95-003635BID Blue Cross And Blue Shield Of Florida, Inc. vs. Agency For Health Care Administration Status: Closed Recommended Order on Wednesday, September 27, 1995.

Summary: Bid award not illegal, etc. and agency's preparation of Request For Proposal, interpretation, weighing, and scoring not arbitrary. Statistical bias does not prove prejudice.

STATE OF FLORIDA

DIVISION OF ADMINISTRATIVE HEARINGS

BLUE CROSS AND BLUE SHIELD OF )

FLORIDA, INC., )

)

Petitioner, )

)

vs. ) CASE NO. 95-3635BID

)

AGENCY FOR HEALTH CARE )

ADMINISTRATION, )

)

Respondent, )

)

UNISYS CORPORATION, )

)

Intervenor. )

_________________________________)

RECOMMENDED ORDER

Pursuant to notice, the Division of Administrative Hearings, by its duly

designated Hearing Officer, Mary Clark, held a formal hearing in the above-

styled case on August 3-4 and 7-9, 1995, in Tallahassee, Florida.

APPEARANCES

For Petitioner: Michael J. Glazer, Esquire

Stephen C. Emmanuel, Esquire

Steven P. Seymoe, Esquire

MACFARLANE, AUSLEY, FERGUSON & MCMULLEN

227 South Calhoun Street

Tallahassee, Florida 32302

For Respondent: Paul Martin, Esquire

Office of the Attorney General

2020 Capital Circle, Southeast

Ervin Building, Suite 102

Tallahassee, Florida 32301

and

James H. Peterson, Esquire

Steven A. Grigas, Esquire

Agency for Health Care Administration

Fort Knox Executive Center

2728 Mahan Drive, Room 3403A

Tallahassee, Florida 32308

For Intervenor: M. Stephen Turner, Esquire

David K. Miller, Esquire

BROAD & CASSEL

First Florida Bank Tower, Suite 400

215 South Monroe Street

Tallahassee, Florida 32301

STATEMENT OF THE ISSUES

Petitioner, Blue Cross and Blue Shield of Florida, Inc. (BCBS), has

challenged the Agency for Health Care Administration's (AHCA) proposed award of

contract pursuant to Request for Proposals No. SHP 95-002 to Unisys Corporation

(Unisys). The ultimate issue in this proceeding is whether that proposed award

is fraudulent, dishonest, arbitrary or illegal. In their pleadings and

presentations the parties have framed these subsidiary issues:

1. Whether the Unisys bid was responsive to and met the mandatory

requirements of the RFP;

2. Whether the allocation of scoring weights was arbitrary and capricious

and likely to result in the state's expenditure of excessive funds for health

care in favor of much smaller savings in administrative costs;

3. Whether the scores assigned by the proposals' evaluators were

unreliable and biased;

4. Whether the evaluation of proposals illegally failed to apply "present-

value methodology" required by section 287.0572, Florida Statutes; and

5. Whether the award is illegal because AHCA is unconstitutionally

structured in violation of Article IV, Section 6, Florida Constitution.

Although all parties concede that determination of the constitutional issue

is beyond the jurisdiction of the hearing officer, BCBS claims that the issue is

"preserved" for judicial determination and AHCA and Unisys argue the issue has

been waived.

PRELIMINARY STATEMENT

On July 19, 1995, this case was referred to the Division of Administrative

Hearings after BCBS filed its formal written protest and petition for formal

administrative hearing. The hearing was scheduled within the deadlines provided

in section 120.53(3), Florida Statutes. Preliminary discovery matters were

disposed of in a telephone hearing on July 24, 1995. Other issues were argued

in a prehearing conference on August 2, 1995 and disposition of those issues was

deferred until the formal evidentiary hearing. Unisys' motion to intervene was

granted.

At the formal hearing BCBS presented the following witnesses: Rick Lutz,

Kate Morgan, V. Sheffield (Chip) Kenyon, Robert W. Nay, Suzanne Gelber Rinaldo,

Ph.D., Judy Hefren, Elton Scott, Ph.D. and James T. McClave, Ph.D. BCBS

exhibits Number 1-6, 8-22 and 24-28 were received in evidence. Exhibit Number 7

was rejected, and ruling on admissibility of exhibit Number 23 was reserved.

Exhibit Number 23 is now rejected as irrelevant, as addressed in the conclusions

of law, below.

AHCA presented the testimony of Rick Lutz and Mark Johnson, Ph.D. AHCA's

exhibits Number 1-8 were received into evidence.

Unisys presented the following witnesses: Doreen Corwin, Carol Lockwood,

and Erwin Bodo, Ph.D. Unisys exhibits Number 1-15 and 17-19 were admitted.

Exhibit Number 16 was marked for identification and rejected.

BCBS motion filed on September 11, 1995, to strike portions of intervenor's

memorandum of law and portions of intervenor's appendix of unpublished

authorities is DENIED. The arguments in the motion and in Unisys' response have

been considered in weighing the authorities' relevance in this proceeding.

The hearing transcript was filed and the parties provided thorough proposed

recommended orders and memoranda. These have been considered and the specific

findings of fact proposed by each have been incorporated or rejected as

described in the attached appendix.

FINDINGS OF FACT

The Parties

1. The Agency for Health Care Administration (AHCA or agency), as provided

in section 110.123(3)(b), Florida Statutes, is responsible for all aspects of

the purchase of health care for state employees under the state group health

insurance plan and the health maintenance organization plans. The

responsibilities include the development of requests for proposals for state

employee health services, the determination of benefits to be provided and

negotiation of contracts for health care and health care administrative

services.

2. Blue Cross and Blue Shield of Florida, Inc. (BCBS) is a large managed

care company providing a wide range of health care services including a full

array of insured products, traditional indemnity products and preferred provider

organization (PPO) products to large and small groups. It has provided health

services to state employees since the 1940's. In 1978, when the state switched

from offering a fully insured product to a self-insured product, BCBS became the

administrator of the self-insured program and has remained the administrator

since that time.

3. Unisys Corporation (Unisys) is a publicly held corporation,

incorporated in the State of Delaware. It has been actively engaged in the

health information services and technology market since 1976, and its health

information management group, headquartered in Reston, Virginia, has experience

in all facets of health care claims processing.

The Request for Proposals (RFP)

4. The state group health insurance plan (plan) is self-insured, which

means that payment for services rendered by health care providers to covered

recipients is paid by the state from a trust fund established for this purpose.

The plan currently covers approximately 235,000 persons, including employees and

their dependents and retired persons. BCBS' contract to administer the plan

expires on December 31, 1995.

5. In July, 1994 AHCA began to develop an RFP for the new contract to run

for four years, commencing January 1, 1996, with four one-year extensions, at

the state's option. Previous RFP's for state employees' health services had

been prepared by the Department of Management Services (DMS) or its predecessor

agency, and this is the first time AHCA has had the responsibility.

6. The principal authors of the RFP were Rick Lutz, AHCA's Director of the

Division of State Health Purchasing, and Kate Morgan, Chief of the Bureau of

State Employee Health Insurance. Ms. Morgan reports directly to Mr. Lutz. Both

individuals had prior experience in the state's Medicaid program and both had

supervised or directly developed RFP's for health services-related procurements.

7. The RFP requested interested offerors to submit proposals to provide

services in one or more of three categories of services: 1) third party

administrative services (TPA); 2) use of a preferred provider organization

network (PPO); and 3) utilization review and case management services (UR). A

single offeror could submit a proposal in one, two or all three categories. A

single offeror could also submit a proposal to provide services in all three

categories, but utilizing subcontractors.

8. The RFP was divided into sections as follows:

(a) Section 10 - introductory information.

(b) Section 20 - description of the RFP process.

(c) Section 30 - contract terms and conditions.

(d) Section 40 - State's obligations.

(e) Section 50 - specifications for third party

administration (TPA) services, including claims

examination and payment, participant relations, and

coordination of benefits.

(f) Section 60 - questions and requests for

information on the offeror's ability to perform TPA

services.

(g) Section 70 - specifications for preferred

provider organization (PPO) services, including

recruiting and maintaining a network of qualified

providers to perform health care services under

pre-negotiated fee schedules.

(h) Section 80 - questions eliciting information

on the offeror's ability to perform the PPO services.

(i) Section 90 - specifications for utilization

review and care management (UR) services to determine

the medical efficacy and necessity of requested

services as a cost-saving and quality enhancing

measure.

(j) Section 100 - questions eliciting information on

the offeror's ability to perform the UR services.

(k) Section 110 - the cost proposal, or the amount

the offeror would charge to provide the TPA, PPO

and UR services.

(l) Section 120 - evaluation procedure.

RFP Sections 50, 70 and 90 contained the specifications; Sections 60, 80 and 100

contained the scoring elements describing the offeror's capability and prospects

for performance.

9. The RFP sought administrative services only. It did not solicit

offerors to provide direct medical services to participants, and the amounts to

be paid to health care providers for medical services to participants were not

determined or covered by the contract.

10. The RFP directed offerors to submit their proposals in two parts. In

the technical part, the offeror certified that it would comply with the

specifications and responded to the questions to be scored. The cost part

contained the offeror's price to perform the TPA, PPO and UR services for the

contract term, calculated at present value according to a provided formula.

11. Before proposals were submitted, potential offerors were informed that

the cost proposal was assigned 4000 points, and the technical proposal was

assigned 6000 points, consisting of 2400 points for TPA services, 2400 points

for PPO network services, and 1200 points for UR services. Potential offerors

also knew the individual scoring questions relating to TPA, PPO and UR services,

but did not know the preassigned internal weights of these individual questions.

These weights were ascribed in advance by the RFP administrators, Mr. Lutz and

Ms. Morgan, but were sealed and locked away in order to assure that both

offerors and scorers would deal diligently with every question and would not

concentrate on heavily weighted questions.

12. The RFP was issued on March 3, 1995. The RFP specifically provided

that potential offerors could protest the contents of the RFP itself. On March

16, 1995, BCBS filed a protest challenging numerous provisions of the RFP. This

protest was resolved by a settlement agreement on March 31, 1995, in which the

agency modified some provisions and BCBS abandoned all other issues that were

raised or might have been raised in the protest.

13. The RFP provided potential offerors an extended opportunity to pose

questions to clarify the specifications and evaluation criteria. BCBS posed

numerous questions, including questions concerning how the agency would weight

and score criteria concerning PPO networks. AHCA responded in a general manner

without disclosing the weights that would be assigned to various questions.

Other potential offerors also posed questions. All responses by AHCA were

incorporated as addenda to the RFP.

14. Four integrated proposals and two component proposals (less than all

three categories) were submitted. Only integrated proposals were evaluated

because the component proposals , considered together, failed to comprise a

complete package of all three services.

The Proposals and Their Scoring

15. The four proposals were by Health Plan Services (later disqualified

after the cost proposals were opened), by Humana, by Unisys and by BCBS. The

technical proposals were opened on May 18, 1995.

16. The BCBS proposal offered to provide all three components, TPA, PPO

and UR. The Unisys proposal described Unisys as the prime contractor and TPA,

with Beech Street, a separate company, providing the PPO component and Cost

Care, another company, providing the UR component.

17. Rick Lutz selected 24 staff personnel to score the technical merits of

the responses to Sections 60, 80, and 100. They were selected based on their

experience in areas involving finance and accounting, management information,

claims processing, customer relations, reporting, network development, and

utilization review. Half of the scorers were from offices supervised by Mr.

Lutz, and half were from other offices within AHCA. In his twenty-five years of

experience Mr. Lutz was familiar with the backgrounds and abilities of the

individuals. He attempted to recruit other scorers from the Department of

Management Services (DMS) but was informed that DMS' workload precluded such

participation.

18. The scorers were assigned questions to score in the same area as their

functional backgrounds. They were also given a one-half day training session

and a workbook containing specific guidance on factors to consider in scoring

each assigned question. The scorers were directed to consider all relevant

information contained in the proposal in scoring each question. They were

allowed to ask written questions concerning the scoring, and written answers

were provided. The RFP schedule provided one month scoring the proposals.

19. The scorers were to score each assigned question on a 0-10 scale and

to record (in the workbooks) their reasons for each score given. Three scorers

were assigned to score each question; however, each scorer was instructed to

form an independent judgement as to the appropriate score and to not discuss the

score given with anyone else.

20. The instructions provided for a debriefing session in which scorers

whose scores were more than three points apart on a particular question could

confer to ensure that each scorer had considered all information relevant to

that question. Scorers were allowed to change their scores on the basis of

information that they had previously overlooked or they were allowed to leave

them unchanged.

21. The three raw scores for each question were averaged, and the averaged

score was multiplied by the predetermined weight to produce a raw score for each

question. Scores were then added and adjusted to the 6000 point scale.

22. Unisys received the best raw scores for the TPA and the UR components.

BCBS received the best raw scores for the PPO component. However, BCBS'

advantage in the PPO component was sufficient to place it ahead in the aggregate

raw score for the technical proposal, so it was awarded the maximum raw score of

6000 points.

23. On June 22, 1995, the agency opened and scored the cost proposals.

Humana had the lowest cost proposal and was awarded the maximum cost score of

4000 points. The Unisys cost proposal was $86,618,919 present value, and

received a prorated score of 3,458.65 points. The BCBS cost proposal was the

highest, $102,200,263.22 present value, and received 2,931.35 points, the lowest

prorated cost score. Scoring the cost proposals was a mechanical, non-

subjective function.

24. Unisys had the highest combined score for the technical and cost

proposals under the framework described in the RFP, as summarized in the

following chart:

==============================================================

ELEMENT UNISYS RAW BCBS RAW UNISYS BCBS ADJUSTED

SCORE SCORE ADJUSTED SCORE SCORE

TPA 1624.97 1602.33

PPO 1274.63 1515.83 5744.97 6000.00

UR 751.28 694.78

Cost 3458.65 2931.35

Total 9203.62 8931.35

==============================================================

25. The agency made a minor math error in its original calculation of the

BCBS score for the technical proposal. When the BCBS score is corrected by

adding 16 points, the effect narrows the gap, but does not materially affect the

result.

26. Based on the results of the overall scoring, Mr. Lutz prepared a brief

report summarizing the evaluation process and sent it to a steering committee

comprised of four senior managerial level employees: Ms. Morgan, Tom Wallace,

the agency's second-in-command; Dr. James Howell, and Mildred Seay of DMS.

27. The committee met with Mr. Lutz on June 26 for approximately two to

three hours. There was general discussion regarding medical costs under a plan

offered by BCBS, as opposed to Unisys' plan. The pricing analysis found in RFP

section 80.10 was explained and discussed; and the committee discussed whether

Unisys, through its subcontractor, Beech Street, could expand the provider

network (PPO) to achieve utilization and prices comparable to those reported by

BCBS. The committee unanimously approved the scorers' ranking and recommended

the contract award to Unisys.

28. The recommendation was forwarded to Douglas Cook, Executive Director

of AHCA and William Lindner, Secretary of DMS, in a brief memorandum. Notice of

intent to award the contract to Unisys was posted by the agency on June 27,

1995.

Responsiveness of Unisys Proposal

29. In creating a mandatory requirement checklist in Section 120.2 of the

RFP, Mr. Lutz sought a simplified process that would assure that proposals were

evaluated on their merits. The agency desired an open competition process that

would score the proposals on the adequacy of the responses, rather than a

process that would eliminate proposals from the evaluation.

30. Mr. Lutz chose two AHCA employees who were not members of the

evaluation team to check the proposals against the checklist and to verify

whether the proposals contained a tabbed section corresponding to that item on

the list. Mr. Lutz anticipated that if the response were wholly deficient, the

evaluators would ascribe a zero score. None of the four proposals was rejected

in this stage of the process.

A. Subcontractors' Certificates of Compliance and

Public Entity Crime Forms Were Not Required

31. RFP Section 30.4 requires a Certificate of Compliance "from each

offeror regardless of whether the offeror submits an integrated proposal or a

component proposal." Section 30.42 requires a Public Entity Crime form to be

submitted by "a[n] offeror submitting a proposal." Section 120.2 contains a

checklist of requirements including the following:

d. Did the offeror submit a signed

certificate of compliance?

* * *

f. Is a completed Public Entity Crimes

Statement included?

(Joint Exhibit Number 1)

These requirements are expressly directed to the "offeror," and do not refer to

subcontractors.

32. Unisys submitted an "integrated" proposal in which it was the sole

offeror and prime contractor responsible for providing all services called for

under the RFP. A Unisys representative signed the Certificate of Compliance and

Public Entity Crime Form, which BCBS concedes was sufficient as to Unisys. The

Unisys proposal specified that it would engage two subcontractors, Beech Street

for the PPO component and Cost Care for the UR component. AHCA did not intend

or expect subcontractors to submit the Certificate of Compliance and Public

Entity Crime form.

33. RFP Section 30.19 reserved the agency's right to approve

subcontractors, while confirming that the prime contractor is responsible for

all contract performance.

34. The purpose of the Certificate of Compliance is to provide assurance

similar to that in PUR 7033 that the offeror is bound to the specifications of

the RFP. PUR 7033 is a form at the front of the RFP, a contractual services

acknowledgment form required only from the "offeror" or prime contractor. The

Certificate of Compliance expressly contemplates that subcontractors are

included in the prime contractor's commitment.

35. The Public Entity Crime form sought assurance that the offeror or "its

officers, directors, executives, partners, shareholders, employees, members or

agents who are active in the management of the entity" (emphasis added) were not

disqualified to contract as a result of a conviction of certain procurement

crimes. The form also sought assurances that "affiliates" of the offeror

entity, meaning its predecessor or successor, or an entity controlled by a

natural person who is not active in the management of the offeror entity, were

not disqualified. BCBS admits that Unisys was qualified and its form was

sufficient as to Unisys itself. (Transcript, p. 769-70)

36. The form does not solicit any information with regard to

subcontractors. None of the subcontractors identified in the Unisys proposal is

active in the management of Unisys or is an affiliate of Unisys. Neither Unisys

nor any of its subcontractors was on the published convicted vendors list

established by section 287.133, Florida Statutes. There is no evidence to

suggest that they are disqualified to contract.

37. At the time the RFP was issued, the controlling statute required

contractors to sign this form only at the time the contract is executed. Section

287.133(3)(a), Florida Statutes (Supp. 1994). Submitting the form with the

proposal was not essential to protect the state's interests, but was a

convenience. While the proposals were under review, this statutory provision

was repealed to eliminate use of this form entirely. Chapter 95-196, Section 33,

Laws of Florida, effective June 8, 1995.

38. This issue arose, in part, out of confusion related to RFP Section

20.14, which described a situation in which two or more offerors combined as a

partnership, and directed that such a partnership designate one partner to act

as the "prime contractor"; in effect, treating that situation like the Unisys

proposal, which involved a prime contractor and subcontractors.

39. In responding to offerors' questions, AHCA initially directed that

each partner in a partnered proposal would be required to submit the forms as

multiple contractors. However, in Addendum 4 of the RFP, the agency later

clarified that only one prime contractor in each proposal was responsible for

contractual issues:

This is to notify all potential offerors of

a correction to an answer that was provided in

Addendum Number 2. Specifically, the answer to

Cost Care's first question is deleted...

When an integrated proposal is submitted in

response to this RFP, one of the partners in

the bid shall be designated in the proposal as

the prime contractor. The other partners in

the integrated proposal shall be subcontractors

and any contract that may result with the state

shall be between the state and the prime

contractor. The state shall hold the prime

contractor responsible for all contractual issues...

(Joint Exhibit Number 1)

40. AHCA intended Addendum 4 to mean that only the prime contractor was

required to submit the forms in question and did not consider the absence of

separate forms for subcontractors to be a defect.

B. Beech Street's Financial Statements

41. Each technical scoring component of the RFP requested the offeror to

furnish two years' audited financial statements. AHCA did not intend this

request to create a precondition for evaluating the proposal, and did not

include these statements as part of the mandatory specifications in Sections 50,

70 and 90. Rather, the agency designed the RFP to treat the presentation of

audited financial statements as a technical scoring issue.

42. In response to the request for financial statements in Question

80.2.g, Unisys presented a narrative statement explaining that Beech Street was

a privately held corporation that kept its financial statements confidential;

but that Beech Street's auditors, Arthur Anderson & Co., had issued unqualified

"clean options" for the two preceding fiscal years, and that the operations had

been profitable in each year, resulting in year end cash reserves exceeding $4.2

million and $5.2 million, respectively. The statement further advised that

current year operations indicated even greater revenue and profit growth. The

proposal also showed Beech Street's longevity and client base and retention

rate, consistent with a financially stable operation.

43. Unisys provided full information available on its subcontractors,

Beech Street and Cost Care. It was not requested to provide any information

concerning Beech Street's subcontractors (who were sub-subcontractors of

Unisys).

44. Judy Hefren, one of the three scorers who graded Question 80.2.g, is a

CPA with several years' accounting experience. Although she reviewed enough

financial information to satisfy herself concerning Beech Street's ability to

perform as a subcontractor, Ms. Hefren strictly interpreted Question 80.2.g and

gave Unisys a zero for that question.

45. Audited financial statement were not mandatory because other

information could show capacity to perform. AHCA looks to the prime contractor

to cover any deficiency in its subcontractors, and required the prime contractor

to post a substantial performance and payment bond. The prime contractor's and

surety's financial stability assures continuing performance of all obligations.

46. The PPO subcontractor is never in possession of any state funds, but

simply is paid an access or rental fee for the term in which its network is

used. BCBS presented no evidence that significant adverse consequences to the

state would ensue from a hypothetical subcontractor bankruptcy, and Mr. Lutz's

and Ms. Hefren's characterization of such a hypothetical event as an

"inconvenience" is accepted.

C. Maternity Counseling Material Not Required

47. RFP Section 90.5.a stated that "The contractor shall provide

educational materials to all pregnant plan participants to include information

about the program, basic prenatal care and reference to specialty physicians and

facilities." The agency considered this specification to be part of the UR

services that the offeror certified it will perform.

48. RFP Question 100.7.b solicits information on how the offeror plans to

meet this specification, including samples of educational materials to be

furnished. Unisys responded that "Cost Care emphasizes direct communication

with both mother and physician, in addition to the educational materials we

provide." The response described direct contacts with the mother and physician.

It offered to produce additional materials for plan participants generally for

additional cost.

49. The RFP treated this as a scoring issue. Although the scorers gave

Unisys relatively low scores for this response (4, 4 and 3), AHCA was satisfied

that there was nothing wrong with this response and that specification 90.5.a

would be met.

Whether the Agency's Allocation of Weights Among the

Questions in the RFP Was Arbitrary and Illegal

50. RFP Question 80.10 required offerors to perform two historical

"pricing analyses" based on data from the period July 1, 1993 to June 30, 1994,

or 1.5 to 2.5 years before the new contract was to commence.

51. Part (a) of Question 80.10 required offerors to price physician costs

for designated medical procedures in each of 19 counties. Offerors were

permitted to report the price available from any physician with whom they had a

negotiated fee schedule in that county; if the offeror had no negotiated fee

with a physician serving that county, then it had to report a state average

charge for that procedure. BCBS reported the lowest aggregate price for

physician services, and was awarded the maximum score of 10 points for Question

80.10(a). Unisys was awarded 9.92 points, reflecting less than one percent

difference in the aggregate reported prices for physician services.

52. Part (b) of Question 80.10 required offerors to price 1,174 claims in

55 specified hospitals, as of specified dates in 1993-94. The hospitals

selected were those that had provided the most services to state employees in

fiscal year 1993-94. If an offeror had a contract with a specified hospital on

the specified transaction date, then that offeror could report its negotiated

fee with that hospital. If the offeror had no contract with that hospital on

that date, then it had to report that hospital's full reported charge for

services, even if a contract was subsequently negotiated. The question did not

allow equivalent hospitals to be substituted.

53. The question favored the incumbent. BCBS was able to report low

prices for the hospitals chosen because almost all of these hospitals were

already in its network in 1993-94. BCBS received the maximum score of 10 points

on Question 80.10(b); Unisys received 3.01 points.

54. Question 80.10 served a limited purpose to help illustrate previous

network development. It was never intended to become a basis for measuring or

comparing future medical costs per employee or medical cost savings to be

realized from selection of a particular offeror, nor would it be accurate for

this purpose.

55. Provider networks are "dynamic," changing over time in response to

evolving client needs. A PPO administrator cannot effectively recruit providers

or achieve favorable prices until it establishes a market share in the

provider's market area. It was intended that during the six-month transition

period the successful offeror would use the increased market share resulting

from the contract award to expand and tailor its network to serve state

employees.

56. Mr. Lutz commented on the agency's reasons for assigning limited

weight to Question 80.10 as follows:

We certainly did assume that other proposers

would be able to come in, develop a network,

and in the process achieve discounts that would

have been greater than the discounts that they

might have had a year ago.

If we didn't believe that -- there is no sense

in going through a competitive procurement to

start with. If we wanted to start with the

premise that the only entity that could

establish a network and achieve discounts was

the one that we had, then why bother? It seems

to me that the conclusion is we don't want a

competitive procurement, we simply want to issue

a new contract.

(Transcript, p. 133)

57. All parties agreed that it is "very hard" or "impossible" to predict

future network growth and its effect on health care prices. There is no

specific formula available to compute the amount of future medical costs. RFP

Question 80.9 asked offerors to predict percentage changes in health care prices

over the eight year potential contract term, and to provide assurances that the

prediction would be accurate. Unisys predicted a percentage change for the

first four years; BCBS referenced various indices for the first four years.

Neither Unisys nor BCBS predicted anything beyond four years or guaranteed its

prediction by sharing substantial risk if health care prices were to exceed

their predicted levels. These responses help confirm that future health care

prices are volatile and unpredictable.

58. Because the network development and other factors affecting the future

cost of medical care are not easily quantified or predicted, the great majority

of RFP questions concerned evaluations of the offeror's experience and expertise

in developing and managing networks, its specific plans to implement the network

contemplated by the RFP, and its provider credentialling, quality assurance and

payment methodology, as well as performance of TPA and UR functions. All of

these questions concern the offeror's capability to provide a satisfactory

network and reflect its ability to control future medical costs. AHCA intended

that the questions in Section 80 would collectively demonstrate the offeror's

capability and prospects for developing a cost-effective PPO network.

59. BCBS, through its State Business Director and expert witness,

Sheffield Kenyon, asserted that the agency should have increased the weight

assigned to Question 80.10 from 240 points (10 percent of the PPO component

weight) to 1000 points. Mr. Kenyon viewed the historical price analysis as the

"single best proxy" for a future health care price prediction, and was surprised

that the agency had not given it greater weight. His opinion was not based on

any mathematical formula; nor did he identify any industry standard concerning

the weight to be given such historical analysis. His opinion, competent though

it was, was based on exactly what Mr. Lutz and Ms. Morgan brought to the process

of ascribing weights: a rich, full, varied background and years of experience.

60. BCBS State Employee Market Director, Robert Nay, prepared medical cost

projections which purported to show that the Unisys proposal could result in

significantly larger expenditures by the state and its covered persons for

medical care than would be the case with BCBS. He acknowledged that preparing

projections was not a part of his normal work. His analysis was limited to two

factors, network utilization rate and reported discount rate. Mr. Nay compared

a projected savings for BCBS with three projected scenarios for Unisys/Beech

Street.

61. The first scenario assumes the network available to Unisys and Beech

Street will remain static from 1993-94. However, the testimony was unrefuted

that network development is driven by the client base. It is unrealistic to

assume that there has been, and would be, no development prior to contract

implementation in January 1996. Even Mr. Nay agreed this was not likely to

occur. (Transcript, p. 417). Scenario 2 also assumes that Unisys would be

unable to achieve a network utilization rate in Florida comparable to BCBS', and

is likewise speculative and unsupported by the weight of the evidence.

62. Scenarios 1 and 2 used Beech Street's 1994 national average discount

rate as stated in Question 80.2 of the Unisys proposal, and scenario 3 assumed a

slightly improved discount rate. However, there was no evidence to show that

the 1994 national average discount rate would be applicable to the proposed

Florida contract. The Unisys proposal in Section 80.2 reported that Beech

Street's national savings averages may be understated, as most of its network

hospitals are nonprofit hospitals which generally charge less than for-profit

facilities.

63. In Section 80.9, Unisys and BCBS provided information showing rate

changes the state should expect to experience. Unisys reported an actual 9

percent decrease in inpatient hospital rates, and a 13 percent decrease in

outpatient rates for 1995. The Unisys proposal also projected that the state

should experience a 4 percent decline in inpatient and outpatient hospital rates

for 1996 and 2 percent or greater decline in those rates for 1997, 1998 and

1999. BCBS projected increases in these rates for the years 1996-99. These

projections were not included in Mr. Nay's analysis.

64. Beech Street representatives, Doreen Corwin and Carol Lockwood,

described successful efforts in adding provider groups to the Beech Street

network. Beech Street has been received favorably in negotiations with

providers. The final award of the contract should enable Beech Street to

finalize its relationships with Unisys and with sub-subcontractors and

providers.

65. The current Beech Street Florida network includes approximately 1.1

million covered lives. The addition of the state plan's approximately 240,000

covered lives will significantly add to Beech Street's bargaining power to

negotiate prices in markets where participants live.

66. Although there is conflicting evidence of whether providers are more

or less anxious now than in the past to negotiate discounted agreements with a

PPO, it is reasonable to expect that most providers who currently have contracts

with BCBS would be very likely to enter into similar arrangements with Beech

Street to avoid losing patients. The plan encourages covered employees to

utilize the less expensive network providers, so loss of network status would be

detrimental to a provider who relies on that employee patient base.

67. Utilization review services can substantially affect cost of health

care. Cost Care representative, Sandra O'Toole, described its independent

utilization review services for state governments in Mississippi, Alabama, and

Georgia, as well as for other clients around the country, based on a clinical

model using board-certified physicians to review cases. The Cost Care average

number of inpatient admissions per thousand plan participants is approximately

60. BCBS, which performs both PPO network and UR functions in-house, reports

approximately 90 inpatient admissions per thousand in Florida, with a decrease

from highs of approximately 109 in 1990 and 1991. (Joint Exhibit Number 5,

exhibit to question Number 60.6.2.a, p. 15).

68. Cost savings or impact on costs to the trust fund and individual

employees are thus reflected throughout the RFP, and not simply in Section

80.10, giving additional credence to the weights ascribed by the RFP framers.

69. BCBS speculates that the state and its plan participants will

inevitably incur substantial extra health care expenses if AHCA's evaluation

turns out to be wrong. However, even if the Unisys-Beech Street network fails

to fully achieve comparable prices, there are safety net features in the

contract. Participants can elect to use HMO's or private insurance in lieu of

the plan, and the Legislature is considering additional options. The agency has

reserved the right to carve out particular health care services for separate

direct contracts with providers or to provide services through Community Health

Purchasing Alliances (CHPA's) in lieu of the plan. The agency also has reserved

the right to terminate the contract entirely for convenience, without obligation

except to pay for services rendered. Finally, the agency will evaluate Unisys-

Beech Street's implementation plan for expanding the PPO network to meet the

plan's needs within 30 days after the contract award and can seek remedies for

any deviation from that plan.

70. The agency's weighting of the technical questions must be considered

in light of all circumstances, including the known administrative costs

reflected in the competing proposals. BCBS's evidence does not prove that the

agency's weighting of the limited purpose historical price analysis in Question

80.10 produced an irrational evaluation of competing proposals, nor that any

potential risk so clearly outweighs known administrative cost savings as to make

the contract award to Unisys irrational. BCBS argued that additional weight

should have been given to other questions of the RFP. However, BCBS presented

no evidence that would indicate the subjective determinations of weighting

calculated by BCBS are any better or worse than the determinations of weighting

made by the agency. Reasonable persons can, and do (as in this case) differ.

The evidence, as developed through the testimony of Mr. Lutz and Ms. Morgan, has

shown that the agency's weighting scheme was a carefully designed, strict

implementation of AHCA's goals and intent.

Statistical Analysis of the Scores

71. BCBS presented statistical analyses of the overall scoring through its

expert, Dr. James T. McClave, along with charts and graphs prepared by Dr.

McClave. The analyses prepared and presented by Dr. McClave included an

analysis of inter-rater agreement, as well as several tests that Dr. McClave

said showed a statistical bias in one of eight groups of evaluators.

72. In order to test inter-rater agreement, Dr. McClave applied a

statistical model called the Kappa method. With this method, Dr. McClave

compared the scores given by three evaluators for each of approximately 235

scored questions on the score sheets generated for BCBS, Unisys and Humana. Dr.

McClave compared the scores on the 0-10 category scale, as well as a series of

"collapsed" scales (i.e. a five-category scale based on 0, 1-3, 4-6, 7-9, 10),

with one of the scales using as few as two categories (i.e., 0-5, 6-10).

73. In order to find an "agreement" between evaluators using the Kappa

method for the 0-10 point scale, Dr. McClave defined agreement as "pure

agreement," in other words, the scores had to be the same. To expect high

agreement or exact agreement for the 0-10 scale was a tough standard from a

statistical point of view and therefore he began to look at "collapsed" scales

which, according to Dr. McClave, would be relatively easier to meet. For the

other "collapsed" scales, the scores still needed to fall into the same category

to be considered perfect agreement.

74. All of the Kappa tests presented by Dr. McClave had percentages of

perfect agreement of less than 40 percent, which, according to the scale picked

from the statistics text used by Dr. McClave, represented "poor" agreement.

Based upon the Kappa method that he employed and the scale set forth in the

text, Dr. McClave concluded that the level of inter-rater reliability was poor,

and that the evaluation cannot be trusted. He conceded that there was no

precise way to identify the reasons why the reliability was so low, but

conjectured that a lack of training or amount of time allowed for the scoring

could have been a cause.

75. Dr. McClave also described a statistical method which he referred to

as the weighted Kappa. The weighted Kappa gives more weight to the level of

agreement, for example, a four versus a three is higher than a one versus a

four. The unweighted Kappa method employed by Dr. McClave assigned the same

"zero" agreement value for a score of four versus three as it did to a score of

one versus four.

76. Dr. McClave did not use the weighted Kappa because in his "review of

the literature the statistical theory behind the weighted Kappa has not been

sufficiently developed to the point where one can use it in the case we have."

(Transcript, p. 635) Weighted Kappa, in his view, compares two evaluators, one

against the other. Dr. McClave admitted that unweighted Kappa was designed for

nominal data, the most basic category of data. The scores in the evaluation

were done in ordinal fashion and according to widely recognized authorities in

the field, weighted Kappa is the appropriate statistical method for analyzing

ordinal data.

77. Dr. McClave has no expertise in any of the substantive areas of the

technical proposal (TPA, PPO or UR), or in the development, weighting or scoring

of RFP's in these substantive areas. He admitted that he had no reason to

believe that any of the scorers was not conscientious and diligent, or that they

used any improper scoring method or standard. BCBS did not offer a single

incident to show scoring was improper, nor any basis to claim that scorers were

not motivated to be conscientious and fair.

78. Unisys presented Erwin Bodo, Ph.D., as its statistical expert witness.

Dr. Bodo reviewed the circumstances in which the scoring was performed, i.e.,

use of 24 evaluators with diverse backgrounds and perspectives; use of questions

involving the application of judgement and subjective standards; and use of the

0-10 scoring scale without any exact or true score for any questions. Under

these circumstances, substantial disagreement is ordinarily expected. The

difference between the highest and lowest scores was two points or less for 50

percent of the questions, and three points or less for 80 percent of the

questions. This constitutes reasonably good agreement among scorers, according

to Dr. Bodo.

79. The question of whether or not scorers on a particular question were

consistent is irrelevant to whether the evaluation was valid. As long as each

particular scorer was internally consistent, the overall scoring would be fair.

The Kappa analysis proves nothing relative to the fairness or validity of the

scoring, but simply reflects that the scorers saw the merits differently.

80. Dr. McClave's second statistical analysis separated the 24 scorers

based on the offices in which they worked. He found that the aggregate mean

scores of five scorers from the Bureau of State Employees Insurance (BSI) were

more favorable to Unisys or less favorable to BCBS in a statistically

significant degree from the aggregate mean scores from each of the other seven

offices.

81. Dr. McClave used the term "statistical bias" to describe the

differences between the five BSI scorers' aggregate scores and the other 19

scorers' aggregate scores, grouped by their respective offices. However, this

analysis does not prove actual prejudice or unfairness because the statistical

tests will not demonstrate such matters.

82. Dr. McClave acknowledged that the disagreement could be related to

differences in scorers' backgrounds and perspectives. He had no knowledge of

the scorers' backgrounds beyond what offices they worked in. He acknowledged

that the questions were subjective, that the scorers applied the scale in

different ways, and that there was no perfect answer because human judgement was

involved. Dr. McClave did not know which scores were right and which were

wrong, and could not say that disagreement among scorers made either score

wrong. He did not analyze individual questions to determine whether they were

properly scored. He had no basis to assume that any BSI scorers were unfair.

He nevertheless proposed disqualifying all BSI scorers and eliminating their

scores, giving BCBS enough additional technical points to win the contract.

83. Dr. McClave's proposed disqualification would effectively eliminate

ten questions that were scored by BSI scorers only. He admitted this was a

problem. It would also reduce the scoring of other questions to one or two

scorers, violating the RFP requirement that at least three persons score each

question.

84. There were numerous questions in which BSI scorers gave BCBS a higher

score than non-BSI scorers, or in which BSI scorers gave Unisys a lower score

than non-BSI scorers. This evidence supports a finding that there was no

systematic prejudice exhibited by the BSI scorers. Give the subjective nature

of the technical proposal, the use of scorers with diverse backgrounds and

perspectives enhanced the fairness of the process.

85. RFP Sections 120.3, 120.3.1 and 120.6 described how the scoring

process would be conducted, resulting in the ranking of proposals by the total

of scores awarded. There was no requirement for any supermajority or any

particular statistical level of agreement among the scorers beyond that which

results in a majority of the points, and BCBS did not challenge the absence of

such a requirement when it challenged the RPP.

SUMMARY OF FINDINGS

86. The disputed issues in this case arise from the differing opinions of

competent and articulate experts rather than from the underlying facts, which

facts are generally uncontroverted.

87. Drawing on the experience of its staff and borrowing some guidance

from its predecessor agency, AHCA developed its first RFP for state employee

health services. The process was designed to enhance competition and the

prospective offerors had ample opportunity for input. The questions they asked

and answers provided by the agency were incorporated into the RFP document. The

agency's preparation of the RFP, its interpretation of the document and its

scoring of the parties' responses were careful, well-intended and fair.

88. Competent experts differ on the agency's interpretation of the RFP as

applied to items not included in the Unisys responses; they differ on the

weights assigned to segments of the RFP. But the agency's interpretation and

weighting were not proven arbitrary or illegal.

89. Competent experts disagreed on whether the scores were statistically

reliable or biased. Their evidence was informative, and even entertaining, but

in the end had little practical application. None suggested that the scorers

colluded, conspired or falsified their scores. Any explanation for near-random

results (assuming that Dr. McClave's methodology was appropriate) is based on

conjecture and not on any real evidence. The scorers were experienced, were

trained and were afforded the time to accomplish their assignments. Statistical

bias by one group is irrelevant in the absence of actual prejudice. The

statistical bias, like the suggested inter-rater unreliability, can be made to

appear or to vanish with simple manipulation of methodology or realignment of

the groups under scrutiny. Such evidence is too tenuous to establish the

agency's misprision.

90. The agency's intended award is appropriate and fair, and not arbitrary

or illegal.

CONCLUSIONS OF LAW

91. The Division of Administrative Hearings has jurisdiction in this

matter pursuant to sections 120.53(5) and 120.57(1), Florida Statutes.

92. The Division of Administrative Hearings has no jurisdiction with

regards to the alleged unconstitutionality of AHCA, and that issue will be

determined in a proper judicial forum. If the agency is found to be

unconstitutional, the court, and not the Division of Administrative Hearings,

will determine whether prior actions of the agency are void or illegal. For

that reason, the exhibit offered by BCBS, the declaratory judgement currently

under appeal, from the Second Judicial Circuit, in and for Leon County, case

Number 94-3128, is irrelevant.

93. In a bid protest such as this, the sole responsibility of the hearing

officer is to ascertain whether the agency acted fraudulently, arbitrarily,

illegally or dishonestly. Department of Transportation v. Groves-Watkins

Constructors, 530 So2d 912 (Fla 1988). Citing Liberty County v. Baxter's

Asphalt and Concrete, Inc., 421 So2d 505 (Fla 1982) the Court reminds:

A public body has wide discretion in soliciting

and accepting bids for public improvements and

its decision, when based on an honest exercise

of this discretion, will not be overturned by

a court [or a hearing officer] even if it may

appear erroneous and even if reasonable persons

may disagree.

(emphasis in original, Groves-Watkins, p. 913)

94. There was not the slightest evidence of fraud or collusion in the

development of the RFP, its weighting or the scoring of the responses. The

evidence likewise falls short of establishing that the agency acted arbitrarily

or illegally.

95. Section 287.0572, Florida Statutes, provides, in pertinent part:

(1) The cost of bids or proposals for state

contracts which require the payment of money

for more than 1 year and include provisions

for unequal payment streams or unequal time

payment periods shall be evaluated using

present-value methodology. ...

96. As found above, the costs that are the subject of this RFP, that is

the administrative costs, were evaluated using present-value methodology. The

health care costs, both to the state's trust fund and individual employees,

while affected by the RFP, are incalculable and unpredictable and cannot be

evaluated using present-value methodology. See, Capital Group Health Services of

Florida, Inc. d/b/a Capital Health Plan v. Department of Administration, and

Healthplan Southeast, Inc., DOAH Case Number 87-5387BID (recommended order

3/9/88, final order 4/28/88).

RECOMMENDATION

Based on the foregoing, it is hereby,

RECOMMENDED:

That the Agency for Health Care Administration issue its final order

awarding the contract to Unisys, as intended.

DONE and ORDERED this 27th day of September, 1995, in Tallahassee, Florida.

___________________________________

MARY W. CLARK, Hearing Officer

Division of Administrative Hearings

The DeSoto Building

1230 Apalachee Parkway

Tallahassee, Florida 32399-1550

[phone]

Filed with the Clerk of the

Division of Administrative Hearings

this 27th day of September, 1995.

APPENDIX

The following constitute specific rulings on the findings of fact proposed

by the Petitioner. The findings proposed by the agency and Intervenor have been

substantially adopted.

Petitioner's Proposed Findings

1. Adopted in paragraph 1.

2 & 3. Adopted in paragraph 4.

4 & 5. Adopted in substance in paragraph 2.

6. Rejected as unnecessary.

7. Adopted in paragraphs 4 and 12.

8. Adopted in paragraph 5.

9. Adopted in paragraph 6.

10 and 11. Adopted in substance in paragraph 7.

12. Rejected as unnecessary.

13. Adopted in paragraph 5.

14. Adopted in paragraph 10.

15. Adopted in part in paragraph 16. The

characterization of the Unisys proposal as a

"Unisys/Beech Street/Cost Care proposal" is

rejected as misleading.

16. Adopted in paragraph 15.

17. Adopted in paragraph 29.

18. Adopted in substance in paragraph 30.

19 & 20. Rejected as unnecessary.

21. Adopted in substance in paragraph 30.

22. Rejected as a conclusion contrary to the evidence.

One proposal was disqualified in the second phase.

23. Adopted in substance in paragraph 17.

24. Adopted in substance in paragraph 8.

25. Adopted in part in paragraph 18. The lack of

"formal training" or interviews is rejected as

immaterial and misleading. The staff were trained

and were amply instructed.

26. Adopted in summary in paragraph 19.

27. Adopted in paragraphs 11 and 21.

28. Adopted in paragraphs 9 and 23.

29. Adopted in substance in paragraph 23.

30. Adopted in paragraph 24.

31 & 32. Adopted in substance in paragraphs 26 and 27.

33. Adopted in paragraph 28.

34. Addressed in Preliminary Statement.

35. Rejected as contrary to the evidence (the

conclusion of "arbitrary and capricious").

36. Rejected as misleading as the experience of both

was found to be appropriate to the task.

37. Adopted in substance in paragraphs 50 - 52.

38. Adopted in part in paragraph 52; otherwise

rejected as misleading argument.

39. Rejected as unnecessary.

40 - 42. Rejected as unnecessary and argument that is not

supported by the weight of evidence.

43 & 44. Adopted in summary in paragraphs 51 and 53.

45 - 61. Rejected as irrelevant or argument that is not

supported by the greater weight of evidence, which

evidence did support the agency's contention that

section 80.10 is only a piece of the financial

outlook picture.

62 - 64. Rejected as unnecessary. The contract is not for

direct medical services and the cost of those

services over the term is incalculable.

65 - 66. Adopted in summary in paragraph 31.

67 - 77. Rejected as unnecessary or argument that is

unsupported by the weight of evidence, which

evidence supports the interpretation by the agency

that the forms were not required from contractors,

and Beech Street and Cost Care were subcontractors

rather than "offerors".

78. Adopted in substance in paragraph 42.

79 - 82. Rejected as unnecessary. The evaluators did

review the financial statements, but not as a

mandatory item, and scored the responses based on

the review. Although it is accepted that the

audited financial statements are important, so

also are other indicia of financial viability and

stability.

83. Adopted in paragraph 47.

84. Adopted in paragraph 48.

85. Adopted in paragraph 71.

86. Adopted in paragraph 72.

87 - 91. Adopted in summary in paragraphs 72 through 76.

92 & 93. Adopted in part in paragraph 74, as to the results

of Dr. McClave's statistical analysis; rejected as

to the conclusions that the agency's evaluation

was unreliable or arbitrary and capricious, as the

statistical analysis does not support that

conclusion.

94. Adopted in paragraph 80.

95 - 98. Adopted in part in paragraphs 80 through 82;

otherwise rejected as irrelevant.

99 - 103. Rejected as irrelevant. See Conclusion of Law

Number 92.

COPIES FURNISHED:

Michael J. Glazer, Esquire

Stephen C. Emmanuel, Esquire

Steven P. Seymoe, Esquire

MACFARLANE, AUSLEY, FERGUSON &

MCMULLEN

Post Office Box 391

227 South Calhoun Street

Tallahassee, Florida 32302

Paul Martin, Assistant Attorney General

Office of the Attorney General

PL-01, The Capitol

Tallahassee, Florida 32399-1050

James H. Peterson, III

Steven Grigas

Agency for Health Care

Administration

2727 Mahan Drive, Suite 3400

Tallahassee, Florida 32308

Stephen Turner, P.A.

David K. Miller, P.A.

BROAD & CASSEL

215 South Monroe Street, Suite 400

Post Office Box 11300

Tallahassee, Florida 32302

Jerome W. Hoffman

General Counsel

Agency for Health Care

Administration

2727 Mahan Drive

Tallahassee, Florida 32309

Mr. Sam Power, Agency Clerk

Agency for Health Care

Administration

Building 3, Room 3431

2727 Mahan Drive

Tallahassee, Florida 32308

STATE OF FLORIDA

DIVISION OF ADMINISTRATIVE HEARINGS

BLUE CROSS AND BLUE SHIELD OF )

FLORIDA, INC., )

)

Petitioner, )

)

vs. ) CASE NO. 95-3635BID

)

AGENCY FOR HEALTH CARE )

ADMINISTRATION, )

)

Respondent, )

)

UNISYS CORPORATION, )

)

Intervenor. )

_________________________________)

NOTICE

The following notice was inadvertently omitted from the Recommended Order

which was issued on September 27, 1995:

NOTICE OF RIGHT TO SUBMIT EXCEPTIONS

All parties have the right to submit written exceptions to the Recommended

Order. All agencies allow each party at least 10 days in which to submit

written exceptions. Some agencies allow a larger period within which to submit

written exceptions. You should consult with the agency that will issue the

Final Order in this case concerning their rules on the deadline for filing

exceptions to this Recommended Order. Any exceptions to this Recommended Order

should be filed with the agency that will issue the Final Order in this case.

DONE and ORDERED this 2nd day of October, 1995, in Tallahassee, Florida.

___________________________________

MARY W. CLARK, Hearing Officer

Division of Administrative Hearings

The DeSoto Building

1230 Apalachee Parkway

Tallahassee, Florida 32399-1550

[phone]

Filed with the Clerk of the

Division of Administrative Hearings

this 2nd day of October, 1995.

COPIES FURNISHED:

Michael J. Glazer, Esquire

Stephen C. Emmanuel, Esquire

Steven P. Seymoe, Esquire

MACFARLANE, AUSLEY, FERGUSON &

MCMULLEN

Post Office Box 391

227 South Calhoun Street

Tallahassee, Florida 32302

Paul Martin, Assistant Attorney General

Office of the Attorney General

PL-01, The Capitol

Tallahassee, Florida 32399-1050

James H. Peterson, III

Steven Grigas

Agency for Health Care

Administration

2727 Mahan Drive, Suite 3400

Tallahassee, Florida 32308

Stephen Turner, P.A.

David K. Miller, P.A.

BROAD & CASSEL

215 South Monroe Street, Suite 400

Post Office Box 11300

Tallahassee, Florida 32302

Jerome W. Hoffman

General Counsel

Agency for Health Care

Administration

2727 Mahan Drive

Tallahassee, Florida 32309

Mr. Sam Power, Agency Clerk

Agency for Health Care

Administration

Building 3, Room 3431

2727 Mahan Drive

Tallahassee, Florida 32308