Opinion
CASE: 1:06-cv-0878-SEB-DML.
November 23, 2010
Order on Discovery Motions
This matter is before the court on motions by both the plaintiffs and defendants regarding the defendants' discovery obligations with respect to extracting computer data. The defendants (hereafter referred to collectively as "FSSA") seek a protective order (Dkt. 264) regarding the data extract and the plaintiffs have filed a motion to compel (Dkt. 267).
FSSA is Indiana's Family and Social Services Administration which, among other things, administers the Medicaid program in Indiana.
Background
A detailed background of the nature of the claims in this case and their procedural history is provided here to give context to the parties' discovery dispute regarding the data extract.
This action, originally filed in Hamilton Superior Court and removed to this court on June 5, 2006, challenges FSSA's administration of certain aspects of Indiana's Medicaid program. In April 2008, with the stipulation of the parties, the court certified this action as a Rule 23(b)(2) class action with two classes defined as follows:
A Rule 23(b)(2) class is appropriate where the defendant has "acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole." Fed.R.Civ.P. 23(b)(2).
1. Class 1: All current and future applicants for or recipients of Medicaid with a "spend down" whose income exceeds program eligibility standards.
2. Class 2: All current and future Medicaid recipients who have received or will receive a notice of action to reduce or terminate benefits.See Dkt. 141 (parties' stipulation); Dkt. 144 (court's approval of stipulation and class certification).
The parties also stipulated that the issues to be resolved in this action for each class are:
1. Class 1: (a) Whether FSSA's standard notices (used to deny, reduce or terminate benefits due to excess income) violate Due Process; and (b) whether FSSA's rules and interpretation of FSSA's standard for counting incurred medical expenses violate Plaintiffs' rights under Due Process and federal and state law.
2. Class 2: Whether FSSA routinely violates the rights of beneficiaries to have benefits continued upon appeal of an adverse action under Due Process and federal law.See Dkt. 141 (parties' stipulation); Dkt. 144 (court's approval of stipulation).
Entry of Summary Judgment on Class 1 Claims
On March 31, 2009, the court granted summary judgment, in part, in favor of the "spend-down" Class 1. "Spend-down" refers to a Medicaid applicant's or enrollee's obligation to spend his "excess" income on medical expenses before he is eligible for Medicaid benefits. Class 1 challenged FSSA's process for determining whether an applicant or current enrollee was eligible for, or continued to be eligible for, the spend-down program, and alleged that FSSA's standard notices of negative determinations for spend-down program enrollment, or continued enrollment, were confusing and ambiguous, violating the due process rights of the class. In its summary judgment entry, the court ruled that FSSA's adoption in 2008 of a change to its enrollment policy for the spend-down program rendered moot the claims of some Class 1 plaintiffs.
The 2008 program change eliminated the requirement that an applicant show proof of his ongoing and anticipated medical expenses in order to enroll in the program. This requirement, according to the plaintiffs, had led to systematically incorrect decisions because FSSA's governing rules and instructions to and training of its caseworkers were inconsistent and indiscernible. The new 2008 policy automatically granted enrollment to a person otherwise eligible for Medicaid without requiring the person to prove his anticipated medical expenses. Instead, FSSA's computer system tracks the person's medical bills and correlates them with his spend-down amount to determine the proper payment. In addition, in the course of this litigation and in compliance with court orders, FSSA had reinstated to the spend-down program all persons whose enrollments had been terminated under the pre-2008 policy.
The court agreed with the plaintiffs that the 2008 policy change and the reinstatement of spend-down enrollees who had been terminated under the pre-2008 policy did not resolve the claims of applicants to the spend-down program, whose enrollment was denied under the old policy between the time the complaint was filed and the new policy was implemented. (Dkt. 185 at pp. 12-13). As to these persons, the court granted summary judgment to the plaintiffs, finding that the pre-2008 spend-down enrollment standards, and FSSA's notices to these applicants determining ineligibility for the spend-down program, violated due process. ( Id. at p. 15). The court withheld ruling on the appropriate remedy to these class members pending further discussions among the parties and the court. ( Id. at p. 16). The Court's Summary Judgment Ruling on Class 2 Claims
The court's March 31, 2009 summary judgment entry denied summary judgment on Class 2 issues. As discussed above, Class 2 concerns FSSA's termination of Medicaid benefits during the pendency of an appeal. The plaintiffs contend that FSSA routinely fails to continue benefits pending timely appeals (which violates the law), and that FSSA's notices fail to adequately inform Medicaid recipients about their appeal rights and the procedures for perfecting an appeal. The court's denial of summary judgment on Class 2 claims was based, in large part, on the fact that FSSA was in the midst of a massive overhaul and modernization of its Medicaid system, which was in part specifically geared to correcting the appeals processing system to ensure that appeals are expeditiously determined and benefits continued pending timely appeals. The court was "reluctant to enjoin an ongoing self-corrective process," a process that had rendered the parties' summary judgment briefing — based on the "old" system — out of date. The parties were directed to develop a plan for further proceedings. (Dkt. 185 at p. 17).
Proceedings Since the Summary Judgment Ruling
Since the summary judgment ruling, the parties and the court have focused their attention on the appropriate relief for Class 1, discovery necessary for litigating the scope of relief for Class 1, and that necessary for litigating liability and relief issues for Class 2. Issues surrounding the defendants' extraction of data from FSSA's computer system (known as ICES) have been the source of disputes between the parties since as early as July 2009. ( See, e.g., Dkt. 191 (entry from July 28, 2009 pretrial conference); Dkt. 192 (entry from August 13, 2009 telephone discovery conference)).
On August 20, 2009, FSSA asked the court for a protective order prohibiting or deferring data extract discovery for Class 1 on the grounds that the discovery was individualized and not necessary for determining the nature of appropriate relief for Class 1 or the composition of the members of Class 1. The plaintiffs had served discovery seeking "a complete data extract from the ICES computer system for all members of Class 1 in a form to be specified by Plaintiffs' data analyst," and certain specifically listed data from the ICES system for each Medicaid applicant who was issued a notice of action that contained reason code 376. Reason code 376 was the code number that FSSA used to indicate the denial of applications that did not satisfy the old, pre-2008 spend-down policy.
The plaintiffs' data discovery requests for Class 2 were similar to Class 1, in that the plaintiffs first asked for "a complete data extract" in the form to be specified by the plaintiffs' data analyst, and then asked for specific data with respect to Class 2 members.
The parties disagree whether Class 1 includes those persons whose Medicaid applications had been denied solely for reason code 376, or a broader group of persons whose Medicaid applications had been denied for reason code 376 and other reason codes. In connection with an earlier discovery dispute, FSSA argued that no discovery was necessary to resolve the issue whether Class 1 consists of "only reason code 376" applicants or a broader group of "reason code 376 plus" applicants. The plaintiffs countered that discovery into the "plus" reason codes was necessary to litigating both the scope of the class and the nature of appropriate relief since, for example, that discovery may reveal a pattern or representative sampling showing that the "plus" codes entered by FSSA case workers were "throw-away," or otherwise unreliable, codes added to denials based principally on code 376. By order entered October 21, 2009 (Dkt. 209), this magistrate judge denied FSSA's motion for protective order, ruling that FSSA must provide responses to the plaintiffs' data extract requests for Class 1, that discovery must encompass applicants denied because of reason code 376, whether alone or in conjunction with other reason codes, and that the discovery was relevant to determining the nature of appropriate relief and whether Class 1 consists of "reason code 376 only" applicants or the broader "reason code 376 plus" group.
On August 26, 2009, before the court had resolved the above discovery dispute, the parties filed a joint status report advising that (without prejudice to FSSA's motion for protective order on the Class 1 data extract discovery issue) their respective technology experts had participated in a telephone conference and FSSA agreed to provide a "mocked up" data extract for Class 1 and Class 2 to allow the parties to confirm with each other the data the plaintiffs are seeking, and to allow expert-to-expert communications to facilitate the data process. According to FSSA, on September 1, 2009, it provided to plaintiffs a proposal for four data extracts that would provide the data the plaintiffs had listed specifically in their document requests with respect to Class 1 and Class 2. Several months later, the plaintiffs responded to FSSA's proposal with additional questions, and numerous communications between counsel and the parties' experts regarding the ICES system and how it works followed. On March 11, 2010, the plaintiffs provided to FSSA their proposed data extract.
In the meantime, the plaintiffs filed on November 26, 2009, a motion for preliminary injunctive relief for Class 2 members, seeking an injunction prohibiting FSSA from terminating or reducing Medicaid benefits for all existing Medicaid beneficiaries. The plaintiffs' motion contends that the Medicaid appeals system is so broken that benefit terminations or reductions should be stopped for everyone on a preliminary basis, whether or not an appeal is involved, while FSSA revamps its notice and appeal system. To litigate the plaintiffs' motion for preliminary injunction for Class 2, this magistrate judge set a deadline of February 5, 2010, to complete discovery necessary for the injunction hearing, and a briefing schedule. The magistrate judge also set a conference to be held before Judge Barker and the magistrate judge, a conference that was held on February 22, 2010, and came to be dubbed "the summit." Its purpose was, in part, to determine how the case should be best moved toward its final resolution. A schedule was developed for the parties to negotiate the language and content of appeal notices, to exchange proposals for Class 1 relief, for the plaintiffs to respond to FSSA's data extract proposal, and for the parties to participate in a settlement conference to address Class 1 relief. The hearing on preliminary injunctive relief for Class 2 was vacated and a stay of discovery was ordered. (Dkt. 238).
Following the court's entry setting these deadlines, the plaintiffs moved to compel the defendants to provide the plaintiffs with direct access to the computerized case records of certain members of Class 2 who had administrative appeals pending. This magistrate judge denied the plaintiffs' motion for several reasons, including that the plaintiffs did not show how these records related to Class 2 issues and because their discovery request (made informally) was for use in individual administrative proceedings that were not part of this lawsuit.
In March 2010, the parties reported to the court that they had not reached agreement regarding the data extract process and proposed to set forth their positions to the court. FSSA filed a statement regarding its position regarding the data extract on April 14, 2010 (Dkt. 247), a document that is similar to FSSA's motion for protective order currently before the court. Over the Spring and Summer 2010, the parties exchanged proposals for Class 1 relief, but reached an impasse. On August 26, 2010, the plaintiffs moved the court to lift the discovery stay and set a trial date for both classes for early 2011. On September 13, 2010, the court granted the motion, set a trial date of April 11, 2011, and ordered that the plaintiffs' motion for preliminary injunctive relief on Class 2 would be consolidated with the merits trial to be held April 11. (Dkt. 257, Dkt. 260).
The setting of a trial date pushed to the forefront again the parties' dispute regarding the data extract for Class 1 and Class 2.
The Present Motions
FSSA's motion for a protective order (Dkt. 264) asks the court to order that FSSA's data extract obligation is limited to the data extract outlined in its September 1, 2009 proposal to the plaintiffs. The plaintiffs' original motion to compel (Dkt. 267) asks the court to order FSSA to provide within 30 days the data extract outlined in the plaintiffs' expert's March 2010 proposal. The plaintiffs' reply brief on its motion to compel seeks different relief in light of new information learned during a deposition of FSSA's data expert, Mr. Jim Perez of Deloitte. The plaintiffs no longer are asking FSSA to provide the exact data extract the plaintiffs' expert outlined in the March 2010 proposal. Rather, based on the new information learned from Mr. Perez, the plaintiffs seek relief that would result in a new data extract proposal from the plaintiffs' expert, which would be based on computer file layouts and other documentation about FSSA's computer system that the plaintiffs contend have been withheld from them.
Governing Principles
Rule 26 of the Federal Rules of Civil Procedure affords litigants liberal discovery of any "nonprivileged matter that is relevant to any party's claim or defense." Fed.R.Civ.P. 26(b)(1). "Relevant" is a broad term for discovery purposes, and includes information admissible at trial and anything "reasonably calculated to lead to the discovery of admissible evidence." Id. While the scope of discovery is broad under Rule 26, that rule confers broad powers on the court to regulate or deny discovery even though the materials sought are otherwise within the scope of Rule 26(b)(1). Rule 26(b) provides that the scope of discovery may be "limited by court order," and Rule 26(b)(2)(C) requires the court to limit discovery if the court determines that the burden or expense of the discovery on one party outweighs its likely benefit to the other party, after considering "the needs of the case, the amount in controversy, the parties' resources, the importance of the issues at stake in the action, and the importance of the discovery in resolving the issues." Fed.R.Civ.P. 26(b)(2)(C)(iii).
FSSA's Contentions
FSSA contends that the data extract proposal it provided to the plaintiffs will extract all of the data that the plaintiffs asked for in the two discovery requests in which the plaintiffs specifically listed information desired for Class 1 and Class 2. FSSA argues that although it expected that the plaintiffs might fine-tune FSSA's proposal after the various expert-to-expert communications and other dialogue between the parties over the ensuing months, the plaintiffs' March 2010 data extract proposal so significantly expanded the scope of the requested work that it would take approximately five months to complete the work at a total cost of nearly $100,000. According to FSSA, it has already spent about $16,000 on the original work and could complete the original work in about six weeks and for an additional $20,000. These time and expense figures are based on estimates provided to FSSA by Deloitte, a consulting firm that has the primary responsibility for carrying out all projects relating to ICES. According to FSSA and Deloitte, custom computer programs must be written to generate data out of ICES, and the plaintiffs' data request is significantly more time consuming (and thus more expensive) than FSSA's proposal because:
1. The plaintiffs' extract proposal was based on presentation screens, rather than on how the data is actually stored, and this approach adds to the amount of computer program coding, testing, and development time because of the increased number of data searches required to produce a given extract record.
2. The plaintiffs' extract proposal seeks the extraction of 160 data elements, compared to 31 data elements in FSSA's proposal.
3. The plaintiffs' extract proposal seeks 15 or 16 different extracts, compared to four in FSSA's proposal.
4. Certain historical data requested in the plaintiffs' extract proposal adds to the complexity of the computer programs that must be written to capture the information.
( See Exhibit B to Adams Declaration, Dkt. 264-1).
The Plaintiffs' Contentions
The plaintiffs' opposition to FSSA's motion for protective order is essentially two-fold. First, they contend that the court has already ordered FSSA to provide the plaintiffs' expert with whatever data he wants. The court disagrees. Although the court denied FSSA's motion for a protective order on Class 1 data extract discovery in October 2009 (Dkt. 209), the order did not purport to adjudicate the contents of a data extract (except to determine that the scope must include applicants denied for reason code 376, alone or in combination with other codes). Nor did the court suggest that FSSA's data extract obligations were to be unilaterally determined by the desires of the plaintiffs' expert. The court rejects the plaintiffs' suggestion that if their expert had desired, FSSA was required to produce extracts of every single data element regarding every potential class member in Classes 1 and 2. ( See plaintiffs' response to motion for protective order, Dkt. 266 at p. 3).
Second, the plaintiffs believe that FSSA's estimate (based on Deloitte's estimate) of the time and expense required to prepare data extracts as requested by their expert is exaggerated. Their expert, Mr. Larry Glatz (the Data Project Director for the Center for Medicare Advocacy, Inc., Data Unit, in Harrison, Maine), who is experienced with databases similar to the ICES system and with its COBOL programming language, stated his surprise at the high cost and time estimate provided by Deloitte. He also explained that the data extract he proposed was designed to minimize the necessary programming tasks, but that his design work was made difficult by the lack of documentation available regarding the ICES system. According to Mr. Glatz, he was told by FSSA (or Deloitte) that standard forms of system documentation were not available for the ICES system, such as file layouts. When he asked for the best available proxy, he was provided with 400 pages of "screen shots" and "help texts," which he studied to learn how data elements may be gathered and stored in the ICES system, information he then used in designing his proposal. ( See Affidavit of Larry Glatz, Dkt. 266-2).
Since the filing of their opposition to FSSA's motion for protective order and their own motion to compel, the plaintiffs took the deposition on November 5, 2010, of Jim Perez of Deloitte, who has overall responsibility for the ICES system as part of Deloitte's contract with FSSA. The plaintiffs contend they learned for the first time at Mr. Perez's deposition that file layouts — the very thing that Mr. Glatz wanted, and asked for, so that he could design a data extract proposal consistent with the way data is stored in ICES — are (and were) available. Mr. Perez thought they had been provided to Mr. Glatz. Mr. Perez testified that at least one file layout — related to the notice database and its segments — was given to Mr. Glatz.
Mr. Perez testified that the "majority" or "biggest" reason the plaintiffs' extract would be more time consuming and expensive than FSSA's proposed extract is the increased number of extracts requested (15 or 16 compared to FSSA's four). (Perez Dep. at p. 51; Dkt. 277-3). Mr. Perez stated that another reason for the increased cost was because the plaintiffs' data extract was based on screen shots rather than actual file layouts:
Q. And if I understand you correctly, a major element of the time that it will take to obtain this data [plaintiffs' proposed extract] and the cost is the fact that it is asking you for data based on screen shots rather than on the actual file layouts and the way that the data is stored.
A. That's part. That's part.
Q. Is that a substantial part of the problem?
A. I would have to guess but I would — it's not the biggest part. The biggest —
Q. What would you say is the biggest part?
A. I would say the biggest part is just the increased number of extracts that would have to be created.
(Perez Dep. at p. 26).
From the plaintiffs' perspective, Mr. Perez's testimony establishes that FSSA (or its counsel) withheld information from the plaintiffs and Mr. Glatz that would have saved Mr. Glatz substantial amounts of time in developing an extract proposal, and would have permitted Mr. Glatz to design his proposal in ways compatible with the file layout structure of the ICES computer system. The plaintiffs have told the court they intend to file a motion for sanctions based on this information.
But even if Mr. Glatz's job had been made easier, that does not mean that the plaintiffs' data extract proposal would be substantially less costly and time-consuming for FSSA to produce. Corroborating what FSSA told the court in its motion for protective order, Mr. Perez testified that every one of the plaintiffs' 15 or 16 data extracts requires the writing of a custom computer program, and testing. Although 16 data extracts (plaintiffs' proposal) may not always require four times as much effort (and cost) as four data extracts (FSSA's proposal), FSSA has provided the court with sworn testimony, worthy of credence, that the plaintiffs' extract would take five months and about $100,000, while FSSA's will take another six weeks and cost a total of about $36,000.
Although the plaintiffs indicated in their filings a disbelief that custom programs must be written, suggesting instead that FSSA needed only to make a "raw dump" of data, the plaintiffs have not offered any basis for the court to reject FSSA's affidavits and Mr. Perez's deposition testimony that a custom program must be written for every extract.
Does the Burden and Expense Outweigh the Benefit?
The time and expense burden the plaintiffs want FSSA to bear is great. And although the plaintiffs may desire different data extracts (as explained in their reply brief on their motion to compel) based on additional file-layouts they want FSSA to provide, the plaintiffs have not indicated any intention to reduce the number of extracts they desire.
When the court compares the heavy time and expense to create the data extracts that the plaintiffs originally proposed (and assuming that their new proposal will request a similar number of extracts) with the benefits of that discovery and its importance to the issues to be resolved in this case, the plaintiffs come up short. The plaintiffs have not provided a clear explanation of how the data from FSSA's extracts is insufficient to allow the plaintiffs to present evidence of the proper scope of Class 1 and/or the proper injunctive relief for Class 1. Simply asserting that their expert would like to have it is not enough. FSSA's extracts for Class 1 provide the following information for each Medicaid applicant who was denied spend-down program eligibility, either alone or in combination with other denial reasons: name, address, telephone number and ID number; case number, date of application, date of denial, reason codes for denial, identifying information for any assistance group for the recipient, spend-down information, any appeals receipt, and appeals decision.
With respect to Class 2, the plaintiffs also do not explain how or why FSSA's proposed extracts provide too little information for proving entitlement to relief or the appropriate scope of relief. FSSA's proposed extracts will identify every Medicaid recipient whose case file was closed or who suffered a reduction in benefits from January 2006 forward. For each person, the information about him will include: name, address, telephone number and ID number; case number, date of application, date of negative action and reasons codes associated therewith, whether closure action was taken or benefits were reduced, identifying information for any assistance group for the recipient, any appeals receipt, and appeals decision. With this information (and other discovery), the plaintiffs should be able to present evidence whether FSSA continues to terminate benefits improperly pending timely appeals and what the scope of the problem is.
The court notes that FSSA apparently prepares monthly reports, and provides them to the plaintiffs, regarding the rate that Medicaid benefits are terminated in the context of appeals. See Exhibit to Plaintiffs' Motion for Status Conference, Dkt. 251-1, filed May 21, 2010.
The plaintiffs discuss only two sets of information within their expert's data extract proposals that are not included in FSSA's proposal: historical data and FIAT information. But the plaintiffs have not shown that having the information will help to resolve the issues in this litigation. With respect to historical information, the plaintiffs contend that a Class 2 member may have wrongfully suffered a termination in benefits more than once during the four-year history of this litigation, and that FSSA's data extract will show information only about the most recent termination. That argument focuses on individualized relief rather than a class-wide determination that injunctive relief is necessary to address a systemic problem at FSSA in terminating benefits when a timely appeal has been filed. The magistrate judge notes that the court previously denied summary judgment for Class 2 to allow FSSA to fix its handling (or mishandling) of timely appeals. The most current information for Medicaid recipients is sufficient evidence for the plaintiffs to use in their attempt to demonstrate whether FSSA is still mishandling timely appeals and whether the scope of the problem demands a court-imposed injunction.
The second type of information discussed by the plaintiffs is "FIAT" information. According to the plaintiffs, FSSA case workers use a so-called FIAT procedure to restore recipients' benefits when benefits were not maintained after the filing of a timely appeal "as well as other ad-hoc eligibility determinations." That explanation does not address the significance of that information as class-wide proof of liability and appropriate relief. Indeed, the apparent "ad-hoc" nature of the FIAT information indicates that the information would not assist in the resolution of class-wide issues. In any event, the value of the information does not justify the cost and burden of providing it.
The magistrate judge has already noted in this case that as a Rule 23(b)(2) class action, this case must focus on proof that FSSA acted (or did not act) on grounds that apply generally to the class so that injunctive relief is appropriate respecting the class as a whole. Data extracts can be a source for identifying common problems and common treatment, and assist in identifying files for sampling, but the purpose of data extracts is not to litigate individual cases or to determine the relief that particular individuals may be accorded if class-wide relief is granted.
The Plaintiffs' Revised Request for Relief on Their Motion to Compel
The plaintiffs' reply brief on their motion to compel states that they no longer are requesting that the court order FSSA to provide in 30 days the data extracts described in the March 2010 proposal. Instead, they want time for their expert to design new data extracts after being provided with file layouts.
The plaintiffs' description in their reply brief of FSSA's withholding of file layout information does not convince the court that the burdens and benefits of data extracts should be measured dramatically differently. The opportunity for the plaintiffs' expert to review file layout information does not seem likely to reduce the scope of extracts that the plaintiffs will want or the amount of data the plaintiffs will want, or to significantly reduce the number of computer programs that would need to be written and tested to create extracts for the plaintiffs. The court therefore denies the plaintiffs' request that FSSA make available all file layouts for the purpose of allowing the plaintiffs' expert to revise his requested data extracts. However, the plaintiffs' reply brief (Dkt. 277) raises serious issues surrounding the production, or withholding, of file layouts requested by the plaintiffs' expert. The court notes the plaintiffs' statement that they intend to file a motion for sanctions for that conduct and will address that motion, if ultimately filed, after briefing by the parties.
Conclusion
FSSA's motion for a protective order (Dkt. 264) is GRANTED. The plaintiffs' motion to compel (Dkt. 267 and Dkt. 277) is DENIED.
FSSA immediately shall proceed in continuing the preparation of the extracts described in its September 1, 2009 proposal and to make its production by January 5, 2011.
So ORDERED.