Delmastro, Blanca Galvez, Elizabeth Dudek, and Daniel Sullivan. Petitioner's exhibits 1A, 1B, 2A, 2B, 3A, 3B and 4 were received into evidence; however, Petitioner's exhibits 1A, 2A, and 3A were admitted in recognition of Intervenor's hearsay objection. Consequently, to the extent that matters contained therein constitute uncorroborated hearsay, they may not form the basis of a finding of fact. Respondent called Elizabeth Dudek as a witness, and its exhibits 2-4 were received into evidence.
The transcript of hearing was filed with the Division of Administrative Hearings on December 11, 1986, and the parties were granted leave until December 29, 1986, to file proposed finding of fact. Those proposed findings have been addressed in the appendix to this Recommended Order.
On October 15, 1985, Petitioner, Mercy Hospital, Inc. (Mercy), filed an application with Respondent, Department of Health and Rehabilitative Services (Department) for a certificate of need (CON) to convert 29 medical/surgical beds into 29 long-term substance abuse beds. On February 27, 1986, the Department denied Mercy's application, and Mercy timely petitioned for formal administrative review.
Mercy is a 538-bed acute care hospital located in Miami, Dade County, Florida. Due to a declining patient census, Mercy is, however, operating only
360 of its 530 licensed beds. Mercy currently offers services in medicine, surgery, psychiatry, obstetrics, gynecology, emergency medical services and outpatient services.
Need
The predicate for the Department's denial of Mercy's application was a perceived lack of need for long-term substance abuse beds in District XI (Dade and Monroe Counties), and the impact such lack of need would exert on the other statutory and rule criteria. Resolution of the need issue is dispositive of Mercy's application.
There currently exists no numeric need methodology for determining the need for long-term substance abuse beds. The Department has, however, adopted Rule 10-5.11(27)(h)1, Florida Administrative Code, which establishes the following occupancy standard:
No additional or new hospital inpatient substance abuse beds shall normally be approved in a Department service district unless the average occupancy rate for all existing hospital based substance abuse impatient beds is at or
exceeds 80 percent for the preceding 12 month period.
District XI has 190 approved long-term hospital impatient substance abuse beds; however, only 30 of those beds are currently licensed. The licensed beds are located in Monroe County at Florida Keys Memorial Hospital (Florida Keys), and are operating well below the 80 percent occupancy standard established by rule. 1/ The remaining beds are to be located in Dade County
where Intervenor, Management Advisory and Research Center, Inc., d/b/a Glenbeigh Hospital (Glenbeigh) holds a CON for a 100-bed unit and Mount Sinai Medical Center (Mount Sinai) holds a CON for a 60-bed unit. Glenbeigh's facility is currently under construction, and Mount Sinai is seeking licensure. While not licensed, Mount Sinai has operated its 60-bed unit under its acute care license, and for the first three quarters of 1985 reported occupancy rates of 49.7 percent, 62 percent, and 48.9 percent.
While the beds approved for District XI do not demonstrate an 80 percent occupancy rate, only one unit, Florida Keys, is licensed and operational. That unit is located in Key West, serves the middle and lower keys, and is not accessible to Dade County residents. The remaining units are not licensed, and their occupancy figures are not representative of a functional substance abuse unit. Accordingly, a failure to demonstrate compliance with the
80 percent occupancy standard is not necessarily dispositive of the question of need.
There currently exists, however, no recognized methodology to calculate need for long-term substance abuse services. Accordingly, to demonstrate a need in 1990 for such services, Mercy relied on a numeric need methodology devised by its health planning expert, Daniel Sullivan 2/ (Petitioner's exhibit 4). Sullivan's methodology was not, however, persuasive.
The First Step in Sullivan's Methodology
The first step in Sullivan's methodology was to derive an estimate of the number of substance abusers in District XI who would seek treatment in an inpatient setting. The figure he calculated (a,170) was derived-through a four- stage refinement process.
Initially, Sullivan estimated the number of problem drinkers within the district for the horizon year by applying the Marden methodology. That methodology, routinely relied upon by health planners, identifies the number of problem drinkers in a given population by multiplying a probability factor to age and sex groupings. By applying the Marden methodology to the age and sex demographics of District XI, Sullivan calculated that an estimated 148,541 problem drinkers would reside within the district in 1998.
Sullivan then strove to estimate the number of problem drinkers who would seek treatment in some formal setting (network treatment). To establish that estimate, Sullivan relied on a report prepared for the National Institute on Alcohol Abuse and Alcoholism (NIAAA) entitled "Current Practices in Alcoholism Treatment Needs Estimation: A State-of-the-Art Report". According to Sullivan, that report estimates the percentage of problem drinkers who will seek network treatment to be 20 percent. Therefore, he calculated that an estimated 29,788 problem drinkers in District XI would seek such treatment in 1990.
Sullivan then strove to estimate the number of problem drinkers who would seek treatment in an inpatient setting. To establish that estimate, Sullivan relied on a survey conducted in 1982 by the NIAAA entitled National Drug and Alcoholism Treatment Utilization Survey". According to Sullivan, that survey indicated that approximately 78 percent of all problem drinkers who sought treatment did 50 on an outpatient basis. Therefore, using a factor of 22 percent, he calculated that an estimated 6,536 problem drinkers in District XI would seek such treatment in 1990.
Sullivan's methodology, at stage two and three of his refinement process, was not persuasive. While Sullivan relied on the factors presented in the reports, there was no proof that health planning experts routinely relied on the reports. More importantly, there was no evidence of the type of survey conducted, the reliability of the percentage factors (i.e.: + 1 percent, 10 percent, 50 percent, etc.), or their statistical validity. In sum, Sullivan's conclusions are not credited.
The final stage at step one of Sullivan's methodology, was to estimate the number of substance abusers (alcohol and drugs) who would seek treatment in an inpatient setting. To derive that estimate, Sullivan relied on a report prepared by the Department's Alcohol, Drug Abuse and Metal Health Office, contained in a draft of its 1987 state plan, which reported that 80 percent of substance abusers abuse alcohol and 20 percent abuse other drugs. Applying the assumption that 80 percent of substance abusers abuse alcohol, Sullivan estimated that 8,170 substance abusers in District XI would seek inpatient treatment in 1990.
Sullivan's conclusion is again not persuasive. To credit Sullivan's methodology, one must assume that substance abusers (alcohol and drugs) seek treatments at the same rate as alcohol abusers. The record is devoid of such proof. Accordingly, for that reason and the reasons appearing in paragraph 12 supra, Sullivan's conclusions are not credited.
The Second Step in Sullivan's Methodology.
The second step in Sullivan's methodology was to estimate the number of hospital admissions, as opposed to other residential facility admissions, that would result from the need for substance abuse services. To quantify this number, Sullivan relied on one 1982 survey conducted by NIAAA. According to that survey, the distribution of inpatient substance abuse clients by treatment setting in 1982 was as follows:
Facility Location | Number | Percent of Total |
Hospital | 17,584 | 34.1 |
Quarterway House | 1,410 | 2.7 |
Halfway Housed/ |
|
|
Recovery Home | 14,648 | 28.4 |
Other Residential |
|
|
Facility | 15,980 | 31.0 |
Correctional Facility | 1,985 | 3.8 |
TOTAL | 51,607 | 100.0 percent |
Therefore, using a factor of 34.1 percent, Sullivan estimated the number of substance abuse hospital admissions to be 2,784 for 1990. For the reasons set forth in paragraph 12 supra, Sullivan's conclusions are, again, not credited. The Third and Fourth Steps in Sullivan's Methodology.
The third step in Sullivan's methodology was to estimate the number of substance abuse hospital admissions that would require long-term, as opposed to short-term, services. To derive this estimate, Sullivan calculated admissions to short-term beds by applying a 28-day length of stay and an 80 percent occupancy standard to the Department's short-term bed need methodology (.06 beds per 1,000 population) contained in Rule 10-5.11(27)(f)1, Florida Administrative Code. Sullivan then subtracted that number (1,182) from the estimated number of
substance abuse hospital admissions for 1990 (2,784), and concluded that the estimated number of hospital admissions in 1990 that would result from the need for long-term substance abuse services would be 1,602.
The final step in Sullivan's methodology was to calculate the need for long-term hospital substance abuse beds. To derive this estimate, Sullivan multiplied the estimated number of long-term substance abuse admissions (1,602) by an average length of stay of 37 days, and divided that total by an occupancy standard of 292 days (80 percent of 365 days). Under Sullivan's methodology, a gross need exists for 203 long-term substance abuse beds in District XI.
To establish net need, Sullivan would reduce the 203 bed district need by the 160 beds approved for Dade County, but ignore the 30-bed unit at Florida Keys because of its geographic inaccessibility. By Sullivan's calculation, a net need exists for 43 beds in Dade County. Sullivan's analysis, at steps three and four of his methodology, is not credited.
Throughout his methodology Sullivan utilized District XI population figures (Dade and Monroe Counties) to develop a bed need for Dade County. Although Monroe County accounts for only 4 percent of the district's population, the inclusion of that population inflated Dade County's bed need. More demonstrative of the lack of reliability in steps three and four of Sullivan's analysis are, however, the methodologies by which he chose to calculate short- term admissions and long-term substance abuse bed need.
Sullivan calculated admissions to short-term beds by applying a 28-day length of stay and an 80 percent occupancy standard to the Department's short- term bed need methodology (.06 beds per 1,000 population) contained in Rule 10- 5.11(27)(f)1, Florida Administrative Code. 3/ By using a 28-day length of stay, the maximum average admission permitted for short-term beds, as opposed to the district's demonstrated average of 24-days, Sullivan inappropriately minimized the number of estimated short-term admissions and maximized the number of estimated long-term admissions. 4/
Sullivan sought to justify his use of a 28-day standard by reference to testimony he overheard in a separate proceeding. According to Sullivan, a Department representative testified that the 28-day standard was used in developing the Department's .06 short-term beds per 1,000 population rule. Sullivan's rationalization is not, however, persuasive. First, Sullivan's recitation of testimony he overheard in a separate proceeding was not competent proof of the truth of those matters in this case. Second, Sullivan offered no rational explanation of how a 28-day standard was used in developing the rule. Finally, the proof demonstrated that the average short-term length of stay in District XI is 24 days, not 28 days.
The difference between a 24-day and 28-day average short-term length of stay is dramatic. Application of Sullivan's methodology to the population of Dade County, and utilizing a 24-day average, would demonstrate a need for 170 long-term beds, as opposed to Sullivan's calculated need of 203 beds.
In addition to the average short-term length of stay factor, long-term bed need is also dependent on an average length of long-term admissions factor. Under Sullivan's approach, the higher the average, the higher the bed need. Accordingly, to derive a meaningful bed need requires that a reliable average length of stay be established. The data chosen by Sullivan to calculate such an average was not, however, reliable.
Sullivan used a 37-day average length of stay to develop his long-term bed need. This average was developed from the CON applications of Mercy, Glenbeigh and Mount Sinai. In the applications, Mercy estimated an expected length of stay of 30-37 days, Glenbeigh 36-38 days, and Mount Sinai 28-49 days.
Use of a simple average, of the expected lengths of stay contained in Mercy's, Glenbeigh's and Mount Sinai's applications, to develop an average long- term length of stay is not persuasive. The figures contained in the applications are "expected length of stay", a minimum/maximum figure. Mercy failed to demonstrate that a simple average of those figures was a reliable indicator of average length of stay. Indeed, Mercy presented evidence at hearing that its average length of stay would be 30-31 days; a figure that is clearly not a simple average of the 30-37 day expected length of stay contained in its application. Mercy's failure to demonstrate a meaningful average length of stay renders its calculated bed need unreliable.
Sullivan's Methodology - An Overview
Each step of Sullivan's methodology was inextricably linked to the other. Consequently, a failure of any step in his analysis would invalidate his ultimata conclusion. Notwithstanding this fundamental fact Mercy, with the exception of the Marden methodology, failed to present a reasonable evidentiary basis to demonstrate the reliability and validity of Sullivan's methodology or any of its parts. Since his methodology was not validated, or each of its inextricably linked parts validated, Sullivan's conclusions are not persuasive or credited.
Other Considerations
If Mercy receives a CON, it will enter into a management contract with Comprehensive Care Corporation (CompCare) to operate the substance abuse unit. The parties anticipate that Mercy will provide its existing physical plant, custodial services, support services, dietary services, complimentary medical services, medical records and pharmacy services, and that CompCare will provide the treatment team, quality assurance, public information, promotion and operational management. Under its proposed agreement with CompCare, Mercy would pay CompCare on a per patient day basis. This fee was not, however, disclosed at hearing nor were the other expenses for patient care established. 5/ Consequently, Mercy failed to establish that its proposal was financially feasible on either a short or long term basis.
Mercy also proposes to provide bilingual staff, and dedicate a portion of its patient days to indigent and Medicaid patients. There was no competent proof to establish, however, that such needs were not met, or would not be met, by the existing facilities.
CONCLUSIONS OF LAW
The Division of Administrative Hearings has jurisdiction over the parties to, and the subject matter of, these proceedings.
Pertinent to this proceeding, Section 38l.494(6)(c), Florida Statutes, and Rule 10-5.11(27), Florida Administrative Code, establish the criteria which must be considered in evaluating applications for a certificate of need. See: Balsam v. Department of Health and Rehabilitative Services, 486 So.2d 1341 (Fla. 1st DCA 1986) and Department of Health and Rehabilitative Services v. Johnson and Johnson Home Health Care, Inc., 447 So.2d 361 (Fla. 1st DCA 1984). The
weight to be accorded each criteria and the consequent balancing of the criteria will vary, however, depending on the facts and circumstances of each case.
Collier Medical Center, Inc., v. Department of Health and Rehabilitative Services, 462 So.2d 83 (Fla. 1st DCA 1985). See: Graham v. Estuary Properties, Inc., 399 So.2d 1374 (Fla. 1981).
In the instant case, need as contemplated by Section 381.494(6)(c)1, 2, 9, and 12, Florida Statutes, as well as Rule 10-5.11(27), Florida Administrative Code, are the key criteria in evaluating the proposal of Mercy. Mercy's failure to satisfy those criteria is dispositive of its request for licensure, and such failure is not outweighed by any other, or combination of any other, criteria.
RECOMMENDATION
Based on the foregoing Findings of Fact and Conclusions of Law, it is RECOMMENDED:
That the application of Mercy for a certificate of need to convert 29 medical/surgical beds to 29 long-term substance abuse beds be DENIED.
DONE AND ORDERED this 4th day of February, 1987, in Tallahassee, Florida.
WILLIAM J. KENDRICK
Hearing Officer
Division of Administrative Hearings The Oakland Building
2009 Apalachee Parkway
Tallahassee, Florida 32301
(904) 488-9675
Filed with the Clerk of the Division of Administrative Hearings this 4th day of February, 1987.
ENDNOTES
1/ The occupancy rate for Florida Keys was 24 percent in 1983, 18.2 percent in 1984, 20 percent in 1985, and 16.1 percent for the first two quarters of 1986.
2/ There is no rule mandated planning horizon for long-term substance abuse beds. The Department does, however, apply a 5- year planning horizon for short- term substance abuse beds and acute care beds. Consequently, use of a five year planning horizon from date of application is reasonable. Mercy also calculated need for 1991, but it had no appreciable effect on the numeric results reached.
3/ .06 x 1,889 x 365 x .80
4/ Rule 10-5.11(27), Florida Administrative Code, defines short-term and long- term hospital inpatient substance abuse services as:
Short term hospital substance abuse services are defined as short-term services not exceeding
an average length of stay of 28 days.
Long term hospital substance abuse services are defined as services exceeding an average length of stay of 28 days. (Emphasis added).
