 41 runs north and south through the District and splits Pasco County into Sub- District 5-1 and Sub-District 5-2. Sub-District 5-1, where Community Hospital and North Bay are located, extends from U.S. 41 west to the Gulf Coast. Sub-District 5-2 extends from U.S. 41 to the eastern edge of Pasco County. Pinellas County is the most densely populated county in Florida and steadily grows at 5.52 percent per year. On the other hand, its neighbor to the north, Pasco County, has been experiencing over 15 percent annual growth in population. The evidence demonstrates that the area known as Trinity, located four to five miles southeast of New Port Richey, is largely responsible for the growth. With its large, single- owner land tracts, Trinity has become the area's fuel for growth, while New Port Richey, the older coastal anchor which houses the applicants' facilities, remains static. In addition to the available land in Trinity, roadway development in the southwest section of Pasco County is further fueling growth. For example, the Suncoast Highway, a major highway, was recently extended north from Hillsborough County through Sub-District 5-1, west of U.S. 41. It intersects with several large east-west thoroughfares including State Road 54, providing easy highway access to the Tampa area. The General Proposals Community Hospital's Proposal Community Hospital's CON application proposes to replace its existing, 401-bed hospital with a 376-bed state- of-the-art facility and relocate it approximately five miles to the southeast in the Trinity area. Community Hospital intends to construct a large medical office adjacent to its new facility and provide all of its current services including obstetrical care. It does not intend to change its primary service area. North Bay's Proposal North Bay's CON application proposes to replace its existing hospital with a 122-bed state-of-the-art facility and also plans to relocate it approximately eight miles to the southeast in the Trinity area of southwestern Pasco County. North Bay intends to provide the same array of services it currently offers its patients and will not provide pediatric and obstetrical care in the proposed facility. The proposed relocation site is adjacent to the Trinity Outpatient Center which is owned by North Bay's parent company, Morton Plant. The Outpatient Center offers a full range of diagnostic imaging services including nuclear medicine, cardiac nuclear stress testing, bone density scanning, CAT scanning, mammography, ultrasound, as well as many others. It also offers general and specialty ambulatory surgical services including urology; ear, nose and throat; ophthalmology; gastroenterology; endoscopy; and pain management. Approximately 14 physician offices are currently located at the Trinity Outpatient Center. The Condition of Community Hospital Facility Community Hospital's core facilities were constructed between 1969 and 1971. Additions to the hospital were made in 1973, 1975, 1976, 1977, 1979, 1981, 1992, and 1999. With an area of approximately 294,000 square feet and 401 licensed beds, or 733 square feet per bed, Community Hospital's gross area-to-bed ratio is approximately half of current hospital planning standards of 1,600 square feet per bed. With the exception of the "E" wing which was completed in 1999, all of the clinical and support departments are undersized. Medical-Surgical Beds And Intensive Care Units Community Hospital's "D" wing, constructed in 1975, is made up of two general medical-surgical unit floors which are grossly undersized. Each floor operates 47 general medical-surgical beds, 24 of which are in three-bed wards and 23 in semi-private rooms. None of the patient rooms in the "D" wing have showers or tubs so the patients bathe in a single facility located at the center of the wing on each floor. Community Hospital's "A" wing, added in 1973, is situated at the west end of the second floor and is also undersized. It too has a combination of semi-private rooms and three-bed wards without showers or tubs. Community Hospital's "F" wing, added in 1979, includes a medical-surgical unit on the second and third floor, each with semi-private and private rooms. The second floor unit is centrally located between a 56-bed adult psychiatric unit and the Surgical Intensive Care Unit (SICU) which creates security and privacy issues. The third floor unit is adjacent to the Medical Intensive Care Unit (MICU) which must be accessed through the medical-surgical unit. Neither intensive care unit (ICU) possesses an isolation area. Although the three-bed wards are generally restricted to in-season use, and not always full, they pose significant privacy, security, safety, and health concerns. They fail to meet minimum space requirements and are a serious health risk. The evidence demonstrates that reconfiguring the wards would be extremely costly and impractical due to code compliance issues. The wards hinder the hospital's acute care utilization, and impair its ability to effectively compete with other hospitals. Surgical Department and Recovery Community Hospital's surgical department is separated into two locations including the main surgical suite on the second floor and the Endoscopy/Pain Management unit located on the first floor of "C" wing. Consequently, the department cannot share support staff and space such as preparation and recovery. The main surgical suite, adjacent recovery room, and central sterile processing are 25 years old. This unit's operating rooms, cystoscopy rooms, storage areas, work- stations, central sterile, and recovery rooms are undersized and antiquated. The 12-bay Recovery Room has no patient toilet and is lacking storage. The soiled utility room is deficient. In addition, the patient bays are extremely narrow and separated by curtains. There is no direct connection to the sterile corridor, and staff must break the sterile field to transport patients from surgery to recovery. Moreover, surgery outpatients must pass through a major public lobby going to and returning from surgery. The Emergency Department Community Hospital's existing emergency department was constructed in 1992 and is the largest provider of hospital emergency services in Pasco County, handling approximately 35,000 visits per year. The hospital is also designated a "Baker Act" receiving facility under Chapter 394, Florida Statutes, and utilizes two secure examination rooms for emergent psychiatric patients. At less than 8,000 total square feet, the emergency department is severely undersized to meet the needs of its patients. The emergency department is currently undergoing renovation which will connect the triage area to the main emergency department. The renovation will not enlarge the entrance, waiting area, storage, nursing station, nor add privacy to the patient care areas in the emergency department. The renovation will not increase the total size of the emergency department, but in fact, the department's total bed availability will decrease by five beds. Similar to other departments, a more meaningful renovation cannot occur within the emergency department without triggering costly building code compliance measures. In addition to its space limitations, the emergency department is awkwardly located. In 1992, the emergency department was relocated to the front of the hospital and is completely separated from the diagnostic imaging department which remained in the original 1971 building. Consequently, emergency patients are routinely transported across the hospital for imaging and CT scans. Issues Relating to Replacement of Community Hospital Although physically possible, renovating and expanding Community Hospital's existing facility is unreasonable. First, it is cost prohibitive. Any significant renovation to the 1971, 1975, 1977, and 1979 structures would require asbestos abatement prior to construction, at an estimated cost of $1,000,000. In addition, as previously noted, the hospital will be saddled with the major expense of complying with all current building code requirements in the 40-year-old facility. Merely installing showers in patient rooms would immediately trigger a host of expensive, albeit necessary, code requirements involving access, wiring, square footage, fireproofing columns and beams, as well as floor/ceiling and roof/ceiling assemblies. Concurrent with the significant demolition and construction costs, the hospital will experience the incalculable expense and loss of revenue related to closing major portions, if not all, of the hospital. Second, renovation and expansion to the existing facility is an unreasonable option due to its physical restrictions. The 12'4" height of the hospital's first floor limits its ability to accommodate HVAC ductwork large enough to meet current ventilation requirements. In addition, there is inadequate space to expand any department within the confines of the existing hospital without cannibalizing adjacent areas, and vertical expansion is not an option. Community Hospital's application includes a lengthy Facility Condition Assessment which factually details the architectural, mechanical, and electrical deficiencies of the hospital's existing physical plant. The assessment is accurate and reasonable. Community Hospital's Proposed Replacement Community Hospital proposes to construct a six- story, 320 licensed beds, acute care replacement facility. The hospital will consist of 548,995 gross square feet and include a 56-bed adult psychiatric unit connected by a hallway to the first floor of the main hospital building. The proposal also includes the construction of an adjacent medical office building to centralize the outpatient offices and staff physicians. The evidence establishes that the deficiencies inherent in Community Hospital's existing hospital will be cured by its replacement hospital. All patients will be provided large private rooms. The emergency department will double in size, and contain private examination rooms. All building code requirements will be met or exceeded. Patients and staff will have separate elevators from the public. In addition, the surgical department will have large operating rooms, and adequate storage. The MICU and SICU will be adjacent to each other on the second floor to avoid unnecessary traffic within the hospital. Surgical patients will be transported to the ICU via a private elevator dedicated to that purpose. Medical-surgical patient rooms will be efficiently located on the third through sixth floors, in "double-T" configuration. Community Hospital's Existing and Proposed Sites Community Hospital is currently located on a 23-acre site inside the southern boundary of New Port Richey. Single- family homes and offices occupy the two-lane residential streets that surround the site on all sides. The hospital buildings are situated on the northern half of the site, with the main parking lot located to the south, in front of the main entrance to the hospital. Marine Parkway cuts through the southern half of the site from the west, and enters the main parking lot. A private medical mall sits immediately to the west of the main parking lot and a one-acre storm-water retention pond sits to the west of the mall. A private medical office building occupies the south end of the main parking lot and a four-acre drainage easement is located in the southwest corner of the site. Community Hospital's administration has actively analyzed its existing site, aging facility, and adjacent areas. It has commissioned studies by civil engineers, health care consultants, and architects. The collective evidence demonstrates that, although on-site relocation is potentially an option, on balance, it is not a reasonable option. Replacing Community Hospital on its existing site is not practical for several reasons. First, the hospital will experience significant disruption and may be required to completely close down for a period of time. Second, the site's southwestern large four-acre parcel is necessary for storm-water retention and is unavailable for expansion. Third, a reliable cost differential is unknown given Community Hospital's inability to successfully negotiate with the city and owners of the adjacent medical office complexes to acquire additional parcels. Fourth, acquiring other adjacent properties is not a viable option since they consist of individually owned residential lots. In addition to the site's physical restrictions, the site is hindered by its location. The hospital is situated in a neighborhood between small streets and a local school. From the north and south, motorists utilize either U.S. 19, a congested corridor that accommodates approximately 50,000 vehicles per day, or Grand and Madison Streets, two-lane streets within a school zone. From the east and west, motorists utilize similar two-lane neighborhood streets including Marine Parkway, which often floods in heavy rains. Community Hospital's proposed site, on the other hand, is a 53-acre tract positioned five miles from its current facility, at the intersection of two major thoroughfares in southwestern Pasco County. The proposed site offers ample space for all facilities, parking, outpatient care, and future expansion. In addition, Community Hospital's proposed site provides reasonable access to all patients within its existing primary service area made up of zip codes 34652, 34653, 34668, 34655, 34690, and 34691. For example, the average drive times from the population centers of each zip code to the existing site of the hospital and the proposed site are as follows: Zip code Difference Existing site Proposed site 34652 3 minutes 14 minutes 11 minutes 34653 8 minutes 11 minutes 3 minutes 34668 15 minutes 21 minutes 6 minutes 34655 11 minutes 4 minutes -7 minutes 34690 11 minutes 13 minutes 2 minutes 34691 11 minutes 17 minutes 6 minutes While the average drive time from the population centroids of zip codes 34653, 34668, 34690, and 34691 to the proposed site slightly increases, it decreases from the Trinity area, where population growth has been most significant in southwestern Pasco County. In addition, a motorist's average drive time from Community Hospital's existing location to its proposed site is only 10 to 11 minutes, and patients utilizing public transportation will be able to access the new hospital via a bus stop located adjacent to the proposed site. The Condition of North Bay Facility North Bay Hospital is also an aging facility. Its original structure and portions of its physical plant are approximately 30 years old. Portions of its major mechanical systems will soon require replacement including its boilers, air handlers, and chillers. In addition, the hospital is undersized and awkwardly configured. Despite its shortcomings, however, North Bay is generally in good condition. The hospital has been consistently renovated and updated over time and is aesthetically pleasing. Moreover, its second and third floors were added in 1986, are in good shape, and structurally capable of vertical expansion. Medical Surgical Beds and ICU Units By-in-large, North Bay is comprised of undersized, semi-private rooms containing toilet and shower facilities. The hospital does not have any three-bed wards. North Bay's first floor houses all ancillary and support services including lab, radiology, pharmacy, surgery, pre-op, post-anesthesia recovery, central sterile processing and supply, kitchen and cafeteria, housekeeping and administration, as well as the mechanical, electrical, and facilities maintenance and engineering. The first floor also contains a 20-bed CMR unit and a 15-bed acute care unit. North Bay's second and third floors are mostly comprised of semi-private rooms and supporting nursing stations. Although the rooms and stations are not ideally sized, they are in relatively good shape. North Bay utilizes a single ICU with ten critical care beds. The ICU rooms and nursing stations are also undersized. A four-bed ICU ward and former nursery are routinely used to serve overflow patients. Surgery Department and Recovery North Bay utilizes a single pre-operative surgical room for all of its surgery patients. The room accommodates up to five patient beds, but has limited space for storage and pre-operative procedures. Its operating rooms are sufficiently sized. While carts and large equipment are routinely stored in hallways throughout the surgical suite, North Bay has converted the former obstetrics recovery room to surgical storage and has made efficient use of other available space. North Bay operates a small six-bed Post Anesthesia Care Unit. Nurses routinely prepare patient medications in the unit which is often crowded with staff and patients. The Emergency Department North Bay has recently expanded its emergency department. The evidence demonstrates that this department is sufficient and meets current and future expected patient volumes. Replacement Issues Relating to North Bay While it is clear that areas of North Bay's physical plant are aging, the facility is in relatively good condition. It is apparent that North Bay must soon replace significant equipment, including cast-iron sewer pipes, plumbing, boilers, and chillers which will cause some interruption to hospital operations. However, North Bay's four-page written assessment of the facility and its argument citing the need for total replacement is, on balance, not persuasive. North Bay's Proposed Replacement North Bay proposes to construct a new, state-of-the- art, hospital approximately eight miles southeast of its existing facility and intends to offer the identical array of services the hospital currently provides. North Bay's Existing and Proposed Sites North Bay's existing hospital is located on an eight-acre site with limited storm-water drainage capacity. Consequently, much of its parking area is covered by deep, porous, gravel instead of asphalt. North Bay's existing site is generally surrounded by residential properties. While the city has committed, in writing, it willingness to assist both applicants with on-site expansion, it is unknown whether North Bay can acquire additional adjacent property. North Bay's proposed site is located at the intersection of Trinity Oaks Boulevard and Mitchell Boulevard, south of Community Hospital's proposed site, and is quite spacious. It contains sufficient land for the facilities, parking, and future growth, and has all necessary infrastructure in place, including utility systems, storm- water structures, and roadways. Currently however, there is no public transportation service available to North Bay's proposed site. Projected Utilization by Applicants The evidence presented at hearing indicates that, statewide, replacement hospitals often increase a provider's acute care bed utilization. For example, Bartow Memorial Hospital, Heart of Florida Regional Medical Center, Lake City Medical Center, Florida Hospital Heartland Medical Center, South Lake Hospital, and Florida Hospital-Fish Memorial each experienced significant increases in utilization following the opening of their new hospital. The applicants in this case each project an increase in utilization following the construction of their new facility. Specifically, Community Hospital's application projects 82,685 total hospital patient days (64,427 acute care patient days) in year one (2006) of the operation of its proposed replacement facility, and 86,201 total hospital patient days (67,648 acute care patient days) in year two (2007). Using projected 2006 and 2007 population estimates, applying 2002 acute care hospital use rates which are below 50 percent, and keeping Community Hospital's acute care market share constant at its 2002 level, it is reasonably estimated that Community Hospital's existing hospital will experience 52,623 acute care patient days in 2006, and 53,451 acute care patient days in 2007. Consequently, Community Hospital's proposed facility must attain 11,804 additional acute care patient days in 2006, and 14,197 more acute care patient days in 2007, in order to achieve its projected acute care utilization. Although Community Hospital lost eight percent of the acute care market in its service area between 1995 and 2002, two-thirds of that loss was due to residents of Sub- District 5-1 acquiring services in another area. While Community Hospital experienced 78,444 acute care patient days in 1995, it projects only 64,427 acute care patient days in year one. Given the new facility and population factors, it is reasonable that the hospital will recapture half of its lost acute care market share and achieve its projections. With respect to its psychiatric unit, Community Hospital projects 16,615 adult psychiatric inpatient days in year one (2006) and 17,069 adult inpatient days in year two (2007) of the proposed replacement hospital. The evidence indicates that these projections are reasonable. Similarly, North Bay's acute care utilization rate has been consistently below 50 percent. Since 1999, the hospital has experienced declining utilization. In its application, North Bay states that it achieved total actual acute care patient days of 21,925 in 2000 and 19,824 in 2001 and the evidence at hearing indicates that North Bay experienced 17,693 total acute care patient days in 2002. North Bay projects 25,909 acute care patient days in the first year of operation of its proposed replacement hospital, and 27,334 acute care patient days in the second year of operation. Despite each applicant's current facility utilization rate, Community Hospital must increase its current acute care patient days by 20 percent to reach its projected utilization, and North Bay must increase its patient days by at least 50 percent. Given the population trends, service mix and existing competition, the evidence demonstrates that it is not possible for both applicants to simultaneously achieve their projections. In fact, it is strongly noted that the applicants' own projections are predicated upon only one applicant being approved and cannot be supported with the approval of two facilities. Local Health Plan Preferences In its local health plan for District 5, the Suncoast Health Council, Inc., adopted acute care preferences in October, 2000. The replacement of an existing hospital is not specifically addressed by any of the preferences. However, certain acute care preferences and specialty care preferences are applicable. The first applicable preference provides that preference "shall be given to an applicant who proposes to locate a new facility in an area that will improve access for Medicaid and indigent patients." It is clear that the majority of Medicaid and indigent patients live closer to the existing hospitals. However, Community Hospital proposes to move 5.5 miles from its current location, whereas North Bay proposes to move eight miles from its current location. While the short distances alone are less than significant, North Bay's proposed location is further removed from New Port Richey, is not located on a major highway or bus-route, and would therefore be less accessible to the medically indigent residents. Community Hospital's proposed site will be accessible using public transportation. Furthermore, Community Hospital has consistently provided excellent service to the medically indigent and its proposal would better serve that population. In 2000, Community Hospital provided 7.4 percent of its total patient days to Medicaid patients and 0.8 percent of its total patient days to charity patients. Community Hospital provided the highest percentage and greatest number of Medicaid patient days in Sub-District 5-1. By comparison, North Bay provided 5.8 percent of its total patient days to Medicaid patients and 0.9 percent of its total patient days to charity patients. In 2002, North Bay's Medicaid patients days declined to 3.56 percent. Finally, given the closeness and available bed space of the existing providers and the increasing population in the Trinity area, access will be improved by Community Hospital's relocation. The second local health plan preference provides that "[i]n cases where an applicant is a corporation with previously awarded certificates of need, preference shall be given to those which follow through in a timely manner to construct and operate the additional facilities or beds and do not use them for later negotiations with other organizations seeking to enter or expand the number of beds they own or control." Both applicants meet this preference. The third local health plan preference recognizes "Certificate of Need applications that provide AHCA with documentation that they provide, or propose to provide, the largest percentage of Medicaid and charity care patient days in relation to other hospitals in the sub-district." Community Hospital provides the largest percentage of Medicaid and charity care patient days in relation to other hospitals in Sub-District 5-1, and therefore meets this preference. The fourth local health plan preference applies to "Certificate of Need applications that demonstrate intent to serve HIV/AIDS infected persons." Both applicants accept and treat HIV/AIDS infected persons, and would continue to do so in their proposed replacement hospitals. The fifth local health plan preference recognizes "Certificate of Need applications that commit to provide a full array of acute care services including medical-surgical, intensive care, pediatric, and obstetrical services within the sub-district for which they are applying." Community Hospital qualifies since it will continue to provide its current services, including obstetrical care and psychiatric care, in its proposed replacement hospital. North Bay discontinued its pediatric and obstetrical programs in 2001, does not intend to provide them in its proposed replacement hospital, and will not provide psychiatric care. Agency Rule Preferences Florida Administrative Code Rule 59C-1.038(6) provides an applicable preference to a facility proposing "new acute care services and capital expenditures" that has "a documented history of providing services to medically indigent patients or a commitment to do so." As the largest Medicaid provider in Sub-District 5-1, Community Hospital meets this preference better than does North Bay. North Bay's history demonstrates a declining rate of service to the medically indigent. Statutory Review Criteria Section 408.035(1), Florida Statutes: The need for the health care facilities and health services being proposed in relation to the applicable district health plan District 5 includes Pasco and Pinellas County. Pasco County is rapidly developing, whereas Pinellas County is the most densely populated county in Florida. Given the population trends, service mix, and utilization rates of the existing providers, on balance, there is a need for a replacement hospital in the Trinity area. Section 408.035(2), Florida Statutes: The availability, quality of care, accessibility, and extent of utilization of existing health care facilities and health services in the service district of the applicant Community Hospital and North Bay are both located in Sub-District 5-1. Each proposes to relocate to an area of southwestern Pasco County which is experiencing explosive population growth. The other general acute care hospital located in Sub-District 5-1 is Regional Medical Center Bayonet Point, which is located further north, in the Hudson area of western Pasco County. The only other acute care hospitals in Pasco County are East Pasco Medical Center, in Zephyrhills, and Pasco Community Hospital, in Dade City. Those hospitals are located in Sub-District 5-2, east Pasco County, far from the area proposed to be served by either Community Hospital or North Bay. District 5 includes Pinellas County as well as Pasco County. Helen Ellis and Mease are existing hospital providers located in Pinellas County. Helen Ellis has 168 licensed beds, consisting of 150 acute care beds and an 18-bed skilled nursing unit, and is located 7.9 miles from Community Hospital's existing location and 10.8 miles from Community Hospital's proposed location. Access to Helen Ellis for patients originating from southwestern Pasco County requires those patients to travel congested U.S. 19 south to Tarpon Springs. As a result, the average drive time from Community Hospital's existing and proposed site to Helen Ellis is approximately 22 minutes. Helen Ellis is not a reasonable alternative to Community Hospital's proposal. The applicants' proposals are specifically designed for the current and future health care needs of southwestern Pasco County. Given its financial history, it is unknown whether Helen Ellis will be financially capable of providing the necessary care to the residents of southwestern Pasco. Mease Countryside Hospital has 189 licensed acute care beds. It is located 16.0 miles from Community Hospital's existing location and 13.8 miles from Community Hospital's proposed location. The average drive time to Mease Countryside is 32 minutes from Community Hospital's existing site and 24 minutes from its proposed site. In addition, Mease Countryside Hospital has experienced extremely high utilization over the past several years, in excess of 90 percent for calendar years 2000 and 2001. Utilization at Mease Countryside Hospital has remained over 80 percent despite the addition of 45 acute care beds in April 2002. Given the growth and demand, it is unknown whether Mease can accommodate the residents in southwest Pasco County. Mease Dunedin Hospital has 189 licensed beds, consisting of 149 acute care beds, a 30-bed skilled nursing unit, five Level 2 neonatal intensive care beds, and five Level 3 neonatal intensive care beds. Its former 15-bed adult psychiatric unit has been converted into acute care beds. It is transferring its entire obstetrics program at Mease Dunedin Hospital to Mease Countryside Hospital. Mease Dunedin Hospital is located approximately 18 to 20 miles from the applicants' existing and proposed locations with an average drive time of 35-38 minutes. With their remote location, and the exceedingly high utilization at Mease Countryside Hospital, neither of the two Mease hospitals is a viable alternative to the applicants' proposals. In addition, the construction of a replacement hospital would positively impact economic development and further attract medical professionals to Sub-District 5-1. On balance, given the proximity, utilization, service array, and accessibility of the existing providers, including the applicants, the relocation of Community Hospital will enhance access to health care to the residents. Section 408.035(3), Florida Statutes: The ability of the applicant to provide quality of care and the applicant's record of providing quality of care As stipulated, both applicants provide excellent quality of care. However, Community Hospital's proposal will better enhance its ability to provide quality care. Community is currently undersized, non-compliant with today's standards, and located on a site that does not allow for reasonable expansion. Its emergency department is inadequate for patient volume, and the configuration of the first floor leads to inefficiencies in the diagnosis and treatment of emergency patients. Again, most inpatients are placed in semi-private rooms and three-bed wards, with no showers or tubs, little privacy, and an increased risk of infection. The hospital's waiting areas for families of patients are antiquated and undersized, its nursing stations are small and cramped and the operating rooms and storage facilities are undersized. Community Hospital's deficiencies will be effectively eliminated by its proposed replacement hospital. As a result, patients will experience qualitatively better care by the staff who serve them. Conversely, North Bay is in better physical condition and not in need of replacement. It has more reasonable options to expand or relocate its facility on site. Quality of care at North Bay will not be markedly enhanced by the construction of a new hospital. Sections 408.035(4)and(5), Florida Statutes, have been stipulated as not applicable in this case. Section 408.035(6), Florida Statutes: The availability of resources, including health personnel, management personnel, and funds available for capital and operating expenditures, for project accomplishment and operation The parties stipulated that both Community Hospital and North Bay have available health personnel and management personnel for project accomplishment and operation. In addition, the evidence proves that both applicants have sufficient funds for capital and operating expenditures. Community Hospital proposes to rely on its parent company to finance the project. Keith Giger, Vice-President of Finance for HCA, Inc., Community Hospital's parent organization, provided credible deposition testimony that HCA, Inc., will finance 100 percent of the total project cost by an inter-company loan at eight percent interest. Moreover, it is noted that the amount to be financed is actually $20 million less than the $196,849,328 stated in the CON Application, since Community Hospital previously purchased the proposed site in June 2003 with existing funds and does not need to finance the land acquisition. Community Hospital has sufficient working capital for operating expenditures of the proposed replacement hospital. North Bay, on the other hand, proposes to acquire financing from BayCare Obligated Group which includes Morton Plant Hospital Association, Inc.; Mease; and several other hospital entities. Its proposal, while feasible, is less certain since member hospitals must approve the indebtedness, thereby providing Mease with the ability to derail North Bay's proposed bond financing. Section 408.035(7), Florida Statutes: The extent to which the proposed services will enhance access to health care for residents of the service district The evidence proves that either proposal will enhance geographical access to the growing population in the service district. However, with its provision of obstetrical services, Community Hospital is better suited to address the needs of the younger community. With respect to financial access, both proposed relocation sites are slightly farther away from the higher elderly and indigent population centers. Since the evidence demonstrates that it is unreasonable to relocate both facilities away from the down-town area, Community Hospital's proposal, on balance, provides better access to poor patients. First, public transportation will be available to Community Hospital's site. Second, Community Hospital has an excellent record of providing care to the poor and indigent and has accepted the agency's condition to provide ten percent of its total annual patient days to Medicaid recipients To the contrary, North Bay's site will not be accessible by public transportation. In addition, North Bay has a less impressive record of providing care to the poor and indigent. Although AHCA conditioned North Bay's approval upon it providing 9.7 percent of total annual patient days to Medicaid and charity patients, instead of the 9.7 percent of gross annual revenue proposed in its application, North Bay has consistently provided Medicaid and charity patients less than seven percent of its total annual patient days. Section 408.035(8), Florida Statutes: The immediate and long-term financial feasibility of the proposal Immediate financial feasibility refers to the availability of funds to capitalize and operate the proposal. See Memorial Healthcare Group, Ltd. d/b/a Memorial Hospital Jacksonville vs. AHCA et al., Case No. 02-0447 et seq. Community Hospital has acquired reliable financing for the project and has sufficiently demonstrated that its project is immediately financially feasible. North Bay's short-term financial proposal is less secure. As noted, North Bay intends to acquire financing from BayCare Obligated Group. As a member of the group, Mease, the parent company of two hospitals that oppose North Bay's application, must approve the plan. Long-term financial feasibility is the ability of the project to reach a break-even point within a reasonable period of time and at a reasonable achievable point in the future. Big Bend Hospice, Inc. vs. AHCA and Covenant Hospice, Inc., Case No. 02-0455. Although CON pro forma financial schedules typically show profitability within two to three years of operation, it is not a requirement. In fact, in some circumstances, such as the case of a replacement hospital, it may be unrealistic for the proposal to project profitability before the third or fourth year of operation. In this case, Community Hospital's utilization projections, gross and net revenues, and expense figures are reasonable. The evidence reliably demonstrates that its replacement hospital will be profitable by the fourth year of operation. The hospital's financial projections are further supported by credible evidence, including the fact that the hospital experienced financial improvement in 2002 despite its poor physical condition, declining utilization, and lost market share to providers outside of its district. In addition, the development and population trends in the Trinity area support the need for a replacement hospital in the area. Also, Community Hospital has benefited from increases in its Medicaid per diem and renegotiated managed care contracts. North Bay's long-term financial feasibility of its proposal is less certain. In calendar year 2001, North Bay incurred an operating loss of $306,000. In calendar year 2002, it incurred a loss of $1,160,000. In its CON application, however, North Bay projects operating income of $1,538,827 in 2007, yet omitted the ongoing expenses of interest ($1,600,000) and depreciation ($3,000,000) from its existing facility that North Bay intends to continue operating. Since North Bay's proposal does not project beyond year two, it is less certain whether it is financially feasible in the third or fourth year. In addition to the interest and depreciation issues, North Bay's utilization projections are less reasonable than Community Hospital's proposal. While possible, North Bay will have a difficult task achieving its projected 55 percent increase in acute care patient days in its second year of operation given its declining utilization, loss of obstetric/pediatric services and termination of two exclusive managed care contracts. Section 408.035(9), Florida Statutes: The extent to which the proposal will foster competition that promotes quality and cost-effectiveness Both applicants have substantial unused capacity. However, Community Hospital's existing facility is at a distinct competitive disadvantage in the market place. In fact, from 1994 to 1998, Community Hospital's overall market share in its service area declined from 40.3 percent to 35.3 percent. During that same period, Helen Ellis' overall market share in Community Hospital's service area increased from 7.2 percent to 9.2 percent. From 1995 to the 12-month period ending June 30, 2002, Community Hospital's acute care market share in its service area declined from 34.0 percent to 25.9 percent. During that same period, Helen Ellis' acute care market share in Community Hospital's service area increased from 11.7 percent to 12.0 percent. In addition, acute care average occupancy rates at Mease Dunedin Hospital increased each year from 1999 through 2002. Acute care average occupancy at Mease Countryside Hospital exceeded 90 percent in 2000 and 2001, and was approximately 85 percent for the period ending June 30, 2002. Some of the loss in Community Hospital's market share is due to an out-migration of patients from its service area to hospitals in northern Pinellas and Hillsborough Counties. Market share in Community's service area by out-of- market providers increased from 33 percent in 1995 to 40 percent in 2002. Community Hospital's outdated hospital has hampered its ability to compete for patients in its service area. Mease is increasing its efforts to attract patients and currently completing a $92 million expansion of Mease Countryside Hospital. The project includes the development of 1,134 parking spaces on 30 acres of raw land north of the Mease Countryside Hospital campus and the addition of two floors to the hospital. It also involves the relocation of 51 acute care beds, the obstetrics program and the Neonatal Intensive Care Units from Mease Dunedin Hosptial to Mease Countryside Hospital. Mease is also seeking to more than double the size of the Countryside emergency department to handle its 62,000 emergency visits. With the transfer of licensed beds from Mease Dunedin Hospital to Mease Countryside Hospital, Mease will also convert formerly semi-private patient rooms to private rooms at Mease Dunedin Hospital. The approval of Community Hospital's relocated facility will enable it to better compete with the hospitals in the area and promote quality and cost- effectiveness. North Bay, on the other hand, is not operating at a distinct disadvantage, yet is still experiencing declining utilization. North Bay is the only community-owned, not-for- profit provider in western Pasco County and is a valuable asset to the city. Section 408.035(10), Florida Statutes: The costs and methods of the proposed construction, including the costs and methods or energy provision and the availability of alternative, less costly, or more effective methods of construction The parties stipulated that the project costs in both applications are reasonable to construct the replacement hospitals. Community Hospital's proposed construction cost per square foot is $175, and slightly less than North Bay's $178 proposal. The costs and methods of proposed construction for each proposal is reasonable. Given Community Hospital's severe site and facility problems, the evidence demonstrates that there is no reasonable, less costly, or more effective methods of construction available for its proposed replacement hospital. Additional "band-aide" approaches are not financially reasonable and will not enable Community Hospital to effectively compete. The facility is currently licensed for 401 beds, operates approximately 311 beds and is still undersized. The proposed replacement hospital will meet the standards in Florida Administrative Code Rule 59A-3.081, and will meet current building codes, including the Americans with Disabilities Act and the Guidelines for Design and Construction of Hospitals and Health Care Facilities, developed by the American Institute of Architects. The opponents' argue that Community Hospital will not utilize the 320 acute care beds proposed in its CON application, and therefore, a smaller facility is a less- costly alternative. In addition, Helen Ellis' architectural expert witness provided schematic design alternatives for Community Hospital to be expanded and replaced on-site, without providing a detailed and credible cost accounting of the alternatives. Given the evidence and the law, their arguments are not persuasive. While North Bay's replacement cost figures are reasonable, given the aforementioned reasons, including the fact that the facility is in reasonably good condition and can expand vertically, on balance, it is unreasonable for North Bay to construct a replacement facility in the Trinity area. Section 408.035(11), Florida Statutes: The applicant's past and proposed provision of health care services to Medicaid patients and the medically indigent Community Hospital has consistently provided the most health care services to Medicaid patients and the medically indigent in Sub-District 5-1. Community Hospital agreed to provide at least ten percent of its patient days to Medicaid recipients. Similarly, North Bay agreed to provide 9.7 percent of its total annual patient days to Medicaid and charity patients combined. North Bay, by contrast, provided only 3.56 percent of its total patient days to Medicaid patients in 2002, and would have to significantly reverse a declining trend in its Medicaid provision to comply with the imposed condition. Community Hospital better satisfies the criterion. Section 408.035(12) has been stipulated as not applicable in this case. Adverse Impact on Existing Providers Historical figures demonstrate that hospital market shares are not static, but fluctuate with competition. No hospital is entitled to a specific or historic market share free from competition. While the applicants are located in health planning Sub-District 5-1 and Helen Ellis and the two Mease hospitals are located in health planning Sub-District 5- 2, they compete for business. None of the opponents is a disproportionate share, safety net, Medicaid provider. As a result, AHCA gives less consideration to any potential adverse financial impact upon them resulting from the approval of either application as a low priority. The opponents, however, argue that the approval of either replacement hospital would severely affect each of them. While the precise distance from the existing facilities to the relocation sites is relevant, it is clear that neither applicants' proposed site is unreasonably close to any of the existing providers. In fact, Community Hospital intends to locate its replacement facility three miles farther away from Helen Ellis and 1.5 miles farther away from Mease Dunedin Hospital. While Helen Ellis' primary service area is seemingly fluid, as noted by its chief operating officer's hearing and deposition testimony, and the Mease hospitals are located 15 to 20 miles south, they overlap parts of the applicants' primary service areas. Accordingly, each applicant concedes that the proposed increase in their patient volume would be derived from the growing population as well as existing providers. Although it is clear that the existing providers may be more affected by the approval of Community Hosptial's proposal, the exact degree to which they will be adversely impacted by either applicant is unknown. All parties agree, however, that the existing providers will experience less adverse affects by the approval of only one applicant, as opposed to two. Furthermore, Mease concedes that its hospitals will continue to aggressively compete and will remain profitable. In fact, Mease's adverse impact analysis does not show any credible reduction in loss of acute care admissions at Mease Countryside Hospital or Mease Dunedin Hospital until 2010. Even then, the reliable evidence demonstrates that the impact is negligible. Helen Ellis, on the other hand, will likely experience a greater loss of patient volume. To achieve its utilization projections, Community Hospital will aggressively compete for and increase market share in Pinellas County zip code 34689, which borders Pasco County. While that increase does not facially prove that Helen Ellis will be materially affected by Community Hospital's replacement hospital, Helen Ellis will confront targeted competition. To minimize the potential adverse affect, Helen Ellis will aggressively compete to expand its market share in the Pinellas County zip codes south of 34689, which is experiencing population growth. In addition, Helen Ellis is targeting broader service markets, and has filed an application to establish an open- heart surgery program. While Helen Ellis will experience greater competition and financial loss, there is insufficient evidence to conclude that it will experience material financial adverse impact as a result of Community Hospital's proposed relocation. In fact, Helen Ellis' impact analysis is less than reliable. In its contribution-margin analysis, Helen Ellis utilized its actual hospital financial data as filed with AHCA for the fiscal year October 1, 2001, to September 30, 2002. The analysis included total inpatient and total outpatient service revenues found in the filed financial data, including ambulatory services and ancillary services, yet it did not include the expenses incurred in generating ambulatory or ancillary services revenue. As a result, the overstated net revenue per patient day was applied to its speculative lost number of patient days which resulted in an inflated loss of net patient service revenue. Moreover, the evidence indicates that Helen Ellis' analysis incorrectly included operational revenue and excluded expenses related to its 18-bed skilled nursing unit since neither applicant intends to operate a skilled nursing unit. While including the skilled nursing unit revenues, the analysis failed to include the sub-acute inpatient days that produced those revenues, and thereby over inflated the projected total lost net patient service revenue by over one million dollars.
Recommendation Based on the foregoing findings of fact and conclusions of law, it is RECOMMENDED that: Community Hospital's CON Application No. 9539, to establish a 376-bed replacement hospital in Pasco County, Sub- District 5-1, be granted; and North Bay's CON Application No. 9538, to establish a 122-bed replacement hospital in Pasco County, Sub-District 5- 1, be denied. DONE AND ENTERED this 19th day of March, 2004, in Tallahassee, Leon County, Florida. S WILLIAM R. PFEIFFER 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 19th day of March, 2004. COPIES FURNISHED: James C. Hauser, Esquire R. Terry Rigsby, Esquire Metz, Hauser & Husband, P.A. 215 South Monroe Street, Suite 505 Post Office Box 10909 Tallahassee, Florida 32302 Stephen A. Ecenia, Esquire R. David Prescott, Esquire Richard M. Ellis, Esquire Rutledge, Ecenia, Purnell & Hoffman, P.A. 215 South Monroe Street, Suite 420 Post Office Box 551 Tallahassee, Florida 32302-0551 Richard J. Saliba, Esquire Agency for Health Care Administration Fort Knox Building III, Mail Station 3 2727 Mahan Drive Tallahassee, Florida 32308 Robert A. Weiss, Esquire Karen A. Putnal, Esquire Parker, Hudson, Rainer & Dobbs, LLP The Perkins House, Suite 200 118 North Gadsden Street Tallahassee, Florida 32301 Darrell White, Esquire William B. Wiley, Esquire McFarlain & Cassedy, P.A. 305 South Gadsden Street, Suite 600 Tallahassee, Florida 32301 Lealand McCharen, Agency Clerk Agency for Health Care Administration 2727 Mahan Drive, Mail Station 3 Tallahassee, Florida 32308 Valda Clark Christian, General Counsel Agency for Health Care Administration 2727 Mahan Drive, Mail Station 3 Tallahassee, Florida 32308 Rhonda M. Medows, M.D., Secretary Agency for Health Care Administration 2727 Mahan Drive, Mail Station 3 Tallahassee, Florida 32308
The Issue Whether the Agency for Health Care Administration should approve the application of Kindred Hospitals East, LLC, for a Certificate of Need to establish a 60-bed, long- term care hospital ("LTCH") to be located in Brevard County, one of four counties in AHCA District 7.
Findings Of Fact The Parties Kindred Hospitals East, LLC, ("Kindred" or the "Applicant") is a subsidiary of Kindred Healthcare, Inc. ("Kindred Healthcare"). Kindred Healthcare operates 84 LTCHs nationwide, including eight in the State of Florida. Twenty-four of Kindred Healthcare's LTCHs are operated by Kindred Hospitals East, LLC, including the eight in Florida. The Agency is the state agency responsible for the administration of the Certificate of Need program in Florida. See § 408.034(1), Fla. Stat., et seq. Pre-hearing Stipulation The Joint Pre-hearing Stipulation between Kindred Hospitals East, LLC, and Agency for Health Care Administration, filed May 25, 2006, contains the following: E. STATEMENT OF FACTS WHICH AREADMITTED AND WILL REQUIRE NO PROOF The CON application filed by Kindred complies with the application content and review process requirements of Sections 408.037 and 408.039, Florida Statutes (2005), and Rule 59C- 1.008, Florida Administrative Code, and the Agency's review of the application complied with the review process requirements of the above-referenced Statutes and Rule. With respect to compliance with Section 408.035(3), Florida Statutes (2005), it is agreed that Kindred has the ability to provide a quality program based on the descriptions of the program in its CON application and based on the operational facilities of the applicant and/or of the applicant's parent facilities which are JCAHO certified. With respect to compliance with Section 408.035(4), Florida Statutes (2005), it is agreed that Kindred has the ability to provide the necessary resources, including health personnel, management personnel, and funds for capital and operating expenditures, for project accomplishment and operation. With respect to compliance with Section 408.035(6), Florida Statutes (2005)it is agreed that the project is likely to be financially feasible. The parties agree there are no disputed issues with respect to Kindred's compliance with Section 408.035(8), Florida Statutes (2005), which relates to an applicant's proposed costs and methods of proposed construction for the type of project proposed. The parties agree there are no disputed issues with respect to Kindred's compliance with Section 408.035(9), Florida Statutes (2005), which relates to an applicant's proposed provision of health care services to Medicaid patients and the medically indigent. Section 408.035(10), Florida Statutes (2005), relating to nursing home beds, is not at issue with respect to the review of Kindred's CON application. With respect to compliance with Rule 59C-1.008(1)(a)-(c), Florida Administrative Code, it is agreed that Kindred complied with the letter of intent requirements contained therein. 9. Rules 59C-1.008(1)(d), (e), (h), (i), and (j) are not at issue with respect to the review of Kindred's CON applications. With respect to compliance with Rule 59C-1.008(1)(f), Florida Administrative Code, it is agreed that Kindred complied with the applicable certificate of need application submission requirements contained therein. The need assessment methodology is governed by Rule 59C-1.008(2)(e)2.a.- d., Florida Administrative Code. With respect to Rule 59C-1.008(2), Florida Administrative Code, except as to Rule 59C-1.008(2)(e)2.a-d and (2)(e)3, Florida Administrative Code, it is agreed that this provision is not applicable to this proceeding, as the Agency did not at the time of the review cycle at issue, and currently does not, calculate a fixed need pool for LTCH beds. With respect to compliance with Rule 59C-1.008(3), Florida Administrative Code, it is agreed that Kindred submitted the required filing fees. With respect to compliance with Rule 59C-1.008(4)(a)-(e), Florida Administrative Code, it is agreed that Kindred complied with the certificate of need application requirements contained therein. Rule 59C-1.008(5), Florida Administrative Code, relating to identifiable portions of a project, is not at issue with respect to the review of Kindred's CON applications. In light of the stipulation, the issues remaining generally concern: the need for Kindred's proposed facility (including the reasonableness of Kindred's need methodology and whether its need assessment conforms to AHCA rules), the accessibility of existing LTCH facilities, and the extent to which the proposal will foster competition that fosters cost-effectiveness and quality. Long-Term Care Services The length of stay in the typical acute care hospital (a "short-term hospital" or a "STACH") for most patients is four to five days. Some hospital patients, however, are in need of acute care services on a long-term basis, that is, much longer than the average lengths of stay for most patients in an short-term hospital. Patients appropriate for LTCH services represent a small but discrete sub-set of all inpatients. They are differentiated from other hospital patients. Typically, they have multiple co-morbidities that require concurrent treatment. Patients appropriate for LTCH services tend to be elderly and frail, unless they are victims of severe trauma. All LTCH patients are generally medically complex and frequently catastrophically ill. Generally, Medicare patients admitted to LTCHs have been transferred from short-term hospitals. At the LTCH, they receive a range of services, including cardiac monitoring, ventilator support, and wound care. Existing LTCHs in District 7 At the time of the CON application there were 12 LTCHs operating in Florida with a total licensed bed capacity of 805 beds. There is one existing LTCH within District 7. Another is approved and under construction. Select Specialty Hospital-Orlando, Inc. ("Select-Orlando") contains 35 beds; it was licensed in 2003. The occupancy rate for this facility for CY 2005 was 73.57 percent. Select-Orlando's history shows few discharges to Brevard County. The majority of its discharges are to Orange, Seminole, and Osceola Counties. Most of the balance are to Volusia, Lake, and Polk Counties. A second LTCH, Select Specialty-Orange, Inc., has been approved and is under construction. It will contain 40 beds. The total licensed capacity of these two LTCHs will be 75 beds. Both of the facilities are located in Orange County and are located in or near Orlando within a few miles of each other. The acuity levels of the patients in the existing LTCH are not known. There are no LTCHs in Brevard County where Kindred proposes to build and operate a new LTCH should its application be approved. Kindred's Proposal Kindred's proposal in Brevard County, AHCA District 7, is for a freestanding 60-bed LTCH, with all private rooms, including an 8-bed intensive care unit (ICU). The proposed LTCH will follow a care model template that is similar to Kindred's other LTCHs. It will be a freestanding, licensed, certified and accredited acute-care hospital with an independent self-governed medical staff under the same model as a short-term acute hospital. The majority of patients in an LTCH typically arrive after discharge from a short-term acute care hospital, most often ending their STACH stay in an ICU. Not surprisingly, Kindred projects that its proposed LTCH will receive the bulk of its referrals from STACHs in the surrounding area. Kindred's LTCH patients will be discharged to either their homes, home health care, or to another post-acute provider on the basis of patient needs, family preference, and geography. There are several levels of care provided within an LTCH such as Kindred's proposed facility. Typically, LTCHs accept stable medical patients but with catastrophically ill patients some are bound to become medically unstable. There are eight ICU beds for the medically unstable patient. Thus, Kindred's patients who undergo changes of condition (such as becoming medically unstable) can be cared for without a transfer, unlike in skilled nursing facilities or comprehensive medical rehabilitation hospitals facilities not suited for the medically unstable patient. The goal of an LTCH is to take acute care hospital patients and provide them with a higher level of medical rehabilitation than they would receive in an STACH, and rehabilitate them so that they can be transferred home, or to a rehab hospital, or to a nursing facility. The "medical rehabilitation" of an LTCH addresses system failures and dependence on machines. This is different from the rehabilitation that takes place in an inpatient or outpatient rehab center, where patients usually have suffered an injury or trauma to a muscular or bone system, and their care is based on physical medicine rather than internal medicine. The Orlando Metropolitan Area and Brevard County In evaluating markets that may need an LTCH, Kindred looks at established metropolitan areas the boundaries of which are determined on the basis of population concentrations and commuting data. District 7 contains most of the metropolitan area associated with the city of Orlando (the "Orlando Metropolitan Area"). Like District 7, the Orlando Metropolitan Area has a presence in four counties. But the counties are different. The Orlando Metropolitan Area encompasses all or part of Orange, Osceola, Seminole, (shared with AHCA District 7) and a county that is not in District 7: Lake County. In addition to the three counties it shares with the Orlando Metropolitan Area, District 7 includes Brevard County. At hearing, Mr. Wurdock explained the following about the Orlando Metropolitan Area: When we talk about the Orlando area, we are not just talking about Orange County. Orange County, Osceola County, and Seminole County are all part of the Orlando metro area. That means they're an integrated economic unit based on commuting patterns. Lake County is also part of [the Orlando metropolitan area.] . . . [W]hen we looked at . . . Orange, Osceola and Seminole and ran an analysis . . ., we found . . . there was a need for approximately 180 more beds beyond the . . . 35 that currently existed. So even after you take out the 40 under construction, there is still a really huge need [in the Orlando metropolitan area.] Tr. 56-57. That Brevard County is not part of the Orlando Metropolitan Area is a consideration in this case. Kindred's evaluation also showed two other factors about Brevard County that distinguish it from the Orlando Metropolitan Area. First, it does not have adequate access to long-term care hospitals. Second, it's population with a significant number of seniors and a high number of discharges from STACHs makes it one of the few markets of its size that does not have at least one LTCH. As Mr. Wurdock continued at hearing: Brevard County has a population of more than a million people and it's got more than 100,000 seniors and they have six short term hospitals that produce more than 60,000 discharges a year. . . . [T]here are very few markets of that size in this country that do not have at least one long term hospital . . . Tr. 57. These two factors led Kindred to pursue the application that is the subject of this proceeding. Kindred's decision to pursue a CON for an LTCH in District 7 also stemmed from the interest of Brevard County physicians who had referred patients to Kindred facilities in Fort Lauderdale and Green Cove Springs in Duval County, a government unit consolidated with the city of Jacksonville. This interest was also supported by evidence that showed a predominate north/south referral pattern along the I-95 corridor. Patients in Brevard County STACHs appropriate for LTCH services are referred to facilities in Duval County (north) and Fort Lauderdale (south), but generally not to the lone District 7 LTCH in Orlando. The number of short-term acute hospitals in an area affects the decision of whether to locate a facility in a particular market. The presence of STACHs in a market is significant because the vast majority of an LTCH's patients are transfers from STACHs. The growing senior population (persons aged 65 and over) in Brevard County was also a factor; the elderly population is a large constituent of an LTCH's patient base. Dr. Richard Baney, who practices with Melbourne Internal Medicine Associates, the largest physician- practice group in Brevard County, holds privileges at Holmes Regional Medical Center, and is familiar with the various health care facilities of all types in Brevard County, including hospitals, inpatient rehabilitation hospitals, and nursing homes. Dr. Baney anticipates serving as either an attending or consulting physician if the Kindred facility in Brevard is approved, as do several of the other physicians in his group, including some "intensivists" such as pulmonologists, critical-care physicians, and cardiologists. Dr. Baney's physician group consists of 45 primary care physicians, including internists, family practitioners and pediatricians. The group also includes OB/GYNs, neurologists, medical sub-specialists such as cardiologists, pulmonologists, endocrinologists, hematologists, and oncologists. Among the oncologists are radiological oncologists. There are general surgeons in the group, surgery sub-specialists, including vascular surgeons, and ENT (ear, nose, and throat) physicians. Dr. Baney summed up his opinion on the need for an LTCH in Brevard County as follows: In our area we have excellent acute- care hospitals, and we have a good network located throughout the area of subacute rehab facilities, as well as nursing homes, and then home care, and then eventually a patient is home. What we don't have in this area is a long-term acute-care facility that would handle the more significantly ill patients who need more intensive medical and nursing and physical therapy support. Right now those patients that would normally benefit from this type of facility have to dwell in the hospital for . . . weeks and weeks at a time until they achieve a point of stability where they can be moved into a subacute rehab. What this does in turn is clog up the hospital beds, ICU beds in particular, and every year we have at our large acute-care hospitals here at Holmes patients who are being quartered in the . . . auditorium at the hospital, in the hallways of the emergency room, since the hospital gets just overwhelmed with patients and cannot move them out. Certainly I believe a facility in this area would have no trouble being able to fill that need of taking many of these patients who need this kind of care out of the [STACH] into a better, more efficient setting. Also, we don't have any place that's nearby that patients and their families can go for this kind of care. . . . [I]t's really not logistically feasible for patients and their families to go 80, 90 miles away or further to . . . have their care for this type of duration. Kindred No. 7, Deposition of Richard Baney, Jr., M.D., at 10-11. When asked about the difficulty presented by the distance to the LTCHs in Fort Lauderdale and Duval County, Dr. Baney answered with regard to one of his patients that administratively there a few if any problems. The problem is for the family: But the family was very hesitant to allow their father to be transported . . . 150, 180 miles away and be there for weeks or months while they were recovering. They were quite resistant to the idea of him so far away, since the family would have to travel back and forth. Eventually they overcame that and the patient did go . . . to the facility down south. * * * But it was quite a hurdle that we had to get over. Id., p. 16, 17. Aside from the logistical problems faced by the families whose loved one is a potential patient at an LTCH at great distance from home, Dr. Baney's testimony accentuates another factor faced by potential LTCH patient in Brevard County. This is a factor favoring approval of an LTCH application recently recognized by AHCA when it approved Select-Orange, a second LTCH in the Orlando Metropolitan Area dominated by two large hospital organizations. Similar to the Orlando area, Brevard County STACHs, for the most part, belong to one of two hospital organizations predominate in the area. Brevard County's Two Main Hospital Organizations There are two main hospital organizations in Brevard County: the Health First system and the Wuesthoff system. Health First includes Holmes Regional Medical Center; Palm Bay Community Hospital, which is about 90 beds; and Cape Canaveral Hospital, which is also about 90 beds, in the central part of the county. Palm Bay is a large community about 15 miles south of Melbourne. Cape Canaveral is about 20 miles from Melbourne, and Rockledge is about 15 miles from Melbourne. The Wuesthoff system consists of Wuesthoff Rockledge and Wuesthoff Melbourne. Wuesthoff Rockledge is a 267-bed acute care facility with 32 ICU beds, 8 cardiac surgery beds, and an active emergency room that sees about 1,500 visits a month. Wuesthoff Melbourne is a 115-bed facility with a 12-bed ICU and an active ER of around 800 visits a month. Wuesthoff currently refers LTCH patients- primarily long-term ventilator patients-to Kindred's facilities in Fort Lauderdale and near Jacksonville. When Wuesthoff refers a patient to Kindred, it calls Kindred's intake coordinator who journeys to Wuesthoff to review the patient's records, meet with the family, and determine if the patient can be placed. Only if a physician from the LTCH signs an admission order concurring that the patient is clinically appropriate for admission to an LTCH is the patient transferred. Often, however, because the Kindred facilities are so far away, just as Dr. Baney pointed out, the families do not want to move the patient out of Wuesthoff. This resistance continues despite increased education about the benefit of LTCHs to potential LTCH patients. LTCH Education When an LTCH comes into a market, an education process begins. It begins with the physicians, and with the case managers and social workers in the STACH. Kindred educates these professionals about what an LTCH is, what its services are, and where it fits into the continuum of care. Kindred's Admission and Patient Evaluation Processes Kindred does not admit every patient that falls within the diagnoses that might produce LTCH-appropriate patients. Patients are pre-assessed before admission using what is nationally known as Interqual criteria for hospital admissions. That set of criteria is based on severity of the patient's illness and the intensity of services required to treat the patient, and then a review committee at the LTCH makes a clinical determination whether or not the patient is appropriate for LTCH services. The sole way that a patient gets referred to a Kindred Hospital is through a physician order. Before a patient comes to a Kindred Hospital, a physician has determined that to the best of his or her judgment the patient requires continued care at the level of an acute care hospital and that the patient's course of treatment will be prolonged. A physician from a Kindred Hospital must write the admission order, concurring that it is appropriate for that patient to be in an LTCH. Prior to obtaining that physician order, potential candidates for transfer are identified through the STACH case management staff, with the assistance of the LTCH staff. The STACH medical staff, nurses, or other personnel initiate the request for Kindred to visit the patient, interview the family, talk with the STACH attending physician, and make a determination of whether transfer and care at Kindred is clinically appropriate. Kindred gathers information on a potential patient to assist in making the admission determination using individuals in the field known as "clinical liaisons," who are primarily licensed registered nurses. The clinical liaison gathers the information, but does not make the ultimate determination as to whether to admit the patient to a Kindred facility. The ultimate determination for admission is made by the physician who will be seeing the patient at the Kindred facility. In order to comply with Medicare reimbursement requirements, Kindred employs such safeguards to make sure only appropriate patients are admitted. Medicare reviews the patients treated into the hospital, and it can and does reduce payment for "short stay outliers" who do not stay at least five-sixths of the geometric mean of the length of stay (GMLOS) for the patient's diagnosis. Mathematically, however, LTCHs will always have some patients who are short stay outliers. Even if GMLOSs rise as result of the elimination of short stay patients, between 35 and 40 percent of patients will always be "short stay outliers" under CMS's current definition. They will just be hospitalized for a stay that is short relative to a longer length of stay. Kindred LTCHs utilize criteria that assure that patients, once admitted, have sufficient severity of illness and need sufficient intensity of service to continue to warrant acute care. Case managers in LTCHs apply discharge screens to patients as they near completion of their LTCH care plan to help physicians make a judgment of when they are ready to be transferred either home or to a lower level of post-acute care. Kindred's CON application included a utilization review plan, using an example from Kindred Hospital North Florida. Every hospital has a utilization review plan designed to assure that appropriate care is given to patients. It serves an oversight function for medical care, nursing care, medication administration, and any other area where resources are expended on behalf of the patient. A PPS for LTCHs Effective October 1, 2002, the federal Centers for Medicare and Medicaid Services (CMS) established a prospective payment system for LTCHs. Through this system, CMS recognizes the patient population of LTCHs as separate and distinct from the populations treated by providers of short-term acute care or other post short-term acute care providers. Under the system, each patient is assigned an LTCH DRG, indicating that the patient's diagnosis is within a certain Diagnostic Related Group. The LTCH is reimbursed the pre-determined payment rate for that DRG, regardless of the cost of care. The creation of separate DRGs for LTCH patients is the mark of the federal government's recognition of the validity of LTCH services and the distinct place occupied by LTCHs in the continuum of care based on the high level of LTCH patient acuity. Despite this recognition, concerns about the identification of patients that are appropriate for LTCH services have been voiced both at the federal level and at the state level. With the rise in LTCH applications over the last several years, AHCA has been consistent in voicing those concerns, particularly when it comes to LTCH population levels of acuity. Acuity The Agency is not convinced that there is not significant overlap between the LTCH patient population and the population of patients served appropriately in healthcare settings other than LTCHs. The Agency has reached the conclusion that there are options (other than an LTCH in Brevard County) available to patients targeted by Kindred. The options depend on such matters as physician preference and the availability of long-term care hospitals in a given geographic area. Kindred answers the concerns, in part, with evidence that relates to acuity. A "case mix index" for the hospital is a measure of its average resource consumption. Resource consumption can be viewed as a surrogate measure of complexity and severity of illness, so case mix index is often cited as a readily available measure of patient acuity. Using that indicator, the case mix index of Kindred hospitals is high compared to the entire LTCH industry, and is higher than the average case mix index for STACHs. The APR/DRG system is a way to further refine the variation of patients' acuity within a DRG. The system assigns not only a DRG, but a severity of illness on a scale of one (minor severity) to four (extreme severity). Using that tool with the Kindred data base (as well as the federal MedPAR data base) confirms that the distribution of severe and extreme severity of illness is skewed toward LTCH patients, meaning that there are more patients with higher severity of illnesses in LTCHs than in STACHs. As is to be expected, and one would hope if LTCHs are appropriately serving their niche in the continuum of care, this is consistent with the empirical observation that patients in LTCHs, are more sick than those in STACHs. A third measure of patient acuity routinely used in Kindred hospitals is an APACHE score, which is a combination of physiologic derangement and concurrent illnesses. The average Kindred patient has an APACHE score of about 45, whereas the average critical care patient in all STACHs has a score about two-and-a-half points higher. Thus, Kindred's LTCHs treat a severely ill population only a few points, on the APACHE measure, below that of critical care units in STACHs across the country. The Agency does not, by rule or order, define the level of acuity at which LTCH patients should be for admission. Information on acuity level of patients in STACHs is not available through the State's health statistics data base, nor is any information that would allow an LTCH applicant to undertake an acuity analysis of potential patients. AHCA acknowledges that it has no reason to believe that Kindred admits lower-acuity patients with the least need for resources among those in LTCH- appropriate DRGs. Family Hardship In those markets that do not have LTCHs, STACH patients typically have no choice of treatment but to stay in the STACHs, unless they are willing to travel long distances. As Dr. Baney pointed out in his deposition testimony, many patients who could benefit from an LTCH are not inclined to travel long distances. One reason is that the patients' families are not able to commute that distance. If the patient is going to be in an LTCH for weeks or months, it creates a hardship on the family to have their loved one that far away. The family either loses contact with their loved one or they actually have to relocate to where the loved one is and abandon their home temporarily. The need for family presence and involvement is more than just an emotional matter of patient and family preference. Families are involved in the treatment of a patient in a long-term care hospital, not only through their presence in the hospital but also because they will participate in patient care after the patient leaves the LTCH. Families have to learn how to get the patients out of bed, feed them, and possibly suction them. The families would be taught how to care for their family members once they leave the LTCH, by nursing and therapy staff, teaching them exercises for the patient, how to regulate the oxygen, and giving medications. Differences between LTCHs and Other Providers LTCHs and STACHs do not have the same purpose, and the gap is widening between the two. Over the last 20 years, STACHs have evolved into settings that are very good at stabilizing patients, diagnosing their conditions, and developing treatment plans. Most admissions to the medical ward of an STACH come in through the emergency room where patients are so acute, so unstable, that emergency care is required to stabilize the patient. In their role as diagnostic centers, STACHs provide imaging and laboratory services, and then develop a treatment plan based on the diagnostic work-up performed. STACHs have moved away from the function of carrying out a treatment plan. This is borne out by shrinking STACH lengths of stay over the last 20 years, which now average four to six days. As a result, STACHs have limited capability to provide a prolonged treatment plan for patients with multiple co-morbidities. In contrast, LTCHs do not hold themselves out to be diagnostic or stabilization centers. They have developed expertise in caring for the small subset of patients that require a prolonged treatment plan. A multi-disciplinary physician- based care plan is provided in LTCHs that is not provided in STACHs or other post-acute settings. LTCH patients meet hospital level criteria, and if there is no LTCH readily accessible to provide a hospital-level discharge option for these patients, then the STACH has no option but to keep them, and manage their treatment and costs as best they can. LTCHs take care of those patients who need to be in a hospital, but for whom reimbursement is not adequate for STACHs to treat. The reimbursement system is driving this to a great extent, because of the incentives it gives to discharge patients as quickly as possible. Not every STACH patient needs LTCH care; as a rule of thumb, about one percent of all non-obstetric patients are potentially LTCH-appropriate. Ms. Woods, Vice President for Wuesthoff Health System which operates STACHs in Brevard County, testified in deposition that Wuesthoff's ICUs in Wuesthoff hospitals often retain patients who could be placed in an LTCH. As the Wuesthoff ICUs remain full, the ability to move patients through the hospital, from the emergency department through the ICU, is significantly impacted. While long-term care hospitals take a team approach to getting patients weaned from ventilators or getting them to a rehab involvement, an acute care hospital ICU deals more with acute crisis situations, such as an acute MI (myocardial infarction) or an acute blood clot to the lungs, or someone who has acute sepsis or infection. The roles that LTCHs play have a significant impact on acute care hospitals such as Wuesthoff. If an acute care hospital has to maintain a patient for 30 to 60 days on a ventilator in order to get them weaned or to meet their needs, that poses the potential to interfere with the acute care hospital from meeting the needs of the community, such as patients who are coming in the emergency room with acute conditions. Most of the stays in Wuesthoff's ICU beds, for example, are five to seven days; they are trauma patients, surgery patients that need support and critical care, and patients coming in with major infections. When ICU beds are unavailable, these patients are being held in the emergency departments; it stops the patient flow if the beds in a community hospital are taken up from a long-term ventilator patient. SNFs and LTCHs are different both in intent and execution. SNFs are appropriate for patients whose primary needs are nursing, who are stable and unlikely to change, and who do not require very much medical intervention. Conversely, LTCHs, being licensed and accredited as acute care hospitals, are appropriate when daily medical intervention is required. LTCHs are able to respond to changes in conditions and changes in care plans much better than SNFs because LTCHs have access to diagnostics, laboratory, radiology, and pharmacy services. Further, there are no skilled nursing facilities in Brevard County that operate beds for ventilator dependent patients, nor are there hospital-based skilled nursing units ("HBSNUs"). Using Kindred's own nursing data base, which consists of 250 SNFs across the country, and Kindred's LTCH data base, consisting of 75 LTCHs, Kindred has discovered that that overlap in patient condition is very small. Where there is overlap, it tends to be at the ends of care in LTCHs and the beginning of care in SNFs. This progression makes sense, since SNFs are a common discharge destination for LTCH patients. LTCHs and rehab hospitals are also distinctly different. Rehab hospitals are geared for people with primarily neurologic or musculoskeletal orthopedic issues, and are driven with a care model based on physical medicine rather than internal medicine; LTCH care requires the oversight of an internist rather than a physical medicine doctor. While rehab is a concurrent component of LTCH care, the patient in an LTCH cannot tolerate the three hours per day of therapy required for admission to rehab hospitals due to their medical conditions. In fact, a common continuum of care is for an LTCH patient to receive treatment and improve to the point where they can tolerate three hours of rehab and so be transferred to a rehab hospital. There is one acute rehab center in Brevard County, and it does not take ventilator-dependent patients. There are no hospital based skilled nursing units in Brevard County. There are no skilled nursing facilities in Brevard County that can accommodate ventilator-independent patients. Often ventilator-dependent patients also have IV antibiotics and tube feedings, and these are complicated conditions that a nursing home will not treat. LTCH care cannot be provided through home health care, because, by definition, LTCH patients meet criteria for inpatient hospitalization. Home health care is designed for patients who are very stable and have such a limited medical need that it can be administrated by a visiting nurse or by families. This is in sharp contrast to an LTCH patient where many hours a day of nursing, respiratory, and other therapies are required under the direct care of a physician. On the basis of regulation alone, STACHs could provide LTCH care. They generally do not do so because they have evolved into centers of stabilization, diagnosis, and initiating a treatment plan. Case studies bear out that when patients who made very little progress in STACHs are transferred to LTCHs, where the multidisciplinary approach takes over from the diagnostic focus, the patients improve in both medical and physical well-being. Those patients that would normally benefit from an LTCH have to dwell in the hospital for weeks until they achieve a point of stability where they can be moved in to a subacute facility; instead of continuing to move efficiently down the continuum they remain in the "upper end of the stream." This, in turn, may overwhelm the short-term acute care hospital, particularly in its ICU, resulting in patients being quartered in the auditorium at the hospital and in the hallways of the emergency room. The LTCHs available along the east coast of Florida in Fort Lauderdale or Jacksonville are at a distance from Brevard County that is an obstacle to referral of a Brevard County patient. Having a long-term care hospital in Brevard County would enhance the continuum of services available to Brevard County residents. On the other end of the referral process from Dr. Baney is Rita DeArmond, the clinical liaison for Kindred Hospital Fort Lauderdale. Her duties include, "patient evaluations on potential admissions to [Kindred Fort Lauderdale], which also involves meeting with families and educating the families, . . . case managers, . . . physicians and other people in the community about our hospital and long-term acute care hospitals in general." Kindred No. 8, at 5. She serves "Palm Beach County, the area around Lake Okeechobee [Okeechobee and Hendry Counties], Martin County, . . ., St. Lucie County, Indian River County and Brevard County." Id. at 11. In Ms. DeArmond's experience in dealing with potential long-term care hospital patients and their families not in the immediate vicinity of an LTCH, the willingness of those patients to travel great distances is the biggest hurdle for the patients admission to an LTCH. Most of the patients and their spouses are elderly, and they do not tend to travel long distances, or on the interstate. Being faced with traveling hundreds of miles round-trip to visit a loved one is very distressing to most of them. Not only would potential Brevard County LTCH patients be more likely to avail themselves of LTCH services if there were an LTCH in Brevard County but so would patients in other counties. For example, according to Ms. DeArmond, Lawnwood Regional Medical Center in Fort Pierce, a St. Lucie County STACH, and Sebastian River Medical Center, an STACH in Indian River County, would definitely send potential LTCH patients to an LTCH in Brevard County rather than the current closest LTCH, Kindred Fort Lauderdale. Having an LTCH would be a positive impact for other Brevard County STACHs as well. For example, Wuesthoff would not experience the backup in its emergency department and in its ICU beds, especially in the winter time where there is a high census due to more cases of pneumonia in the winter. If a patient who might be clinically appropriate for an LTCH remains in the ICU in an acute care hospital such as Wuesthoff, that patient does not receive the same care that he or she would receive at an LTCH. Acute care hospitals do not provide the medical rehabilitation work that LTCH's do, such as a plan of care just for the rehab of ventilator patients. An acute care hospital can deal with the pneumonia, and can wean the patient, but does not have the same plans or care or the same focus that an LTCH does with those types of patients. If the patient does not go to an LTCH, they will stay in the acute care hospital using the hospital resources. Wuesthoff has had patients there up to 65 days. The hospitals and physicians visited by Kindred- Fort Lauderdale clinical liaison Ms. DeArmond on a regular basis are located in Brevard County in District 7, as well as Indian River and St. Lucie counties. The hospitals within Brevard County that she contacts include Holmes Regional and Wuesthoff Melbourne Hospital; within Indian River County, Indian River Memorial Hospital in Vero Beach and Sebastian River Medical Center, and within St. Lucie County, St. Lucie Medical Center in Port St. Lucie and Lawnwood Hospital in Fort Pierce. In gathering letters of support that were submitted with Kindred's CON application for a long-term care hospital in Brevard County, Ms. DeArmond met with case managers and physicians and informed them of Kindred's intention to apply for a CON to build a hospital in Brevard County. The physicians and case managers who provided letters of support had previously referred patients to Kindred Hospital in Fort Lauderdale, so they were familiar with the services that Kindred can offer in an LTCH. It is reasonable to assume that such physicians and case managers would refer patients to a Kindred LTCH in Brevard County, if approved. MedPAC Concerns In denying Kindred's application, AHCA relied on reports issued to Congress annually by the Medicare Payment Advisory Committee (MedPAC) that discuss the placement of Medicare patients in appropriate post-acute settings. The June 2004 MedPAC report state the following about LTCHs: Using qualitative and quantitative methods, we find that LTCHs' role is to provide post-acute care to a small number of medically complex patients. We also find that the supply of LTCHs is a strong predictor of their use and that acute hospitals and skilled nursing facilities are the principal alternatives to LTCHs. We find that, in general, LTCH patients cost Medicare more than similar patients using alternative settings but that if LTCH care is targeted to patients of the highest severity, the cost is comparable. AHCA Ex. 7, at 121. The June 2004 MedPAC report, therefore, concludes that LTCHs should "be defined by facility and patient criteria that ensure that patients admitted to these facilities are medically complex and have a good chance of improvement." Id. Despite the above language in the June 2004 MedPAC report, discussion in the SAAR of portions of the MedPAC report shows that AHCA may have misread some of the subtleties of the MedPAC findings. The MedPAC report makes statements that LTCHs and SNFs substitute for one another. While there is some gross administrative data to support that hypothesis, that conclusion cannot yet be drawn due to limitations in data and the wide variation of patient conditions that may be represented by a single administrative grouping such as a DRG. An example of patients in different settings who would appear to be similar are those under DRG-475, which means they were on ventilator life support for at least 96 hours. Such patients may be discharged in conditions that vary greatly. These conditions range from an "alert, talking patient, no longer on life support," to a patient who is "not on life support but is making no progress." There is no administrative data that describes patients at the time of their discharge. MedPAC analysis, therefore, lacks the data to determine why some of those patients went to a higher versus a lower level of care. The SAAR also concludes, based on a letter from the MedPAC Chairman, that LTCH patients cost more on average than patients in other settings. This conclusion is based on an analysis that is unable to differentiate patients within a DRG based on their severity at the time of discharge. The limitation in the DRG is that it is designed to describe the patient's need at the time of admission rather than discharge, so the DRG classification alone does not identify whether the patient was healthy or ill at the time of discharge. Furthermore, MedPAC found that patients who tended to be more severe based on DRG assignment tended to be cared for at similar cost between LTCHs and other settings. In fact, for the tracheostomy patient, which is the extreme of severity and complexity, there was evidence of lower cost of care for patients whose case included an LTCH stay. MedPAC Chairman Glenn Hackbaith, in his March 20, 2006 letter, agreed that CMS's proposed change to the short stay outlier policy was "too severe"; that it affects a "substantial percentage of LTCH patients"; and that it would continue to affect a large percentage of admissions "regardless of the admission policies of LTCHs." MedPAC's March 2006 Report to Congress notes that the total Medicare payments to LTCHs nationwide -- $3.3 billion in 2004 -- represented less than one percent of all Medicare spending. Need Analysis in the Absence of an AHCA Need Methodology The Agency does not have a rule that sets out a formula for determining the need for LTCH beds. Accordingly, AHCA does not publish a fixed need pool for LTCH beds. As the parties agree, this case is governed, therefore, by Florida Administrative Code Rule 59C- 1.008(2)(e)2.a-d (the "Needs Assessment Rule"). Application of the Needs Assessment Rule makes Kindred responsible for demonstrating need through a needs assessment methodology that covers specific criteria listed by the rule as detailed below, following the sections of this Order devoted to Kindred's Need Methodology and AHCA's criticisms of it. Kindred's Need Methodology Kindred bases its need methodology (the "Kindred Methodology") in this case on long-stay patients in short- term hospitals. A description of the Kindred Methodology, supported and proved by the testimony of Mr. Wurdock at hearing, appears in Kindred's CON application under a section entitled "Bed Need Analysis," see Exhibit K-1, at 14. It begins with the statement: "Long-term care hospital bed need can be estimated directly based on the acute care discharges and days occurring in the market." Id. There follows a chart that lists the six Brevard County STACHs and shows the number of patient discharges in the six months ending March 2004 and the patient days for the same period. These total 68,710 and 309,704, respectively. To identify the number of patient days appropriate for LTCH care, the Kindred Methodology takes into account patient diagnosis at discharge, patient age and length of stay. Some types of patients (burn patients, obstetric and pediatric patients or behavioral patients) are not appropriate for LTCH admission. Likewise, patients with short-term rehabilitation diagnoses typically are not appropriate for LTCH care. The first step in the Kindred Methodology, therefore, is to identify and omit those diagnoses which represent patients not appropriate for long-term care admission. Those include all DRGs in the Major Diagnostic Categories (MDC) of 13-Female Reproductive System; 14-Pregnancy, Childbirth and Puerperium; 15- Newborns and Other Neonates; 19-Mental Diseases and Disorders; 20-Alcohol and Substance Abuse; 22-Burns; and 23-Factors Influencing Health Status. Two additional groups of DRGs are omitted by the Kindred Methodology: DRGs specific to patients less than 18 years of age and DRGs for organ transplant patients who are usually required to remain in the STACH for specialized care. The end result of the first step in the Kindred Methodology is a list of 387 short-term acute care DRGs ("LTCH Referral DRGs") that represent patients who potentially could be eligible for LTCH admission. The Kindred Methodology's second step is to identify discharges that are assigned to one of the LTCH Referral DRGs and are aged 18 or older and whose length of stay exceeds a threshold number of days. This threshold is described in the application as follows: The length of stay threshold is defined in terms of the national geometric mean length of stay (GeoMean). That statistic is calculated annually by the federal Centers for Medicare and Medicaid Services (CMS) for each DRG. The number of long-term hospital patients and patient days is affected by the timing of the referrals. Referrals usually occur after the patient's length of stay has become longer than average. It is commonly accepted that many patients who stay in the acute care hospital beyond the geometric mean length of stay would be best cared for in a specialized, long- term environment. Therefore, in this analysis it is assumed that referral to Kindred Hospital Brevard will occur five days after a patient has passed their DRG-specific geometric mean length of stay. This allows time for patient assessment and transfer arrangements. Another important factor affecting the potential number of long-term hospital patients and patient days is the length of time a patient stays in the LTCH. In order to qualify for Medicare certification, long-term care hospitals must maintain a minimum average length of stay of twenty-five days or greater among their Medicare patients. Admission criteria, therefore, are used to minimize the number of Medicare patients requiring just a few days of care. To reflect this in the analysis, patients are considered to be LTCH appropriate only if they would have a long-term hospital length of stay of ten days or more. Exhibit K-1, at 16. Discharged patients, therefore are considered appropriate for LTCH care by the Kindred Methodology if they are discharged from a Brevard County STACH, are at least 18 years of age, are assigned to one of the 387 Referral DRGs, and have a hospital length of stay that exceeds the geometric mean by at least 15 days, the sum of a referral period of five days and an LTCH minimum length of stay of ten days. The third step in the Kindred Methodology is to sum the potential LTCH days produced by the appropriate patients. For these patients, potential LTCH days include all days after the "'transfer day' (i.e., all days that exceed the GeoMean + five days)." Id. For the 12-month period ending March 2004, this calculation yielded approximately 18,400 hospital days in the six Brevard County hospitals, for an average daily census (ADC) of 50.4. The fourth step in the Kindred Methodology is to identify the number of patient days that are leaving Brevard County for LTCH care, due to the absence of an LTCH in the county. During the 12-month period ending in March 2004, 41 Brevard County residents were discharged from Kindred Hospital North Florida in Green Cove Springs and Kindred Hospital Fort Lauderdale. Those patients received 2,229 days of LTCH care, equaling an average daily census of 6.1. Adding that to the 50.4 ADC un-served patients in Brevard County, yields a potential LTCH ADC of 56.5. The fifth step is to account for population growth. This is especially important when there is rapid growth in senior population as there is in Brevard County. According to AHCA projections, the population 65 and over will increase 9.2 percent during the next five years, while the total population will increase 10.8 percent. It is appropriate to increase LTCH ADC at least by 9.2 percent during this time period, since the proposed project will not open until 2007 at the earliest, and will not achieve full utilization until at least 2011. This step produces an LTCH ADC of 61.7. The sixth and final step is to calculate LTCH "bed need" by assuming 85 percent occupancy. Dividing the LTCH ADC of 61.7 by 0.85 yields a bed need of 72 LTCH beds. The Kindred Methodology does not account for the five percent or more of referrals that come from sources other than LTCHs such as nursing homes. Nor does it take into account the admissions from Indian River County currently served by Kindred Hospital Fort Lauderdale, some of which are sure to come to the proposed project if approved. AHCA Criticism The methodology is criticized by AHCA on the bases, among others, that it does not account for beds available elsewhere in District 7, and that it determines need solely within Brevard County, a departure from the statutory mandate which requires Agency review of CON applications with regard to "availability, quality of care, accessibility, and extent of utilization of existing health care facilities in the service district of the applicant." § 408.035(2), Fla. Stat. (emphasis supplied). The Agency's argument with regard to the un- utilized beds at the one existing LTCH in District 7, Select-Orlando, is undermined by recent action of the agency in approving a second Select facility in Orange County, a 40-bed freestanding facility: Select Specialty Hospital-Orange, Inc. ("Select-Orange"). The Agency approved the 40-bed Select-Orange facility, not open at the time of hearing, by way of a Settlement Agreement (the "Select-Orange Settlement Agreement") with the applicant. The two parties to the agreement, AHCA and Select-Orange, jointly stated in that document: [T]he Agency, in recognizing that there are two distinct health systems in the Orlando area, believes that this LTCH is needed for the Orlando Regional Healthcare System due to that unique situation . . . Kindred Ex. 4, at 2. The two distinct health systems in the Orlando area are Orlando Regional Healthcare System, Inc., which has a number of STACHs in the Orlando area including Orlando Regional Medical Center (ORMC), a tertiary medical facility with more than 500 beds, and the Adventist Health System, Inc. (Adventist), a hospital organization with a nationwide presence that as of 2002 operated seven acute care campus systems under a single license held by Adventist d/b/a Florida Hospital in the Orlando Metropolitan Area. See Orlando Regional Healthcare System, Inc. vs. AHCA, Case No. 02-0449 (DOAH November 18, 2002), pp. 8-10. The Select-Orange Settlement Agreement was entered in the midst of administrative litigation over AHCA's preliminary agency action with regard to a CON application. The meaning and impact of AHCA's statement quoted above from the Select-Orange Settlement Agreement were not fully elaborated upon at hearing by any direct evidence. Kindred established through the testimony of Mr. Wurdock and through cross-examination of Ms. Rivera that although Select-Orange was originally approved as a "hospital-in- hospital" or "HIH," that Select-Orange obtained a modification of its CON to become a freestanding facility. Had the facility remained an HIH, federal regulations would have limited the percentage of Medicare referrals that could come from its host hospital, ORMC. As a freestanding facility, Select-Orange has no such limitations. It can fill its beds with referrals from ORMC. Whatever the impact of the freestanding nature of Select-Orange, the Agency's recognition of the unique situation in the Orlando area created by two distinct health systems, such that there is support for a new LTCH when the existing LTCH has available beds, gives rise in this case to an inference in Kindred's favor. If two distinct hospital systems in the Orlando area can support the addition of 40 LTCH beds, then it is highly likely that Brevard County can support a 60-bed LTCH. The county is not a part of the Orlando Metropolitan Area. LTCH referral patterns are north-south along the I-95 corridor (not to Select-Orlando). There are geographic and roadway access issues from Brevard County to the Orlando area demonstrated by commuting patterns that exclude Brevard County from the Orlando Metropolitan Area. And most significantly, the methodology reasonably established need for more than 60 beds in Brevard County. The Needs Assessment Rule The need for any health care service or program regulated by CON Law for which AHCA has not provided a specific need methodology by rule is governed by Florida Administrative Code Rule 59C-1.008(2)(e)(the "Need Assessment Rule"), which states in part: . . . If an agency need methodology does not exist for the proposed project: . . . If an agency need methodology does not exist for the proposed project: The Agency will provide to the applicant, if one exists, any policy upon which to determine need for the proposed beds or service. The applicant is not precluded from using other methodologies to compare and contrast with the agency policy. If no agency policy exist, 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; Market conditions; and Competition. The Agency does not publish a fixed need pool for LTCH beds because it does not have a specific need formula or methodology for LTCH beds. The Agency, furthermore, has not provided Kindred with any policy upon which to determine need in this case. Accordingly, Kindred used its own methodology for determining need in Brevard County and elsewhere in the district (the Orlando Metropolitan Area). Finally, since no agency policy exists with regard to an LTCH need methodology, Kindred is required to prove the existence of need for its proposed project on the basis of the five categories of criteria (referred to in the rule as "topics") listed in sub-paragraphs "a. through e.," of paragraph 2., in subsection (2)(e) of the Rule. Population Demographics and Dynamics In assessing an area's population and demographics for the purpose of evaluating LTCH need, special attention is paid to the elderly population because the majority of LTCH patients are Medicare patients. The elderly are also more likely to produce LTCH patients because they are more likely to be medically complex and catastrophically ill with co-morbidities and dependent on medical equipment like ventilators. Brevard County, while home to only an approximate one-quarter of District 7's population, accounts for more than a third of its seniors. While Brevard County's elderly population is experiencing average or slightly below average growth in relation to the rest of the state, there is no question that Brevard County's elderly population is on the increase and reasonably projected to increase in the future. Availability, Utilization, and Quality of Like Services in the District "[B]y definition, putting a long term hospital in Brevard County will increase accessibility [make LTCH services more available] because . . . the people in Brevard County will no longer have to go all the way to Orlando, or Jacksonville, or Ft. Lauderdale for [LTCH] care." Tr. 48. Mr. Wurdock elaborated on the point of district availability at hearing: We did look at the entire district. . . . [T]here [are] only four counties in the district, three of which orbit around the Orlando and then there is the Palm Bay/Melbourne metropolitan area, which is Brevard. And when we looked at the district as a whole, what we discovered was that there is a need really for two new long term hospitals in the district. There is clearly a need for another one in Orlando [beyond the existing Select- Orlando and the approved not yet operating Select-Orange] and there is also a need for one in Brevard County. . . . [You] could build . . . two new long term care hospitals, both of them in Orlando, but that doesn't . . . make . . . sense when you've got a very large concentration of seniors significantly removed from the Orlando area with six short term hospitals in . . . [Brevard C]ounty comprising essentially its own market. So logically, you . . . put one long term hospital in Brevard and then another long term hospital in Orlando. Tr. 48-49. The presence of six STACHs in Brevard County and the large senior population is significant. The closest LTCH is Select-Orlando more than an hour's drive away. The distance to Select-Orlando and Select-Orange's future site from the municipality in which Kindred proposes to site its proposed LTCH, Melbourne, is more than 60 miles, in a direction not favored by Brevard County residents oriented to driving north or south along the I-95 corridor, but not to the west into the Orlando Metropolitan Area. Furthermore, and most significantly, family members rarely fully understand and accept that their catastrophically ill elderly loved one should be shipped 60 miles away when the patient is in a hospital with a good reputation. Their resistance to a referral at such a distance is unlikely to increase utilization at the Orlando area LTCHs no matter how convinced are their physicians and other clinical practitioners that such a move is required for better care. Medical Treatment Trends LTCHs are recognized as a legitimate part of the health care continuum by the federal government and CON approvals of LTCHs in Florida have been on the upswing throughout this decade. At the federal level, in recognition of their treatment of a small but important subset of patients, Medicare has adopted LTCH DRGs, that is, DRGs specific to LTCHs, for reimbursement under Medicare's PPS. At the state level, the Agency recognizes that "[t]he trend is for LTCHs to be increasingly used to meet the needs of patients in other settings who for a variety of reasons are better served in LTCHs." Respondent Agency for Health Care Administration's Proposed Recommended Order, at 15. This recognition is made by AHCA despite MedPAC's concerns, many of which were tempered and adequately addressed by Kindred in this proceeding. Market Conditions At first blush, market conditions might not seem to favor Kindred's application. The occupancy rate in the District indicates that there are available beds. In AHCA's view, the occupancy rate at the one existing LTCH in District 7, the 35-bed Select-Orlando facility, an H-I-H in a converted nursing home at Florida Hospital Orlando, is not optimal. Select-Orlando opened in 2003, only a few years ago, and it is operating at a high occupancy rate that is approaching optimal. Kindred, moreover, did not confine its need case to its Brevard County methodology. It also presented evidence of need in the Orlando Metropolitan Area consisting, in part, of the three other counties in District 7. Competition While the Agency asserts that it did not give competition much weight in this application, AHCA has not taken the position that Kindred's proposed facility would not foster competition. Having an LTCH in Brevard County would foster competition in the traditional sense in that the only LTCHs in the District, one existing and one approved, are those of Select Medical Corporation, Kindred's chief competitor. A Reasonable Methodology for Brevard County In short, Kindred's methodology is reasonable for determining need in Brevard County and it appropriately includes the topics required by the Needs Assessment Rule. The Agency's argument that there is no need for LTCH beds in Brevard County when there are LTCH beds available elsewhere in the district is defeated by its approval of the Select-Orange facility. Whether Kindred's methodology in this case carries the day for Kindred, given the Agency's approach on a district-wide basis to the need for LTCHs, is addressed in the section of this Order devoted to conclusions of law.
Recommendation Based on the foregoing Findings of Facts and Conclusions of Law, it is recommended that the Agency for Health Care Administration issue CON No. 9835 to Kindred Hospitals East, LLC, for a 60-bed, long-term acute care hospital in AHCA Health Planning Service District 7, to be located in Brevard County. DONE AND ENTERED this 29th day of November, 2006, in Tallahassee, Leon County, Florida. S DAVID M. MALONEY 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 29th day of November, 2006. COPIES FURNISHED: Alan Levine, Secretary Agency for Health Care Administration Fort Knox Building III, Mail Stop 3 2727 Mahan Drive, Suite 3116 Tallahassee, Florida 32308 Christa Calamas, General Counsel Agency for Health Care Administration Fort Knox Building III, Mail Stop 3 2727 Mahan Drive, Suite 3116 Tallahassee, Florida 32308 Richard Shoop, Agency Clerk Agency for Health Care Administration Fort Knox Building III, Mail Stop 3 2727 Mahan Drive Tallahassee, Florida 32308 M. Christopher Bryant, Esquire Oertel, Fernandez, Cole & Bryant, P.A. 301 South Bronough Street Tallahassee, Florida 32302 Sandra E. Allen, Esquire Agency for Health Care Administration Fort Knox Building III, Mail Stop 3 2727 Mahan Drive Tallahassee, Florida 32308
The Issue Whether Respondent should recoup Medicaid payments made to Petitioner for health care services provided to eight patients.
Findings Of Fact Petitioner, Indian River Memorial Hospital, Inc., (Hospital), has contracted with Respondent, Agency for Health Care Administration (AHCA), to provide services to Medicaid patients. The parties have agreed that there is a dispute for Medicaid reimbursement for goods and services provided to eight patients: S.G., J.D., R.J., C.A., G.M., S.S., M.P., and C.T. The Agency has paid the Hospital for the services rendered to these eight patients and seeks to recoup the payment based on a retrospective review by a peer review organization, Keystone Peer Review Organization (KePro). The Agency claims that either the admission or a portion of the length-of-stay for the eight patients was medically unnecessary. Services were provided to C.T. in 1994 and to the remainder of the patients at issue in 1995. Payment for Medicaid services is on a per diem basis. The rate for 1994 is $473.22 per day, and the rate for 1995 is $752.14. The Agency contracted with KePro to do a review of the Medicaid payments to the Hospital. KePro employs nurses to review the patient files based on criteria on discharge screens. If the services meet the criteria, there is no further review and the payment is approved. If the nurse determines that the services do not meet the criteria on the discharge screens, the patient's files are reviewed by a board certified physician, who in this case would be a psychiatrist. If the physician determines that the services are not medically necessary, a letter is sent to the Medicaid provider, giving the provider an opportunity to submit additional information. Additional information submitted by the provider is reviewed by a board certified physician. If the doctor concludes that the services are still medically unnecessary, the provider is notified that that services do not qualify for reimbursement and the provider may ask for a reconsideration of the denial. If the provider seeks reconsideration, the file is reviewed by a physician, and the provider has an opportunity to be present during the review. If the physician determines that the services are medically unnecessary, KePro sends a letter to the Agency stating the reasons for denial. The denial letters that KePro sends to the Agency are reviewed by the Medical Director of KePro, who is not a psychiatrist. Dr. John Sullenberger, the Agency's Medicaid physician, reviews the KePro denial letters sent to the Agency, and 99.9 percent of the time he agrees with the findings of KePro regarding whether the services were medically necessary. Dr. Sullenberger does not review the patient's charts when he does this review. The Agency sends a recoupment letter to the Medicaid provider requesting repayment for services provided. Patient S.G., a 12 year-old boy, was being treated pursuant to the Baker Act. He was admitted to the Hospital on March 8, 1995, and discharged on March 25, 1995. The Agency denied Medicaid reimbursement for the admission and the entire length-of-stay for S.G. based on KePro's determination that it was not medically necessary for the services to S.G. to be rendered in an acute care setting because the patient was neither suicidal nor homicidal. Three to five days prior to his admission to the Hospital, S.G. had attempted to stab his father. He also had further violent episodes, including jumping his father from behind and choking him and pulling knives on his parents. S.G. had a history of attention deficit and hyperactive disorder. He had been using multiple substances, such as alcohol, LSD, cocaine, and marijuana, prior to his admission. His behavior was a clear reference that he was suffering from a psychosis. A psychosis is a significant inability to understand what is reality, including delusions of false beliefs, hallucinations, hearing and seeing things which do not exist, and ways of thinking that are bizarre. Psychosis is a reason to admit a patient, particularly combined with substance abuse. S.G.'s treating psychiatrist noted that S.G. had tangentiality, which means that his thoughts did not stay together. He did not have a connection between thoughts, which is a sign of a psychosis. The chart demonstrated that S.G. had disorder thinking, which includes the possibility of a psychosis. There was also a reference in the charts to organic mental disturbance which could infer brain damage as the cause for the mental disturbance. Two days after admission, there was an issue of possible drug withdrawal because S.G. was agitated and anxious and showed other symptoms. Drug withdrawal, psychosis, and a demonstration of overt violence require a stay in an acute care facility. There was some indication that S.G. was suicidal. While in the Hospital he was placed under close observation, which is a schedule of 15-minute checks to determine if the patient was physically out of harm's way. S.G. was started on an antidepressant, Wellbutrin, because the treating physician thought S.G. was becoming increasingly depressed and was having trouble organizing his thoughts. Antidepressants, as contrasted to a medication such as an antibiotic, may take a minimum of two to three weeks before the patient will benefit from the full effect of the drug. It is difficult to stabilize the dosage for an antidepressant on an outpatient basis. S.G. was taking Ritalin, which is commonly used for children with attention deficit, hyperactivity disorders. During his stay at the Hospital, S.G. was engaging in strange behavior, including absence seizures. On March 16, 1995, he was still lunging and threatening harm. On March 20, 1995, he was still unstable and at risk. The dosage of Wellbutrin was increased. On March 21 and 22, 1995, S.G. was still threatening and confused. S.G. was discharged on March 25, 1995. The admission and length-of-stay for S.G. were medically necessary. Patient J.D. was a 16 year-old boy who was admitted to the Hospital on March 7, 1995, and discharged on March 14, 1995. The Agency denied the admission and entire length-of-stay based on KePro's determination that the patient was not actively suicidal or psychotic and services could have been rendered in a less acute setting. J.D. was admitted from a partial hospitalization program pursuant to the Baker Act because he was observed by a health care professional banging his head against the wall and throwing himself on the floor. He had a history of depression and out-of-control behavior, including being a danger to himself and running away. At the time of his admission, he was taking Prozac. Banging his head against the wall can mean that the patient is psychotic, can cause brain damage, and can be dangerous if the cause of the behavior is unknown. Admission to the Hospital was justified because the patient was extremely agitated and self abusive, requiring restraints and medication to decrease his agitation and self abusiveness. One of the tests administered during his hospital stay indicated that J.D. was a moderate risk for suicidal behavior. During his hospital stay, it was discovered that J.D. had threatened to kill himself while at school. He had been in a partial treatment program during the day, but that environment was not working. There was violence in the home, and J.D. was becoming overtly depressed. During his stay at the Hospital, J.D. was placed on close observation with 15-minute checks. His dosage of Prozac was increased. The admission and length-of-stay for J.D. were medically necessary. R.J., a 10 year-old male, was admitted to the Hospital on January 1, 1995, and discharged on February 9, 1995. The Agency denied Medicaid reimbursement based on a determination by KePro that the treatment in an acute care facility was not medically necessary because R.J. was not psychotic, not suicidal, and not a threat to others; thus treatment could have been provided in an alternate setting. R.J. had been referred by a health care professional at Horizon Center, an outpatient center, because of progressive deterioration over the previous fourteen months despite outpatient treatment. His deterioration included anger with temper outbursts, uncontrollable behavior at school, failing grades, sadness, depressed mood, extreme anxiety, extensive worrying and a fear of his grandmother. R.J. also suffered from encopresis, a bowel incontinence. He was agitated, lacked energy, neglected his hygiene, experienced crying spells, and had difficulty concentrating. R.J. needed to be admitted for an evaluation to rule out a paranoid psychosis. It was necessary to do a 24-hour EEG as opposed to a 45-minute EEG. In order to do a 24-hour EEG, the patient is typically placed in an acute care facility. The EEG showed abnormal discharge in the brain, which could be contributing to a psychiatric illness. At school R.J. had smeared feces on the walls, behavior that could be seen in psychotic persons. There was evidence that he had been hitting and throwing his stepbrother and 3 year-old brother. He was fearful of his grandmother and, based on his family history, there was reason to fear her. R.J. was placed on Buspar, a medication which generally takes two weeks to take effect. Contrary to the Agency's determination, R.J. was disorganized. He was also violent in terms of threatening danger and extreme anger. The admission and length-of-stay for R.J. at the Hospital were medically necessary. Patient C.A., a 9 year-old male, was admitted to the Hospital on June 1, 1995, and discharged on June 12, 1995. The Agency disallowed one day of the length-of-stay based on a determination by KePro that the services provided on June 11, 1995, could have been provided in a less restrictive setting. C.A. was admitted for violent and disruptive behavior. He also had an attention deficit, hyperactivity disorder and was taking Lithium and Depakote. These medications are used for patients who experience serious mood swings and abrupt changes in mood, going from depression to anger to euphoria. To be effective, medicating with Lithium and Depakote requires that the blood levels of the patient be monitored and the dosage titrated according to blood level. C.A. also was given Wellbutrin during his hospital stay. On June 11, 1995, C.A. was given an eight-hour pass to leave the hospital in the care of his mother. The physician's orders indicated that the pass was to determine how well C.A. did in a less restrictive setting. He returned to the Hospital without incident. He was discharged the next day to his mother. The treatment on June 11, 1995, could have been provided in an environment other than an acute facility; thus the stay on June 11, 1995, was not medically necessary for Medicaid reimbursement purposes. Patient G.M., an 11 year-old male with a history of being physically and sexually abused by his parents, was admitted to the Hospital on March 21, 1995, and was discharged on April 3, 1995. The Agency denied Medicaid reimbursement for inpatient hospital treatment from March 28 to April 3, 1995, based on KePro's determination that the length of hospital stay exceeded health care needs at an inpatient level and could have been provided in a less acute setting. At the time of admission, G.M. had suicidal ideation. His school had reported that G.M. had mutilated himself with a pencil, banged himself on the knuckles, and told the school nurse that he wanted to die. Prior to admission, G.M. had been taking Ritalin. His treating physician took G.M. off the Ritalin so that she could assess his condition and start another medication after a base-line period. The doctor prescribed Clonidine for G.M. Clonidine is a drug used in children to control reckless, agressive and angry behavior. Clonidine must be titrated in order to establish the correct dosage for the patient. During his hospital stay, G.M. was yelling and threatening staff. He was placed in locked seclusion, where he began hitting the wall. G.M. was put in a papoose, which is similar to a straitjacket. The papoose is used when there is no other way to control the patient. The patient cannot use his arms or legs while in a papoose. This type of behavior and confinement was occurring as late as March 31, 1995. G.M. was given a pass to go to his grandparents on April 2, 1995. He did well during his pass, and was discharged on April 3, 1995. Treatment in an acute facility was medically necessary through April 1, 1995. Treatment on April 2, 1995, could have been provided in a less acute setting. Patient S.S., a 5 year-old male, was admitted to the Hospital on March 9, 1995, and was discharged on April 3, 1995. The Agency denied Medicaid reimbursement for the admission and entire length of his hospital stay based on a determination by KePro that S.S. was not psychotic or an immediate danger to himself or others and the evaluation and treatment could have been rendered in a less acute setting. Prior to admission to the Hospital, S.S. was threatening suicide, ran into a chalk board at school, scratched his arms until they bled, and showed aggressive intent toward his sister, saying that he would kill her with a saw. S.S.'s condition had been deteriorating for approximately three months before his admission. At the time of admission, he had been suicidal, hyperactive, restless, and experiencing hallucinations. The hallucinations imply a psychosis. S.S. was put on Trofanil, an antidepressant which needs to be titrated. The patient's blood level had to be monitored while taking this drug. During his hospital stay, S.S. was on close observation. All objects which he could use to harm himself were removed from his possession. After he ate his meals, the hospital staff would immediately remove all eating utensils. On March 28, 1995, S.S. threatened to kill himself and became self-abusive. His blood level on March 31, 1995, was sub-therapeutic, and his medication dosage was increased. On April 1, 1995, S.S. had a temper tantrum. The admission and length-of-stay for the treatment of S.S. were medically necessary. Patient M.P., a 10 year-old male, was admitted to the Hospital on April 27, 1995, and was discharged on May 6, 1995. The Agency denied Medicaid reimbursement for the admission and entire length-of-stay based on a determination by KePro that the patient functions on an eighteen to twenty-four month level but is not psychotic and the treatment could have been provided in a less acute setting. M.P.'s IQ is between 44 and 51. He was diagnosed with a pervasive development disorder, which is a serious lack of development attributed to significant brain damage. His condition had deteriorated in the six months prior to his admission. He had episodes of inappropriate laughter, fits of anger, hit his head, hit windows, and put his arm in contact with the broken glass through the window. At the time of his admission, he had a seizure disorder. An EEG and an MRI needed to be performed on M.P. in order to evaluate his condition. M.P. had to have a regular EEG, a 24-hour EEG, and a neurological examination. The patient was aggressive, restless, and uncooperative. In order for the MRI to be performed, M.P. had to be anesthetized. The admission and length-of-stay for M.P. were medically necessary. Patient C.T., a 34 year-old female, was admitted to the Hospital on November 11, 1994, and was discharged on November 26, 1994. The Agency denied the treatment from November 17, 1994, to November 26, 1994, based on a determination by a peer review organization that the patient was stable by November 17, 1994, and psychiatric follow-up could have been performed in an outpatient setting. C.T. was admitted for kidney stones. She did pass the kidney stones but continued to have severe pain. Her doctor asked for a psychiatric consult. The psychiatrist diagnosed C.T. as having a personality disorder, chronic psychogenic pain disorder, and an eating disorder. Her depressive disorder exacerbated pain. C.T. had been given narcotics for the pain associated with the kidney stones. In order to assess her mental status, the physicians needed to taper the dosage of Demerol which she had been receiving. She was started on Sinequan, which is an anti-depressant given to alleviate the psychological condition and to help with the physical complaints. C.T. was later put on Vicodin, an oral narcotic, which seemed to bring the pain under control. The drugs used could cause a drop in blood pressure; therefore, they had to be titrated slowly. Her treating physician was trying to find an appropriate anti-depressant, while weaning the patient from intramuscular narcotics. On November 17, 1994, C.T. left her room and went to the hospital lobby, where she was found by nursing staff. C.T. was crying and saying that she was in pain and wanted to die. During her hospital stay, C.T. was in much distress; she would scream out that she was in pain. On November 18, 1994, she was found crying on the floor of the hospital chapel and had to be returned to her room. It was the opinion of Dr. Bernard Frankel, an expert retained by the Hospital, that C.T. probably could have been discharged a day earlier. The hospital stay for C.T. from November 17, 1994, to November 25, 1994, was medically necessary. The last day of her stay was not medically necessary.
Recommendation Based on the foregoing Findings of Fact and Conclusions of Law, it is RECOMMENDED that a Final Order be entered requiring Indian River Memorial Hospital, Inc., to pay to the Agency $752.14 for one day of service provided to G.M., $752.14 for one day of service provided to C.A., and $473.22 for one day of service provided to C.T. and finding that the Hospital is not liable for payment for any of the other services at issue in this proceeding. DONE AND ENTERED this 2nd day of November, 1998, in Tallahassee, Leon County, Florida. SUSAN B. KIRKLAND 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 Filed with the Clerk of the Division of Administrative Hearings this 2nd day of November, 1998. COPIES FURNISHED: Thomas Falkinburg, Esquire Agency for Health Care Administration Fort Knox Building 3 2727 Mahan Drive, Suite 3431 Tallahassee, Florida 32308 John D. Buchanan, Jr., Esquire Henry, Buchanan, Hudson, Suber & Williams, P.A. 117 South Gadsden Street Tallahassee, Florida 32302 Sam Power, Agency Clerk Agency for Health Care Administration Fort Knox Building 3 2727 Mahan Drive, Suite 3431 Tallahassee, Florida 32308 Paul J. Martin, General Counsel Agency for Health Care Administration Fort Knox Building 3 2727 Mahan Drive, Suite 3431 Tallahassee, Florida 32308
