State of FloridaStatewide Grand Jury Presentment, No True Bill, and Recommendations, OSWP No. 96-0292-NFB
June 2, 1998
IN THE SUPREME COURT OF THE STATE OF FLORIDA-CASE NUMBER 86,726
(This document has been formatted for the Internet)
During the past eighteen months, we have examined numerous criminal cases involving fraud against the government, with a special emphasis on Medicaid fraud. Most of the cases were investigated by the Attorney General's Medicaid Fraud Control Unit and the Florida Department of Law Enforcement. We issued indictments against 75 criminal defendants for Medicaid fraud; an additional 70 defendants have been charged by direct information. These cases represent $25-30 million dollars in thefts from the Medicaid program.
During our term, certain evidence was brought to our attention concerning the activities of Unisys Corporation, which serves as the Fiscal Agent for the State in the Medicaid Program. We learned that two employees of the company had been arrested by FDLE for involvement in schemes to defraud the State out of a significant amount of money: more than $380,000 in one instance and $1.3 million in another.
During the course of this examination, we were also apprised of serious allegations against the company with respect to its operations under another health-related State contract: the Florida State Employees Health Self Insurance Trust Fund ("State Health Fund".)
Taken together, these contracts give Unisys responsibility and control over billions of taxpayer dollars held in trust by the State. Both contracts are worth millions in administrative fees to Unisys. We were concerned with how the contracts were obtained, how they were managed, and how the company performed. Our task was to determine if any criminal activity occurred in any aspect of these relationships between the State and Unisys.
We dealt with these issues during every one of our monthly sessions between September 1996 and June 1997. In all, we heard testimony under oath from numerous witnesses, we reviewed hundreds of pages of documents, and we were instructed on the application of dozens of relevant State laws. The Directors and key employees of the Agency for Health Care Administration ("AHCA") and the Department of Management Services ("DMS") testified before us voluntarily, as did representatives of the Governor's Office for Planning and Budgeting. We heard from former and current employees of Unisys, as well as representatives of the Mercer Company, a health care benefits and consulting group, and experts in using the particular claims processing software ("claims software") used by Unisys under this contract. We are told that our legal advisers and agents of the Florida Department of Law Enforcement interviewed dozens more witnesses in this investigation.
We divided our inquiry into four parts: (1) management of the Medicaid Trust Fund contract; (2) procurement of the State Health Fund contract; (3) management of the State Health Fund contract; and (4) performance under the State Health Fund contract.
We were looking for evidence of theft, fraud, unlawful compensation, bribery, and computer crimes. Our investigation has not uncovered any evidence that Unisys authorized the thefts from the Medicaid Trust Fund, or that it directed the criminal behavior of its employees. (For further discussion, see pages 8-9 herein.) We found no evidence that Unisys lied to obtain the State Health Fund contract. We did not find evidence that the bid process was unlawfully skewed in favor of Unisys (see pages 10-25), or that anyone was bribed or received unlawful compensation to award the contract to Unisys. We find insufficient evidence at this time to prove that Unisys, as a corporate entity, lied to the State in order to keep the contract (see pages 25-34). In summary, we have not found sufficient evidence at this time of criminal activity to charge the company, its employees, or any State employee with criminal violations. While we find certain circumstances to be suspicious, and these give us great pause, we do not have proof sufficient to overcome a reasonable explanation. Circumstantial evidence that does not exclude every reasonable hypothesis of innocence cannot support a conviction. Thus, we cannot issue an Indictment. Based upon our examination of the facts and the law, it would be wrong to do so, even though there is more than enough blame to go around. Instead we issue this No True Bill.
However, pursuant to our authority to comment upon civil administration, we issue this document. Given the responsibilities Unisys has for the proper administration of the Trust Funds, we view Unisys as the equivalent of a State employee, acting on behalf of State officials. We find: (1) that the contractual relationship between the State and Unisys in these two critical programs is not in the best interests of the taxpayers; (2) that the State's contract procurement and management process is in need of some "common sense" revision; (3) that AHCA and DMS Agency heads are responsible for the state of affairs we noted in each of these programs; and (4) that the laws of this State do not provide criminal penalties for this situation: where employees of a corporation engage in a series of legally and ethically questionable actions without the direct knowledge and consent of senior officers of the corporation.
With respect to the Medicaid Trust Fund, we conclude that Unisys could have prevented some portion of the fraud by the implementation of greater security and controls, and a more pro-active approach to claims evaluation and payment. However, we do not find sufficient evidence in this mismanagement to justify criminal charges against Unisys.
With respect to the State Health Fund contract, we conclude that Unisys was awarded the contract as a result of serious deficiencies in the procurement process. First, AHCA changed the terms of the bid from prior awards in an effort to downplay Blue Cross and Blue Shield's ("BC/BS") traditional advantages in provider networks and discount rates, with the avowed intention to "level the playing field" for other bidders. Second, there was some evidence that the Director of AHCA harbored animosity toward the incumbent vendor, BC/BS, but the bid evaluation process was conducted under extremely sterile conditions. Third, before awarding the contract, AHCA tried to require the vendors to perform a processing test of State-owned claims. When BC/BS threatened to file a protest, arguing that it should not be required to release proprietary information, AHCA relented and allowed the language to be deleted from the bid. Fourth, AHCA evaluators scored the proposals based simply on the responses supplied by the bidders, and interviewed only the references provided, rather than seeking more independent opinions regarding the abilities of the bidders to perform.
The allegations of impropriety and changes from the former bid weighting process are troubling. And, the bid evaluation process, no matter how far removed from the Director, did allow an inexperienced and unprepared bidder to prevail in the scoring. However, we find no evidence of bribery or corruption, nor do we find evidence of misrepresentations in the procurement process sufficient to justify criminal charges against Unisys or any State employee.
After the vendor was selected, AHCA was required by law to pass responsibility for management of the contract to DMS. Because DMS adopted a "hands off" policy during system implementation and start-up, AHCA tried to perform this function, relying on its experience managing the Medicaid contract which, by comparison, was vastly less complicated than the State Health Fund. DMS played little part in the procurement of the contract, and appeared to exhibit little responsibility for the contract's successful management. During the implementation process, the lines of authority for contract management became blurred, essentially leaving the vendor without effective supervision and control. While we hold the Agency heads accountable for the events that ensued, we do not find any evidence of criminal behavior.
The performance of Unisys under the State Health Fund contract was uniformly unacceptable, consistently failing to meet contractually mandated benchmarks. Lower-level employees were not effectively trained or properly managed; senior-level employees were not kept informed of lower level problems. State employees who depended on Unisys to process their health care claims were repeatedly subjected to late and inaccurate payment, and the consequent embarrassment and damage to their reputations and personal credit records. The State imposed the maximum "error rate" penalties against Unisys allowed by the contract and eventually, seeing the handwriting on the wall, Unisys has agreed to a termination of the contract. While the damage Unisys caused the citizens of this State by its sub-standard performance is severe, we find insufficient evidence at this time to justify the issuance of criminal charges against Unisys.
We heard testimony about very suspicious events: manipulation of claims processing error rates and other contract benchmarks, and concealment of large numbers of claims from State oversight, both of which arguably could have allowed the company to evade liquidated damages penalties under the contract. Some of the witnesses were highly informed and credible, some less informed or credible; and we outline their allegations at length in this report. On balance, we found no allegation for which there was not a countervailing reasonable explanation. Given the State's standard of proof in a criminal prosecution, we find that there is not a substantial likelihood of conviction. Therefore we decline to issue criminal charges at this time against Unisys or any of its employees.
II. INVESTIGATIVE FINDINGS
A. Medicaid Program
Unisys obtained the contract to serve as the State's fiscal agent under the Medicaid program in 1993. The contract requires Unisys to handle provider enrollment, process and pay claims (both paper and electronic), and to handle provider inquiries. The contract gives them control over $6.7 billion dollars annually. The value of the contract to Unisys is $14 - $15 million dollars per year. The contract expires in 1998, and the State has notified Unisys that it will not automatically renew the contract.
During our term, we heard testimony about the arrests of two Unisys employees who are charged with committing serious criminal acts on the Medicaid program. One employee stole several huge Medicaid checks, while working in the banking section as a temporary employee. He has entered a plea of guilty to the charges and is cooperating with the State in the case against his sister, with whom he split the proceeds of the unlawful activity. He has also told us how he was able to commit his crimes. He said that the checkbook wasn't secured, that many employees had access to it, and that the signature stamp was also easily accessible. He said he committed his thefts by stealing blank checks and by modifying checks that had been returned as undeliverable. He testified the returned checks were just thrown in a box in a corner until someone reconciled the checkbook. He said staff was months behind in check ledger reconciliations, and that is why he could commit crimes over an extended period of time without fear of being caught. He said if he thought he was going to be caught by his supervisors, he never would have done it. It was only when his sister tried to cash one of the altered checks, that the illegal activity was put to an end. When the crime was discovered, AHCA deducted the sum, more than $380,000, from Unisys' next monthly payment.
Another Unisys employee is charged with participating in a scheme to illegally divert Medicaid funds to fictitious providers in the South Florida area. It is alleged that he facilitated this scheme by altering provider records with fictitious change of address information. As a result, more than $1.3 million dollars was sent by Unisys to commercial post office facilities. The checks were retrieved by runners who then cashed the checks at commercial check cashing stores and other facilities that didn't require proof of identification. We are told that this case is still under investigation by FDLE. Our interest in this was, again, how could this happen? Again, we were told that change of address information can be entered by many employees, and that follow-up confirmation with the provider is not conducted.
This same scenario was played out in many of the cases which we have investigated over our term. Checks were sent to commercial check cashing stores, to empty store-fronts, and to fictitious providers. We heard no evidence that Unisys routinely verifies endorsements on the checks issued. Nor did we hear that Unisys monitors account activity for unusual billing activity. Not once did we hear that Unisys stopped payment on any checks unless told to do so by AHCA.
We heard that they conducted no criminal background checks on their banking employees, and that they once placed a State work-release prisoner on their data entry staff.
We cannot comprehend why Unisys ignored normal financial security controls or didn't take a pro-active approach to guard against fraud on the Medicaid fund. We are convinced that Unisys is more concerned with its technical capabilities than its fiscal duties to the State. Even if the contract doesn't address this specific issue, the State should nevertheless insist that Unisys exercise normal fiscal prudence. The absence of relatively simple internal controls resulted in opportunities for Unisys employees to enrich themselves at public expense.
B. State Health Fund Contract
The State Health Fund Contract requires the administrator to process insurance claims. Because the State does not have its own health claims processing system, the administrator is required to: (1) design its own customized computer hardware system from the ground up; (2) create, lease, or purchase its own computer software specifically tailored to process claims under Florida's unique Benefits Plan; and (3) set up and maintain preferred provider organizations of physicians, hospitals and other health care professionals, and negotiate discounts with them based on the volume of the State's client base. Florida has never required the administrator to share ownership of the processing system as a condition of the contract. Historically, the contract administrator pays out approximately $300 million dollars per year in claims. The administrator has historically received in excess of $10 million dollars per year for this service.
1. Procurement of the State Health Fund Contract
In 1991, the then-existing Department of Administration, which included the Division of State Employees' Insurance (DSEI), was combined with the then-existing Department of General Services to form a new agency: the Department of Management Services. This was a part of Governor Chiles' "Right-sizing Government" initiative. Many agencies were combined and duplication of services was eliminated throughout the government under this policy.
At the time, the incumbent vendor for the State Health Fund contract was BC/BS, which had administered the program since its conversion to a self-insurance fund in 1978. Blue Cross' management of the fund was supervised throughout that time by DSEI. We were told that BC/BS helped design the Florida State Employees' Benefit Plan over the intervening years, with the result that either the Plan mirrored BC/BS computer capabilities or BC/BS computer capabilities mirrored the Plan, or both. Whichever occurred, witnesses testified that the BC/BS system apparently handled employee health claims very quickly and efficiently.
When BC/BS was awarded the 1991 State Health Fund contract, a two-step bid evaluation process was involved: (1) vendors were asked for written assurances that they possessed the technical expertise necessary to administer the trust fund; and (2) they were asked for a written cost proposal, including both (a) per member per month fees and (b) the discounts anticipated from the vendor's Preferred Provider Organizations. For the purposes of the 1991 procurement, discount rates were computed by reference to a standard: claims cost figures released by the Health Care Cost Containment Board. These were industry averages of claims reported by providers to the Board.
Once a vendor alleged that it could meet the technical benchmark, the competition shifted to the cost factor without reference to technical expertiese. The two cost elements were to be blended in a ratio of 85% to 15% (pmpm fees / anticipated discounts). Once the final point totals were reached, the Division would make a decision based on the point totals, unless sufficient other factors intervened to justify ignoring the low bid. For example, in 1991, the winner, BC/BS, did not submit the lowest cost bid. The BC/BS bid was recommended for acceptance by a committee which suggested that BC/BS' agreed higher discount numbers would save the State far more money than the amount saved by the lower administrative fees sought by losing bidders.
When Unisys was awarded the contract in 1995, this bid-weighting scheme was changed. This served, in part, to eliminate BC/BS as the incumbent.
In 1992, AHCA was created to centralize governmental health policy and purchasing. Its various responsibilities were to take effect on a staggered basis throughout 1992 - 1994. The purchase of administrator services for the Medicaid Fund and the State Health Fund became its responsibility as of July 1, 1993. The responsibility to manage the resulting administrator's contract remained a duty of DSEI at DMS.
We heard testimony that this split of authority allowed the State to take advantage of the "greater expertise" of AHCA in regard to health matters. Contract management was left with DMS, which managed all other State employees' benefits. If management of the health plan were left with DMS, it was reasoned, State employees could get all their benefits questions answered at the same source. Historically, DMS gives State employees the opportunity to enroll in various benefit plans yearly during an open enrollment period starting in September.
During 1992-94, Governor Chiles was advocating several new and controversial legislative initiatives concerning health related matters: Community Health Purchasing Alliances, the "Healthy Kids" program, and the Florida Health Security Act. The Director of AHCA, Douglas Cook, acted as Governor Chiles' spokesman and the Governor's leading lobbyist on these issues. Director Cook testified that he possesses a forceful personality and that he has been known to argue strongly on behalf of these issues. Blue Cross found itself on the other side of these issues.
We heard testimony that Director Cook and Blue Cross' main lobbyists engaged in several heated arguments concerning Blue Cross' position on these issues. Blue Cross' lobbyists alleged that Mr. Cook emphasized that he was very disappointed that Blue Cross opposed the Governor's initiatives. We learned from an FDLE agent that a Blue Cross official/lobbyist said that Mr. Cook complained that Blue Cross had extra access to legislators as the result of its position with the State group, stating, "I find it absolutely incredible that you all people have the State group and you are the largest insurer and you will not support this government." More troubling, in his statement under oath to FDLE, the Blue Cross official quoted Mr. Cook as saying that Blue Cross,"...will pay a price for not doing this." During his testimony, Mr. Cook denied making these statements, but he agreed that heated arguments are not uncommon in the halls of the Capitol.
When, in 1995, the contract for administration of the State Health Fund came up for re-bidding, Blue Cross was concerned that AHCA might try to prevent it from winning in the new bidding. However, Mr. Cook testified under oath that he did not attempt to influence the bid process in any way, that he did not make his displeasure with BC/BS known to the bid evaluation team or his deputy in charge of procurement, nor did he say anything in the presence of these individuals from which they might reasonably have concluded that he did not want Blue Cross to win the 1995 contract. Richard Lutz, AHCA deputy in charge of the procurement, confirmed these statements under oath. We heard no other testimony which might lead us to believe that Mr. Lutz was aware of any "bad blood" between Cook and Blue Cross.
In major governmental purchasing decisions, Florida law tries to insulate the decision-maker from all outside forces. Purchases of goods and services used by government and its agencies in performance of its duties are made under guidelines contained in Chapter 287, Florida Statutes. Chapter 287 provided the skeleton for the procedures which Mr. Lutz used to purchase Third Party Administration (TPA) services for the State Health Fund.
According to Mr. Lutz, AHCA outlined its requirements in a Request for Proposal (RFP), took bids from all interested vendors, evaluated vendor responses to the RFP in an environment insulated from contact with anyone who might attempt to influence the decision inappropriately, and weighed the evaluations as objectively as possible. More than twenty Agency employees were involved over a period of 3-4 weeks in the evaluation of vendor responses to the RFP. Although Mr. Lutz invited DMS employees to participate, he testified that DMS did not contribute any evaluators to the process. DSEI Director, Ms. Millie Seay, testified that she had insufficient personnel to contribute to such a lengthy evaluation process.
According to Mr. Lutz and other witnesses from AHCA, the evaluators were chosen based on their knowledge of the issues involved in the procurement, prior expertise and experience. For example, an employee with a background in accounting and financial management was given the responsibility to evaluate responses to RFP sections concerning the financial arrangements in the RFP; someone with a background in computers and systems analysis was tasked to evaluate responses to RFP sections concerning the computer systems to be used, and so on. Evaluators worked on the same sections of all of the vendor responses, so that the same persons judged the issue for every vendor.
Several witnesses testified that they evaluated vendor responses based upon whether the system described on paper looked like it could work. There was no site visit or any attempt to ensure that any bidder could actually assemble and efficiently operate a claims processing system which could handle the State Employees' Health Self Insurance Fund. The only attempt made to investigate the vendors' expertise and ability to carry through with their promises was to contact the vendors' self-reported references. Witnesses testified that the evaluators were not allowed to follow up any other "leads" to determine the suitability or credibility of the vendors. They were instructed to evaluate only the information supplied by the vendor in its response. Mr. Lutz testified that these restrictions on information gathering were designed to ensure that evaluators were completely objective.
There was no requirement in the RFP that the vendor demonstrate or prove that its system could process claims quickly and efficiently using the State of Florida Benefit Schedule. Witnesses testified that the evaluators did not give extra weight to the vendor with the working system because they believed that would give BC/BS an unfair advantage over the other bidders.
As discussed previously, the bid-weighting process was re-designed in 1995. Mr. Lutz testified that he did so in order to neutrally "level the playing field": to remove BC/BS' historic advantages, thereby allowing all bidders, including BC/BS, to have an equal chance to succeed.
The way in which Mr. Lutz "leveled the playing field" was to alter the bid weighting process. Mr. Lutz testified to concerns that the "technical skills benchmark, then cost only" process, used in 1991 could arguably have been seen as arbitrary (although no protests succeeded on that ground in 1991). He therefore changed the scoring so that the old process was abandoned. Instead, the technical component of the vendors' bids were scored at the same time as the cost component. Thus, an award decision which in 1991 was completely based on cost (after the vendors had reached a technical benchmark), was in 1995 based on both the vendors' technical and cost submissions. The technical/cost split in 1995 was 60% - 40%. In the 1995 calculation, the weight allowance for discount rates was 2.4% of the newly combined technical/cost total, as opposed to the 15% weight given discounts under the "technical skills benchmark, then cost only" 1991 protocol.
Mr. Lutz testified that the lower weight given to discount rates in 1995, resulted from his decision that the importance of discount rates had been overstated previously. He testified that discount rates in and of themselves are inherently misleading. He testified that if the original bill for medical services is inflated, even a large discount may be less beneficial to the trust fund than a smaller discount from a reasonable bill. (For example, in 1991, BC/BS promised to obtain a 40% discount from its providers starting January 1, 1992. Prior to that date, BC/BS was obtaining a 27% discount from its providers. While it is possible that BC/BS had managed to negotiate its own providers down another 13% between December 31 and January 1, it is equally possible that abuses might occur).
Several witnesses testified that each point of discount from established rates is worth approximately $3 million dollars. The 1991 BC/BS contract included a written agreement that claims under the State Employees' Health Self Insurance Fund would receive a discount of 40% from BC/BS's Preferred Provider Organization (PPO). Agency witnesses testified, however, that the BC/BS reported discount rate of 40% was unprovable. They pointed out that between 1991 and 1995, neither the Department of Administration nor its successor, DMS, contracted for an outside audit of the State Employee Health Self Insurance Trust Fund, as would have been allowed under the 1991 BC/BS contract. Such an audit would have shown whether BC/BS was living up to its contract promises concerning discount rates. Therefore, in the 1995 Request for Proposal, AHCA did not seek any similar specific discount rate from vendors.
As it turned out, when AHCA attempted to obtain from BC/BS certain historical claims documents belonging to the State in order to design a claims test during the 1995 RFP, BC/BS refused to deliver them to AHCA. We heard testimony that BC/BS argued that the claims might be used to compute BC/BS' discount rates, and would therefore disclose proprietary information. We heard that BC/BS threatened to challenge the provision in administrative hearings. Mr. Lutz and Mr. Cook both testified that they decided not to force the issue of proprietary language with Blue Cross because the litigation would have delayed the contract process, and they needed to have a new administrator by January 1, 1996.
In the absence of this discount rate information, Mr. Lutz designed a test of the vendors' ability to command a low price. In RFP section 80.10, entitled, Pricing Analysis, vendors were asked to "cost out" a known claims base. Witnesses testified that the claims test package included hundreds of claims from fifty-six different hospitals. The vendors were asked to evaluate the claims historically and specify the fees they would have paid for health care at that facility on that date. Mr. Lutz pointed out that BC/BS had a built-in advantage on this section because the claims were at Blue Cross hospitals with Blue Cross providers. However, as mentioned above, the points available under this section only constituted 2.4% of the total points on the whole contract, down from 15% under the 1991 contract.
Following the bid process and evaluations described above, Unisys won the bid. Mr. Cook testified that Mr. Lutz "turned pale" when the results of the evaluation and weighting were compiled, and they found that BC/BS would not be awarded the contract. Mr. Cook testified that both he and Mr. Lutz were well aware of the implications and the difficulties of such a change-over of third party administrators, after the Fund had been handled by BC/BS for more than seventeen years.
Blue Cross protested the results of the bid evaluation, arguing that the process was improperly weighted against it. In the Final Order, issued October 27, 1995, the administrative hearing officer ruled there was"...not the slightest evidence of fraud or collusion in the development of the RFP, its weighting or the scoring of responses. The evidence likewise falls short of establishing that the agency acted arbitrarily or illegally."
When we learned that Mr. Lutz had changed the bid-weighting process, it sounded unusual to us. However, without credible evidence that the rules were changed for some nefarious purpose, it is nothing more than that: unusual. Just as the hearing officer found no evidence of fraud, we have found no credible evidence which would lead us to believe that the evaluation process was intentionally designed so that Unisys would win or that Blue Cross would lose the contract. Even assuming that Mr. Cook made the statements attributed to him by Blue Cross, we have received no evidence to refute Mr. Lutz's testimony that he was unaware of them or acted in furtherance of any possible grudge held by Mr. Cook. Moreover, Mr. Lutz' painstaking effort to remove any subjective input from the evaluation process is evidence that the outcome of the process was not affected by any outside influences.
Why did the process that was so meticulously designed to insulate the evaluators from inappropriate influences lead to such a disastrous result for the citizens of this State? The answer is that in its eagerness to neutralize the selection process, AHCA mistook its obligation not to favor any particular vendor, with an obligation to overlook the incumbent vendor's natural advantage in having a working system. That fact has value, a value which was not taken into account sufficiently in the RFP process. With the State Health Fund, the continued smooth, even seamless, operation of the fund is a paramount value in itself. The importance of obtaining competent administrator services is as important as the value of integrity in the bidding process protected by Chapter 287's rigorous provisions. To make this point is not to devalue the importance of Chapter 287 in State purchasing. It has occurred to us that as bad as the outcome for the State has been in this matter, at least the damage did not spring from corruption or graft.
Nevertheless, this is not the purchase of a fungible commodity like automobiles, furniture, inmate clothing or law enforcement supplies. If the services purchased do not measure up, there is no quick fix solution available. The State cannot just go to a retail store and buy a working product. If the State Health Fund doesn't function correctly, more than 200,000 State employees and dependents have no back-up system to take care of them while an acceptable alternative is sought. They would be left out in the cold -- much as they have been since January 1, 1996.
2. Management of the State Health Fund Contract
As discussed previously, the management of the State Health Fund contract remained with DMS when the procurement function was moved to AHCA in 1993. The difficulties inherent in this bifurcated structure did not surface until they had to handle the new procurement of State Health Fund administration services, which did not occur for several years after the change in duties. Witnesses told us that this reopened old wounds. DMS lost personnel and budget to AHCA at the same time that it lost the procurement function.
These tensions manifested themselves in several ways: DMS employees testified that they were not closely connected with the drafting of the RFP; neither did AHCA seek out DMS' advice, in spite of its seventeen years of experience in these matters. In fact, AHCA had former DSEI employees on staff and didn't assign them to leading roles in the new procurement.
When the 1995 RFP was completed by AHCA, it was shared with a Steering Committee consisting of members from DMS, AHCA, and the Department of Health and Rehabilitative Services, among others, including DSEI Director, Ms. Millie Seay. Ms. Seay testified that after she heard the bid package described at the meeting, she told her supervisor at DMS that the new discount ratio formula could have a large impact on the soundness of the trust fund. According to Ms. Seay, she was told that the procurement of this RFP was AHCA's business, and that DMS would deal with the outcome.
While Ms. Seay had a point, we also heard testimony that the changeover to Unisys was followed by a large exodus of State employees from the State Health Fund to Health Maintenance Organizations ("HMOs"). We heard that these employees were the youngest and healthiest in the plan. They took their premiums with them. Therefore, the older, less healthy employees remaining in the State Health Fund plan were covered by a smaller Trust Fund. The witnesses termed this phenomenon, "adverse selection" and said it would erode the soundness of the Trust Fund. In March, 1997, we learned from DMS employee Jeff Dykes, that the Trust Fund must be increased in the next few years to meet this unanticipated need. He told us that as of the February 1997 Legislative Estimating Conference estimate, the Trust needs more than $100 million dollars to replace the revenue lost to adverse selection. The money will come from increased employee contributions, reductions in employee benefits, or new taxes.
After Unisys won the bid in July, there was limited contact between DMS and Unisys, because managers of both Departments believed it would be inappropriate in light of the bid protest by BC/BS. We heard testimony that between October and December, 1995, Unisys obtained office space, hired supervisory staff, purchased the "claims software" and attempted to hire claims processing and data entry staff. We also heard testimony that some limited training on the "claims software" was conducted in December.
AHCA employee Jay TerLouw testified that he was asked by Director Cook to visit the Unisys Tallahassee site around December 18, 1995. Mr. Cook asked Mr. TerLouw to give his opinion on whether Unisys could comply with its contract. Mr. TerLouw had directed the Medicaid implementation team in 1994, which had a six month implementation period. Mr. TerLouw recognized that two weeks was substantially less start up time than Unisys had on the Medicaid contract. But, Mr. TerLouw testified that he thought that Unisys would be ready to start the State Health Fund contract on January 1. Having been impressed with the Unisys Medicaid team, he assumed that the State Health Fund team would be as competent and professional. Also, Mr. TerLouw believed that Unisys could expect a slower start up period or a "breather" while the claims "ramped up" to expected levels, because the incumbent would be responsible for processing pre-1996 claims. He testified that he expected that it would take a month or two for claims activity to fill in. After the visit, Mr. Cook asked Mr. TerLouw to head the State Health Fund implementation team, as well.
Even though AHCA and DMS shared procurement and contract management duties under F.S. 110.123, they had no interagency agreement until long after the contract was in force. William Lindner, the Secretary of the Department of Management Services, testified that there was no consensus between AHCA and DMS from the beginning. The Secretary felt that DMS wanted more guidance than was provided in the State Health Fund contract about the duties of each agency under the contract. Mr. Lindner also said that his sense was that AHCA felt like the statute gave clear guidance. Nevertheless, according to Mr. Lindner, no one at DMS started to draft an interagency agreement until at least March or April, 1996.
During his four months on-site, Mr. TerLouw determined that the "claims software" used by Unisys was not well-suited to the task of processing the State's claims. He testified that certain provisions in the State's Benefit Plan ( payments for certain services "in connection with" a hospital or doctor visit) were difficult for the system to handle. Interestingly, when FDLE contacted the Louisiana procurement official about their use of the same "claims software" used by Unisys, he said he worked closely with the Louisiana legislature to remove similar language from their law, with the consequent effect that their claims processing system worked much more quickly. TerLouw noted in January, 1996 that Unisys was having difficulty loading "provider files" into the system; and after January, he noted that claims were not being processed as quickly as they should have been. Mr. TerLouw suggested to Cook and DMS that penalties be assessed against Unisys.
The contract authorized liquidated damages penalties as follows: if more than 1% (1) of the claims processed in a month contained a financial error rate, a graduated scale of penalties kicked in. If errors exceeded 1%, then 1% of the monthly fee (approx. $1 million dollars) would be forfeited by Unisys. Similar scaled penalties existed for failure to comply with contract standards on payment errors and overall payment and procedureal errors.(2) Overall, the highest penalties possible under the RFP based on all of the error rates combined amounted to approximately $80,000 per month (8% of monthly fees).
However, Mr. TerLouw told us that Unisys was unable to provide, and the State could not obtain reliable data on the number of claims processed. Even though an AHCA-led implementation team was on-site, DMS as contract manager was the only entity that could penalize Unisys, so the penalty letter came from DMS. DMS docked Unisys' next monthly payment for the highest forfeitable amount for errors: $80,000. Because the claims figures received from Unisys were not reliable, there was no attempt to penalize Unisys at that time for performance.(3) Subsequently, after more reliable information came out, Unisys was fined more than $4.4 million dollars for performance penalties.
According to a letter from AHCA Director Cook to DMS Secretary Lindner, dated April 11, 1996, AHCA formally handed over authority for all "day-to-day operational matters to [DMS]" on that date. By this time, DMS had retained Coopers and Lybrand to investigate whether the Unisys system was functioning well.
Coopers and Lybrand conducted an audit based on interviews and data samples. The audit revealed problems with error rates, report generation, and system security, and found no fraud detection programs. In June, DMS sent a "cure letter" to Unisys, attaching a copy of the audit. The letter placed Unisys on notice of DMS' right to terminate the contract if Unisys failed to cure the deficiencies noted. We were told that AHCA's general counsel helped draft the letter.
We learned that Unisys responded with a "corrective plan of action." But, we also learned that DMS never enforced the termination provisions, even though DMS staff accumulated mounds of evidence that Unisys was not in compliance. William Lindner, DMS Secretary, testified that it was "time to make a change," but he "felt like [he] needed more players in the mix to help make a decision of that--that stature."
DMS has a documented history of failing to properly supervise the State Health Fund contract. In 1994, the Auditor General criticized DMS for failing to audit BC/BS and in 1995, stated that DMS was exercising even less supervision than before. The Auditor General's findings are set forth in Audit Report No.12221 and 12433.
This "hands-off" approach to managing a critical contract disturbs us. If the State were going to let the administrator run itself, then prior State health claims experience was even more important. Contrary to Mr. Lindner's assertion that aggressive implementation by the contract manager was a "gray area" as far as he was concerned, we find it makes perfect sense. Mr. Lindner testified that he relied on his managers for day-to-day operations and that he was comfortable with Mr. Cook's abilities and leadership skills. Lindner said he "didn't feel it was [his] job to second guess at that point because [he] had confidence, both in him and in the Governor, where they -- you know the direction they were moving."
In September, 1996, the Governor's Office of Planning and Budgeting began holding inter-agency management team meetings to handle the Unisys problem. The team met daily for three months. Unisys witnesses testified that these meetings gave them consistent direction for the first time. In spite of these months of daily meetings, Unisys was never able to substantially improve its claims processing performance.
The Legislature opened up new hearings in the matter and accomplished some critical changes. First, procurement and contract management functions were merged again in DSEI's successor agency: the Division of State Group Insurance. The dual agency contract manager should be a thing of the past. The legislation also provides for the Division Director to be appointed by the Governor, confirmed by the Senate and not accountable to the Secretary of DMS.
More importantly, the Legislature appeared to spur the Agencies to action. In order to respond to legislators' concerns, DMS contracted with Mercer and Co., to evaluate Unisys' claims processing abilities and the chances of Unisys being able to reach contract standards consistently. Witnesses testified that Unisys might be able to raise its performance in one category for a month, but seldom for two months and never across the board. When Mercer representatives testified in legislative hearings, the discussion shifted. The seemingly unimaginable had been mentioned: Unisys might never be able to do the job. The State needed to prepare for Unisys' departure. Shortly after Mercer's presentation to the Legislature, the Governor named a Commission and assigned it to find a way out of the mess.
The management of the State Health Fund Contract was the most troubling aspect of this investigation. It was not until the situation deteriorated to an alarming extent, that AHCA and DMS attempted to work together, even though they were required to do so under the terms of the State Health Fund contract and F.S. 110.123. If the managers had behaved otherwise, it is possible that this disaster might have been avoided.
The legislative split of authority in F.S. 110.123 was an unfortunate idea, one that could only have worked with close cooperation between the two Agencies. In its absence, neither Agency truly owned the contract or issues involved in it. The split of authority also unfortunately allowed both Agency heads to put off making the hard but necessary decision to terminate the Unisys contract. Instead, both Agencies and Unisys spent lots of time and lots of money throwing money at the problem, hoping it would go away.
It is unfortunate that DMS was not more pro-active in its treatment of Unisys. If DMS had been involved earlier on, they might have been able to limit the damage. By allowing AHCA to take the lead in helping Unisys during its implementation stage, DMS sowed the seeds of the confusion which allowed Unisys to get away with sub-standard performance. The State should have been speaking with one strong voice.
3. Performance under the State Health Fund contract
Complaints about Unisys' performance have plagued them since the early days of 1996. Witnesses testified that the company was understaffed in January and that positions had to be filled with "temporaries" hired from outside the Tallahassee area. The anticipated implementation period was cut substantially by the BC/BS bid protest. They had 2 ½ months from the administrative hearing officer's ruling in October, 1995, to get up and running by January, 1996. During this period, Unisys had to supplant the entire BC/BS system which had been operating seventeen years, supervise the formation of a Preferred Provider Organization, hire and train employees and staff to handle claims processing, find office space, and create and test its projected computer system. A computer professional hired by Unisys as an independent contractor testified that the computer system was not connected and functioning as of Christmas, 1995.
Another reason Unisys started out slowly is that it had absolutely no experience in setting up or operating a self-insurance fund. Although Unisys has years of experience in claims processing (particularly in Florida, Medicaid claims processing), Florida was to be Unisys' maiden effort in this area. In order to compete for the contract on an even basis with BC/BS, Unisys had to contract with other companies which had already created Preferred Provider Organizations. In order to cover the same geographic area covered by Blue Cross, Unisys had to contract with two such companies. This led to additional complications in processing claims, because two schedules had to be compared every time a claim was processed to determine which was more beneficial to the State. Unisys had a tremendous learning curve to overcome.
Despite the initial reports from Mr. TerLouw, which TerLouw now describes as too optimistic, Mr. Cook testified that he nevertheless still had qualms about Unisys' readiness to proceed. He testified that he felt he had no choice given the administrative hearing officer's ruling on the bid protest. He contacted Blue Cross to see whether they would be willing to assist Unisys in the short term, but Blue Cross was unwilling unless the "stand-in" contract was for a least one year. Unisys would not agree to a one-year contract, and, instead counter-offered with an agreement to waive the RFP provisions which held them harmless for liquidated damages penalties for the first six months of the contract. Thus AHCA/DMS could fine Unisys immediately if they failed to perform. Cook agreed and recommended that DMS, as contract manager, sign the contract. DMS Secretary Lindner signed the contract on December 20, 1995. Mr. Cook says he still had lingering concerns that Unisys could do the job.
Claims processing did not begin until mid-January, 1996, because of difficulties in loading the provider information into the computer system. Unisys' error rates (the percentage of claims payments containing an error) were already high in January and February, when the claims volume was very low. This led DMS to hire the Coopers and Lybrand accounting firm to test the effectiveness of Unisys' claims processing system. The error rates discovered by the April, 1996, Coopers and Lybrand report, (8.5% for processing errors and 5.8% for financial errors) were much higher than industry standards.
At the same time as these shortcomings came to light, anecdotal evidence was being widely reported in the media. Newspapers wrote about State employees whose claims weren't being paid and who faced serious financial difficulty as a result, including collection action and negative credit ratings. Additionally, several former employees of Unisys came forward with allegations of perceived misconduct. The grand jury heard testimony from several of these former employees.
Scott Youngstrand testified that sometime between January and March, 1996, while he was working as a claims adjudicator at Unisys , his supervisor directed him to move mass quantities of claims from one electronic mailbox to another within the claims processing system for the purpose of "renewing" the claims to avoid liquidated damages penalties. He told us that on many occasions, he moved many such claims at one time between mailboxes.
We spoke to many witnesses in order to learn enough about Unisys' "claims software" so that we could judge the credibility of this testimony. Experts in the "claims software" system testified that the system itself does not allow claims to be moved in blocs from one mailbox to another. In order to move several claims (or thousands, as alleged), each claim must be individually opened, manually routed, closed, and a new claim opened by the system.
Experts in the use of Unisys' "claims software", Unisys employees, and DMS employees have all testified that moving claims from one electronic mailbox to another in this system has no effect on the age of the claim within the system. Furthermore, we learned from the testimony of DMS employees that claims still in mailboxes are not subject to the performance penalties under the liquidated damages penalty system. Only claims which were paid or denied are counted in the calculation of performance penalties. Claims remaining in the mailboxes are not considered "paid" or "denied." Penalties are assessed on the percentage of claims that are more than thirty days old when they are paid. The existence of individual claims substantially older than thirty days would have almost no effect on the penalties, and that effect would only occur when the claims are paid or denied, and they leave the mailbox system.
Other former Unisys employees testified that they were instructed to change the "batch and reference number" (B&R #) on hard copies of claims which were already entered into the claims processing system. We learned that before a claim is paid by the "claims software" system, it does not have a unique claim number for tracking purposes. Instead, it is located within the batch of claims within which it was entered by a B&R#, (which is a composite built up out of the date, a code for the type of claim form used, the batch number, and the serial number of the claim within that batch). Former Unisys mail room employees testified that on many occasions, claims were delivered from the processing area and given a B&R# which they did not already have. They thought this was odd because all claims were received and opened in the mail room. Sometimes, the witnesses were instructed to give claims a new B&R# on an opaque tape which was placed over the original B&R#. The dates on the new B&R# were frequently changed to the date it was altered.
We heard testimony from Unisys employees that sometimes when adjudicators dealt with claimants and providers seeking additional information in order to process claims, the additional information was sent back to the adjudicator by name, so it did not go through the original process which would have given it a B&R#. These claims were then delivered back to the mail room to be entered into the system. We heard that sometimes a claim which had received a B&R#, was given an improper form code within it, and it was the normal practice to send such miscoded claims back for alteration with a taped over, corrected B&R#. Witnesses testified that, until earlier this year, there was no procedure in place to ensure that the second B&R# included the original date.
Employees of Unisys and DMS testified that there would be little sinister purpose served by either coding procedure. The claim would not have been renewed in the system, only doubled. Experts in the use of Unisys' "claims software" testified that the original claim would have remained in the system at the same time that the claim with the new B&R# was entered.
We also heard the testimony of another former Unisys employee that he was instructed by his supervisor to manipulate their "correspondence response rate" figures, the data which showed how fast Unisys was responding to inquiries from State employees and providers. He testified that his supervisor told him to selectively count complaints, which were maintained in several different places around the office, excluding those which had not been handled. We learned that FDLE questioned the supervisor about this information and he denied the accusations.
We learned that correspondence response rate information was not subject to a liquidated damages penalty no matter how unresponsive the company looked as a result of the rates. If the figures were indeed "fixed", the sole purpose was to make Unisys look good, not to avoid any money damages.
When DMS assumed management of the State Health Fund contract in April 1996, Jeff Dykes was assigned to determine whether Unisys was performing as required. Mr. Dykes told us that from the beginning he found it very difficult to comprehend what was happening to State employee claims, because the claims processing numbers given to DMS by Unisys were so unreliable. After several months attempting to manage Unisys using their self-reported numbers, Mr. Dykes actually re-created inventory numbers from Unisys' basic "claims received", "claims paid/denied", and "claims returned to sender" numbers. When he finally had an inventory number that he felt he could trust (approximately September, 1996), it disclosed a difference of some 50,000 claims from the inventory numbers he was receiving from Unisys. When he approached Unisys about this difference, Mr. Dykes learned that there were claims in electronic mailboxes which were not typically disclosed to DMS because they were awaiting some form of correspondence or other action by a provider or claimant. Without regard to the propriety or legality of holding back information of this magnitude, it is troubling to think that an administrator of State funds would not feel an obligation to advise DMS regarding the status of all of the State's health claims. We learned that all of the claims "pending" (awaiting adjudication) in the electronic system were located in electronic mailboxes.
In October, 1996, we began to look into the matter of how Unisys treated the claims in these mailboxes. We heard testimony and received evidence from Unisys employees. We subpoenaed documents and computer tapes from Unisys. Because the State does not have an ownership interest in Unisys' claims processing system, or a system similar to it, Unisys was asked to download their back-up tapes in a format the State could read, onto a storage medium compatible with the State's equipment. FDLE retained an expert to interpret the tapes and compared hard-copy claims to the electronic data. This was a time consuming process.
From the analysis, we learned that Unisys segregated claims awaiting provider correspondence into two mailboxes, called P5 and P6. We also learned that they had segregated claims awaiting "pre-existing medical condition" information into a mailbox called PX7. Unisys employees testified that these claims were kept in these mailboxes as a way of keeping track of them.
We learned that Unisys supervisors kept track of the number of claims in these electronic mailboxes by means of a daily report, the Mailbox Inventory Report, which contained the numbers of claims in each mailbox. This Report was first compiled in March, 1996, by a Unisys employee at the request of a Unisys supervisor, Mark Sewell. Mr. Sewell wanted to be able to see at a glance the inventory in all of the mailboxes his employees were assigned to, so he could concentrate their efforts on mailboxes with larger inventories. Mr. Sewell and his division, the claims adjudication section, had no particular interest in mailboxes P5, P6 or PX7, because they were not "active." Unisys adjudicators could not work on them in the absence of the information the claims were waiting on. When the Mailbox Inventory Report was created, the programmer included every mailbox. When the Report was printed out after March, 1996, every mailbox was contained on it.
We have learned that in May 1996, DMS became familiar with the existence of Mr. Sewell's Mailbox Inventory Report. On May 24, DMS representatives asked the Unisys Project Director for access to the Reports. There is no evidence that DMS was specifically seeking information about mailboxes P5, P6 or PX7 or claims awaiting receipt of correspondence. The Mailbox Inventory Reports were not required reports under the contract. According to Unisys witnesses, they were intended to be internal in nature, to allocate manpower resources. We received evidence that on June 2-3, there was E-Mail traffic between Mr. Sewell, his supervisor, and several subordinates to the effect that Mailboxes P5, P6 and P7 should be removed from the Mailbox Inventory Report, because they contained non-active claims for which Mr. Sewell's adjudicators had no responsibility. The mailboxes were then removed from the Report. The claims continued to exist in the system subject to the monthly standard reports provided for in the contract. However, the altered Reports were not routinely sent to the State.
Several weeks later, Unisys assigned a new Deputy Project Manager to the Tallahassee site. When he learned that the Mailbox Inventory Reports did not contain a complete listing of all claims in the electronic database, he directed the programmer to put them back in the Report. After July 19, 1996, Mailbox reports included P5, P6 and PX7 again. Again, the newly complete Mailbox Reports were not routinely sent to the State, either. In September, when Mr. Dykes began to ask questions about missing claims and learned about these mailboxes, he learned that the majority of the claims he could not account for were located in mailboxes P5, P6, and PX7.
On October 9, 1996, Ms. Millie Seay, DSEI Director, wrote a detailed letter to Unisys asking for an explanation of the missing claims. Instead of replying to Ms. Seay's requests, Unisys performed an internal "one-time quality control audit." In this audit, every claim in the claims processing system was individually scrutinized to make sure it was being handled properly. This audit was performed to "ensure the accuracy of provider and enrollee data." Unisys did not indicate that it was performed in response to Ms. Seay's letter. The audit took ten days (during which time no new claims were processed), and as a result, there was a substantial reduction in claims processed during that month. Therefore, in October, the "claims error rates" went down substantially (although still remaining several times the contractually acceptable error rates), but "claims paid" declined substantially, too. At the same time, "claims inventory" rose by over 100,000 claims.
Several outside consultants were hired to determine if Unisys' performance would likely achieve acceptable levels. Unisys retained an outside contractor to assist in managing the State Health Fund. DMS retained Mercer Company, an Atlanta health consulting group, to evaluate Unisys' performance. A representative of Mercer testified that his team's opinion was that it was not likely that Unisys would be able to reach contractually acceptable levels of performance by July 1, 1997. He suggested to the Legislature and the Governor that another vendor be sought on an alternative basis. The new vendor would either subcontract to assist Unisys to achieve acceptable levels or replace Unisys, in the event that Unisys was not in compliance by July.
In the end, Unisys never was able to bring their claims processing performance up to an acceptable standard and maintain that level for a substantial period of time. After seventeen months, Unisys and the State agreed to terminate the State Health Fund claims processing contract. The contract requires Unisys to continue to operate the plan until December 30, 1997, at which time another vendor will take over. The new vendor will be chosen by the newly-created Division of State Group Insurance, headed by a director appointed by the Governor and confirmed by the Senate. The new vendor will not be chosen under the provisions of Chapter 287, but under an emergency executive order by Governor Chiles. We have learned that the selection process for the new vendor has begun, and that the Division has conducted site visits to examine the claims processing capabilities of the three potential vendors.
Based on the testimony and evidence we received, we conclude that Unisys was unprepared to undertake the State Health Fund contract. Given the short lead time to implement the contract, resulting from the Blue Cross bid protest, Unisys was handicapped. But, instead of admitting this when given the opportunity by AHCA to delay taking over claims processing from BC/BS for a year, Unisys insisted on undertaking responsibilities outside their experience, with subtleties and complexities they had not predicted.
What followed was probably foreseeable, and definitely unnecessary. The shortcomings of Unisys' claims processing system would have been immediately evident in a claims test. That test should have been mandated under the bid provisions. From this date forward, every major State contract competition should include a requirement that no system be bought on speculation. The paper-only evaluation during the bid process is as responsible for Unisys' failure as Unisys' own performance.
As a result of our inquiry, we determined that the State should implement the following concepts into its contracting procedures. If these ideas are adopted, we believe the government will be in a better position to prevent similar disasters in the future.
1. Due Diligence Inquiry
During our investigation, we learned that responding vendors were not asked to report any negative character information or conduct which might not have resulted in criminal convictions, but which could have had an impact on the decision to deal with the vendors. Nor did the State contact any third parties, other than those listed as references by the vendor, to determine if there was any negative information on the company. If vendors had been required to produce a due diligence report attesting to their competence, or had the State hired a third party to do so, it is likely that the evaluation team would have learned the same information revealed during a brief search by FDLE. In New Jersey, Massachusetts, and Kentucky, significant questions were raised and reported in the media about Unisys' performance on other Medicaid contracts. Reported problems completing contracts on time and in full compliance concerned us and likely would have caused concern to AHCA, if known. Another way to reach this issue might be to require the vendor to seek a bond to cover expenses to the State if the vendor needs to be replaced. Whichever idea is adopted, the point is to assure the State that it is placing trust in a reputable, competent company.
2. Proven Functional System
Any contract concerning day-to-day operations of crucial State functions should only be awarded to bidders who prove that their proposed system can work. The procuring agency must make it very clear that the cost competition will not be open to any bidders who can not demonstrate the effectiveness of their proposed systems. In 1991, DSEI used a "technical skills benchmark, then cost" process. That was as flawed as the 1995 process, given that it did not require a working system either. But it could serve as a model: rather than accepting a bidder's assurances that it will comply with the requirements, the process should be structured to exclude vendors whose systems are not able to process test claims. AHCA came close, by trying to require such a test under their 1995 RFP. However, by removing important language from the RFP to avoid a bid protest by BC/BS, AHCA in effect left the door open for an inexperienced, unprepared vendor with a plausible bid package to win.
If Unisys' system and "claims software" had been tested prior to January 1, 1996, its inadequacies would have been apparent. AHCA Director Cook could then have recommended against signing the contract with Unisys' with no fear of censure or lawsuit. If only AHCA's 1995 idea had been blended with DMS' idea from 1991, then we would have had a provision requiring that a cost bid would not be accepted unless the vendor demonstrates that the proposed system will work.
If a pre-contract test provision is unworkable, perhaps an escape clause might be fashioned. This could be achieved through the use of a probationary term in the contract. If the vendor is not able to do the job, it could then be dismissed. If a contract probation is unworkable, perhaps the new vendor might be required to set up its system during the last quarter of the incumbent vendor's contract. The two systems could then run parallel. If the new system is not effective, the incumbent vendor might be paid to train the new one.
In its efforts to replace Unisys, the State has approached five different vendors with the experience to put a health claims processing system together. We have also learned that the decision-makers have made actual site visits to view examples of working systems. It is unfortunate this was not done by AHCA prior to its awarding the 1995 State Health Fund contract to Unisys.
3. Proactive Fraud Units
If the way Unisys handled the State Health Fund and Medicaid contracts is any indicator of a common thread, then perhaps the administration of State trust fund dollars is not taken as seriously as it ought to be. From the lax fiscal controls on the Medicaid side to the miserable performance on the State Health Fund side, it is clear that Unisys does not appreciate the magnitude of the trust given it by the State in entering into these contracts.
During this investigation, we learned that Florida law mandates that insurance companies with premium fees earned in Florida in excess of $10 million dollars annually must have internal Special Investigative Units to investigate possible fraudulent claims, must conduct fraud training for employees, and must report the results of the investigations to the Department of Insurance.
We recommend that all trust fund administrators be required to provide and maintain self-funded Special Investigative Units, similar to those required of insurance companies by F.S. 626.9891. The Units would be required by law to report to the State any allegations of criminal wrongdoing or fraud, and to fully assist police agencies in investigating the allegations, providing documents and access to witnesses within the corporation. While Unisys has informed us that they have a "Corporate Ethics Program," this is not sufficient in itself. This is only window-dressing, unless all information developed is freely shared with the police authorities, and responsible employees act as a liaison with police, helping to obtain testimony from witnesses and producing relevant documents. Only when a third party administrator is required to view the taxpayers as the victims of inappropriate or criminal employee conduct, and obligated to act as a good corporate citizen by cooperating with the authorities, will we have any faith that the company is acting within our best interests.
4. Corporate Crimes
Although the evidence we discovered may be sufficient under the law to support civil litigation, Florida law as presently constituted would not support a criminal conviction of the corporation in the absence of proof of corporate managers' knowledge of (or consent to) lower level employees' offenses. Theft or fraud charges require proof of a wilful and intentional act. We have found insufficient evidence that top Unisys managers knew about the lax supervision over the check book or address change system on the Medicaid side or the mailbox report issues on the State Health Fund side.
We do find, however, that when performance problems emerged on the State Health Fund contract, no effective measures were taken to deal with them. Initially, senior management was inattentive and middle management was very inexperienced. This led to inconsistent enforcement of standards concerning employee conduct. While Unisys corporate officials did not direct such offenses to occur, we find that there existed a corporate culture which tolerated the employees' misdeeds. The lack of internal controls at the Medicaid project gave free rein to opportunistic employees. It appears to us also that there was reckless indifference or grossly careless disregard for the financial transactions of State Employees under the State Health Fund. However, Unisys' State Health Fund employees did not themselves benefit from their sharp practices, the corporation did. Unisys was able to keep these contracts even though its employees were mishandling certain documents and financial transactions.
Simple negligence with respect to State funds is not a crime. "Culpable negligence" applies only where personal physical injury is inflicted or threatened (F.S. 784.05). "Culpable negligence" is defined as "recklessness of a gross and flagrant character which evinces a reckless disregard for the safety of others; it is that entire want of care which raises a presumption of indifference to consequences." Killingsworth v. State, 584 So. 2d 647 (Fla. 1st DCA 1991). There is no law in Florida criminalizing culpable negligence with respect to property or economic damage. "Criminal Mischief" (F.S. 806.13), requires proof of a willful and malicious act resulting in damage to real or personal property. Because we have no such proof, there is no other option under the law for the State to seek criminal remedies, such as incarceration of corporate managers or criminal fines against the corporation.
In situations like this, where the State has contracted for services to be performed on behalf of its citizens, the Legislature should consider criminalizing culpable negligence against State property, including financial assets, resulting in economic damage.
Alternatively, the Legislature could adopt a law allowing juries to consider an inference based on "corporate culture evidence" that would replace the need for direct evidence of upper level management knowledge. The law could allow juries to assume that unacceptable levels of employee management and discipline, or serious breaches of common fiscal security controls, constitute a degree of knowledge or intent on the part of the corporation. In other words, if corporate management does not take reasonable steps to properly protect the State's assets entrusted to their care by contract, then criminal theft or fraud charges may lie. Otherwise, in complex and multi-layered corporate structures, corporations will continue to find shelter in the artful defense of ignorance at the top.
Given what we have learned in this investigation about the inadequacies of Unisys' claims processing system, we are pleased that the State has terminated the State Health Fund contract with Unisys. We are also relieved that the State will not automatically renew the contract with Unisys for the Medicaid fund. Finally, we are convinced that the creation of the Division of State Group Insurance will greatly simplify the procurement and management of State Health Trust fund administration contracts.
We learned that the State expected to "save" $8 million dollars a year as a result of the award to Unisys for the State Health Fund contract. If there is anything else to be learned from our investigation, it is that there is truth in the old adage: "You get what you pay for."
THIS PRESENTMENT, NO TRUE BILL, AND RECOMMENDATIONS IS RESPECTFULLY SUBMITTED to the Honorable F. E. Steinmeyer, III, Presiding Judge of the Thirteenth Statewide Grand Jury, this ________ day of June, 1997.
ALVAND R. WILLIAMS
Thirteenth Statewide Grand Jury
I, MELANIE ANN HINES, Statewide Prosecutor and Legal Adviser, Thirteenth Statewide Grand Jury of Florida, and I, JAMES J. SCHNEIDER, Assistant Statewide Prosecutor and Assistant Legal Adviser, Thirteenth Statewide Grand Jury of Florida, hereby certify that we, as authorized and required by law, have advised the Grand Jury which returned this PRESENTMENT, NO TRUE BILL, AND RECOMMENDATIONS, this _______ day of _______, 1997.
MELANIE ANN HINES
Thirteenth Statewide Grand Jury
JAMES J. SCHNEIDER
Assistant Statewide Prosecutor
Assistant Legal Adviser
Thirteenth Statewide Grand Jury
THE FOREGOING Presentment, No True Bill, and Recommendations of the Thirteenth Statewide Grand Jury was returned before me this _____ day of _________, 1997, and is hereby sealed until further order of this Court, upon proper motion of the Statewide Prosecutor.
Further, upon oral motion for disclosure of the Presentment, No True Bill, and Recommendations, the Statewide Grand Jury Legal Adviser is authorized to disclose the testimony and proceedings recounted in the document in furtherance of the criminal investigation and civil responsibilities of the Thirteenth Statewide Grand Jury.
F. E. STEINMEYER, III
Thirteenth Statewide Grand Jury of Florida
1 If errors exceeded 1.5% but were less than 2%, Unisys would forfeit 2% of monthly fees.
If errors exceeded 2% but were less than 2.5%, Unisys would forfeit 3% of monthly fees.
If errors exceeded 2.5%, Unisys would forfeit 4% of monthly fees (no matter how much more than 2.5% errors were observed.) See Section 50.5.6, 1995 State Health Self Insurance Fund RFP.
2 Payment error penalties kicked in at 2% errors (1% of monthly fees) through 3% errors (2% of monthly fees).
Paymt and proc. error penalties kicked in at 5%(combined rate) (1% of monthly fees) through 6% (combined rate)( 2% of monthly fees).
3 Mr. Dykes and other DMS employees also described the performance penalties available to the State under the contract for failing to meet the Claims Processing and Claims Final Action standards. In the event that Unisys failed to process 95% of the claims within 14 days, penalties could be assessed of $5,000 per percentage point for every point over 5% processed late. Additionally, penalties could be assessed for failure to pay 95% of paid claims within 30 days. If more than 5% of paid claims during a thirty-day period were over 30 days old, penalties of $5,000 per percentage point could be assessed for each point over 5% paid within 30 days.