Arizona Health Care Cost Containment System v. Centers For Medicare and Medicaid Services

• Court: District Court, D. Arizona
• Decided: February 18, 2020
• Docket: 2:17-cv-04462
• Status: Unknown

Opinion

1 WO 2 3 4 5 6 IN THE UNITED STATES DISTRICT COURT 7 FOR THE DISTRICT OF ARIZONA

9 Arizona Health Care Cost Containment No. CV-17-04462-PHX-DJH System, 10 ORDER Plaintiff, 11 v. 12 Centers For Medicare and Medicaid 13 Services, et al.,

14 Defendants. 15 16 This matter is before the Court on an appeal by Plaintiff, Arizona Health Care Cost 17 Containment System (“AHCCCS”), against Defendants, the United States Department of 18 Health and Human Services, by and through its Centers for Medicare and Medicaid 19 Services, and the Secretary of the United States Department of Health and Human Services, 20 in his official capacity (collectively “HHS”). AHCCCS is appealing from a decision by 21 the Health and Human Services Departmental Appeals Board disallowing $11,716,850 in 22 federal financial participation for claimed school-based administrative costs. (Doc. 1). 23 AHCCCS filed an Opening Brief (Doc. 30), HHS submitted an Answering Brief 24 (Doc. 34), and AHCCCS filed a Reply Brief (Doc. 35). 25 I. Background 26 Under the Social Security Act, states are eligible to receive federal reimbursement, 27 referred to as “federal financial participation,” for school-based administrative activities 28 that support Medicaid-eligible school children under the Individuals with Disabilities 1 Education Act. (Doc. 17-3 at 1-3). For the period between January 2004 and September 2 2008, AHCCCS claimed $30,584,822 as federal financial participation for school-based 3 administrative activities (Doc. 17-3 at 4); this amount was originally fully paid to AHCCCS 4 (Doc. 30 at 8). 5 In October 2008, the Health and Human Services Office of the Inspector General 6 informed AHCCCS that it was conducting an audit of Arizona’s contingency fee payment 7 arrangements with consultants for claiming school-based administrative costs; in March 8 2011, the objective of the audit was “revised” to focus on AHCCCS’s actual claims for 9 school-based administrative costs. (Doc. 17-5 at 109-110; Doc. 17-5 at 143-144). 10 Following these audits, the Office of the Inspector General disallowed $11,716,850 in 11 federal funding for claimed school-based administrative costs. (Doc. 17-5 at 152). It 12 disallowed $6,295,139 of this amount based upon AHCCCS’s data-collection method for 13 school-based administrative costs. (Doc. 17-5 at 152, 163-166). It disallowed the 14 remaining $5,421,711 because it found that AHCCCS failed to maintain all needed claim 15 substantiation documents for the first quarter of 2004 and the second quarter of 2005. (Doc. 16 17-5 at 152, 161-163). AHCCCS appealed this decision to the Department of Health and 17 Human Services Departmental Appeals Board (the “Appeals Board”), which affirmed. 18 (Doc. 17-3 at 1-26). 19 This Order will first discuss relevant context, including the applicable terms of the 20 2003 Claiming Guide, which provided guidance to states regarding submitting claims for 21 federal financial participation of school-based administrative costs; AHCCCS’s approach 22 to documenting the amounts requested for federal financial participation of school-based 23 administrative costs; the audits of AHCCCS’s claims; and the administrative review 24 process in this matter. This Order will next address whether the Appeals Board properly 25 disallowed $6,295,139 based upon AHCCCS’s data-collection methodology. Finally, this 26 Order will address whether the Appeals Board properly disallowed $5,421,711 for 27 AHCCCS’s failure to fully substantiate its claims during the 2011 audit. 28 … 1 A. The Claiming Guide 2 In a May 2003 publication, the Centers for Medicare and Medicaid Services 3 (“CMS”) issued a Medicaid School-Based Administrative Claiming Guide (the “Claiming 4 Guide”) (Doc. 17-5 at 1-61). The Claiming Guide outlines acceptable methods for 5 accurately assessing time spent on administrative activities. (Doc. 17-5 at 44). It noted 6 that “one of the most commonly used sampling methodologies for time studies is random 7 moment sampling” (“RMS”) and that the “RMS method represents an acceptable method 8 for accurately assessing the time spent on administrative activities.” (Doc. 17-5 at 44). 9 The Claiming Guide acknowledged that there was some flexibility in the sampling 10 methodology, but stated that the methodology must remain statistically valid and that the 11 methodology must be acceptable to CMS:

12 Flexibility is afforded within the bounds of statistical validity. However, the validity and reliability of the sampling methodology must be acceptable to 13 CMS. That is, the state must include details of how its time study methodology will be validated. 14 15 (Doc. 17-5 at 45). 16 The Claiming Guide next addressed the use of over-sampling and non-responses in 17 the time study methodology. (Doc. 17-5 at 45). Although it recognized that oversampled 18 responses are sometimes substituted for responses that were not received, it cautioned 19 against substituting oversampled responses for completed responses when there were few 20 reported Medicaid activities:

21 To ensure an adequate number of responses, many schools oversample and/or factor in a non-response rate in their time study methodology. Under 22 this methodology, oversampled responses are sometimes substituted for responses not received. However, oversampled responses should not be 23 substituted for completed responses in which there are no or few reported Medicaid activities in order to increase the Medicaid reimbursable portion of 24 the claim. No completed responses should be deleted or ignored. 25 (Doc. 17-5 at 45). 26 Finally, with respect to claiming methodology, the Claiming Guide indicated that it 27 is potentially problematic for employees who do not perform many Medicaid activities to 28 fail to complete the time-study; in order to avoid such problems, non-responses must be 1 coded to non-Medicaid time study codes:

2 Another potential problem is employees who are instructed to not complete the time study if they typically do not perform many Medicaid activities. To 3 avoid this, all non-responses should be coded to non-Medicaid time study codes. 4 (Doc. 17-5 at 45 (emphasis added)). 5 B. Arizona’s Approach 6 In a January 2004 document entitled, “Medicaid Administrative Claiming Program 7 Guide; DRAFT – Pending CMS Approval,” AHCCCS described its proposed methodology 8 for collecting the sampling to establish the amount of time spent on administrative 9 activities. (Doc. 17-5 at 63-73). The proposed methodology stated that forms would be 10 marked invalid if they contained missing or inaccurate information or if the form was not 11 approved; invalid forms would then be removed from the sample pool of observation 12 forms: 13 Forms that cannot be validated, due to missing or inaccurate information, or 14 failure [to] return the updated form will be marked invalid. Once all invalid forms have been extracted from the sample pool of observation forms, all 15 valid forms are included in the tabulation. 16 (Doc. 30 at 12; Doc. 17-5 at 72-73). 17 On March 23, 2004, via e-mail, AHCCCS submitted its proposed claiming plan for 18 the Medicaid School-Based Program for CMS’s review and approval. (Doc. 17-5 at 100). 19 The transmittal e-mail stated that “[t]he guide has been prepared in accordance with the 20 CMS May 2003 Guide and in response to AHCCCS’[s] change of contractor.” (Doc. 17- 21 5 at 100). Apparently not hearing back from CMS regarding the proposed plan, AHCCCS 22 e-mailed CMS in November 2004; AHCCCS asked whether there was “any information 23 regarding the status of the AHCCCS claiming guide review/approval process.” (Doc. 17- 24 5 at 99). In response, Kenneth Adams of CMS indicated possible, but not official, approval 25 of the plan, based upon AHCCCS’s representation that the plan was consistent with the 26 Claiming Guide:

27 I forwarded Arizona’s School-Based Admin claiming plan to CO1 right after I received it from you on March 23, 2004. I have not heard from CO and 28 1 The record is unclear regarding the identity of “CO.” 1 confess I also have not had an opportunity to review it closely – but other staff indicated it looks good, but they have a few questions; 2 Until such time as we formally approve AHCCCS’[s] plan we are relying on 3 your statement in the transmittal e-mail that says AHCCCS’s program was prepared in accordance with the May 2003 CMS guide. 4 5 (Doc. 17-5 at 99). CMS never approved or disallowed AHCCCS’s 2004 proposed plan, 6 which remained in place until AHCCCS modified the plan in May 2008. (Doc. 17-5 at 7 107). 8 AHCCCS made the May 2008 modification in response to an April 22, 2008, 9 recommendation by its contractor, Maximus. In its recommendation, Maximus noted that, 10 although the Claiming Guide was “arguably . . . contrary” to “other federal guidance” 11 regarding non-responses, “AHCCCS treats ‘non-responses’ differently than the CMS 12 Guide requires.’” (Doc. 17-5 at 103, 107). 13 C. AHCCCS Claims Audits–Sampling Errors and Unretained Documents 14 On October 20, 2008, the Health and Human Services Office of the Inspector 15 General (the “OIG”) notified AHCCCS that it intended to audit AHCCCS’s contingency 16 fee agreements with consultants for claiming school-based administrative costs (the “2008 17 audit”). (Doc. 17-5 at 109-110). The specific stated objectives were to:

18 determine the (1) extent to which the Arizona Health Care Cost Containment System (AHCCCS) has contracted with consultants through contingency fee 19 payment arrangements and (2) impact of these arrangements on the submission of improper claims to the Federal Government. 20 (Doc. 17-5 at 109). The audit period was from January 1, 2004, through June 30, 2008. 21 (Doc. 17-5 at 109). An attachment to the notification letter detailed the documents 22 requested. (Doc. 17-5 at 111-112). Examples of the requested documents include 23 AHCCCS policies regarding school-based administrative claims and copies of contracts 24 between AHCCCS and Maximus. (Doc. 17-5 at 111-112). The 2008 audit also requested 25 “Quarterly amount claimed by AHCCCS for Federal financial participation for Medicaid 26 school-based administrative expenditures (in an Excel file).” (Doc. 17-5 at 112). The 2008 27 audit notification letter further stated that, “we will also need access to additional 28 documents and records”; the letter did not specify which additional documents and records 1 would be needed. (Doc. 17-5 at 109). 2 The OIG issued a second intent to audit notification letter to AHCCCS on March 3 11, 2011 (the “2011 audit”). (Doc. 17-5 at 143-144). The “purpose” of that letter was to 4 notify AHCCCS of the OIG’s “intention to conduct an audit of Medicaid school-based 5 administrative costs claimed by the State of Arizona.” (Doc. 17-5 at 143). The stated 6 objective was to “determine whether [AHCCCS] claimed Medicaid administrative costs 7 for the school-based program in accordance with Federal regulations and guidance.” (Doc. 8 17-5 at 143). The letter then linked the school-based administrative costs audit to the 9 October 2008 audit of AHCCCS’s contingency fee agreements with consultants for 10 claiming school-based administrative costs:

11 We initially started this review as part of a nationwide survey of Medicaid contingency fee payment arrangements. . . . After discussion with Centers 12 for Medicare & Medicaid Services officials and OIG management, we have revised our objective to focus on AHCCCS’[s] claiming of Medicaid 13 administrative costs for the school-based programs. 14 (Doc. 17-5 at 143). 15 In a January 2013 Report, the OIG disallowed $11,716,850 from the $30,545,822 16 that AHCCCS claimed as the Federal share of school-based administrative costs. (Doc. 17 17-5 at 152). The OIG found that $6,295,139 of this amount was unallowable based upon 18 AHCCCS’s Random Moment Time Sampling method. (Doc. 17-5 at 152). The OIG 19 determined that AHCCCS “inappropriately discarded sample items when calculating the 20 statewide Medicaid percentages.” (Doc. 17-5 at 152). The OIG reasoned that AHCCCS 21 reduced the sample size when it discarded items; the reduced sample size resulted in higher 22 Medicaid percentages, which therefore increased the amount of Federal reimbursement. 23 (Doc. 17-5 at 164). 24 The OIG disallowed the remaining $5,421,711 because, for two out of the nineteen 25 quarters considered—the first quarter of 2004 and the second quarter of 2005—AHCCCS 26 did not maintain documentation to support “(1) the universes of total available moments in 27 time and RMTS participants and/or (2) the sample of random moments for selected 28 participants.” (Doc. 17-5 at 152). In reaching this conclusion, the OIG stated that 1 AHCCCS provided copies of completed observation forms,2 but “did not provide data files 2 to support the sample universe determination and the sample selection.” (Doc. 17-5 at 3 162). The OIG stated that, without the files, it could not determine whether the provided 4 observation forms were for the sample items selected for those two quarters. (Doc. 17-5 5 at 163). As a result, the OIG disallowed all claimed reimbursement for the first quarter of 6 2004 and the second quarter of 2005. (Doc. 17-5 at 161-163). 7 D. The Administrative Review Process 8 On March 4, 2013, AHCCCS responded to the OIG report, disagreeing with the 9 recommendation that AHCCCS refund $11,716,850 in federal reimbursement for school- 10 based administrative costs. (Doc. 17-5 at 179-180). By letter dated October 20, 2016, 11 HHS formally disallowed the $11,716,850 in claimed federal financial participation for 12 school-based administrative costs. (Doc. 17-5 at 182-187). AHCCCS submitted a Request 13 for Reconsideration to the Secretary of the Department of Health and Human Services on 14 December 14, 2016. (Doc. 17-5 at 189-201). In its Request for Reconsideration, 15 AHCCCS: (1) argued that the disallowance for purported methodological issues was 16 “unfounded and unfair” because it was based upon AHCCCS’s lack of explicit 17 methodology (Doc. 17-5 at 194), and (2) acknowledged that Maximus “was unable to 18 provide the supporting documentation” for the two quarters at issue, but argued that 19 disallowing the entire amount claimed was “unreasonable and unduly punitive” (Doc. 17- 20 5 at 193). On February 14, 2017, CMS issued its decision on AHCCCS’s request for 21 reconsideration. (Doc. 17-5 at 204-205). CMS affirmed the disallowance. (Id.). 22 AHCCCS filed a Notice of Appeal with the Department of Health and Human 23 Services on April 3, 2017. (Doc. 17-5 at 207-208). On October 2, 2017, the Appeals Board 24 issued its decision. (Doc. 17-3 at 1 – 26). The Board upheld the $11,716,850 disallowance. 25 (Doc. 17-3 at 1). 26 AHCCCS now asks this Court to reverse the Departmental Appeals Board decision 27 2 For the first quarter of 2004, AHCCCS provided copies of 3,559 filled-out observation 28 forms and, for the second quarter of 2005, AHCCCS provided copies of 3,730 filled-out observation forms. (Doc. 17-5 at 162-163). 1 and set aside the disallowances that were upheld by that decision. (Doc. 1). Based upon 2 the parties’ Stipulation (Doc. 10), this Court ordered that, as an administrative appeal, this 3 case is excluded from the Mandatory Initial Discovery Pilot and that no discovery would 4 be conducted. (Doc. 11). AHCCCS filed an Opening Brief (Doc. 30), HHS filed an 5 Answering Brief (Doc. 34), and AHCCCS filed a Reply Brief (Doc. 35) 6 II. Discussion 7 A. Standard of Review 8 This Court reviews the HHS Departmental Appeals Board decision under the 9 Administrative Procedure Act (the “APA”), 5 U.S.C. § 551 et seq. Under the APA, the 10 reviewing court shall “hold unlawful and set aside agency action, findings, and conclusions 11 found to be arbitrary, capricious, an abuse of discretion, or otherwise not in accordance 12 with law.” 5 U.S.C. § 706(2)(A). The agency’s decision is “presumptively valid; the 13 plaintiff bears the burden of showing otherwise.” Texas Tech Physicians Assocs. v. United 14 States Dep’t of Health and Human Services, 917 F.3d 837, 844 (5th Cir. 2019). 15 The Court must defer to agency regulations that present a reasonable interpretation 16 of an ambiguous statute. Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 17 467 U.S. 837, 842-44 (1984). An agency’s interpretation of its own regulation is similarly 18 entitled to deference. Auer v. Robbins, 519 U.S. 452, 461 (1997). However, agency 19 interpretations that “lack the force of law,” such as those in opinion letters, policy 20 statements, agency manuals, and enforcement guidelines, “do not warrant Chevron-style 21 deference.” Christensen v. Harris County, 529 U.S. 576, 587 (2000). Such interpretations 22 are “‘entitled to respect’ . . ., but only to the extent that those interpretations have the ‘power 23 to persuade.’” Id. (internal citations omitted). 24 B. Disallowance Based Upon Reporting Method 25 The Appeals Board upheld the $6,295,139 sampling methodology disallowance 26 because “CMS did not unreasonably interpret the [Claiming Guide] after the fact” and 27 because “Arizona discarded responses when the [Claiming Guide] then in effect did not 28 1 specifically permit such action and without CMS approval.”3 (Doc. 17-3 at 16). AHCCCS 2 argues that this Court should reverse this disallowance because the disallowance rationale 3 was based on an unannounced and subjective standard that CMS has not always applied; 4 because CMS failed to provide a reasoned basis for the disallowance; and because 5 AHCCCS should have been permitted to rely on its proposed methodology in light of 6 CMS’s failure to notify AHCCCS that the methodology was not acceptable. (Doc. 30 at 7 12-25). In response, HHS states that this Court should uphold the disallowance because 8 the Claiming Guide required AHCCCS to include non-responses in its sample and because 9 CMS never approved the sampling methodology. (Doc. 34 at 8-12). As explained below, 10 the Court affirms this disallowance because, based upon the administrative record, the 11 Appeals Board decision was not arbitrary, capricious, an abuse of discretion, or otherwise 12 not in accordance with the law. 13 The Claiming Guide was provided to state agencies and schools to explain the 14 acceptable process and methodology for claiming federal reimbursement for school-based 15 administrative costs. Specifically, the Claiming Guide stated that its purpose was to 16 “inform schools, state Medicaid agencies, and other interested parties on the appropriate 17 methods for claiming federal reimbursement for the costs of Medicaid administrative 18 activities performed in the school setting.” (Doc. 17-5 at 5). The Claiming Guide further 19 noted that “[s]tate Medicaid Agencies are responsible for ensuring . . . that claims are 20 submitted to CMS in conformance with such requirements.” (Doc. 17-5 at 6). Finally, the 21 Claiming Guide stated that it “does not supersede any statutory or regulatory 22 requirements.” (Doc. 17-5 at 6). Instead, “it clarifies and consolidates CMS’[s] guidance 23 on how to meet these statutory and regulatory requirements and explains the application of 24 such requirements in the context of current practices.” (Doc. 17-5 at 6). 25 Regarding capturing the time for administrative Medicaid costs, the Claiming Guide 26 required states to “develop an allocation methodology that is approved” by HHS. (Doc.

27 3 The Appeals Board also affirmed the $6,295,139 sampling methodology disallowance because it found that Arizona failed to demonstrate that it actually interpreted the Claiming 28 Guide language to mean that it could exclude non-responses without obtaining approval from CMS (Doc. 17-3 at 12-15). 1 17-5 at 12). Different methodologies could be used, including random moment sampling, 2 contemporaneous time sheets, or other “quantifiable measures of employee effort.” (Doc. 3 17-5 at 12). Regardless of the time study method used, the study had to reflect “all of the 4 time and activities (whether allowable or unallowable under Medicaid) performed by 5 employees participating in the Medicaid administrative claiming program.” (Doc. 17-5 at 6 12 (emphasis in original)). 7 Although the Claiming Guide provided for some “flexibility” in terms of claiming 8 methodology, there were express limitations on that flexibility: first, the sampling 9 methodology had to be “within the bounds of statistical validity”; second, “the validity and 10 reliability of the sampling methodology must be acceptable to CMS.” (Doc. 17-5 at 45). 11 Other limitations included that “[n]o completed responses should be deleted or ignored” 12 and that “all non-responses should be coded to non-Medicaid time study codes.” (Doc. 17- 13 5 at 45). 14 AHCCCS takes issue with this final limitation requiring all non-responses to be 15 coded to non-Medicaid time study codes. It argues that it rightfully believed that it was 16 acceptable to exclude non-responses based upon the language of the Claiming Guide, as 17 well as past practices, and that its belief was justifiable because it timely communicated its 18 plan to CMS, which never suggested that the plan was unacceptable. (Doc. 30 at 21). The 19 Appeals Board rejected this argument, reasoning that AHCCCS’s proposed plan did not 20 conform to the Claiming Guide because its proposal excluded “certain types of responses 21 considered invalid” and “essentially sought approval to deviate from the general rule of 22 including all responses.” (Doc. 17-3 at 17). The Appeals Board further found that the 23 Claiming Guide could be interpreted to permit the exclusion of non-responses, but only 24 “with an approved protocol to support it.” (Doc. 17-3 at 16 (emphasis added)). Because 25 CMS never approved the plan and because the plan excluded certain non-responses, the 26 Appeals Board concluded that the claims were “properly disallowed for improperly 27 excluding non-responses or incomplete or inaccurate responses.” (Doc. 17-3 at 19). This 28 Court agrees. 1 First, when looking at the language of the Claiming Guide, AHCCCS’s program 2 was not prepared in accordance with the Guide. Under the Claiming Guide, the default for 3 incomplete forms was to include the form as non-Medicaid: “Another potential problem is 4 employees who are instructed to not complete the time study if they typically do not 5 perform many Medicaid activities. To avoid this, all non-responses should be coded to 6 non-Medicaid time study codes.” (Doc. 17-5 at 45 (emphasis added)). Under AHCCCS’s 7 plan, however, if there was a problem with a form, the default was to exclude the form 8 from consideration. The plan stated that forms would be marked invalid if there was 9 missing information, inaccurate information, or a failure to return the form; these “invalid 10 forms” would then be “extracted from the sample pool of observation forms.” (Doc. 17-5 11 at 72-73). In other words, if a form was not returned, was missing information, or contained 12 inaccurate information, that form would be excluded from the tabulation. After those 13 invalid forms were excluded, the remaining valid forms would be used for the tabulation. 14 (Doc. 17-5 at 72-73). Therefore, AHCCCS’s methodology of excluding invalid forms 15 contradicts the Claiming Guide’s express language. 16 Second, CMS did not approve AHCCCS’s proposed plan. AHCCCS correctly 17 states that it followed appropriate procedures by presenting CMS with the proposed plan. 18 AHCCCS similarly correctly states that CMS never disapproved the plan based on the way 19 the plan handled non-responses; lack of disapproval, however, is not approval. In the 20 communication from Kenneth Adams of CMS to AHCCCS, Mr. Adams acknowledged 21 that he had forwarded the claiming plan once he received it, that he had “not heard” back 22 about the plan, and that he has “not had an opportunity to review it closely.” (Doc. 17-5 at 23 99). While he did say that “other staff indicated it looks good,” he tempered this statement 24 with the caveat that “they have a few questions.” (Doc. 17-5 at 99). Notably, he indicated 25 that, until CMS formally approves the plan, CMS is “relying on your statement in the 26 transmittal e-mail that says AHCCCS’s program was prepared in accordance with the May 27 2003 CMS guide.” (Doc. 17-5 at 99). As discussed above, however, AHCCCS’s plan 28 directly contradicted the express language of the Claiming Guide; therefore, contrary to 1 AHCCCS’s representation in its transmittal email to Mr. Adams, the plan could not have 2 been “prepared in accordance” with the Claiming Guide. 3 Further, while the reasons for CMS’s delay in either approving or disapproving the 4 proposed plan are unclear,4 CMS’s silence cannot be considered a concession that the plan 5 was acceptable because the Claiming Guide stated that CMS approval was required: 6 “Flexibility is afforded within the bounds of statistical validity. However, the validity and 7 reliability of the sampling methodology must be acceptable to CMS.” (Doc. 17-5 at 45). 8 Based on this language, the Appeals Board correctly found that AHCCCS was not entitled 9 to interpret CMS’s silence as approval of its proposed plan. 10 Finally, AHCCCS suggests that it nonetheless properly used its claiming 11 methodology because some approved state methodologies, including AHCCCS’s 2010 12 plan, allow non-responses to be discarded. (Doc. 30 at 14-16). As recognized by the 13 Appeals Board, the problem with this argument is that “the parallel goals of statistical 14 validity and claim integrity” only allow exclusion of non-responses when there is “an 15 approved protocol.” (Doc. 17-3 at 16). Arizona’s revised, approved protocol is much more 16 detailed regarding non-responses than the proposed 2004 plan and therefore, as found by 17 the Appeals Board, “differs materially from the draft 2004 plan, to ensure a valid sample 18 size.” 5 (Doc. 17-3 at 18). Accordingly, because AHCCCS’s proposed plan treated non- 19 responses differently from the Claiming Guide instructions, because CMS approval was 20 required to ensure statistical validity, and because AHCCCS implemented the plan without 21 obtaining CMS approval, this Court affirms the Appeals Board’s decision disallowing 22 $6,295,139 based upon Arizona’s Random Moment Time Sampling method.

23 4 The Appeals Board suggests that “[s]ince CMS never approved Arizona’s 2004 plan, as drafted, it is reasonable to infer that CMS had concerns or reservations” about the 2004 24 plan. (Doc. 17-3 at 18 (emphasis in original)). Further, the Appeals Board found that, CMS’s approval of Arizona’s modified plan “strongly suggests that CMS had concerns 25 about a matter relevant to statistical validity that the 2004 plan did not adequately or specifically address.” (Doc. 17-3 at 18). 26 5 The 2010 plan states “[Arizona] will require an 85% return rate. Non-responsive 27 moments, moments not returned or not accurately completed and subsequently resubmitted . . . will not be included in the results unless the return rate for valid moments is less than 28 85%. If the return rate of valid moments is less than 85%, then, non-returned moments will be included and coded as a non-allowable.” (Doc. 17-3 at 18). 1 C. Disallowance Based Upon Failure to Retain Documentation 2 The Appeals Board also upheld a $5,421,711 disallowance based upon AHCCCS’s 3 failure to retain supporting claim documentation from the first quarter of 2004 and the 4 second quarter of 2005. It reasoned that AHCCCS improperly failed to maintain 5 documentation to support its claims for those quarters and that, without the documentation, 6 AHCCCS could not substantiate any of its claims for those quarters. (Doc. 17-3 at 23-25). 7 AHCCCS asserts this Court should reverse that decision because, although underlying 8 documents were admittedly missing at the time of the 2011 audit, AHCCCS was not 9 required to retain those documents until the 2011 audit; therefore, according to AHCCCS, 10 the Appeals Board decision is arbitrary and capricious and should be reversed. (Doc. 30 11 at 27-29). HHS argues that the disallowance was rational because AHCCCS was obligated 12 to substantiate its claims regardless of any document retention guidelines and because, in 13 any event, AHCCCS had sufficient notice that its claims were being challenged and 14 therefore should have retained the records. (Doc. 34 at 12-14). As explained below, the 15 Court affirms this disallowance because, based upon the administrative record, AHCCCS 16 should have retained the needed supporting documentation until all audits were resolved; 17 because AHCCCS failed to retain the documents and could not substantiate its claims for 18 those two quarters, the Appeals Board decision was not arbitrary, capricious, an abuse of 19 discretion, or otherwise not in accordance with the law. 20 In general, claim documents must be retained for a minimum of three years; if there 21 is an audit of a claim, this period is extended until the audit is resolved. 42 C.F.R. § 433.32; 22 45 C.F.R. § 75.361. The Code of Federal Regulations, regarding State Fiscal 23 Administration and Federal Matching and General Administration Provisions, provides 24 that state Medicaid agencies, as well as local agencies administering the plan, must 25 “[r]etain records for 3 years from date of submission of a final expenditure report.” 42 26 C.F.R. § 433.32(b). Likewise, regarding Post Federal Award Requirements, the Code of 27 Federal Regulations states that “[f]inancial records, supporting documents, statistical 28 records, and all other non-Federal entity records pertinent to a Federal award must be 1 retained for a period of three years from the date of submission of the final expenditure 2 report or, for Federal awards that are renewed quarterly or annually, from the date of the 3 submission of the quarterly or annual financial report.” 45 C.F.R. § 75.361. However, if 4 an “audit is started before the expiration of the 3-year period, the records must be retained 5 until all litigation, claims, or audit findings involving the records have been resolved and 6 final action taken.” 45 C.F.R. § 75.361(a); see also 42 C.F.R. § 433.32(c) (stating that 7 agencies must “[r]etain records beyond the 3-year period if audit findings have not been 8 resolved”). 9 In order to determine the ending date of the three-year documentation retention 10 period for the first quarter of 2004 and the second quarter of 2005, and to therefore 11 determine whether AHCCCS was obligated to maintain the missing documentation at the 12 time of the 2011 audit, the Court must first determine the starting dates of those three-year 13 periods. Based on the record, it is unclear exactly when AHCCCS filed its claims for the 14 first quarter of 2004 and the second quarter of 2005. Pursuant to 42 U.S.C. § 1320b-2(a), 15 states must file claims “within the two-year period which begins on the first day of the 16 calendar quarter immediately following such calendar quarter.” 42 U.S.C. § 1320b-2(a). 17 Therefore, claims for the first quarter of 2004 must have been submitted by April 1, 2006, 18 and claims for the second quarter of 2005 must have been submitted by July 1, 2007.6 19 Based upon the three-year document retention period from those dates, AHCCCS must 20 have retained the documents from the first quarter of 2004 until April 1, 2009, and must 21 have retained documents from the second quarter of 2005 until July 1, 2010.7 22 If an audit is started before the end of the three-year document retention period, the 23 records must be retained until the audit is resolved and final action has been taken. 45 24 6Because AHCCCS did not submit evidence of the exact claim submission dates, the Court 25 will analyze this issue using the latest dates on which AHCCCS could have supported the claims. 26 7 CMS asserts that a five-year, not three-year, document retention policy applies based 27 upon AHCCCS’s own 2004 plan, which required Maximus to retain the time study records for no less than five years. (Doc. 34 at 13). Based on this Court’s decision, it is unnecessary 28 to address whether the federal three-year document retention requirement applies or whether the claimed internal five-year document retention policy applies. 1 C.F.R. § 75.361(a) (“the records must be retained until all litigation, claims, or audit 2 findings involving the records have been resolved and final action taken”); see also 42 3 C.F.R. § 433.32(c) (agencies must “[r]etain records beyond the 3-year period if audit 4 findings have not been resolved”). In the present case, the 2008 audit was initiated on 5 October 20, 2008, which was before the end of the three-year document retention period 6 for both the first quarter of 2004 and the second quarter of 2005. Therefore, AHCCCS was 7 required to retain all documents related to the 2008 audit until that audit was resolved. In 8 contrast, the 2011 audit was initiated after expiration of the three-year document retention 9 period. Therefore, unless the 2008 audit was broad enough to encompass the missing 10 documents that were required for the 2011 audit, AHCCCS was not obligated to maintain 11 those documents beyond the three-year period. Accordingly, resolution of this issue turns 12 upon whether the 2008 audit was sufficiently broad enough to encompass the 2011 audit. 13 The Court agrees with the Appeals Board, which found that the 2008 audit notice 14 “plainly” announced that it would examine “whether federal funds were properly claimed” 15 and therefore found that the 2008 audit notice put AHCCCS on notice that it would need 16 to retain substantiating claim documentation. (Doc. 17-3 at 21-22). The stated purpose of 17 the 2008 audit was to “determine the (1) extent to which [AHCCCS] has contracted with 18 consultants through contingency fee payment arrangements and (2) impact of these 19 arrangements on the submission of improper claims to the Federal Government.” (Doc. 20 17-5 at 109 (emphasis added)). This essentially means, with respect to AHCCCS, that the 21 audit was intended to examine the contingency fee payment arrangements between 22 AHCCCS and its contractor Maximus and was intended to determine if the contingency 23 fee arrangement impacted the submission of improper claims. Attached to the audit notice 24 was a list of documentation requested in connection with the audit; those documents 25 included “Quarterly amounts claimed by AHCCCS for Federal financial participation for 26 Medicaid school-based administrative expenditures (in an Excel file).” (Doc. 17-5 at 112). 27 On March 11, 2011, the OIG presented a second audit letter to AHCCCS, this time 28 indicating that it was specifically examining the school-based administrative costs claimed 1 by the State of Arizona. (Doc. 17-5 at 143). In doing so, it stated that it had “revised” its 2 prior audit to focus on the claimed administrative costs, rather than the contingency fee 3 payments. (Id.). 4 Although the 2008 audit did not specifically focus on the validity of the school- 5 based administrative claims, the language of the notice letter was broad enough to put 6 AHCCCS on notice that it needed to retain the administrative claim supporting documents 7 until the 2008 audit was resolved. First, the 2008 audit encompassed Arizona’s 8 contingency fee agreements with consultants for claiming “school-based administrative 9 costs.” (Doc. 17-5 at 109). The claims at issue are for these school-based administrative 10 costs. Second, the 2008 audit notice stated that one of its purposes was to examine the 11 impact of contingency fee arrangements “on the submission of improper claims to the 12 Federal Government.” (Doc. 17-5 at 109). When considering this language together, the 13 audit notice indicated that the audit would include school-based administrative costs and 14 would necessarily need to encompass whether claims were improperly submitted for those 15 costs. Thus, as recognized by the Appeals Board (Doc. 17-3 at 21-22), once AHCCCS 16 received the notice, it was required to retain all documents related to the audit until the 17 audit was resolved. See 45 C.F.R. § 75.361; 42 C.F.R. § 433.32(c). 18 AHCCCS suggests that it did not need to retain the documents underlying the claims 19 submissions because those documents were not included in the list of documents attached 20 to the 2008 audit notice. Instead, the notice merely asked for a spreadsheet of quarterly 21 amounts claimed by AHCCCS during the period from 2004 to 2008. (Doc. 30 at 27). 22 Although the list of items did request such a spreadsheet, the audit further stated that “we 23 will also need access to additional documents and records.” (Doc. 17-5 at 109, 112). This 24 language demonstrates that the provided list was not an exclusive list. Therefore, the 2008 25 audit letter placed AHCCCS on notice that all documents related to the audit needed to be 26 retained until that audit was resolved. Because that audit was never resolved, but was 27 instead modified into the 2011 audit, AHCCCS should have continued to maintain the 28 documents through resolution of the 2011 audit. 1 Finally, AHCCCS argues that, even if it had been required to maintain the missing 2 documents, it is “unreasonable and punitive” for CMS to disallow all amounts claimed for 3 the first quarter of 2004 and the second quarter of 2005. (Doc. 30 at 29-32). AHCCCS 4 asks this Court to instead only permit disallowance of an amount proportionate to the 5 percentage of claims disallowed during the 17 quarters for which AHCCCS provided the 6 necessary supporting documentation. (Id.). It reasons that, because only a small 7 percentage was disallowed from the remaining quarters, it is disproportionately punitive to 8 disallow 100% of the claims from the two quarters in question. (Id.). The Appeals Board 9 rejected this argument, finding that the “disallowance of [federal financial participation] 10 for failure to substantiate the claims is not intended to punish the state agency.” (Doc. 17- 11 3 at 24). Instead, the “issue is whether federal funds have been properly paid to a claimant 12 in accordance with applicable requirements.” (Doc. 17-3 at 24). Therefore, the 13 disallowance was based on CMS’s finding that AHCCCS could not substantiate any of its 14 claims for those two quarters because it admittedly failed to retain and to produce the 15 underlying random moment time sampling records for those quarters. (Doc. 17-3 at 25). 16 Accordingly, the Appeals Board concluded that “Arizona has not carried its burden to show 17 that the audit findings were not substantiated.” (Doc. 17-3 at 25). Finally, the Appeals 18 Board stated that, to the extent that Arizona’s request for a reduced disallowance may be 19 considered a request for equitable relief, it could not provide such relief because it is not 20 authorized to issue equitable relief. (Doc. 17-3 at 25). 21 This Court finds that, based on the record provided, the Appeals Board decision was 22 not arbitrary, capricious, an abuse of discretion, or otherwise contrary to law. As stated by 23 the OIG, and as recognized by the Appeals Board, the disallowance was based upon the 24 OIG’s inability to determine the proper amount for reimbursement because AHCCCS 25 failed to substantiate its claims for those two quarters during the audit:

26 Without these files, we could not determine whether the observation forms that the State agency provided were for the sample items selected for those 27 two quarters. Because the State agency was unable to provide required documentation of the sample universe determination and/or sample 28 selection, the Federal reimbursement for school-based administrative costs for those quarters was unallowable. (Doc. 17-5 at 163; see also Doc. 17-5 at 183-184). Therefore, because AHCCCS was 2|| obligated to maintain all required supporting documents once it had notice of the audit and || because the OIG could not determine the proper claim amount without all supporting 4|| documents, the Court affirms the Appeals Board decision finding that that the $5,421,711 5 || was properly disallowed. 6 II. Conclusion 7 The Court finds that the Departmental Appeals Board’s decision was neither 8 || arbitrary and capricious nor an abuse of discretion. Accordingly, 9 IT IS ORDERED affirming the October 2, 2017, Decision of the Department of 10 || Health and Human Services Departmental Appeals Board. 11 IT IS FURTHER ORDERED dismissing the administrative appeal and directing the Clerk to enter judgment accordingly. 13 Dated this 14th day of February, 2020. 14 15 oC. . je _ \Le 16 norable' Diang/4. Hunfetewa 17 United States District Judge 18 19 20 21 22 23 24 25 26 27 28

-18-