Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 1 of 79 Page ID #:1291 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 CHAD A. READLER Acting Assistant Attorney General, Civil Division SANDRA R. BROWN Acting United States Attorney DOROTHY A. SCHOUTEN DAVID K. BARRETT LINDA A. KONTOS JOHN E. LEE (CBN 128696) Assistant United States Attorneys 300 N. Los Angeles Street, Room 7516 Los Angeles, California 90012 Tel: (213) 894-3995; Fax: (213) 894-7819 Email: john.lee2@usdoj.gov MICHAEL D. GRANSTON DANIEL R. ANDERSON CAROL L. WALLACK JESSICA KRIEG JUSTIN DRAYCOTT PAUL PERKINS Attorneys, Civil Division United States Department of Justice P.O. Box 261, Ben Franklin Station Washington, D.C. 20044 Tel: (202) 307-0486; Fax: (202) 307-3852 E-mail: carol.wallack@usdoj.gov JAMES P. KENNEDY, JR. Acting United States Attorney KATHLEEN ANN LYNCH Assistant United States Attorney (Admitted PHV) 138 Delaware Avenue Buffalo, New York 14201 Tel: (716) 843-5830; Fax: (716) 551-3052 E-mail: kathleen.lynch@usdoj.gov Attorneys for the United States of America UNITED STATES DISTRICT COURT FOR THE CENTRAL DISTRICT OF CALIFORNIA WESTERN DIVISION UNITED STATES OF AMERICA ex rel. BENJAMIN POEHLING, Plaintiffs, 24 v. 25 UNITEDHEALTH GROUP, INC., a Delaware corporation; UNITED HEALTHCARE SERVICES, INC., a Minnesota corporation; UNITED HEALTHCARE, INC., a Delaware corporation; UNITEDHEALTHCARE INSURANCE COMPANY, a 26 27 28 No. CV 16-08697 WMF (SSx) UNITED STATES’ COMPLAINT-INPARTIAL-INTERVENTION AND DEMAND FOR JURY TRIAL Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 2 of 79 Page ID #:1292 1 2 3 4 Connecticut corporation; UHIC HOLDINGS, INC., a Delaware corporation; OVATIONS, INC., a Delaware corporation; OPTUM, INC. & OPTUMINSIGHT, INC., Delaware corporations; and Defendants listed on Exhibit 1, Defendants. 5 6 7 8 9 This is a civil fraud action brought by the United States of America (“United States” or “Government”) to recover treble damages and civil penalties under the False 10 Claims Act (“FCA”), 31 U.S.C. §§ 3729-3733, as well as for restitution and common 11 law damages, for monies unlawfully obtained and/or retained from the federal Medicare 12 Program by Defendant UnitedHealth Group Inc. and various of its direct and indirect 13 subsidiaries involved in the Medicare Advantage Program (“United” or the “United 14 Defendants”). Having filed a notice of intervention pursuant to 31 U.S.C. § 3730(b)(4), 15 the United States alleges for its complaint-in-partial-intervention (the “Government’s 16 Complaint” or “Complaint”) as follows: 17 INTRODUCTION 18 1. 19 their Medicare benefits through the Medicare Advantage Program. A central, 20 distinguishing feature of the Medicare Advantage Program is the provision of Medicare 21 benefits by private healthcare insurance organizations. Medicare beneficiaries enroll in 22 managed healthcare insurance plans called Medicare Advantage Plans (“MA Plans”) that 23 are owned and operated by these private organizations, called Medicare Advantage 24 Organizations (“MA Organizations”). This case involves conduct by United – the 25 nation’s largest owner of MA Organizations – to improperly obtain or avoid returning 26 payments under the Medicare Advantage Program that it was not entitled to receive. 27 2. 28 Medicare beneficiary enrolled in its plans. The Government adjusts these payments for Millions of elderly and disabled individuals throughout the United States receive The Government pays each MA Organization a fixed monthly payment for each 2 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 3 of 79 Page ID #:1293 1 various risk factors that affect expected healthcare expenditures, including the health 2 status of each enrollee. The adjustments are intended to ensure that MA Organizations 3 are paid more for those enrollees expected to incur higher healthcare costs and less for 4 healthier enrollees expected to incur lower costs. 5 3. 6 submit diagnosis codes to the Government for the beneficiaries in their MA Plans. 7 These diagnosis codes are from the beneficiaries’ medical encounters (e.g., office visits 8 and hospital stays). Using these diagnosis codes, the Government calculates a risk score 9 for each beneficiary. The beneficiary’s risk score is then used to calculate monthly To obtain payments based on adjustments for health status, MA Organizations 10 payments to the MA Organization for that beneficiary for the following year. In general, 11 the more numerous the conditions, and the more severe the conditions, the higher the 12 risk score for a beneficiary and, thus, the greater the risk-adjusted payments made to the 13 MA Organization for that beneficiary. 14 4. 15 report diagnosis codes in order to exaggerate the expected healthcare costs for their 16 enrollees. In order to combat these incentives and protect the Government from making 17 erroneous payments to MA Organizations, the Government requires that submitted 18 diagnoses be supported and validated by the beneficiaries’ medical records. It is a well- 19 established requirement that all diagnosis codes submitted to the Medicare Program for 20 risk adjustment payments must be unambiguously supported by information included in 21 the beneficiaries’ medical records. United knew that these medical records are the 22 “source of truth” for the purpose of receiving and retaining risk adjustment payments. 23 5. 24 it has provided are accurate and truthful. 42 C.F.R. § 422.504(l)(2). Each MA 25 Organization must also “[a]dopt and implement an effective compliance program, which 26 must include measures that prevent, detect, and correct non-compliance with [the 27 Government’s] program requirements as well as measures that prevent, detect, and 28 correct fraud, waste, and abuse.” 42 C.F.R. § 422.503(b)(4)(vi). This payment model creates powerful incentives for MA Organizations to over- In addition, each MA Organization must expressly certify that the diagnosis codes 3 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 4 of 79 Page ID #:1294 1 6. Millions of elderly and disabled Medicare beneficiaries are enrolled in MA Plans 2 that are owned and operated by United throughout the United States. United is the 3 nation’s largest owner and operator of MA Plans. Furthermore, in March 2017, 4 approximately 229,000 Medicare beneficiaries in the Central District of California were 5 enrolled in United’s MA Plans, including those of Defendants UHC of California 6 (previously known as PacifiCare of California) and Sierra Health and Life Insurance 7 Company. 8 7. 9 Medicare beneficiaries enrolled in its MA Plans. Risk adjustment payments account for 10 a substantial amount of these dollars. The diagnoses submitted by United drive a large 11 percentage of the payments it receives from the Medicare Program. It is not surprising 12 then that United is not a passive conduit of diagnoses from healthcare providers to the 13 Medicare Program. Rather, for many years, United has conducted programs and 14 engaged in other activities to increase the amount of risk adjustment payments from 15 Medicare. This includes programs and other efforts to directly influence both the 16 number of diagnoses and the severity of the medical conditions reported by providers. 17 This also includes programs and efforts which do not involve the providers. 18 8. 19 Review Program to increase the risk adjustment payments it receives from Medicare. 20 For many years, this was United’s biggest effort aimed at increasing risk adjustment 21 payments. During the last ten years, United increased the amount of risk adjustment 22 payments that it received from the Medicare Program by collecting millions of medical 23 records (also known as “charts”) from providers and then employing diagnosis coders 24 (also known as “chart reviewers”) to review the medical records in order to mine for 25 diagnoses that the providers themselves did not report to United for their patients in 26 United’s MA Plans. United then submitted these additional diagnosis codes (“ADDS”) 27 to the Medicare Program for billions of dollars of additional risk adjustment payments. The Government pays billions of taxpayer dollars each year to United for the In particular, for many years, United has conducted a very large national Chart 28 4 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 5 of 79 Page ID #:1295 1 9. United’s national Chart Review Program was strictly a one-sided revenue- 2 generating program. United did not review the beneficiaries’ medical records in good 3 faith in order to obtain a true and accurate picture of the health status of the beneficiaries 4 in its MA Plans or to submit truthful and accurate risk adjustment data to the 5 Government. United used the results of the chart reviews to only increase government 6 payments (i.e., submit additional codes not reported by the providers) while in bad faith 7 systemically ignoring other information from the chart reviews which would have led to 8 decreased payments (i.e., information about diagnoses reported by providers to United 9 and then submitted by United to Medicare which were not supported and validated by 10 the medical records). 11 10. 12 diagnoses reported by providers to it (hereinafter “provider-reported diagnoses”) are 13 invalid because the beneficiaries’ medical records do not substantiate that the 14 beneficiaries had the medical conditions identified by the diagnosis codes reported by 15 the providers. It knew this very early on from audits conducted by the Government and 16 its own internal medical record reviews. Despite this knowledge, United knowingly 17 avoided “looking both ways” as part of its national Chart Review Program, except for a 18 very limited time period when it “looked both ways” at some of its chart review results 19 as part of its Claims Verification Program. That is, United knowingly and improperly 20 avoided comparing the diagnoses reported by the providers and submitted by it to the 21 Government with the results of its coders’ chart reviews to identify those provider- 22 reported codes that were not supported by the beneficiaries’ medical records. United 23 could and should have done this comparison and deleted its prior submission of these 24 unsupported diagnoses, that is, made “DELETES.” If United had done so, the Medicare 25 Program would not have made risk adjustment payments based on these unsupported 26 diagnoses or, if it had already made the payments, it would have recovered them from 27 United. Yet, since at least 2005, United has known that a significant percentage of 28 5 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 6 of 79 Page ID #:1296 1 11. By failing to “look both ways,” United improperly generated and reported skewed 2 data artificially inflating beneficiaries’ risk scores, avoided negative payment 3 adjustments, and retained payments to which it was not entitled. The Government has 4 conservatively estimated that, if United had “looked both ways,” it would not have 5 submitted or, if submitted, it would have deleted hundreds of thousands of invalid 6 diagnoses and the Medicare Program would not have erroneously paid or would have 7 recovered at least over a billion dollars in risk adjustment payments to which United was 8 not entitled. 9 12. By failing to “look both ways,” United violated the FCA. United knowingly 10 presented or caused to be presented false or fraudulent claims to the Medicare Program; 11 knowingly made or used or caused to be made or used false records or statements 12 material to these false or fraudulent claims and to obligations to pay (i.e., return) monies 13 to the Medicare Program; knowingly concealed obligations to pay (i.e., return) monies 14 owed to the Medicare Program; and knowingly and improperly avoided or decreased 15 obligations to pay (i.e., return) monies owed to the Medicare Program. 16 13. 17 disregarding information from its Risk Adjustment Coding Compliance Review 18 (RACCR) Program about invalid diagnoses reported to it by certain of its “incentivized” 19 providers, including certain capitated and gainsharing providers. 20 14. 21 large provider groups on a “capitated” basis. It paid these capitated providers a “fixed” 22 fee per beneficiary cared for by these providers; these fees generally were not dependent 23 on the amount of services rendered by these providers. Often the “fixed” fees were 24 based on a percentage share of the payments that United received from the Medicare 25 Program for the beneficiaries cared for by the “capitated” providers. United’s other 26 providers were paid on a fee-for-service basis for each service (e.g., office visit) they 27 provided. United, however, also entered into “gainsharing” agreements whereby it made 28 incentive payments to some of its fee-for-serve providers. These incentive payments In addition, United violated the FCA by deliberately ignoring or recklessly United paid its providers through a variety of arrangements. United paid many 6 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 7 of 79 Page ID #:1297 1 were based in whole or part on total revenues that United received from the Medicare 2 Program for the beneficiaries cared for by these gainsharing providers. 3 15. 4 these providers to increase the number of diagnoses that they reported to United and to 5 report diagnoses for more severe medical conditions. The more risk adjustment 6 payments obtained by United for the beneficiaries cared for by these providers, the more 7 money United paid to these providers pursuant to the gainsharing and capitation 8 agreements. 9 16. United’s agreements with gainsharing and with capitated providers incentivized United knew that these gainsharing and capitated providers had a financial 10 incentive increasing the risk of their reporting invalid diagnoses in order to increase their 11 own revenues. In fact, based on the results of its own data analyses and medical record 12 reviews as part of its RACCR Program, United knew which incentivized providers were 13 actually or likely engaged in over-reporting diagnoses, including some providers located 14 in this District. But it knowingly continued to submit diagnoses from these incentivized 15 providers to Medicare and knowingly and improperly avoided repaying Medicare for 16 risk adjustment payments based on invalid diagnoses from these providers, all in 17 violation of the FCA. 18 JURISDICTION AND VENUE 19 17. This Court has subject matter jurisdiction over this action pursuant to 28 U.S.C. 20 § 1345 because the United States is the Plaintiff. In addition, the Court has subject 21 matter jurisdiction over the FCA claims for relief under 28 U.S.C. §§ 1331 and 1345 and 22 31 U.S.C. § 3732(a)-(b) and supplement jurisdiction to entertain the common law and 23 equitable claims for relief under 28 U.S.C. § 1367(a). 24 18. 25 § 3732(a) because at least one of the Defendants can be found in, resides in, transacts 26 business in, or has committed the alleged acts in the Central District of California. 27 19. 28 § 3732(a) because at least one of the Defendants can be found in, resides in, and This Court has personal jurisdiction over Defendants pursuant to 31 U.S.C. Venue also lies in this District pursuant to 28 U.S.C. § 1391(b)-(c) and 31 U.S.C. 7 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 8 of 79 Page ID #:1298 1 transacts business in this District, a substantial part of the events or omissions giving rise 2 to the claims occurred in this District, and/or all of the Defendants are subject to the 3 Court’s personal jurisdiction under the FCA. 4 PARTIES 5 I. Plaintiffs 6 20. 7 Department of Health and Human Services (“HHS”), which includes its operating 8 division, the Centers for Medicare and Medicaid Services (“CMS”). At all times 9 relevant to this Complaint, CMS administered and supervised the Medicare Program and 10 made risk adjustment payments to MA Organizations, including United and its affiliates, 11 under Parts C and D of the Program. The United States filed its notice of partial 12 intervention in this action on February 14, 2017. 13 21. 14 Finance for UnitedHealthcare Medicare & Retirement (and its predecessor Ovations), 15 which was the group at United that managed its MA Plans and its Medicare Part D 16 Prescription Drug Programs. From mid-2007 until he left United at the end of 2012, 17 Poehling ran the risk adjustment team at UnitedHealthcare Medicare & Retirement. He 18 was one of United’s management employees responsible for United’s submission of 19 claims to the Medicare Program for risk adjustment payments. He was also one of 20 United’s management employees responsible for United’s risk adjustment revenue- 21 generating activities, including, but not limited to, United’s national Chart Review 22 Program. Poehling expressed concerns to United’s executives about United’s failure to 23 “look both ways” as a part of its Chart Review Program. In March 2011, Poehling 24 initiated this action by filing a complaint against United pursuant to the qui tam 25 provisions of the FCA. 31 U.S.C. § 3730(b)(1). 26 Plaintiff is the United States of America, suing on behalf of the United States The qui tam plaintiff (“Relator”) is Benjamin Poehling, the former Director of II. Defendants 27 22. Defendant UnitedHealth Group Inc. (“UHG”) is a publicly traded Delaware 28 corporation. It is the parent company for all other United Defendants in this action. 8 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 9 of 79 Page ID #:1299 1 UHG, the other United Defendants, and their affiliates have offices in various locations 2 throughout the United States, including in the Central District of California. UHG’s 3 healthcare insurance products, including those under Parts C and D of the Medicare 4 Program, are offered by, and UHG’s MA Plans are managed by, various entities that are 5 UHG’s direct or indirect subsidiaries, including, but not limited to, the other United 6 Defendants identified below. UHG controls all of these entities. 7 23. 8 fifty states and the District of Columbia. As of December 31, 2008, United had 9 approximately 1.5 million Medicare beneficiaries enrolled in its plans under Part C of UHG and its direct and indirect subsidiaries and affiliates operate MA Plans in all 10 the Medicare Program and millions of additional beneficiaries enrolled in its prescription 11 drug benefit plans under Part D of the Medicare Program. As of December 31, 2009, 12 United had approximately 1.8 million Medicare beneficiaries enrolled in its plans under 13 Part C of the Medicare Program and millions of additional beneficiaries enrolled in its 14 prescription drug benefit plans under Part D of the Medicare Program. As of December 15 31, 2010, United had approximately 2.1 million beneficiaries in its plans under Part C of 16 the Medicare Program and millions of additional beneficiaries in its drug benefit plans 17 under Part D. As of December 31, 2011, United had 2.2 million beneficiaries in its plans 18 under Part C and millions of additional beneficiaries in its plans under Part D. As of 19 December 31, 2012, United had approximately 2.6 million beneficiaries in its Part C 20 plans and millions of additional beneficiaries in its Part D plans. In 2013, 2014, and 21 2015 United had approximately 3 million beneficiaries in its Part C plans and 22 approximately 8 million in its Part D plans. In 2013, United’s revenues from Part C and 23 D of the Medicare Program were approximately $44 billion. In 2014, United’s revenues 24 from Parts C and D of the Medicare Program were approximately $46 billion. In 2015, 25 United’s revenues from Parts C and D of the Medicare Program were approximately $50 26 billion. In 2016, United had approximately 3.6 million beneficiaries in its Part C plans, 27 and approximately 8.6 million beneficiaries in its Part D plans. For 2016, United’s 28 revenues from Parts C and D of the Medicare Program were approximately $56 billion. 9 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 10 of 79 Page ID #:1300 1 24. United’s Medicare Part C and D managed healthcare insurance products are 2 offered by it through various entities that are direct and indirect subsidiaries of UHG, 3 including, but not limited to, Defendants UnitedHealthcare Insurance Company, 4 Defendant UnitedHealthcare, Inc., Defendant United HealthCare Services, Inc., 5 Defendant UHIC Holdings, Inc., and the Defendant MA Plans. 6 25. 7 direct subsidiary of Defendant UHIC Holdings, Inc., and an indirect subsidiary of 8 Defendant UHG. 9 26. Defendant UnitedHealthcare Insurance Company is a Connecticut corporation, a Defendant UHIC Holdings, Inc. is a Delaware corporation, a direct subsidiary of 10 Defendant United HealthCare Services, Inc., and an indirect subsidiary of Defendant 11 UHG. 12 27. 13 Defendant United HealthCare Services, Inc., and an indirect subsidiary of Defendant 14 UHG. 15 28. 16 direct or indirect subsidiary of Defendant UHG. Defendant United HealthCare Services, 17 Inc. is also the successor to PacifiCare Health Systems, LLC and PacifiCare Health Plan 18 Administrators, Inc., which were the direct or indirect parents of PacifiCare of California 19 and the other PacifiCare MA Plans acquired by United in 2005. 20 29. 21 Defendant United HealthCare Services, Inc. and an indirect subsidiary of Defendant 22 UHG. Ovations, including its subgroups such as Secure Horizons & Evercare, provided 23 managed healthcare insurance coverage under Part C of the Medicare Advantage 24 Program. Another Ovations subgroup called Ovations Part D provided the prescription 25 drug benefits under Part D of the Medicare Program. 26 30. 27 MA Organizations and MA Plans. These groups were located within Defendant 28 UnitedHealthcare, Inc. They included, depending on the time period, Secure Horizons, Defendant UnitedHealthcare, Inc. is a Delaware corporation, a direct subsidiary of Defendant United HealthCare Services, Inc. is a Minnesota corporation and a Defendant Ovations, Inc. is a Delaware corporation. It is a direct subsidiary of United had one or more groups which had some management or oversight over its 10 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 11 of 79 Page ID #:1301 1 Evercare, the Public and Senior Market Group (which included two subgroups: 2 Ovations and AmeriChoice and which was also referred to as the Public Sector Market 3 Group), and, more recently, UnitedHealthcare Medicare & Retirement and 4 UnitedHealthcare Community & State. Among other things, these groups oversaw 5 United’s risk adjustment activities such as the submission of risk adjustment data and 6 claims to the Medicare Program and the Chart Review, Claims Verification (CV), and 7 RACCR Programs. However, the actual data and claim submission and program work 8 was conducted by Defendants Optum, Inc. and OptumInsight, Inc. and their 9 predecessors, including Ingenix, from offices in this District and elsewhere. 10 31. Defendants Optum, Inc. and OptumInsight, Inc. (collectively “Optum”) are 11 Delaware corporations. Optum is a direct or indirect subsidiary of Defendant UHG. 12 Optum and its predecessor, Ingenix, Inc., were the entities that were responsible for the 13 submission of risk adjustment data and claims to the Medicare Program, the deletion of 14 invalid diagnoses and claims, and the management and operation of the Chart Review, 15 Claims Verification, RACCR and other risk adjustment programs for United. Optum 16 (and formerly Ingenix) also performed this risk adjustment work for third-parties which 17 owned and operated MA Organizations and MA Plans. It referred to these third-parties 18 as commercial clients. Optum (and formerly Ingenix) performed a significant amount of 19 its risk adjustment work for United and its commercial clients from its offices in the 20 Central District of California. 21 32. 22 by acquiring them. In 2004, United acquired Oxford Health Plans LLC (doing business 23 as Oxford Health Plans, Inc.) and Oxford’s plans. Also, in 2004, United acquired Mid- 24 Atlantic Medical Services, Inc. and its plans. 25 33. 26 PacifiCare’s and its affiliates’ MA Plans, including Defendants PacifiCare of Arizona, 27 Inc., incorporated in Arizona; PacifiCare of California, incorporated in California; 28 PacifiCare of Colorado, Inc., incorporated in Colorado; PacifiCare of Nevada, Inc., United became the largest owner of MA Organizations and MA Plans in large part In 2005, United acquired PacifiCare Health Systems (“PacifiCare”) and 11 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 12 of 79 Page ID #:1302 1 incorporated in Nevada; PacifiCare of Oklahoma, Inc., incorporated in Oklahoma; 2 PacifiCare of Oregon, Inc., incorporated in Oregon; PacifiCare of Texas, Inc. 3 incorporated in Texas; and PacifiCare of Washington, incorporated in Washington. Both 4 before and after the acquisition, PacifiCare of California and possibly other PacifiCare 5 plans referred to themselves or to their brand of MA Plans as Secure Horizons. Since 6 2005, these PacifiCare plans have been indirect subsidiaries of and controlled by UHG. 7 Several years after the acquisition, these PacifiCare plans were re-named or re-branded 8 as United plans or merged into other United plans. For instance, in 2011, PacifiCare of 9 California became Defendant UHC of California. After the acquisition, Pacificare 10 Health Systems and one or more entities affiliated with it were merged with and into 11 Defendant United Healthcare Services, Inc. All PacifiCare entities and their successors 12 were direct or indirect subsidiaries of Defendant UGH. 13 34. 14 responsibilities relating to the submission of risk adjustment data and claims to Medicare 15 and to other risk adjustment-related activities worked at a PacifiCare office in Cypress, 16 California, within this District. Sometime after the acquisition, United moved this office 17 to Santa Ana, California, within this District. A substantial part of the events or 18 omissions relevant to this litigation occurred at these and other locations within this 19 District. 20 35. 21 acquired Sierra Health Services, Inc. and its MA Plans, including Defendants Health 22 Plan of Nevada, Inc. and Sierra Health and Life Insurance, Inc. Sierra Health Services 23 Inc. is or was a Nevada corporation located in and around Las Vegas. 24 36. 25 (“WMMI”). United’s acquisition of WMMI included its subsidiaries and affiliates, 26 including, but not limited to, WMMI’s MA Plans, Physician’s Health Choice of Texas, 27 LLC and Citrus Health Care, Inc., which operated in Texas, Florida, New Mexico and 28 Arkansas. Citrus Health Care, Inc. was a Florida corporation and a subsidiary of PHC Before United’s acquisition of PacifiCare, the PacifiCare employees with In 2008, United acquired Unison Health and its MA Plans. Also, in 2008, United In January 2011, United acquired WellMed Medical Management, Inc. 12 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 13 of 79 Page ID #:1303 1 Holdings of Florida, Inc. Sometime after the acquisition, United re-named or re-branded 2 these plans as United plans or merged them with or into United’s other MA 3 Organizations or plans in these states. 4 37. 5 Community Improvement Plus. XL Health (formerly known as Diabetex Corporation) is 6 a Maryland corporation with its principal place of business in Elkridge, Maryland. 7 XLHealth is now part of UnitedHealthCare Medicare Solutions. 8 38. 9 after 2005 are Defendants in this action. These entities are listed in Exhibit 1 to this 10 Complaint, in the Risk Adjustment Attestations submitted by United to Medicare for 11 2005 and subsequent payment years, and/or in the “Subsidiaries of the Company” exhibit 12 to United’s Annual Reports (Forms 10-K) for 2005 and subsequent years. All of these 13 Defendants are directly or indirectly owned and controlled by UHG. 14 39. 15 care market by acquiring and/or operating large groups or networks of direct providers of 16 healthcare services and other entities that manage the provision of such services to 17 beneficiaries enrolled in United’s MA Plans. For instance, as part of its acquisition of 18 Sierra Health Services in 2008, United acquired Southwest Medical Associates, Inc. 19 (SMA), which was owned by Sierra. At the time, SMA was the largest physician group 20 in Las Vegas, Nevada. Currently, Defendant Optum, through its groups called 21 OptumHealth and OptumCare, owns and/or operates large physician groups and large 22 integrated healthcare delivery systems in Arizona, California, Connecticut, Florida, 23 Nevada, New York, Texas, and Utah. This includes SMA in Nevada. 24 40. 25 that United acquired WMMI in 2011. For many years, WMMI had subsidiaries and 26 other affiliates that directly managed the provision of or directly provided healthcare 27 services. These affiliates included WellMed Networks, Inc., WellMed Networks Inc. of 28 Florida, WellMed Medical Management of Florida Inc., and WellMed Medical Group, In 2012, United acquired XLHealth Corporation and its MA Plans, including All MA Organizations and MA Plans acquired, owned, and controlled by United Over the last decade, United has also sought to vertically integrate in the health Vertical integration was also one of the primary reasons, if not the primary reason, 13 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 14 of 79 Page ID #:1304 1 PA. After the acquisition, WellMed became part of the United group called OptumCare. 2 After the acquisition, WellMed also significantly expanded by acquiring more than 50 3 medical practices in Texas and Florida. WellMed included more than 10,000 physicians 4 that provided healthcare to hundreds of thousands of Medicare beneficiaries in Texas 5 and Florida, including beneficiaries enrolled in United’s MA Plans. 6 41. 7 all of the Defendants identified above and in Exhibit 1 to this Complaint. All references to “United” and the “United Defendants” in this Complaint include 8 THE LAW 9 I. The False Claims Act 10 42. 11 recover losses as a result of fraud against the Government.” S. Rep. No. 99-345, at 1 12 (1986), available at 1986 U.S.C.C.A.N. 5266. First, a defendant violates the FCA when 13 it “knowingly presents, or causes to be presented, a false or fraudulent claim for payment 14 or approval.” 31 U.S.C. § 3729(a)(1)(A). Under the FCA, a claim includes a request for 15 money. Id., § 3729(b)(2). Further, a claim is “false or fraudulent” under the FCA if the 16 entity or person submitting the claim was not entitled to payment. 17 43. 18 and Recovery Act of 2009 (“FERA”), Pub.L. 111-21 (May 20, 2009), a defendant 19 violates the FCA when it “knowingly makes, uses, or causes to be made or used, a 20 false record or statement material to a false or fraudulent claim.” 31 U.S.C. 21 § 3729(a)(1)(B). Prior to FERA, a defendant violated this provision of the FCA 22 when it “knowingly [made], use[d], or cause[d] to be made or used, a false record 23 or statement to get a false or fraudulent claim paid or approved by the 24 Government.” 25 44. 26 it “knowingly makes, uses, or causes to be made or used, a false record or statement 27 material to an obligation to pay or transmit money or property to the Government, or 28 knowingly conceals or knowingly and improperly avoids or decreases an obligation to The FCA reflects Congress’s objective to “enhance the Government’s ability to Second, after the 2009 amendments to the FCA by the Fraud Enforcement Third, after FERA’s enactment in May 2009, a defendant violates the FCA when 14 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 15 of 79 Page ID #:1305 1 pay or transmit money or property to the Government.” 31 U.S.C. § 3729(a)(1)(G). 2 Prior to FERA, this provision of the FCA, commonly referred to as the “reverse false 3 claims act” provision of the statute, provided that a defendant violates the FCA when it 4 “knowingly makes, uses, or causes to be made or used, a false record or statement to 5 conceal, avoid, or decrease an obligation to pay or transmit money or property to the 6 Government.” 7 45. 8 had actual knowledge of or acted in deliberate ignorance or reckless disregard of 9 information relating to the truth or falsity of its claims for payment or its false records or 10 statements. Id. § 3729(b)(1)(A). Proof that the defendant had specific intent to defraud 11 the Government is not required. Id. § 3729(b)(1)(B). Congress included “deliberate 12 ignorance” in its definition of the terms “knowing” and “knowingly” to hold a defendant 13 accountable for failing to make the inquiry that a reasonable and prudent person or entity 14 would have made under the circumstances to be reasonably certain that he, she, or it was 15 entitled to the money that he, she, or it sought from the Government. S. Rep. No. 99- 16 345, at 21 (1986), as reprinted in 1986 U.S.C.A.N. 5266, 5286. The terms “knowing” 17 and “knowingly” used in this Complaint have the meaning ascribed to them by the FCA. 18 Similarly, the terms “knowledge,” “knows” and “knew” are used in this Complaint to 19 have the same meaning. 20 46. 21 “obligation.” See FERA, Pub. L. 111-21, 123 Stat. 1617, 1621-25 (2009). It defined the 22 term to mean “an established duty, whether or not fixed, arising from an express or 23 implied contractual … relationship, from a fee-based or similar relationship, from statute 24 or regulation, or from the retention of any overpayment.” 31 U.S.C. § 3729(b)(3). 25 Congress promulgated this definition to reflect its long-held view that an “obligation” 26 under the FCA’s reverse FCA provision, 31 U.S.C. § 3729(a)(1)(G), encompasses non- 27 fixed and contingent duties to pay or repay monies to the Government. S. Rep. 111-10, 28 14, 2009 U.S.C.C.A.N. 430, 441. Under the FCA, the terms “knowing” and “knowingly” mean that the defendant In 2009, Congress also amended the FCA to provide a definition of the term 15 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 16 of 79 Page ID #:1306 1 47. 2 capable of influencing, the payment or receipt of money or property.” Id. § 3729(b)(4). 3 48. 4 damages which it sustained because of a defendant’s violation of the statute and, for each 5 act by the defendant violating the statute, a civil penalty. For violations that occurred 6 before November 2, 2015, the FCA imposes a penalty for each violation of not less than 7 $5,500 and not more than $11,000. For violations occurring after November 2, 2015, all 8 civil statutory penalties, including the FCA, are subject to an annual adjustment for 9 inflation pursuant to Section 701 of the Bipartisan Budget Act of 2015, Public Law 114- 10 74 (No. 2, 2015) (“BBA”). At this time, by operation of the BBA, for all FCA penalties 11 assessed after February 3, 2017, whose associated violations occurred after November 2, 12 2015, the penalty for each violation is not less than $10,957 and not more than $21,916. 13 Under the FCA, “material” means “having a natural tendency to influence, or Under the FCA, the Government is entitled to recover three times the amount of II. The Medicare Statute 14 49. 15 Medicare benefits individuals age 65 and older and the disabled. 42 U.S.C. § 1395c et 16 seq. Parts A and B of the Medicare Program are known as “traditional” Medicare. 17 Medicare Part A covers inpatient and institutional care. Medicare Part B covers 18 physician, hospital outpatient, and ancillary services and durable medical equipment. 19 50. 20 hospitals and physicians) using what is known as a “fee-for-service” (“FFS”) payment 21 system. Under a FFS payment system, healthcare providers submit claims to CMS for 22 reimbursement for each service, such as a physician office visit or a hospital stay. CMS 23 then pays the providers directly for each service. 24 51. 25 beneficiaries can opt out of the traditional Medicare Program (Parts A and B) and instead 26 enroll in and receive managed health care services from MA Plans. MA Plans must 27 provide Medicare beneficiaries all the services that they are entitled to receive from the 28 traditional Medicare Program. Medicare is a federally-operated health insurance program administered by CMS. Under Medicare Parts A and B, CMS reimburses healthcare providers (e.g., Under Medicare Part C (the “Medicare Advantage Program”), Medicare 16 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 17 of 79 Page ID #:1307 1 52. Under Medicare Part D, Medicare beneficiaries can elect to enroll in either a 2 Prescription Drug plan (known as a PD Plan) or an MA Plan that provides prescription 3 drug coverage in addition to the physician office visit and hospital outpatient and 4 inpatient coverage provided under Part C (known as an MAPD Plan). For simplicity, in 5 this Complaint, the Government refers to all MA and MAPD Plans as Medicare 6 Advantage Plans or MA Plans. 7 53. 8 enrollee in that plan. In this Complaint, the terms beneficiaries, members, enrollees, and 9 patients are used interchangeably, but mean the same thing, that is, individuals enrolled Medicare beneficiaries who enroll in an MA Plan are considered a member of and 10 in MA plans. 11 54. 12 them to participate in the Program are set forth in CMS regulations and, each year, the 13 MA Organizations agree in writing to comply with those regulations. 42 C.F.R. 14 §§ 422.504 & 422.505 (Part C); 42 C.F.R. §§ 423.504 & 423.505 (Part D). In addition, 15 MA Organizations must comply with requirements set forth in statutes, such as the FCA, 16 and guidance documents, such as the Medicare Managed Care Manual, the Medicare 17 Prescription Drug Benefit Manual, and Medicare Advantage operating instructions. 18 MA Organizations’ obligations to the Medicare Program and the requirements for III. Medicare Parts C and D Risk Adjustment Payments 19 55. Under Part C, the Medicare Program pays each MA Organization a predetermined 20 monthly amount for each Medicare beneficiary in the plan. This monthly payment is 21 known as a “per-member, per-month” payment. This capitated payment for each plan 22 varies depending on various factors, including amounts set forth in the plan’s bid 23 submitted to CMS. Since 2000, Congress has also required that the payments be risk 24 adjusted for each beneficiary based on demographic factors (e.g., gender, age) and health 25 status. By risk adjusting for health status, Congress required that more be paid for 26 beneficiaries with higher risk scores than be paid for beneficiaries with lower risk scores. 27 CMS currently employs a health-based risk adjustment model – known as the 28 17 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 18 of 79 Page ID #:1308 1 Hierarchical Conditions Category (“HCC”) model – that takes into account diagnoses 2 from inpatient hospital stays, outpatient encounters, and physician office visits. 3 56. 4 medical conditions assigned to beneficiaries by their physicians in one year (referred to 5 by CMS as the “data collection” year but also generally known as the “date of service” 6 or “DOS” year) to set the payment for each beneficiary for the following year (often 7 referred to as the “payment year” or “PY”). The medical conditions included in the 8 model are grouped into HCCs, which are categories of clinically-related medical 9 diagnoses. See 42 C.F.R. § 422.2. The diagnoses grouped into HCCs include major, The HCC model is prospective, meaning that it relies on diagnoses for certain 10 severe, and/or chronic illnesses. Related groups of diagnoses are ranked on the basis of 11 disease severity and the cost associated with their treatment. Between 2004 and 2013, 12 the CMS-HCC model included 70 HCCs. Starting in 2014, the CMS-HCC model 13 included 79 HCCs. 14 57. 15 health status. As with Part C, Part D employs a health-based risk adjustment model – 16 known as the Rx Hierarchical Condition Categories (“RxHCC”) model. Like HCCs, 17 RxHCCs are also groups of clinically-related medical diagnoses that are ranked by 18 disease severity and the cost associated with pharmaceutical drugs used to treat them. 19 58. 20 group that correlates to the predicted incremental costs of care associated with treating 21 the medical conditions in each category. It determines the relative values based on the 22 amounts that it paid to fee-for-service providers to treat these major, severe, and chronic 23 medical conditions under Parts A and B of the Medicare Program. Higher relative 24 values are assigned to HCCs and RxHCCs that include diagnoses with greater disease 25 severity and greater costs associated with their treatment. 26 59. 27 prospective and a beneficiary’s risk score for a particular payment year is determined by 28 his or her medical conditions during the previous year (i.e., the date of service year). Under Medicare Part D, payments to MAPD Plans are also risk adjusted based on The Government assigns a relative numerical value to each HCC and RxHCC As previously stated, the HCC and RxHCC risk adjustment models are 18 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 19 of 79 Page ID #:1309 1 These medical conditions must be documented by a qualified healthcare provider (e.g., a 2 doctor) in the beneficiary’s medical record during the previous year. 3 60. 4 example, a beneficiary’s risk score for payment year 2012 is determined by the 5 diagnoses that his or her qualified healthcare providers documented in his or her medical 6 records during face-to-face medical encounters during date of service year 2011. 7 61. 8 the beneficiaries in their plans. Healthcare providers can transmit diagnosis codes to 9 MA Organizations with claims for payment for services rendered, in encounter records 10 reporting the services rendered, or by alternative means. In this Complaint, the United 11 States refers to diagnosis codes reported by providers through any means as “provider- 12 reported diagnoses.” 13 62. 14 using CMS’ Risk Adjustment Processing System (“RAPS”). Each RAPS submission 15 must include the following information: the Medicare beneficiary’s identification 16 number (called a “HIC number” or “HICN”); the date(s) of the medical encounter; the 17 type of provider (physician or hospital); and the diagnosis code(s) reported by the 18 provider for the encounter. Medical encounters include physician office visits, hospital 19 outpatient visits, and hospital inpatient stays. 20 Each beneficiary’s risk score is calculated anew for each payment year. For MA Organizations obtain diagnosis data from the healthcare providers that treat MA Organizations submit risk adjustment data, including diagnoses, to CMS IV. Legal Obligation to Submit Valid Risk Adjustment Data 21 63. MA Organizations are entitled to risk adjustment payments based on the diagnosis 22 codes that they submit to CMS only if the codes are from face-to-face medical 23 encounters between the Medicare beneficiary and provider, the encounter occurred 24 during the relevant date of service year, the provider was of a type and specialty 25 acceptable for risk adjustment purposes, and at the time of the encounter, the provider 26 documented the medical conditions identified by the diagnosis codes in the medical 27 record based on acceptable documentation. In addition, codes should be based on 28 documented conditions that require or affect patient care treatment or management. See 19 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 20 of 79 Page ID #:1310 1 2008 Risk Adjustment Data Technical Assistance for Medicare Advantage Organizations 2 Participant Guide (“2008 RA Participation Guide”) at § 6.4.1. 3 64. 4 valid only to the extent that the diagnosis codes submitted by the MA Organizations are 5 valid. The diagnoses must be coded according to the International Classification of 6 Diseases (ICD) Clinical Modification Guidelines for Coding and Reporting (“ICD-9- 7 CM” & “ICD-10-CM”) and documented with sufficient clinical specificity. All 8 diagnosis codes submitted by MA Organizations must be supported by medical record 9 documentation. If the medical record is ambiguous, it cannot be relied on for diagnosis 10 information for risk adjustment payments. See 2008 RA Participation Guide at § 7.2.4.1 11 (stating that risk adjustment claims and payments cannot be based on questionable 12 diagnoses). 13 65. 14 incentive for MA Organizations to report diagnoses that are not validated by the 15 beneficiary’s medical records or to not delete previously-submitted invalid diagnoses so 16 that they can increase their payments. Thus, CMS engages in a variety of program 17 integrity activities, including audits of diagnoses submitted by MA Organizations, 18 known as Risk Adjustment Data Validation (“RADV”) audits. To support these audits, 19 MA Organizations and their providers are required, when requested, to provide medical 20 records to validate the diagnoses that they submitted for risk adjustment payments. See 21 42 C.F.R. § 422.310(e). 22 66. 23 compliance programs to ensure the integrity of their payment data, 42 CFR 24 § 422.503(b)(4)(vi) (Part C compliance program regulation); 42 C.F.R. 25 § 423.504(b)(4)(vi) (Part D compliance program regulation); (ii) annually attest to the 26 accuracy and truthfulness of the diagnosis data that they submit for risk adjustment 27 payments, 42 C.F.R. § 422.504(l) (Part C regulation); 42 C.F.R. § 423.505(k) (Part D 28 regulation); and (iii) “comply with . . . Federal laws and regulations designed to prevent Risk adjustment claims are true and the resulting risk adjustment payments are CMS recognizes that risk adjusting based on health status creates a strong In addition, MA Organizations must (i) establish and implement effective 20 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 21 of 79 Page ID #:1311 1 or ameliorate fraud, waste, and abuse, including, but not limited to, applicable provisions 2 of Federal criminal law [and] the False Claims Act (31 USC §§ 3729 et seq.).” 42 3 C.F.R. § 422 (Part C regulation); 42 C.F.R. § 423 (Part D regulation). 4 A. MA Organizations Must Have Effective Compliance Programs 5 67. The implementation of an effective compliance program is a prerequisite to an 6 MA Organization’s obtaining and retaining payments under both Parts C and D of the 7 Medicare Program. Id. §§ 422.503(a) (Part C) & 423.504(b)(4)(vi) (Part D). One 8 purpose of requiring a compliance program is to ensure that MA Organizations submit 9 accurate and truthful information to CMS. 65 FR 40170-01 at 40264 (June 29, 2000). 10 68. 11 compliance program, which must include measures that prevent, detect, and correct non- 12 compliance with CMS’ program requirements as well as measures that prevent, detect, 13 and correct fraud, waste, and abuse.” 42 C.F.R. § 422.503(b)(4)(vi) (Part C); 42 C.F.R. 14 § 423.504(b)(4)(vi) (Part D). The compliance program “must, at a minimum, include 15 [certain] core requirements,” including (but not limited to): 16 Specifically, each MA Organization must “[a]dopt and implement an effective (F) Establishment and implementation of an effective system for routine 17 monitoring and identification of compliance risks. The system should 18 include internal monitoring and audits and, as appropriate, external audits, to 19 evaluate the MA organization[’s], including first tier entities’, compliance 20 with CMS requirements and the overall effectiveness of the compliance 21 program. 22 (G) Establishment and implementation of procedures and a system for promptly 23 responding to compliance issues as they are raised, investigating potential 24 compliance problems as identified in the course of self-evaluations and 25 audits, correcting such problems promptly and thoroughly to reduce the 26 potential for recurrence, and ensuring ongoing compliance with CMS 27 requirements. 28 21 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 22 of 79 Page ID #:1312 1 (1) If the MA organization discovers evidence of misconduct 2 related to payment or delivery of items or services under the 3 contract, it must conduct a timely, reasonable inquiry into that 4 conduct. 5 (2) The MA organization must conduct appropriate corrective 6 actions (for example, repayment of overpayments, disciplinary 7 actions against responsible employees) in response to the potential 8 violation referenced in paragraph (b)(4)(G)(1) of this section. 9 (3) The MA organization should have procedures to voluntarily 10 self-report potential fraud or misconduct related to the MA 11 program to CMS or its designee. 12 69. A compliance program is not effective unless the MA Organization devotes 13 adequate resources to the program. 14 70. 15 other things, MA Organizations are responsible for deleting RAPS data submissions if 16 the diagnoses that they submitted are invalid. Deletion of invalid diagnoses allows CMS 17 to recalculate the beneficiaries’ risk scores and ensure that the Medicare Program does 18 not make improper risk adjustment payments to MA Organizations or that the Program 19 recovers improper payments that were already made. 20 71. 21 otherwise fully complying with all terms and conditions of its contract with CMS,” 22 regardless of any relationship it may have with a downstream or related entity. 42 23 C.F.R. § 422.504. Thus, an MA Organization cannot delegate away its ultimate 24 responsibility for its obligations to the Medicare Program. 25 72. 26 the end of the payment year at issue. For example, for the 2012 payment year, MA 27 Organizations could submit diagnosis codes relating to 2011 date of service medical 28 encounters until February 15, 2013. MA Organizations must ensure the validity of the diagnoses they submit. Among An MA Organization “maintains ultimate responsibility for adhering to and The final deadline for RAPS data submissions is generally four to six weeks after 22 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 23 of 79 Page ID #:1313 1 73. 2 invalid diagnoses in order to withdraw them. See 42 C.F.R. § 422.310(g)(2)(ii) 3 (codifying pre-existing process permitting, after the final deadline, only corrections to 4 delete diagnoses from previously-submitted risk adjustment data). Accordingly, MA 5 Organizations can delete invalid diagnoses both before the deadline for RAPS data 6 submissions for a payment year (known as “open-period deletes”) and after the deadline 7 for RAPS data submissions for a payment year (known as “closed-period deletes”). 8 74. 9 monthly payments made during the payment year are interim payments. After the final 10 submission deadline (February 15, 2013 in the example given above), CMS determines 11 if any adjustments to these interim monthly payments are necessary based on all 12 diagnoses submitted for each beneficiary up until the final submission deadline 13 (excluding those diagnoses that were deleted prior to the deadline) and re-calculates each 14 beneficiary’s risk score for the payment year to determine if it has changed and whether 15 a plus or minus adjustment to the payment for the beneficiary is necessary. If the 16 beneficiary’s risk score is higher because of the submission of additional diagnoses for 17 that beneficiary, CMS makes a final reconciliation payment of any additional payment 18 owed to the plan for that beneficiary for that payment year. Conversely, if the 19 beneficiary’s risk score is lower because of the deletion of diagnoses for that beneficiary 20 prior to the final submission deadline, CMS recovers the funds associated with the 21 deleted diagnoses as part of this final reconciliation payment process. 22 The final deadline is only a submission deadline; it does not pertain to deleting Because the final submission deadline is after the completion of the payment year, B. MA Organizations Must Attest to the Validity of Their Data 23 75. After the final submission deadline but before their receipt of the final 24 reconciliation payments, MA Organizations must attest to the validity of their risk 25 adjustment data, including diagnoses, in a Risk Adjustment Attestation submitted to 26 CMS. Specifically, the chief executive officer, chief financial officer, or an individual 27 delegated with authority to sign on behalf of one of these officers, and who reports 28 23 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 24 of 79 Page ID #:1314 1 directly to such officer, must certify that the risk adjustment data that the MA 2 Organization submitted to CMS was accurate, complete, and truthful. 3 76. 4 Attestation and the submission of this Attestation to CMS is a condition of receiving 5 Risk Adjustment payments. 6 77. 7 including diagnoses, submitted for risk adjustment payments under the prescription drug 8 program. Under the applicable Part D regulation, these attestations are referred to as 9 certifications. 42 C.F.R. § 423.505(k). 10 78. An MA Organization must request payment on a document that contains this The Part D regulations include a similar attestation for risk adjustment data, Every year, each MA Organization agrees in writing that: 11 [a]s a condition for receiving a monthly payment under paragraph B of this 12 article, and 42 CFR Part 422 Subpart G, the MA Organization agrees that its 13 chief executive officer (CEO), chief financial officer (CFO), or an 14 individual delegated with the authority to sign on behalf of one of these 15 officers, and who reports directly to such officer, must request payment 16 under the contract on the form[] attached hereto as . . . Attachment B (risk 17 adjustment data) which attest to (based on best knowledge, information and 18 belief, as of the date specified on the attestation form) the accuracy, 19 completeness and truthfulness of the data identified on these attachments. 20 ... 21 2. Attachment B requires the CEO, CFO, or an individual delegated with 22 the authority to sign on behalf of one of these officers, and who reports 23 directly to such officer, must attest to (based on best knowledge, 24 information and belief, as of the date specified on the attestation form) that 25 the risk adjustment data it submits to CMS under 42 CFR § 422.310 are 26 accurate, complete, and truthful. The MA Organization shall make annual 27 attestations to this effect for risk adjustment data on Attachment B and 28 according to a schedule to be published by CMS. If such risk adjustment 24 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 25 of 79 Page ID #:1315 1 data are generated by a related entity, contractor, or subcontractor must also 2 attest to (based on best knowledge, information, and belief, as of the date 3 specified on the attestation form) the accuracy, completeness, and 4 truthfulness of the data. [422.504(l).] 5 79. MA Organizations have an obligation to acquire knowledge, information, and 6 belief about their risk adjustment data, including diagnoses, in order to both submit such 7 data and attest to the accuracy and truthfulness of the data. Nearly 17 years ago, CMS 8 put MA Organizations on notice that they were “responsible for making good faith 9 efforts to certify the accuracy, completeness, and truthfulness of the encounter [i.e., risk 10 adjustment] data submitted” for payments from the Medicare Program. 65 Fed. Reg. 11 40,170, 40,268 (June 29, 2000) (emphasis added); see also Medicare Managed Care 12 Manual, Chapter 7, at § 111.7 (February 2004). When MA Organizations fail to act in 13 good faith and turn a blind eye to their submission of inaccurate or untruthful data, their 14 Risk Adjustment Attestations are false. 15 THE FACTS 16 80. Since at least 2005, United knew that diagnoses submitted to Medicare for risk 17 adjustment payments had to satisfy various criteria and be supported and validated by the 18 medical records of the beneficiaries its MA Plans. United also knew that many provider- 19 reported diagnoses were not supported and validated by the beneficiaries’ medical 20 records and that it was obliged to undertake good faith efforts to identify and delete those 21 unsupported and invalid diagnoses. Moreover, United knew that it was obligated to 22 “look both ways” at the results of its chart reviews and delete unsupported provider- 23 reported diagnoses. Nonetheless, United conducted millions of medical record reviews 24 as part of its revenue-generating national Chart Review Program, turned a blind eye to 25 the negative results of those reviews showing hundreds of thousands of unsupported 26 diagnoses that it had previously submitted to Medicare, and knowingly and improperly 27 avoided repaying Medicare for at least over a billion dollars in risk adjustment payments 28 to which it was not entitled. Similarly, United disregarded information from its RACCR 25 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 26 of 79 Page ID #:1316 1 Program about invalid coding practices by its incentivized capitated and gain-sharing 2 providers and failed to repay Medicare for additional erroneous payments based on their 3 invalid diagnoses. 4 I. United Knew That Many Provider-Reported Diagnoses Were Invalid 5 And That It Was Obligated To Undertake Good Faith Efforts To 6 Identify And Delete Them 7 81. In 2005, as part of its acquisition of PacifiCare, United retained PacifiCare 8 employees who knew the requirements for the submission of valid diagnoses, the 9 obligation to identify and delete invalid codes, and the various problems relating to the 10 invalidity of provider-reported diagnoses. 11 82. 12 former PacifiCare employees knowledgeable about risk adjustment. Dumcum had been 13 PacifiCare’s Chief Financial Officer and became United’s Vice President of Finance. 14 Will had been a Principal Analyst at PacifiCare who designed risk adjustment programs 15 and joined United as the Program Manager for United’s national Chart Review Program. 16 Holt had been a Project Manager for Network Management Operations at PacifiCare and 17 became the Manager of United’s Provider Outreach for its Risk Adjustment Program. 18 Leal had been an Executive Director of Provider Training and Development for 19 PacifiCare and became United’s Regional Vice President for Market Consultation. 20 83. 21 PacifiCare office in the Central District of California. Thereafter, Holt and Leal worked 22 at the Ingenix office in the Central District of California. 23 84. 24 various sources, including CMS. PacifiCare also had conducted various risk adjustment 25 programs, including a chart review program, and had provided training to healthcare 26 professionals concerning medical record documentation and diagnosis coding. 27 85. 28 Industry Collaboration Effort (“ICE”) and ICE’s Risk Adjustment Data Acquisition & For example, Jeffrey Dumcum, Stephanie Will, Pam Holt, and Pam Leal were all From 2005 to 2007, Dumcum, Will, Holt, and Leal worked for United at the The PacifiCare employees obtained their knowledge about risk adjustment from In addition, PacifiCare had been a member of an industry association called the 26 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 27 of 79 Page ID #:1317 1 Reporting (“RADAR”) team. ICE was an association of MA Organizations and their 2 MA Plans as well as provider groups in California that served beneficiaries in MA Plans. 3 ICE focused on risk adjustment issues and other matters of particular importance to the 4 managed care industry. After the PacifiCare employees were retained by United, they 5 continued to participate in ICE and its RADAR team. In the mid-2000s, Leal was the 6 President and a member of the Board of Directors of ICE. 7 86. 8 the “source of truth” for purposes of providing valid diagnosis data for risk adjustment 9 payments. For example, in a March 2006 email, Will acknowledged that diagnosis data The former PacifiCare employees knew that a beneficiary’s medical record was 10 had to be “fully supported by medical record documentation.” Others at United also 11 understood this, including Patty Brennan who, when she was Director of Retrospective 12 Services (including chart review services) at Ingenix, acknowledged that CMS’ Risk 13 Adjustment Participant Guide established that the medical record was the “one source of 14 truth” for MA Organizations to ensure that they were submitting accurate data to CMS 15 for risk adjustment payments. In fact, United sent notices to physician groups 16 instructing them that they should only report to United “diagnosis codes that can be 17 supported by the documentation in the medical record.” 18 87. 19 records are the “source of truth.” In 2010, ICE’s RADAR team issued a guidance 20 document highlighting that CMS requires complete and accurate documentation of 21 medical conditions for the submission of diagnoses, that only diagnoses depicting 22 documented medical conditions which required care or affected patient care are valid, 23 and that diagnosis codes cannot be submitted “until [the provider] is sure the patient has 24 the condition.” The ICE RADAR Physician Education Work Group also issued a similar 25 document called “Best Practices for Risk Adjustment,” which advised that “ICD-9-CM 26 coding requires documentation of the diagnosis in the medical record as well as 27 evaluation and management. Documentation should indicate how this diagnosis 28 impacted this episode of care.” In 2012, ICE also issued a Medical Record In addition, United also knew from its involvement in ICE that enrollees’ medical 27 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 28 of 79 Page ID #:1318 1 Documentation Tips sheet, which once again warned MA Organizations and providers 2 not to code diagnoses that are probable, suspected, questionable, or “working diagnoses” 3 and also not to code diagnoses when medical records use “other similar terms indicating 4 uncertainty.” More recently, in 2013, ICE issued a “Documentation Newsletter” that 5 repeated earlier advice and also cautioned that “[c]oding guidelines prohibit coders 6 from making assumptions” regarding whether a diagnosis is or is not substantiated by a 7 patient’s medical record. (Emphasis in the original.) That is, the medical record must 8 “clearly reflect” the medical condition. 9 88. However, the former PacifiCare employees were aware that provider-reported 10 diagnoses often did not comply with CMS requirements and were often inconsistent with 11 the information in their patients’ medical records. According to Dumcum, when he was 12 the Chief Financial Officer of PacifiCare, he and others there knew “in Medicare 13 Advantage that the claims did not always match the medical record documentation. So 14 … we were concerned that it should be a place that we try to improve, that we try to 15 educate and try to identify things to make that better.” In addition, Will, Holt and other 16 former PacifiCare employees were aware of common diagnosis coding errors made by 17 providers. They learned of these problems from PacifiCare employees working in the 18 field with physicians, reports of physician-coding trends, and reports from PacifiCare- 19 employed certified coders. For example, a June 2003 PacifiCare PowerPoint 20 Presentation by Will, Holt, and other PacifiCare employees identified diabetes as a 21 medical condition that was often miscoded. 22 89. 23 data validation review conducted by the Government of diagnosis codes previously 24 submitted for medical encounters that occurred in 2003 (i.e., encounters with 2003 dates 25 of service). The PacifiCare employees were aware that the results of CMS’ medical 26 record reviews showed that approximately 30 percent of the provider-reported diagnoses 27 were invalid. The results from this review also put them on notice that providers were 28 reporting codes that were just plain wrong, were coded from laboratory reports, and did In 2005, the PacifiCare employees, including Will and Holt, were also aware of a 28 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 29 of 79 Page ID #:1319 1 not reflect current medical conditions – all of which was in breach of the fundamental 2 rules that the diagnosis codes submitted to Medicare must be accurate and truthful, based 3 on face-to-face visits (e.g., not lab reports), reflect current conditions, and otherwise be 4 valid. 5 90. 6 are not entitled to risk adjustment payments based on diagnoses that are unsupported by 7 the beneficiaries’ medical records, and that CMS expected health plans to delete 8 incorrect diagnosis codes submitted for risk adjustment payments. They also knew, 9 based on their experiences at PacifiCare, that CMS could audit the diagnoses MA Moreover, the former PacifiCare employees were aware that MA Organizations 10 Organizations submitted for risk adjustment payments. 11 91. 12 explained that it expected MA Organizations to correct invalid diagnoses submitted to 13 Medicare for risk adjustment payments. Holt reported this to Will and suggested 14 creating a spreadsheet to give to providers for them to use to inform PacifiCare of invalid 15 diagnoses and allow PacifiCare to delete them. As part of this discussion, Leal 16 explained to Will that, if provider groups “during their chart audits find that physicians 17 have documented rule-out or history-of but coded as if the member had [the medical 18 condition,] they want to be able to fix it so when we get audited again by CMS it is 19 fixed.” Leal further stated that “[o]bviously, as issues are identified there will need to be 20 education to physician[s] on changing their practice of coding incorrectly (as you 21 remember Dr. Norman mentioned habits doctors have, that we will need to break).” Holt 22 agreed that provider groups In April 2005, Holt participated in a “CMS data validation call” in which CMS 23 need something ‘standardized and formalized’ so they know what fields to 24 report if and when they find any obvious discrepancies. They are the type of 25 thing that Pam [Leal] stated in her email below, the code of the actual disease 26 when the documentation clearly only supports ‘history of’ or ‘suspected’ (and 27 then it was not confirmed), or an obvious miscode; the things that Melissa 28 29 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 30 of 79 Page ID #:1320 1 [Ferron] is finding in the data validation. Provider groups will find this 2 during their own chart audits. 3 Holt then reiterated that CMS “was very firm that we need to be doing this, so I would 4 expect [CMS] will look for our process when they get around to more formally auditing 5 our oversight of this process.” 6 92. 7 and Holt knew that CMS had confirmed, in responses to questions about its RADV 8 audits, that it would invalidate diagnosis codes submitted to Medicare that were not 9 supported by the beneficiaries’ medical records. In addition, in August 2006, a year after the April 2005 CMS validation call, Will 10 93. After United’s acquisition of PacifiCare, the former PacifiCare employees began 11 educating others at United about risk adjustment. In particular, Dumcum made formal 12 presentations to various United employees, including senior executives. Dumcum gave a 13 series of presentations where he explained to other United employees that “[p]rovider 14 coding is highly inaccurate and incomplete” and that “more than 30% of coded 15 conditions are not supported by CMS validation findings.” United senior management 16 such as Jerry Knutson, the Chief Financial Officer of the group that managed United’s 17 MA Plans from 2003 to 2009, participated in meetings in which Dumcum made these 18 presentations. 19 94. 20 reported diagnoses. United tracked the risk scores for Medicare beneficiaries cared for 21 by its providers and, for various providers, saw increases in risk scores that were 22 significantly above the norm. United also generated reports that identified the providers 23 with abnormally high average risk scores. Prevalence reports also identified specific 24 medical conditions that were reported by various providers at rates significantly above 25 average. In September 2006, Will, Holt, Leal, and others generated a list of providers 26 that were outliers, which, at that time, they defined as providers with significant 27 increases in their patients’ risk scores. This information made them “question the 28 validity” of these providers’ codes. Furthermore, United’s own data revealed and confirmed problems with provider- 30 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 31 of 79 Page ID #:1321 1 95. In 2007, United assigned to Dumcum and the other former PacifiCare employees 2 the responsibility for creating and then managing a risk adjustment service group within 3 Ingenix (later known as Optum). This enabled United to move the risk adjustment 4 operations from Ovations (the predecessor of UnitedHealth Medicare & Retirement) to 5 Ingenix. Ovations (then UnitedHealth Medicare & Retirement) became Ingenix’s 6 internal “client.” This move also enabled Ingenix to offer its risk adjustment services to 7 other MA Organizations which it referred to as its “commercial clients.” Dumcum 8 became Senior Vice President and Will became Vice President of Risk Adjustment 9 Programs for this new risk adjustment service group. After 2007, Ingenix hired 10 additional employees and increased the size of the Ingenix risk adjustment group, 11 including the size of the group in this District. The Ingenix risk adjustment group in this 12 District was responsible for, among other things, risk adjustment data/diagnosis codes 13 submissions, risk adjustment data remediation and the deletion of invalid diagnoses, risk 14 adjustment data analytics and finance (e.g., tracking the results and financial impact of 15 the Chart Review and Claims Verification Programs), provider outreach and programs 16 relating to risk adjustment, and the Chart Review Program operations. 17 96. 18 provider-reported diagnosis codes. Dumcum knew that United had providers who were 19 paid on either a fee-for-service or capitated basis and who were reporting unsupported 20 diagnoses and that both needed to improve their validation rates. 21 97. 22 discussions about conducting an Internal Data Validation (“IDV”) Program focused on 23 the validity of provider-reported diagnoses. The purpose of the IDV Program was to 24 determine if the physicians’ medical records supported the diagnoses that they reported 25 to United and United submitted to Medicare for risk adjustment payments. 26 98. 27 Submissions at Ingenix, were informed of specific provider-reported codes that were 28 reported based on “faulty coding.” For example, Sharp Community Medical Group in In January 2007, Dumcum told United that it had to improve the validation of Prior to this, in 2006, Dumcum, Will, Holt and others had already participated in In February 2007, Will, Holt, and Patricia Rasmussen, Manager of Encounter 31 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 32 of 79 Page ID #:1322 1 California submitted a diagnosis tracking to HCC 155 (Major Head Injury) but there was 2 “no documentation to support intracranial injury.” Likewise, Monarch/South Coast, 3 located in this District, submitted a diagnosis tracking to HCC 17 but there was “[n]o 4 documentation of diabetes or diabetic complication[.]” Will, Holt, and Rasmussen were 5 requested to remove, or delete, these codes before the final submission deadline but 6 Rasmussen replied that they did not have resources and could not do this. 7 99. 8 diagnoses were invalid based on audits performed by CMS and similar internal audits or 9 reviews that United performed. For example, one such audit was performed on In addition, United knew that a significant percentage of provider-reported 10 diagnoses submitted for 2004 date of service medical encounters (e.g. physician office 11 visits) that mapped to 1,231 HCCs. The results were reported in a September 2007 Risk 12 Adjustment Programs presentation made by Dumcum to others at United. The 13 presentation showed that no support was found in the beneficiaries’ medical records for 14 32 percent of the HCCs at issue. That is, the records did not confirm the diagnoses 15 mapping to 32 percent of the HCCs under review. The same presentation showed the 16 results of another audit of diagnoses submitted for 2005 date of service medical 17 encounters that mapped to 1,160 HCCs. It showed that, as of the date of the 18 presentation, 18 percent of the HCCs were “NOT supported” and another 8 percent were 19 most likely not supported. 20 100. In October 2007, Ingenix also emphasized the fact that more than 30 percent of 21 provider-reported diagnoses were unsupported by the beneficiaries’ medical records as a 22 selling point for the risk adjustment services, including data validation services, that it 23 marketed to another MA Organization. 24 101. During the 2007 and 2008 time period, United implemented the IDV Program. 25 The program focused on physicians in California, Missouri, Texas, and Washington who 26 reported more than three times the number of diagnoses than the average. The program 27 was very small in scope, but the results confirmed that providers reported many invalid 28 diagnoses. In January 2008, the results for medical encounters in 2006 showed a 30 32 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 33 of 79 Page ID #:1323 1 percent invalidation rate. That is, approximately 30 percent of provider-reported 2 diagnoses were invalid, which was consistent with what the PacifiCare employees had 3 known since at least 2005. Accordingly, by at least 2007, the former PacifiCare 4 employees imparted their knowledge to others at United that approximately 30 percent 5 of provider reported conditions were not supported by medical records and, by at least 6 2008, United’s own medical record reviews had confirmed that fact. 7 102. Additional evidence demonstrates that, by at least the 2007 and 2008 time period, 8 the invalidity of provider-reported diagnoses was an on-going concern. For example, in 9 September 2008, one of United’s own actuarial consulting subsidiaries, Reden & 10 Anders, identified unsupported diagnosis codes as a “Potential Compliance Risk Area” 11 and warned that “[t]here is no such thing as minimally compliant.” 12 103. In addition, by at least 2008, the former PacifiCare employees had also imparted 13 their knowledge that United was obligated to identify and delete invalid provider- 14 reported diagnoses. For example, a January 2008 United training guide entitled “CMS- 15 HCC Risk Adjustment Training Module,” which was created for provider outreach, 16 recognized that it is the accuracy of medical record documentation and coding that 17 supports entitlement to risk adjusted payments from the Medicare Program. The 18 presentation recognized that accurate medical record documentation is key to accurate 19 risk adjustment payments and necessary to validate payments. In addition, in June 2008 20 emails, Patty Brennan, the Compliance Manager for Ingenix, and Karen Wagor, a Senior 21 Coder and National Trainer for Ingenix, both of whom worked at Ingenix’s Santa Ana 22 office, recognized that United was not entitled to payment based on diagnoses that were 23 not validated by beneficiaries’ medical records and that United risked having to return 24 money to the Medicare Program for risk adjustment payments based on invalid 25 diagnoses: “We could be at risk of losing $$ if there isn’t another piece in the 26 documentation before or after this date of service for the HCC if it can’t be verified with 27 the most accurate validation. The medical record must support all diagnoses coded 28 33 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 34 of 79 Page ID #:1324 1 for the date of service and must be able to stand alone for an audit on those reported 2 diagnosis codes.” (Emphasis in the original.) 3 104. Moreover, United knew it was required to effectively address compliance issues. 4 A United presentation from November 2009 entitled “Audit Management Overview” 5 reflects United’s knowledge that it was legally required to implement an effective 6 compliance program and that “[i]n order to have an effective Compliance Program, an 7 organization must have a robust internal monitoring and auditing process in place.” In 8 addition, in approximately May 2010, Larry Renfro, who was then the Chief Executive 9 Officer of both the Public and Senior Market Group (PSMG) and Ovations, and David 10 Orbuch, who was then the Executive Vice President and Chief Compliance Officer of 11 PSMG, met with CMS regarding United’s Medicare compliance program and 12 represented to CMS that United would “meet and exceed” CMS’ expectations. 13 105. Kimberly Halva, UnitedHealth Medicare & Retirement’s Finance Compliance 14 Officer from 2010 to 2013, also was keenly aware that one of the issues that United had 15 to effectively address was the validity of provider-reported diagnosis. Halva understood 16 that United had an obligation to identify and delete diagnoses that were not supported by 17 its beneficiaries’ medical records. In November 2010, she sent a memorandum to 18 Relator Poehling with recommendations for compliance problems focused on identifying 19 and deleting invalid provider-reported diagnoses. In January 2011, she sent Relator 20 Poehling “a few suggestions for risk mitigation type programs UnitedHealthcare 21 Medicare and Retirement could take to more equally distribute resources between 22 programs dedicated to improper coding or billing and those focused on identifying 23 additional reimbursement opportunities through improved coding/documentation.” One 24 of Halva’s suggestions was for United to perform “General Coding Accuracy Audits,” 25 which were essentially “look both ways” medical record reviews that would identify 26 both incomplete coding (which she referred to as “under-coding”) and inaccurate coding 27 (which she referred to as “over-coding”). Another of her suggestions was for United to 28 “Specifically Identify Traditionally Over-Coded or Incorrectly Coded Conditions,” “such 34 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 35 of 79 Page ID #:1325 1 as stroke or COPD,” and conduct medical record reviews to determine if those 2 traditionally over-coded or incorrectly coded conditions could be validated. 3 106. United was also aware from audits performed for it by its public accounting firm 4 that chart reviews that “looked both ways” was the only way to achieve a full and 5 complete picture of a beneficiary’s health status and that United was obligated to delete 6 invalid provider-reported diagnoses. 7 107. In addition, as part of its compliance efforts, United had, since at least 2008, 8 required financially-incentivized capitated provider groups to submit attestations to it 9 that certified that the diagnoses that they reported to United were valid and met CMS’s 10 requirements. United sent notices to these providers describing these “CMS DATA 11 ACCEPTANCE GUIDELINES.” (Emphasis in the original.) These notices stated that 12 “[a]ll diagnoses must be documented at the time of the patient encounter or after 13 receipt and confirmation of the diagnosis (i.e. Lab or Radiology report) and must 14 be documented in the medical record,” and “[o]nly report diagnosis codes that can 15 be supported by the documentation in the medical record,” and “[a]ll diagnosis 16 codes should be valid . . . .” (Emphasis in the original.) 17 108. In 2010, Holt and Leal were a part of an email chain discussing California 18 provider groups and risk adjustment data validation. Based on Holt’s involvement with 19 ICE’s RADAR team (and possibly other sources), Holt noted that California provider 20 groups are “acutely aware of the importance of data validation” and that risk adjustment 21 data validation was “a subject discussed frequently at the ICE meetings.” In the same 22 email chain, Holt also assured Relator Poehling that provider groups in California would 23 understand the importance of CMS’s RADV audits. 24 109. In November 2009, the results of the IDV audit for one of Ingenix’s commercial 25 clients, another company that owned and operated MA Plans, further confirmed what 26 United already knew, that is, that providers were reporting invalid diagnoses at an 27 alarming rate. The results of the audit showed a 45 percent invalidation rate. Providers 28 in the commercial client’s plans were also in United’s plans. 35 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 36 of 79 Page ID #:1326 1 110. In June 2010, the results of United’s IDV reviews of its California providers once 2 again confirmed an on-going problem with the validity of provider-reported diagnoses. 3 The results showed a 40 percent invalidation rate. Higher than the 30 percent 4 invalidation rate reported in 2005, these results confirmed that a significant percentage of 5 provider-reported diagnoses were invalid. 6 111. Also, in June 2010, an Ingenix Health Reform Implementation (HRI) Report 7 prepared for UHG’s Chief Executive Officer, Steve Hemsley, and other members of the 8 UHG executive team identified compliance as an important issue of immediate concern 9 to United, particularly compliance with the FCA. The Report also shows that United 10 was aware of a new statute making MA Organizations liable under the FCA for reporting 11 and refunding overpayments in an untimely manner. 12 112. Furthermore, in mid-2010, additional Government audits confirmed the problem 13 with invalid provider-reported diagnoses. At that time, the HHS Office of Inspector 14 General (“OIG”) sent United executives draft reports of audits of the risk adjustment 15 data that two of United’s MA Plans (then called PacifiCare of California and PacifiCare 16 of Texas) submitted for payment year 2007. In the draft reports, the OIG concluded that 17 the diagnoses for half the beneficiaries in the California audit and 44 percent of the 18 beneficiaries in the Texas audit were invalid, and that both plans’ practices were not 19 effective for ensuring that the diagnoses they submitted to CMS complied with CMS 20 requirements. After consideration of United’s responses to the draft audit reports, the 21 OIG issued final reports concluding that the health risk scores for 45 percent of the 22 beneficiaries in the California audit and 43 percent of the beneficiaries in the Texas audit 23 were invalid because the diagnoses were not supported by the beneficiaries’ medical 24 records or were uncertain or unconfirmed diagnoses. In both the draft and final reports, 25 the OIG provided examples of unsupported diagnoses. 26 113. The results of prior RADV audits were also of concern to United. When CMS 27 announced in February 2011 that it planned to extrapolate the sample audit results, 28 Charles Hansen, the Vice President of Finance and Underwriting at UnitedHealth 36 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 37 of 79 Page ID #:1327 1 Medicare & Retirement, left a voicemail message for Relator Poehling stating: “I 2 haven’t been close to this for . . . three or four years now probably, but back that far I 3 know the results of our RADV audits were concerning, to say the least. That we had 4 significant . . . errors or undocumented RAFs [i.e., undocumented diagnoses] in our 5 claim submissions.” He asked Poehling: “I’m sure we’re careful about how we 6 communicate these findings, but can you give me a sense of kind of more recent RADV 7 audits, what impact could this have? So just generally, are we still sort of, I’ll say as bad 8 as we were a few years ago? . . . So, what’s the potential impact on revenue of 9 extrapolating these RADV audit findings . . . ?” 10 114. Because of the problem with provider-reported diagnoses, in 2010, United started 11 its RACCR Program, described more fully at Section VI below of the Government’s 12 Complaint, and its CV Program, described more fully at Section IV below. Mary 13 Hammond, one of the UnitedHealthcare Medicare & Retirement employees that oversaw 14 the RACCR Program, described it as “an important part of Medicare & Retirement’s 15 efforts to meet CMS requirements to submit accurate risk adjustment data.” The CV 16 Program was also supposed to satisfy this basic payment rule. 17 115. However, the major effect of the CV and RACCR Programs was that they 18 highlighted that United was not entitled to a significant amount of the risk adjustment 19 payments that it had received from Medicare because significant percentages of the 20 diagnoses it had submitted were unsupported and invalidated by its beneficiaries’ 21 medical records. For instance, as explained in more detail in paragraph 162 below, in 22 December 2010, the results of the first pilot phase (Phase I) of the CV Program showed 23 that provider-reported diagnoses were unsupported for over 50 percent of the medical 24 records reviewed. Over a year later, as explained in more detail in paragraph 164 below, 25 United completed the second pilot phase (Phase II) of the CV Program on 17,398 charts 26 and the results were even more striking. United identified 4,786 invalid diagnoses to be 27 deleted, which was greater than the number of additional diagnoses (ADDS) identified 28 by United’s coders based on their review of the same medical records. In February 37 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 38 of 79 Page ID #:1328 1 2012, Dumcum, who was responsible for the CV Program, met with Relator Poehling 2 and Theisen, who was the Chief Financial Officer of UnitedHealth Medicare & 3 Retirement at that time, and reported: “I’m deleting as much as I’m adding at the end of 4 the day. And I don’t know if—you know, what I want to do.” He also stated: “we’re 5 getting more and more red here,” and that CV “eats in very substantially” to the revenue 6 from ADDS from chart reviews. 7 116. In addition, by February 2012, United had “identified and deleted 5,519 bad 8 codes” as the result of its RACCR reviews of the medical records for only 13,451 9 beneficiaries. On a beneficiary basis, the results showed more than a 40 percent 10 invalidation rate. 11 117. Over time, United also obtained more information about incorrect diagnosis 12 coding by providers, some of which information confirmed what United already knew 13 years before about this problem. By at least 2013, Tracey Bradberry, a United employee 14 who was a certified coder, was aware of various reasons for invalid provider-reported 15 diagnoses, including, for example, the “[c]linical findings and/or treatment does not 16 support the diagnosis” and “[n]ot a current condition.” In addition, in March of 2014, 17 Melissa Ferron, a coding consultant, sent United a list identifying various HCC codes 18 and diagnosis codes that were unlikely to validate based on medical record reviews 19 because providers frequently used the codes incorrectly. These included HCCs 17 20 (diabetes with acute complications), 96 (specified heart arrhythmias), 104 (monoplegia, 21 other paralytic syndromes), and 111 (chronic obstructive pulmonary disease). 22 118. More recently, other United employees have also recognized United’s obligation 23 to ensure that diagnoses United submits for risk adjustment payments are supported and 24 validated by its beneficiaries’ medical records. They also understood that United was 25 obligated to correct invalid data provided by it to Medicare for risk adjustment 26 payments. For example, in an April 2013 email, Marybeth Meyer, the new Director of 27 Finance for UnitedHealthcare Medicare & Retirement who succeeded Relator Poehling, 28 stated that “as a Medicare Advantage Plan, if we become aware of a coding issue that 38 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 39 of 79 Page ID #:1329 1 impacted the revenue received from CMS for risk adjustment, we are obligated to 2 investigate and correct the submissions to CMS.” In response to Meyer’s email, Halva, 3 the Compliance Officer, agreed. She also stated that United had “an obligation to go 4 back to CMS and correct submissions that impact the revenue received for risk 5 adjustment purposes.” Halva did not think that anything was wrong with the 6 requirement that United delete unsupported diagnoses. Moreover, she believed that 7 “arguably” this obligation existed for ten years because of the ten-year statute of 8 limitation for the Government to pursue FCA claims against United for its failure to 9 delete unsupported diagnoses. 10 119. As another example, in February 2014, Melissa Sedor, a Director at 11 UnitedHealthcare Medicare & Retirement, told Marybeth Meyer that United had 12 ultimate responsibility to ensure that the diagnoses that it submitted to Medicare for 13 payment were valid. She stated that it was United’s “responsibility to submit accurate 14 records to CMS. Even if the provider submits bad data to us, the responsibility is on us 15 to be submitting the correct data.” Sedor further explained that this was the reason 16 United implemented compliance programs like RACCR and CV. 17 120. Finally, other United employees in addition to Halva were aware of United’s 18 potential FCA liability for submitting invalid diagnoses to CMS for risk adjustment 19 payments or for failing to delete them. For example, in 2010, Carol Thompson, the 20 National Manager for Ingenix’s Risk Adjustment Programs who worked at the Ingenix 21 office in this District, provided the following reason as justification for funding for a 22 coding vendor who was assisting United with a CMS RADV audit: “Politics and 23 Congressional scrutiny put an exclamation point on our need to demonstrate that the data 24 submitted to CMS for risk adjustment payment is valid. Data that drives payment from 25 the government can be reviewed to determine if the claims are ‘valid.’ Improper claims 26 submission can fall under the False Claims Act.” Also, as alleged more fully below, 27 Dumcum informed Relator Poehling that the FCA was a consideration in deciding 28 39 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 40 of 79 Page ID #:1330 1 whether United should implement the CV Program and “look both ways” at the results 2 of its Chart Review Program. 3 II. United’s “Look One Way” Chart Review Program 4 121. Over the last decade, United has obtained billions of dollars of risk adjustment 5 payments from Medicare from its national Chart Review Program because it only looked 6 one way at the results of its coders’ chart reviews. As part of this program, United’s 7 coders performed “blind” reviews of the beneficiaries’ medical records. That is, the 8 coders did not know which, if any, diagnoses had been reported by the providers who 9 treated the beneficiaries and created the medical records. Thus, rather than confirming 10 diagnoses already reported by the providers or identifying only additional diagnoses 11 supported by the records, the coders were instructed to look for all medical conditions 12 purportedly documented in the records, record all diagnosis codes identifying those 13 conditions, and give all of those codes to United. But, United utilized results of these 14 chart reviews – that is, the coders’ list of diagnosis codes – for the sole purpose of 15 identifying diagnosis codes that the providers had not reported and submitting ADDS for 16 additional risk adjustment payments. United did not utilize the coders’ lists of diagnosis 17 codes to determine if the providers had reported codes that were not supported by their 18 own medical records. If United had looked both ways and deleted these unsupported 19 and, thus, invalid provider-reported codes, United’s national Chart Review Program 20 would have been far less lucrative. 21 122. United’s national Chart Review Program was its largest risk adjustment revenue- 22 generating program. United started this program in 2006. During the first few years of 23 the program, United reviewed hundreds of thousands of charts each year in order to mine 24 them for additional diagnosis codes. 25 123. Dumcum, Will, and other former PacifiCare employees became key participants in 26 designing, implementing, and managing United’s revenue-generating national Chart 27 Review Program. Will was the Program Manager for the program. 28 40 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 41 of 79 Page ID #:1331 1 124. In 2009, United’s senior management significantly expanded the size of United’s 2 national Chart Review Program by providing Ingenix with the funds to acquire a coding 3 vendor called AIM Healthcare Services, Inc. and integrating Ingenix’s operations in 4 Santa Ana, California, and the AIM coding operations in Tennessee, and by providing 5 Ingenix with the funds to develop an in-house chart review computer database system 6 called ChartSync so that Ingenix’s in-house coders and third-party coding vendors could 7 electronically review medical records and electronically record the coding results of their 8 reviews. At or about the same time, United also expanded the size of its Chart Review 9 Program by providing Ingenix with the funds to open offices in foreign countries (the 10 Philippines and India) where it could hire foreign workers to review beneficiaries’ 11 medical records in order to try to find additional diagnosis codes. 12 125. Due to the increased resources devoted to United’s Chart Review Program, the 13 number of medical records reviewed in the hunt for additional codes significantly 14 increased over time. For the first few years of its Chart Review Program, United 15 reviewed between 500,000 and 600,000 charts each year. According to a presentation by 16 Dumcum, United reviewed 600,000 charts in 2006 and, as of September 21, 2007, had 17 completed 400,000 chart reviews so far that year with 200,000 charts remaining for 18 review during the fourth quarter of 2007. 19 126. By 2010, the number of chart reviews had increased substantially to 20 approximately 850,000 charts for that year. For 2011 to 2014, United reviewed 21 approximately 1.5 million charts a year. The Government believes that, after 2014, 22 United’s national Chart Review Program was most likely similar in size. 23 127. As part of this national Chart Review Program, United focused primarily on its 24 providers paid on a fee-for-service basis. United or its vendors obtained medical records 25 from thousands of these providers throughout the United States. United sent these 26 medical records to coders that it employed in Tennessee, India, and the Philippines. It 27 also hired coding vendors to review these medical records. Those vendors were located 28 in various locations within and outside of the United States. 41 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 42 of 79 Page ID #:1332 1 128. United’s own physician groups also had chart review programs. WellMed had a 2 group called DataRaps that conducted reviews of its physicians’ medical records for 3 patients in United’s MA Plans. Southwest Medical Associates conducted some of its 4 own reviews though it also relied on United’s coding operations in Tennessee and coding 5 vendors for its chart reviews. 6 129. In 2006, United obtained approximately $270 million in additional risk adjustment 7 payments from the ADDS that it submitted based on its national Chart Review Program. 8 The return on investment was substantial (approximately $250 million) as it cost United 9 only approximately $18 million to review the charts in 2006. 10 130. Not unexpectedly, as the number of medical records reviewed significantly 11 increased over the years so did United’s earnings from its Chart Review Program. 12 United received approximately $426 million in additional risk adjustment payments from 13 ADDs that it submitted to Medicare based on the medical record reviews conducted as 14 part of its Chart Review Program for the 2011 payment year, approximately $455 15 million for the 2012 payment year, approximately $758 million for the 2013 payment 16 year, and approximately $882 million for the 2014 payment year. 17 131. For payment years 2010 to 2015 combined, United obtained over $3 billion in 18 additional risk adjustment payments from Medicare due to the ADDs which it submitted 19 based on medical record reviews conducted as part of its national Chart Review 20 Program. 21 132. United’s Chart Review Program was conducted according to an annual cycle. For 22 example, in the spring of 2012, United started its collection and review of medical 23 records relating to medical encounters in 2011 (i.e., with 2011 dates of service). These 24 efforts then intensified through the remainder of the year until the final deadline for 25 submitting diagnosis codes to Medicare for payment year 2012, which was February 15, 26 2013. 27 133. Prior to the start of each annual cycle, United’s senior executives set a revenue 28 target for the program. After chart reviews started, United’s senior executives then 42 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 43 of 79 Page ID #:1333 1 closely monitored the progress of the program and, if the forecast did not look like the 2 program would achieve the revenue target, United made changes to the diagnostic 3 coding being performed by the coders. For instance, when United was not achieving the 4 return on investment it expected from chart reviews for the 2012 payment year, it 5 “liberalized” its coding policies and engaged in a “Recode Project” consisting of re- 6 reviewing 900,000 charts that had already been mined once for additional diagnoses. 7 United liberalized its coding policies to enable the coders to identify more diagnosis 8 codes purportedly supported by the beneficiaries’ medical records. In this second-round 9 review, the coders did identify more codes purportedly supported by the records based 10 on the liberalized coding policies and United submitted those to Medicare for additional 11 risk adjustment payments. 12 134. Despite its aggressive coding to submit as many ADDS as possible to meet its 13 annual revenue targets for its Chart Review Program, United attested to the validity of 14 all of these ADDS in its annual Risk Adjustment Attestations submitted to Medicare. If 15 the results of United’s chart reviews were so reliable that United could attest to the 16 validity of all of the ADDS, then the results were of equal reliability for United to have 17 deleted all previously-submitted diagnoses invalidated by the reviews. To the extent that 18 United calls into question the results of its own chart reviews in not finding support for 19 and not validating hundreds of thousands of diagnoses, it also calls into question the 20 overall reliability of its chart reviews and validity of the ADDS. Under those 21 circumstances, United acted with reckless disregard for the truth by submitting the 22 ADDS and attesting to their validity, knowing it was not entitled to payments by 23 Medicare based on such unreliable data and the United States is entitled to recover those 24 payments in this action. 25 26 III. United Knew It Was Obligated To Look Both Ways At The Results Of Its Chart Reviews And Make DELETES As Well As ADDS 27 135. By 2008, Relator Poehling and others began to question United’s practice of 28 ignoring the negative results of its blind chart reviews invalidating many provider43 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 44 of 79 Page ID #:1334 1 reported diagnoses. In May 2008, Relator Poehling participated in a meeting in which 2 he discussed this issue with Will. She explained to Relator Poehling that, as part of its 3 Chart Review Program, United did not look at whether provider-reported diagnoses were 4 not substantiated by the results of the coders’ blind chart reviews. Poehling and Will 5 discussed the idea of United changing this process to compare the results of the blind 6 chart reviews to the provider-reported diagnoses that United submitted to CMS and 7 deleting the unsupported diagnoses. 8 136. United’s need to address problems with the diagnoses reported by providers paid 9 on a fee-for-service basis continued to be an issue. In March 2009, this question arose 10 during a discussion among Ingenix employees about United’s IDV Program. During 11 that discussion, Ronnie Grower, Vice President of Market Consultation for Ingenix, 12 noted that United’s IDV program excluded fee-for-service providers and questioned 13 whether United would have another program to address diagnoses reported by fee-for- 14 service providers given that “there are common coding errors that get reported across all 15 providers.” In fact, other than its short-lived CV Program, United never had a medical 16 record review program like the IDV or RACCR Programs to address the validity (or lack 17 thereof) of diagnoses reported to it by its fee-for-service providers. 18 137. In 2009, these internal discussions continued. In early April 2009, Relator 19 Poehling, Dumcum, Will, Halva, Janice Redmond (the Senior Vice President of Market 20 Outreach for Ingenix’s risk adjustment group, who worked in Santa Ana, California) and 21 others participated in a discussion about compliance risks, including provider over- 22 coding (i.e., invalid coding). One of the items on the agenda that they discussed 23 included “[a]uditing under & over coded conditions” as part of the chart review process, 24 that is, “looking both ways.” A few weeks later, Relator Poehling again met with Will 25 and others to discuss what United should do when provider-reported diagnosis codes 26 were inconsistent with the results of the coders’ blind chart reviews of the providers’ 27 medical records. At this second meeting in April 2009, the participants began to design 28 a methodology for “looking both ways.” This led to subsequent discussions about 44 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 45 of 79 Page ID #:1335 1 creating a CV Program in order to look both ways at the coders’ blind chart review 2 results and make DELETES as well as ADDS. 3 138. A few months later, in August 2009, in an internal presentation, Redmond noted 4 that chart reviews only looked for additional revenue and suggested that the process 5 should also include a certain amount of charts that are reviewed to determine the validity 6 of providers’ codes. Redmond was concerned that providers paid by United on a fee-for- 7 service basis were not being audited to validate their diagnoses. Her presentation states: 8 “Known problematic codes are not audited across plans/providers raising the risk of 9 error.” 10 139. In early 2010, Will sent Dumcum a presentation proposing the CV Program. The 11 stated goal of the program was to improve the accuracy of the diagnosis data that United 12 submitted to CMS. Will believed that CV would achieve this goal. 13 140. In May 2010, Relator Poehling discussed the potential creation of the CV Program 14 with Dumcum. Dumcum referenced the Department of Justice’s enforcement of the 15 FCA as a consideration. 16 141. According to Halva, while she was the UnitedHealthcare Medicare & Retirement 17 Compliance Officer from 2010 to 2013, the “general consensus from the 18 [UnitedHealthcare Medicare & Retirement] point of view was . . . that any time we 19 opened a chart we should be looking both ways.” 20 142. 21 acknowledged that United should “look both ways” at the results of its blind chart 22 reviews. By at least that time, however, United should and could have compared the 23 results of all chart reviews to the provider-reported diagnoses and deleted all of the 24 invalid provider-reported diagnoses that it previously had submitted to the Medicare 25 Program. United also should and could have done this contemporaneously with 26 submitting ADDS based on the same chart reviews. Instead, United embarked on a very 27 slow, phased development of its CV Program. The senior executives authorized only a 28 pilot test program to look at the negative results (i.e., the results showing that provider- In the fall of 2010, after two years of discussions, United senior executives 45 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 46 of 79 Page ID #:1336 1 reported diagnoses were invalid) from only a very small sample of United’s chart 2 reviews. Moreover, the CV process was designed to “save,” – that is, avoid reporting – 3 the provider-reported diagnoses invalidated by the blind chart reviews conducted as a 4 part of the Chart Review Program. United attempted and sometimes did save some of 5 these invalid codes by re-reviewing the beneficiaries’ medical records, sometimes 6 multiple times, to try to glean any support, even doubtful or ambiguous support, for the 7 provider-reported diagnoses at issue. 8 143. In August 2011, Relator Poehling made clear to Scott Theisen, the Chief Financial 9 Officer of UnitedHealthcare Medicare & Retirement, that Poehling did not believe it was 10 appropriate to conduct chart reviews unless and until CV was fully implemented and 11 United “looked both ways.” Theisen was one of the United senior managers charged 12 with making decisions regarding the implementation and design of the CV Program. At 13 that time, in an email to another United employee, Poehling wrote: “You (and Scott 14 [Theisen]) know where I stand on chart reviews without full CV in place … I wouldn’t 15 do them. Scott, though, is the decision maker . . . .” 16 144. In September 2011, Mary Hammond, Associate Director of Strategy and Support 17 for UnitedHealthcare Medicare & Retirement’s risk adjustment team, attended an 18 industry conference in Washington, D.C. on risk adjustment. She reported that it was 19 “great to get a perspective on what the activity in DC may mean for risk adjustment. 20 More audit protection (looking both ways compliance programs), . . . and less reliance 21 on chart review are the recommendations.” (Emphasis added.) 22 145. In approximately November 2011, a “factual and unbiased” presentation was 23 created for UHG’s Chief Executive Officer Steve Hemsley, to provide him information 24 about Optum’s and its competitors’ risk adjustment programs and other risk adjustment 25 services. The presentation described “Compliance” as “the True Value of Claims 26 Verification.” The presentation further noted that the medical record is the “source of 27 truth” and that looking at this “source of truth” had a negative revenue impact because 28 46 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 47 of 79 Page ID #:1337 1 comparing provider-reported diagnoses with the information in the providers’ medical 2 records resulted in having to delete some of their diagnoses. 3 146. In December 2012, Mike Jacobson, Program Business Analyst/Project 4 Management at Optum, sent Patty Brennan an updated version of Optum’s Business 5 Vision Document (BVD) for the marketing of the CV Program to commercial clients 6 (i.e., third-party MA Organizations). Under the section titled “Business Segment 7 Strategies and Tactics,” the document stated that “Optum has an industry compliance 8 duty and responsibility to ensure that each HCC code is accurate and can be 9 substantiated within the medical charts.” Under the section titled “Competitive Analysis 10 – Market Research,” the document stated: “The marketplace will soon recognize the 11 need and importance of performing due diligence on HCCs added during the chart 12 review process but also verifying HCCs submitted to CMS can be substantiated within 13 the medical chart according to CMS guidelines. For HCCs that cannot be substantiated 14 within the medical chart, clients will need to perform the appropriate deletes in order to 15 remain compliant with CMS guidelines.” At this time, Optum had already begun 16 marketing CV or “looking both ways” chart reviews to commercial clients. 17 147. In September 2012, the Chief Executive Officer of OptumInsight, Bill Miller, 18 informed the Government that United was developing a CV Program to ensure the 19 accuracy of the diagnosis data it submitted to CMS. He explained that the purpose of the 20 program was to determine if United could find support for provider-reported diagnoses 21 that were not identified as a result of the blind chart reviews during the Chart Review 22 Program. Miller further stated that unsupported diagnoses would be deleted. United 23 knew that this information was important to the Government. United led the 24 Government to believe that it was not deliberately ignoring or recklessly disregarding the 25 negative results of its Chart Review Program showing that numerous provider-reported 26 diagnoses that it had submitted for payment were invalid. 27 148. Until 2012, United allowed coders to review medical records at providers’ offices 28 if the providers did not want United to copy the records. However, in 2012, in 47 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 48 of 79 Page ID #:1338 1 accordance with its acknowledgement of its obligation to “look both ways,” United 2 changed its policy and restricted on-site reviews because it needed copies of the charts in 3 order to conduct CV. Hammond, in an April 2012 email, explained this decision: 4 “M&R has made a decision on the chart reviews where providers are requiring onsite 5 coding. We will ask the [United] Provider Advocates to talk to the groups to try to talk 6 them out of it. If they won’t budge, then we will allow onsite coding if OptumInsight 7 has a solution for doing claims verification on those charts.” 8 149. Similarly, at around the same time, United also decided that Optum’s commercial 9 clients were required to retain Optum to perform CV (i.e., to “look both ways”) if Optum 10 performed chart reviews for them and submitted risk adjustment data, including 11 diagnoses, to CMS on their behalf. According to Dumcum, United decided “that if we 12 did chart review and submissions, that we then must do the two-way look. It became our 13 policy of how we executed business at the time. So if they bought both, then we 14 required that piece [i.e., CV] be implemented.” 15 150. United has also consistently demanded that CMS and HHS OIG, when performing 16 their audits, credit United for any additional medical conditions that United believed 17 were supported by the medical records but that had not been reported by the providers. 18 United took the firm position with the Government that, in order to accurately reflect a 19 patients’ true health status, it was necessary to review patients’ medical records for both 20 the under-reporting (not reporting diagnoses supported by the beneficiaries’ medical 21 records) and over-reporting (the reporting of codes unsubstantiated by the beneficiaries’ 22 medical records) of medical conditions by providers. In fact, CMS RADV audits have 23 historically credited MA Organizations, including United, for additional diagnoses found 24 during such audits. 25 151. For example, in a February 2011 letter from Thomas Paul, the Chief Executive 26 Officer of UnitedHealth Medicare & Retirement, to CMS concerning the parties’ dispute 27 about the preliminary results of CMS’ pilot RADV audit of one of United’s MA Plans, 28 United argued that the dispute process “incorrectly exclude[d] consideration of 48 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 49 of 79 Page ID #:1339 1 additional CMS-HCCs” supported by the beneficiaries’ medical records. United stated 2 that it identified additional HCCs (i.e., additional diagnoses mapping to additional 3 HCCs) in the medical records and, if CMS refused to give it credit for them, it would 4 challenge CMS legally. In September 2012, in a letter from Theisen, the Chief Financial 5 Officer of UnitedHealthcare Medicare & Retirement, to CMS about the same pilot 6 RADV audit, United continued to complain that the audit incorrectly excluded 7 consideration of additional HCCs and argued that United should receive credit for the 8 additional HCCs. United stated: “If [the MA Plan] is not credited for these incremental 9 HCCs, then any adjustments made by CMS do not accurately reflect an enrollee’s 10 comprehensive medical conditions. The goal of RADV audits should be to determine 11 the full extent of enrollees’ medical conditions, and make overpayment and 12 underpayment adjustments so that [MA Plans] are paid commensurate with 13 enrollees’ health status.” (Emphasis added.) Similarly, in a September 2012 letter 14 from Thomas Paul to HHS OIG, United stated that “the OIG should correct the invalid 15 HCCs and credit [the MA Plan] with the incidental HCCs documented in the submitted 16 medical records.” 17 152. ICE, the industry group to which United belonged and which it financially 18 supported, also opined that “looking both ways” was a “Best Practice.” It issued a Best 19 Practices document encouraging the industry to conduct chart reviews that “looked both 20 ways.” It described the advantages of such chart reviews as “[p]romoting validation 21 functions for diagnostic codes previously submitted by providers” and “[p]roviding the 22 ability to submit code corrections forward (additions and deletions) to health plans upon 23 the completion of review.” 24 IV. United’s Short-lived “Look Both Ways” Claims Verification Program 25 153. In September 2010, Lee Valenta, Ingenix’s Chief Operating Officer, sent Thomas 26 Paul, the Chief Executive Officer of UnitedHealthcare Medicare & Retirement, a 27 memorandum “summariz[ing] Ingenix’s plans for implementing a claims verification 28 program for charts reviewed by Ingenix on behalf of Ovations [i.e., United’s] Medicare 49 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 50 of 79 Page ID #:1340 1 Advantage Plans.” Valenta explained that the “overarching aim” of the CV Program 2 was “improving the quality of member-level diagnosis information submitted to CMS.” 3 He also explained that the purpose of the program was to identify the provider-reported 4 diagnoses that were not validated by the blind medical record reviews conducted as part 5 of the Chart Review Program and for another coder to conduct a second non-blind 6 review to determine if the provider-reported diagnoses were missed by the first blind 7 coder. 8 154. Valenta’s memorandum also set forth a three-phased approach for the 9 development of the CV Program. The first phase, Phase I, was supposed to focus on a 10 random sample of 850 beneficiaries in United’s MA Plans whose medical records for 11 encounters (i.e., provider office visits) in 2009 were included in the Chart Review 12 Program in 2010 and had already been subject to a blind review as part of that Program. 13 Phase I was supposed to start in October 2010 and be completed by January 31, 2011. 14 The second phase, Phase II, was supposed to focus on a larger sample of beneficiaries 15 and, thus, a larger number of medical records for encounters in 2010. This group was 16 supposed to include all beneficiaries who had medical encounters with only one provider 17 during 2010. Phase II was supposed to start by May 2011. The final phase, Phase III, 18 was supposed to focus on all beneficiaries in United’s MA Plans whose medical records 19 for medical encounters in 2011 were included in United’ national Chart Review Program 20 in 2012 and had already been subject to a blind review as part of that Program. Phase III 21 was supposed to be implemented in 2012. 22 155. In October 2010, Cindy Polich, the President of the group that managed United’s 23 MA Plans at the time (and who worked at United’s Santa Ana office), responded to 24 Valenta’s memorandum. Polich agreed to his proposal. Polich also acknowledged that 25 United needed to improve its risk adjustment programs, including its chart review 26 strategy. She stated that “[w]hile Ingenix is implementing Phase II [of CV], PSMG will 27 conduct a comprehensive review of its current risk adjustment strategies, including our 28 50 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 51 of 79 Page ID #:1341 1 chart review strategy. The purpose of this review is to determine the future programs 2 and approaches to be used to improve the accuracy of our risk scores.” 3 156. In a CV project management document, United also described CV as “[a] risk 4 adjustment chart audit service designed to ensure that qualified patient diagnoses and 5 conditions are identified and supported in the physician medical record documentation. 6 This audit service identifies any discrepant coding patterns contained within the medical 7 record documentation and the associated claims and encounters. This service includes 8 the submission and delete process to CMS, financial reporting, and training/education for 9 providers. Claims Verification will further supplement our efforts to assess coding 10 accuracy and our ability to drive prospective improvements by engaging and educating 11 providers.” 12 157. According to Theisen, United decided to “look both ways” and implement CV in 13 order to defend its Chart Review Program in light of increased scrutiny on risk 14 adjustment. 15 158. United, however, took over three years to develop its CV Program. Furthermore, 16 the manner in which United developed and then implemented its CV Program shows that 17 United was never committed to honoring its obligation to undertake good faith efforts to 18 ensure the validity of the risk adjustment data that it submitted to the Medicare 19 Advantage Program. United did not automatically delete the provider-reported 20 diagnoses that were not supported by its medical record reviews conducted as part of its 21 Chart Review Program. Rather, United considered these invalid diagnoses as mere 22 “potential deletes” and instructed its coders to re-review the medical records to try to 23 avoid deleting them. United trained the CV coders that the goal of CV, above all else, 24 was to “validate” or “save” the potential deletes through finding any support for the 25 diagnosis anywhere in the beneficiary’s chart. Additionally, United instructed its coders 26 to save these invalid diagnoses even if the information in the medical records was 27 ambiguous. 28 51 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 52 of 79 Page ID #:1342 1 159. When performing CV, United also did not consult providers even when medical 2 records were ambiguous concerning whether the beneficiaries actually had the medical 3 conditions depicted by the diagnosis codes. 4 160. United also saved “potential deletes” by simply just accepting the diagnoses 5 reported by providers even when the medical conditions were not unambiguously 6 documented in the beneficiaries’ medical records. United characterized this as deferring 7 to the judgment of the provider or the provider’s administrative staff who assigned the 8 diagnosis codes. For example if a chart was unclear, illegible, or missing, and even 9 though United could not identify any medical records documenting the medical 10 conditions identified by the provider-reported codes, it just accepted the codes and did 11 not delete them. United did this with either deliberate ignorance or reckless disregard for 12 the truth in light of the mountain of knowledge it possessed about the significant 13 percentage of invalid provider-reported diagnoses. 14 161. However, despite all its flaws, the CV Program confirmed what United already 15 knew about the significant error rate associated with provider-reported diagnoses that it 16 submitted to Medicare for risk adjustment payments. 17 162. At the end of 2010, United conducted the first pilot phase, Phase I, of its 18 development of the CV Program. As specified in Valenta’s memorandum, Phase I 19 included a very small sample of medical records that United had reviewed as part of its 20 Chart Review Program for encounters (e.g., office visits) that beneficiaries had with 21 providers in 2009. A “CV Dashboard” summary from December 2010 shows that out of 22 the many hundreds of thousands of medical records included in its Chart Review 23 Program for 2009 encounters, only 843 records were included in Phase I of the CV 24 Program. By December 2010, the review for 728 of those medical records had been 25 completed. The results showed that 224 of the 728 medical records re-reviewed as part 26 of CV (i.e., reviewed once as part of the Chart Review Program and then again as part of 27 Phase I of the CV Program) had both at least one ADD (a diagnosis code that the 28 provider had not reported) and at least one DELETE (a diagnosis code that was reported 52 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 53 of 79 Page ID #:1343 1 by the provider but not validated by the medical record) and 167 of the 728 medical 2 records re-reviewed as part of CV had at least one DELETE but no ADDS. 3 Furthermore, of the HCCs and the RxHCCs for which United looked for validation in 4 these 728 records that it had re-reviewed, 19 percent of the HCCs and 15 percent of the 5 RxHCCs did not validate. At the time that the summary was prepared, United was 6 conducting “follow-up” on 39 medical records. Of the HCCs and the RxHCCs for which 7 United looked for validation in these 39 medical records that it had re-reviewed, 47 8 percent of the HCCs and approximately 45 percent of the RxHCCs did not validate. 9 163. As part of Phase I, but only as part of Phase I, United sometimes contacted the 10 providers who gave it the medical records. It did this when the second non-blind review 11 of the records in CV failed to validate the diagnoses that the first blinded review failed to 12 validate as part of the Chart Review Program. For some beneficiaries, the providers 13 responded that they should not have submitted their claims (i.e., the claims which 14 included the diagnoses) to United. Consequently, the diagnoses should not have been 15 submitted by United to CMS. For some other beneficiaries, the providers did not have 16 additional records or their additional records did not support the diagnoses in question. 17 164. In mid-2011, United began conducting the second pilot phase, Phase II, of its CV 18 Program. Phase II began in the summer of 2011 and was completed in early 2012. Out 19 of the many hundreds of thousands of medical records included in the Chart Review 20 Program for 2010 medical encounters, United included only approximately 17,000 charts 21 in Phase II. United’s coders tried to save as many “potential deletes” as possible when 22 reviewing these medical records as part of Phase II. Nonetheless, the results of their CV 23 reviews confirmed more invalid diagnoses (DELETES) than the number of additional 24 codes (ADDS) gleaned from the same records. 25 165. In mid-2012, United began conducting a preliminary test or pilot for Phase III of 26 the CV Program. This pilot included approximately 5,000 medical records relating to 27 encounters (e.g. office visits) that beneficiaries had with providers during calendar year 28 2011. In September 2012, Theisen explained to others at United, including Relator 53 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 54 of 79 Page ID #:1344 1 Poehling, that UnitedHealthcare Medicare & Retirement, relying on the results of this 2 sample review, had increased its estimate of the financial impact of CV deletes. In 3 October 2012, Theisen sent Daniel Schumacher, the CFO of Defendant 4 UnitedHealthcare, Inc., and others a “CV III analysis – based on 2011-2012 Chart 5 review activity for 2011 DOS.” The document showed validation rates based on the 6 review of HCCs because the coders were instructed to find support for any diagnosis that 7 mapped to the HCC under review even if it was not the same diagnosis reported by the 8 provider that originally mapped to that HCC. The document showed that the validation 9 rate “[b]ased on results of CV3 pilot (5,000 chart sample)” was 66.4 percent. In other 10 words, only 66.4 percent of the HCCs reviewed were supported by the beneficiaries’ 11 medical records. That meant that 33.6 percent of the HCCs were unsupported by the 12 beneficiaries’ medical records even after United reviewed them twice, once as part of the 13 Chart Review Program and again as part of the pilot CV Phase III process. Based on 14 these results, United estimated that the negative financial impact of CV in 2012 would be 15 $226 million based on the number of estimated “potential deletes” that it could not save 16 and would have to be made. United estimated that diagnoses mapping to 120,147 HHCs 17 would have to be deleted and that, on average, each delete would result in a $1,924 18 negative financial impact. 19 166. In or about October 2012, United recorded in its financial records a $208 million 20 accrual for potential revenue reductions due to deletes that would need to be made as 21 part of the CV Program (hereinafter referred to as a “CV liability accrual”). 22 167. Until late 2012, United did not complete its pilot tests and start to implement its 23 CV Program for charts relating to 2011medical encounters (i.e., with 2011 dates of 24 service). Even after that, it never fully implemented the program. It also continually 25 changed the program in order to limit its scope and created arbitrary rules to avoid 26 looking at the negative results of many of the blind chart reviews conducted as part of 27 the Chart Review Program. In addition, when United learned that its second review of 28 the medical records in CV was not saving a significant number of diagnoses from 54 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 55 of 79 Page ID #:1345 1 deletion, it created another level of review. That is, it created a re-re-review or third 2 review of the beneficiaries’ medical records in order to keep trying to save the diagnoses 3 from deletion. United also limited the scope of CV by excluding certain providers, 4 including providers that it owned and operated. In 2014, United then terminated CV 5 without completing the program for charts relating to 2012 medical encounters (with 6 2012 dates of service). 7 168. United imposed several arbitrary exclusionary rules to improperly disqualify many 8 medical records reviewed as part of its Chart Review Program from its CV Program. 9 For medical records for encounters in 2011 and 2012, United arbitrarily and improperly 10 excluded numerous medical records from CV. By arbitrarily and improperly excluding 11 all these medical records from CV, United deliberately ignored or recklessly disregarded 12 a lot of the negative results from its Chart Review Program and knowingly failed to 13 delete all invalid provider-reported diagnoses. For example, United excluded numerous 14 charts from CV because the image of the chart was purportedly “unavailable.” Yet, 15 when making ADDS in the Chart Review Program, the image of the chart or chart itself 16 must have been available. But, instead of locating the images or obtaining the charts for 17 CV, United decided not to re-review the charts to save those diagnoses invalided by the 18 results of its Chart Review Program. It also did not delete those invalid diagnoses. 19 169. The re-reviews conducted as part of United’s CV Program did not save as many 20 deletes as United would have liked and, in 2013, Theisen and others at United became 21 increasingly concerned about the financial impact of the deletes. United decided that its 22 coders were not saving enough “potential deletes.” Accordingly, sometime in 2013, 23 United decided that a third review or a re-re-review of the medical records had to be 24 conducted to try to save more deletes. 25 170. Thus, if United’s internal coders were unable, despite their best efforts, to “save” a 26 diagnosis code, United sent that code to a coding consultant for re-re-review. United 27 knew, however, that the consultant engaged in a pattern of “saving” diagnoses without 28 supporting medical records in several circumstances, including when the chart was 55 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 56 of 79 Page ID #:1346 1 scanned illegibly, when pages were missing from a chart and the diagnosis could not be 2 validated, and when the reported date of service did not appear in the chart. By 3 accepting validation of these diagnoses without supporting medical records, United 4 knowingly and improperly avoided its obligation to return monies to the Medicare 5 Program to which it was not entitled. 6 171. In November 2013, Donald James, the Director of Program Strategy for Optum in 7 Santa Ana, California, reported to senior management that the CV Program had not yet 8 started for charts reviewed as part of the 2013 Chart Review Program. In December 9 2013, Patty Brennan, also in the Optum Santa Ana office, reported to Dumcum that 10 “[h]alf of the CV volume for 2012 DOS has been completed but deletes are on hold.” 11 She also informed him that UnitedHealthcare Medicare & Retirement “[r]equested all 12 CV deletes be held until further noticed so Ops is not going to complete the second half 13 of the volume at this time.” 14 172. When Steve Nelson became the Chief Executive Officer of UnitedHealthcare 15 Medicare & Retirement in early 2014, he spoke with Steve Hemsley, the Chief 16 Executive Officer of UHG, about whether United should continue the CV Program. 17 Hemsley encouraged Nelson to look into whether or not United should do so, formulate 18 an opinion, and report back to him. 19 173. In February 2014, Marc Beckmann, in the Finance – Risk Adjustment Analysis 20 Group at UnitedHealthcare Medicare & Retirement, sent information about CV liability 21 accruals to Daniel Schumacher, the Chief Financial Officer of Defendant 22 UnitedHealthcare, Inc., and Brian Thompson, the new Chief Financial Officer for 23 UnitedHealthcare Medicare & Retirement. He estimated that the effect of CV on 24 United’s revenue would be $208 million for payment year 2012, $125 million for 25 payment year 2013, and either $125 million or $175 million (depending on the number 26 of charts reviewed) for payment year 2014. 27 174. On March 3, 2014, Jon Bird, the Senior Vice President of Risk and Quality 28 Analytics who worked at Optum in Santa Ana, California, participated in a meeting with 56 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 57 of 79 Page ID #:1347 1 Thompson about the CV liability accrual for 2013. Thompson wanted to “move toward 2 the more conservative range of the confidence interval (from 50% to 80%) resulting in 3 [a] $29M higher CV estimate” for the liability accrual. By March 3, 2014, United’s 4 estimate of its CV liability accrual for 2013 already had been increased by over $50 5 million dollars from $125 million to $180 million. 6 175. Also, in February or March 2014, the financial managers at UnitedHealthcare 7 Medicare & Retirement, including Brian Thompson, performed a comparison of their 8 then-expected revenues for 2014 with the revenue estimated in UnitedHealthcare 9 Medicare & Retirement’s annual budget for 2014. They determined that there was going 10 to be a shortfall in their financial performance relative to that budget and they started to 11 think about ways to eliminate the shortfall. 12 176. In March 2014, Thompson sent Nelson “a current brain dump of ‘shut off/stop 13 doing’ that is not yet valued/included in the road back to plan.” Part of the “brain dump” 14 was to “shut off” or reduce compliance efforts. Thompson asked others at 15 UnitedHealthcare Medicare & Retirement for other “shut offs.” 16 177. Subsequently in March 2014, Marybeth Meyer, the United employee who ran the 17 risk adjustment team at UnitedHealthcare Medicare & Retirement after Relator Poehling 18 left United, reported to Thompson that the “CV estimate of $125M may be light 19 (estimate for 2012 DOS/2013 payment year of $167M).” 20 178. In late March or early April, Nelson met with other Chief Executive Officers at 21 Defendant UnitedHealthcare, Inc. to discuss UnitedHealthcare Medicare & Retirement’s 22 financial performance. A very detailed slide deck was created for that meeting. After 23 the meeting, the slide deck was sent by Schumacher to senior executives at United, 24 including executives who reported to Hemsley. The slide deck highlighted that 25 UnitedHealthcare Medicare & Retirement was projecting that its actual revenues for 26 2014 were going to miss the target set forth in the annual budget by half of a billion 27 dollars. It stated: “Best estimate of $500 million budget miss.” It also stated that, 28 because of that projected miss, UnitedHealthcare Medicare & Retirement’s management 57 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 58 of 79 Page ID #:1348 1 was making a commitment to the senior executives to “find $250 million to cut miss in 2 half.” UnitedHealthcare Medicare & Retirement referred to this “Management 3 Commitment” as a $250M “good guy.” 4 179. Nelson and others at United, including Hemsley, knew that, if United terminated 5 the CV Program, it could cut the $500 million miss by $250 million by reversing the CV 6 liability accruals and not deleting the provider-reported diagnoses invalidated by its chart 7 reviews. This was their “good guy.” But, United was concerned about the consequences 8 of terminating the program and reverting to ignoring the negative results of its chart 9 reviews. It decided to ask CMS about the retroactivity of a proposed regulation 10 requiring MA Organizations to design all medical record reviews to validate diagnoses 11 submitted to CMS. 12 180. Under Hemsley’s direction, Larry Renfro, the Chief Executive Officer of Optum, 13 contacted senior government employees at CMS, including the Administrator of CMS, 14 to ask whether United had a legal obligation to perform CV before the effective date of 15 the proposed rule. These employees were not government attorneys and could not 16 render the requested legal advice. Renfro, moreover, knew nothing about United’s Chart 17 Review or CV Programs and could not impart any meaningful information about these 18 programs to the government employees. Accordingly, Renfro could not and did not 19 provide the senior government employees with any description of United’s CV Program 20 or its purpose. Hemsley and his attorneys, however, continued to push Renfro to make 21 further contacts with these employees when United did not initially obtain from them the 22 legal opinion it wanted. 23 181. On or before April 8, 2014, Renfro asked Karen Erickson, an Optum employee 24 who worked directly for him, for speaking points for a call with the Administrator of 25 CMS. On April 8, 2014, Erickson sent Renfro’s assistant, Juliet Domb, “aspirational” 26 talking points, that is, things that they wanted Renfro to get the CMS employees to say, 27 including “CV is not currently required” and that United was “allowed to stop any CV 28 activities (including delete submission) currently underway.” (Emphasis in the original.) 58 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 59 of 79 Page ID #:1349 1 In her email, Erickson told Domb: “I sent this to Marianne [Short] and Matt Shors for 2 editing – they will send final directly, and they know it has to be today.” The same day, 3 Renfro spoke to Hemsley and Short to obtain direction about what he should say to the 4 Administrator of CMS on the call that he had scheduled with the Administrator. Renfro 5 then purportedly spoke with the CMS Administrator and purportedly dictated notes of 6 the call to Domb who wrote the notes by hand on Renfro’s notepad and sent them to 7 Short. According to Renfro, he dictated these notes from memory and he does not 8 usually take notes, but was asked to do so by Short. 9 182. On April 26, 2014, Short asked Renfro to again contact the Administrator of CMS. 10 According to Thompson (the Chief Financial Officer of UnitedHealthcare Medicare & 11 Retirement in 2014), Renfro had not obtained the “clarity” United wanted from the 12 Administrator about whether it was obligated to perform CV. Accordingly, on April 27, 13 2014, Renfro sent an email to the Administrator asking the same questions he had 14 purportedly asked her on April 8 and to which she purportedly had responded on April 8. 15 He also asked for a meeting with other employees at CMS who were responsible for 16 operating the Medicare Advantage Program. On April 27, 2017, Renfro reported to 17 Hemsley and Short that CMS was arranging for United to meet with these employees. 18 183. On April 29, 2014, Hemsley sent United’s attorney, Thad Johnson, to Washington, 19 D.C. to speak with those CMS employees responsible for the Medicare Advantage 20 Program, including Cheri Rice, the Director of the Medicare Plan Payment Group at 21 CMS. Nelson, Schumacher, and an Optum employee, Karen Erickson, also attended the 22 meeting. Long before this meeting, United knew that the Department of Justice was 23 conducting a FCA investigation relating to United’s Chart Review and CV Programs. 24 Yet, United did not notify the Department of Justice that it was sending an attorney and 25 others to speak with government employees about matters under investigation. 26 Accordingly, the government employees at the meeting were unrepresented by 27 government counsel and were not authorized to provide legal advice about United’s 28 obligations under the FCA or any other laws. 59 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 60 of 79 Page ID #:1350 1 184. At the meeting on April 29, 2014, when United informed CMS about the 2 possibility of terminating its CV Program, CMS told United that it had a statutory 3 obligation to report and repay Medicare for erroneous risk adjustment payments and that 4 there were FCA implications if it failed to do so. 5 185. CMS also told United that it could not ignore information in its possession 6 showing that diagnoses may be invalid and that United was obligated to delete invalid 7 diagnoses. According to Erickson, CMS told United that “if there was reason to have 8 knowledge that something had been begun and we were [i.e., United was] far enough 9 along that there was knowledge that something might not be supported that we needed to 10 continue the investigation into those numbers, or those codes.” Erickson further recalled 11 CMS stating that, if United had “knowledge of things that might not be supported we 12 [i.e., United] needed to continue the investigation.” According to Schumacher, CMS 13 also told United that it did not have sufficient information about United’s CV Program to 14 provide further guidance. 15 186. On April 30, 2014, Nelson sent an email to Cheri Rice at CMS about the meeting 16 between United and CMS on April 29, 2014. He stated: “[A]s we discussed yesterday, 17 CMS recently issued a proposed rule that would, if finalized, require MA plans to design 18 any medical record reviews to determine the accuracy of risk adjustment diagnoses 19 associated with those records. During our conversation yesterday and other recent 20 conversations, CMS confirmed to us that these requirements do not apply until the 21 effective date of the rule, and that MA plans are thus not currently required to design 22 their medical record reviews to determine the accuracy of risk adjustment diagnoses. 23 We currently have a process through which we review certain medical records to 24 determine the accuracy of risk adjustment diagnoses and submit appropriate deletes. 25 This process already has resulted in the identification of and, in some instances, the 26 submission of deletes for 2012 dates of service. But based on the proposed rule, 27 including the preamble, and recent conversations with CMS, we suspended that process 28 for 2012 dates of service while we consider whether to make changes. Pursuant to our 60 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 61 of 79 Page ID #:1351 1 discussion, however, we will soon submit for deletion those diagnosis codes that have 2 undergone a complete review and that we have therefore identified as appropriate 3 deletes. In the near future, we will determine whether to continue our review process for 4 the diagnosis codes which were still under review at the time we suspended our process. 5 In the meantime we will not delete these codes.” 6 187. On May 2, 2014, Cheri Rice replied to Nelson’s email: “[R]egardless of the 7 effective date of the proposed requirement related to medical record reviews, there are 8 other laws that do impose standards, requirements and responsibilities on MA plans in 9 connection with the federal payments they receive from CMS. We cannot provide 10 advice to United about the scope of those other laws. Nor can we provide advice on 11 whether United’s plan[ned] course of action and/or purported limits on the scope of its 12 [Risk Adjustment Attestation, submitted April 30, 2014] are compliant with such other 13 laws. Your statement concerning the data submissions that have already been made and 14 United’s plans for future action will be included in our records and we will proceed with 15 our evaluation and use of the risk adjustment data consistent with 42 CFR § 422.308, § 16 422.310, and other applicable law.” 17 188. According to Schumacher and Nelson, Rice’s May 2, 2014 email did not say 18 anything that was inconsistent with what CMS said in the meeting on April 29, 2014. 19 And, as CMS mentioned during the April 29 meeting, the FCA is one of the “other laws” 20 that imposes standards, requirements, and responsibilities on MA Organizations and 21 their MA Plans in connection with the federal payments they receive from CMS. After 22 Rice sent her May 2 email, the Department of Justice sent a letter to United’s counsel 23 emphasizing that point. 24 189. On May 5, 2014, UnitedHealthcare Medicare & Retirement informed Jeffrey 25 Putman, UHG’s Controller who reported to Hemsley, that “M&R CV delete, good guy, 26 is ~$250M for the year (note this is just the M&R piece and there is an incremental 27 component at C&S).” According to Schumacher, the $250 million was the estimated 28 amount of the deletes from CV for 2014 and prior years and, thus, the CV liability 61 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 62 of 79 Page ID #:1352 1 accruals that United had recorded for 2014 and prior years. The “good guy” was the 2 hoped-for release or reversal of those accruals. United knew the accruals were likely 3 underestimated and the financial impact likely greater if it continued CV. 4 190. Despite CMS’ warnings in the April 29 meeting and Cheri Rice’s May 2 email, 5 Nelson, Schumacher, and Thompson decided to terminate the CV Program. This 6 decision was reported to Hemsley and, according to Nelson, Hemsley could have 7 reversed this decision but instead he supported it. 8 191. United also decided not to delete or otherwise report to CMS at least 100,000 9 invalid diagnoses about which it had actual knowledge based on more than one review 10 of the patients’ medical records for encounters in 2011 and 2012 (i.e., encounters with 11 2011 and 2012 dates of service). The single damages to the Medicare Program arising 12 from United’s submission and failure to delete just these invalid diagnoses is 13 approximately $190 million under Part C alone. 14 192. After it terminated CV, United reverted to “looking one way” at the results of its 15 chart reviews, making only ADDS, and knowingly and improperly failing to delete 16 invalid provider-reported diagnoses and repay the Medicare Program for them. 17 193. At the time they decided to terminate CV, Steve Nelson, Dan Schumacher, and 18 Brian Thompson knew that their decision would enable United to reverse its CV liability 19 accruals by more than $250 million dollars. Hemsley also was aware that terminating 20 CV would enable United to achieve this financial benefit. This was important to all of 21 them because they wanted to represent to investors that UnitedHealthcare Medicare & 22 Retirement’s actual revenues were on target. An internal document relating to United’s 23 second quarter 2014 “earnings release” issued in July 2014 states: “Q2 was on track and 24 we are building momentum that will take us through the year and into 2015. Internal: 25 Important to note that 2014 benefits from the one-time claims verification policy change 26 which investors are unaware of.” (Emphasis in the original.) 27 28 62 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 63 of 79 Page ID #:1353 1 2 V. United Failed to Delete At Least Over A Billion Dollars of Diagnoses Invalidated By Its Own Chart Review Program 3 194. The results of United’s chart reviews provided it with information about a 4 significant number of invalid provider-reported diagnoses that should not have been, but 5 were, submitted by United to the Medicare Program for risk adjustment payments. For 6 example, for the 2011 payment year (involving payments based on diagnoses with 2010 7 dates of service), United submitted at least 197,000 diagnoses that were invalidated by 8 its medical record reviews conducted as part of its Chart Review Program. But, for that 9 year, United deleted only a very few (approximately 1,800) of these invalid diagnoses 10 based on the results of Phase II of its CV Program. For the 2012 payment year 11 (involving payments based on diagnoses with 2011 dates of service), United submitted at 12 least 222,329 diagnoses that were invalidated by its medical record reviews conducted as 13 part of its Chart Review Program. Based on the results of its CV Program for charts 14 with 2011 dates of service, United deleted approximately 120,000 of these invalid 15 diagnoses despite its arbitrary exclusionary rules and attempts to save these deletes. For 16 the 2013 payment year (involving payments based on diagnoses with 2012 dates of 17 service), United submitted at least 285,122 diagnoses that were invalidated by its 18 medical record reviews conducted as part of its Chart Review Program. Based on the 19 results of its CV Program for charts with 2012 dates of service, United deleted 20 approximately 27,000 of these invalid diagnoses despite its arbitrary exclusionary rules, 21 its multiple attempts to save these deletes, and its failure to complete the program. For 22 the 2014 payment year (involving payments based on diagnoses with 2013 dates of 23 service), United submitted at least 199,039 diagnoses that were invalidated by its 24 medical record reviews conducted as part of its Chart Review Program. Because United 25 terminated the CV Program, it did not delete any of these invalid diagnoses. These 26 numbers apply to invalid diagnoses relating to risk adjustment payments under Part C 27 only and not also Part D. 28 63 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 64 of 79 Page ID #:1354 1 195. Accordingly, United knowingly and improperly failed to delete or otherwise repay 2 Medicare for most of the diagnoses invalidated by its Chart Review Program over the 3 last decade. For example: for the 2011 payment year, United damaged the Medicare 4 Program by at least $377,734,792 by failing to look both ways and delete diagnoses 5 invalidated by its Chart Review Program; for the 2012 payment year, United damaged 6 the Medicare Program by at least $213,978,134 by failing to look both ways and delete 7 diagnoses invalidated by it Chart Review Program; for the 2012 payment year, United 8 damaged the Medicare Program by at least $317,329,602 by failing to look both ways 9 and delete diagnoses invalidated by its Chart Review Program; and for the 2014 payment 10 year, United damaged the Medicare Program by at least $234,159,775 by failing to look 11 both ways and delete diagnoses invalidated by its Chart Review Program. These 12 numbers apply to damages relating to risk adjustment payments under Part C only and 13 not also Part D. 14 196. Examples of beneficiaries with invalid diagnoses about which United knew but 15 failed to delete based on its Chart Review Programs for payment years 2011 through 16 2014 are set forth in Exhibit 2 to this Complaint. 17 197. Under the FCA, the United States is entitled to treble damages and penalties for 18 the invalid diagnoses that United failed to delete for payment years 2011 through 2014 19 plus treble damages and penalties for the additional diagnoses invalidated by United’s 20 Chart Review Programs for the years before 2011 and after 2014. 21 22 VI. United Knowingly and Improperly Avoided Repaying Medicare For Invalid Diagnoses Reported By Its Financially-Incentivized Providers 23 198. As alleged below, the financial arrangements that United entered into with 24 capitated and gainsharing providers were tied to the risk adjustment payments that 25 United received from the Medicare Program. These providers benefitted financially 26 from any increase in risk adjustment payments resulting from the diagnoses they 27 reported to United for beneficiaries enrolled in United’s MA Plans. United knew that 28 these compensation arrangements created a strong financial incentive for the provider 64 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 65 of 79 Page ID #:1355 1 groups to increase the number of diagnoses reported to United for each beneficiary, and 2 in some cases to report invalid diagnosis codes. 3 199. Furthermore, United recognized that it had an obligation to review diagnoses 4 reported by these incentivized providers to determine their validity. Thus, early on, it 5 performed the Internal Data Validation (“IDV”) reviews discussed earlier in this 6 Complaint. Then, in 2010, United implemented its Risk Adjustment Coding and 7 Compliance Reviews (“RACCR”) Program with the stated goal of safeguarding against 8 improper diagnosis coding by these incentivized providers. 9 200. Although the IDV and RACCR Programs were extremely limited in scope and 10 utility, they confirmed what United already knew: that there were serious problems with 11 the diagnoses being reported by a number of its financially-incentivized providers, 12 including WellMed. The programs also show how United knowingly avoided 13 identifying and deleting invalid diagnoses reported by these providers and repaying 14 Medicare for risk adjustment payments to which United and these providers were not 15 entitled. They also provide additional facts showing the scope of the false claims 16 submitted by United to Medicare for risk adjustment payments and the falsity of 17 United’s Risk Adjustment Attestations. 18 201. United scoured millions of medical records through its Chart Review Program in 19 an effort to identify additional diagnoses and increase revenue, but it only reviewed 20 thousands of charts to identify invalid coding as part of the RACCR Program. Like with 21 CV, United implemented RACCR in such a way as to drastically limit the program’s 22 scope and utility and to avoid deleting invalid diagnoses. 23 202. First, United excluded from its RACCR Program any incentivized providers with 24 fewer than 500 beneficiaries in United’s MA Plans. This resulted in the exclusion of 25 approximately 40 percent of the financially-incentivized providers from the RACCR 26 Program for medical encounters in 2008, 2009, and 2010 (i.e., with dates of service in 27 those years). 28 65 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 66 of 79 Page ID #:1356 1 203. Second, for the incentivized provider groups with 500 beneficiaries or more, 2 United did not review the medical documentation for any diagnoses unless the provider 3 group was an extreme outlier in reporting diagnoses that mapped to one or more HCCs. 4 Combined with the exclusion of providers with less than 500 beneficiaries, this resulted 5 in a total exclusion of over 80 percent of the financially-incentivized providers for 6 medical encounters in 2008, 2009, and 2010. 7 204. For medical encounters in 2008, 2009 and 2010 (i.e., with dates of service in those 8 years), United defined an outlier as a provider that reported diagnoses mapping to a 9 particular HCC more than three times as often as the average national prevalence rate for 10 that HCC for all beneficiaries in United’s MA Plans. For example, if 15 percent of the 11 beneficiaries in United’s MA Plans nationwide were reported by providers to have a 12 diagnosis that mapped to HCC 52 (Drug/Alcohol Dependence), United did not consider 13 the incentivized provider an outlier unless it reported diagnoses mapping to HCC 52 for 14 more than 45 percent of its patients enrolled in United’s MA Plans. 15 205. For medical encounters in 2011 and 2012, United further limited which providers 16 would qualify as an outlier in order to reduce the number of providers and HCCs subject 17 to review. Instead of using a national average prevalence rate for each HCC based on 18 diagnoses reported by all of its providers, United used the average rate at which its 19 financially-incentivized providers reported diagnoses mapping to each HCC. Thus, an 20 incentivized provider was considered an outlier only if it reported diagnoses mapping to 21 an HCC more than three times as often as other incentivized providers, that is, providers 22 who also had a financial incentive to invalidly code. 23 206. A large percentage of the outliers were located in the Central District of 24 California. They are listed in Exhibit 3 to this Complaint. 25 207. Third, after limiting the program to only extreme outliers, United conducted an 26 initial review of just a small sample of beneficiaries for whom the provider had reported 27 diagnoses mapping to the problematic HCC or HCCs (i.e., a problematic HCC being one 28 over 300 percent of the average prevalence rate). For medical encounters in 2008, 2009, 66 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 67 of 79 Page ID #:1357 1 and 2010, the initial sample size for even the largest provider groups was never more 2 than 30 beneficiaries per problematic HCC, and was often as few as 10 beneficiaries per 3 problematic HCC. United purposefully kept the sample size small to ensure the samples 4 were not statistically significant, enabling it to later argue that it was unable to 5 extrapolate the results of its sample reviews to all of the diagnoses reported by the 6 provider that mapped to the problematic HCC. For 2011 and 2012 medical encounters, 7 United increased the initial sample size to 50 beneficiaries per provider for each 8 problematic HCC, which still ensured the results were not statistically significant for 9 many providers. 10 208. Fourth, after reviewing the medical records for the beneficiaries in the sample to 11 determine if the diagnoses were valid, United gave a provider a “passing” grade if 12 anything less than 20 percent of the diagnoses were determined to be invalid. 13 Accordingly, if 19 percent of the diagnoses were invalid, the provider passed this 14 extraordinary lenient test and no further review was conducted. 15 209. Fifth, if a provider failed the initial review for a particular problematic HCC, 16 United often added only a few additional beneficiaries to the sample (i.e., additional 17 beneficiaries for whom the provider had reported diagnoses mapping to the problematic 18 HCC). United called this an “incremental sample.” If United was successful in 19 decreasing the invalidation rate to below 20 percent based on increasing the sample size, 20 United considered the provider group to have a passing grade for that HCC. 21 Accordingly, if 19 percent of the diagnoses in the “incremental sample” were invalid, the 22 provider passed and no further review was conducted. 23 210. Sixth, if United determined that diagnoses in a sample review were not supported 24 by the medical records, United did not always delete them. For example: 25 For 2008 medical encounters, United conducted a sample review for Edinger 26 Medical Group of diagnoses mapping to Drug/Alcohol Dependence (HCC 52) and 27 Major Complications of Medical Care & Trauma (HCC 164). United’s coders 28 determined that “[t]here was no clinical documentation to support the diagnoses” 67 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 68 of 79 Page ID #:1358 1 mapping to HCC 52 for six beneficiaries and that the diagnosis mapping to HCC 2 164 was the “wrong code” for one beneficiary. (For all examples, further 3 information to identify or aid in identifying the beneficiaries will be separately 4 provided to the Defendants.) 5 For 2008 medical encounters, United conducted a sample review for Family 6 Practice Medical Group’s diagnoses mapping to Drug/Alcohol Dependence (HCC 7 52). For two beneficiaries, United’s coder concluded there was “no clinical 8 documentation to support the diagnoses.” 9 In a sample review for Sharp-Rees Sealy for Vascular Disease (HCC 105) for 10 2011, United failed to delete a diagnosis for Beneficiary AA even though its own 11 coder determined it was the “wrong code.” 12 In a sample review for HealthCarePartners for Protein-Calorie Malnutrition (HCC 13 21) for 2011, United failed to delete a diagnosis for Beneficiary BB even though 14 its own coder determined it was the “wrong code.” 15 In a sample review for Mercy Physicians Medical Group for Protein-Calorie 16 Malnutrition (HCC 21) for 2011, United failed to delete a diagnoses for 17 Beneficiary CC even though its own coder determined that it was the “wrong 18 code.” 19 In a sample review for WellMed for Major Complications of Medical Care and 20 Trauma (HCC 164) for 2011, United did not delete a diagnoses for Beneficiary 21 DD even though its coder noted the “dx [was] not documented” in the record. 22 In a sample review for HealthCarePartners for Pneumococcal Pneumonia, 23 Emphysema, Lung Abscess (HCC 112) for 2012, United did not delete a diagnosis 24 code for Beneficiary EE even though its coder determined the “dx [was] not 25 documented.” 26 In a sample review for WellMed for Disorders of Immunity (HCC 45) for 2011, 27 United did not delete a diagnosis code for Beneficiary FF even though its coder 28 determined the diagnosis in the record was “not a current condition.” 68 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 69 of 79 Page ID #:1359 1 211. Seventh, when United determined that an outlier provider failed the incremental 2 sample review, United’s RACCR policy specified that, with limited exceptions, all 3 diagnoses reported by that provider mapping to the problematic HCC should be 4 reviewed. In other words, once it was established that a provider reported diagnoses 5 mapping to a particular HCC three times or more than the average and that 20 percent or 6 more of those diagnoses were not supported by the beneficiaries’ medical records in the 7 sample review, even United recognized that, with limited exceptions, a complete review 8 of the medical records for all beneficiaries for whom the provider had reported diagnoses 9 mapping to the problematic HCC was essential. But, generally, United did not conduct 10 these 100 percent reviews. Instead, as part of the RACCR Program, United purported to 11 require its providers – the very outliers who had reported the diagnoses at issue – to 12 conduct these 100 percent reviews. Not surprisingly, the providers were very resistant to 13 performing these 100 percent reviews. Sometimes, the providers reviewed the records 14 for only some, but not all, additional beneficiaries for whom they had reported diagnoses 15 mapping to the problematic HCC. Sometimes, they did nothing. 16 212. With respect to WellMed, United also knew that it could not rely on it to review 17 its own diagnoses. For example, in a “WellMed RACCR Audit Status Summary as of 2- 18 9-12,” it was reported that WellMed did a self-audit for 8 HCCs, but United determined 19 that “70% of the codes [WellMed] indicated were properly documented actually were 20 not supported in the medical record.” The same summary stated that, based on the 21 sample reviews for 2008, 2009, and 2010, “WellMed failed to achieve an acceptable 22 validation rate for multiple HCCs …, as defined by the RACCR audit program policy 23 and procedure.” 24 213. For medical encounters in 2008 to 2010, at least 58 provider groups should have 25 performed 100 percent reviews because they failed United’s sample validation test for at 26 least one HCC. Combined, these providers should have conducted 100 percent reviews 27 relating to 192 problematic HCCs. However, the providers only conducted reviews of 28 143 of the 192 HCCs. Thus, a quarter of all HCCs that failed United’s validation test 69 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 70 of 79 Page ID #:1360 1 were not further reviewed by either the provider groups or United. Similarly, for 2011 2 encounters, 100 percent reviews were conducted for only 69 of 82 HCCs that failed 3 United’s sample validation test. 4 214. For example, for 2010 encounters, Edinger Medical Group coded Drug/Alcohol 5 Dependence (HCC 52) at a rate greater than three times the national rate. United 6 determined that only 70 percent of the diagnoses in its sample were supported by 7 medical record documentation. Because 30 percent of the sampled diagnoses were 8 invalid, a review should have been conducted for the additional 66 Edinger beneficiaries 9 with a diagnosis mapping to HCC 52 that were not included in the sample. Nonetheless, 10 United did not require Edinger to do this review and United did nothing further to 11 examine Edinger’s additional diagnoses mapping to HCC 52 that were not included in 12 the sample. 13 215. Additional examples of provider groups that failed to conduct 100 percent reviews 14 for particular HCCs are shown in the chart attached as Exhibit 4 to this Complaint. 15 United also did not conduct the 100 percent reviews of these problematic HCCs. 16 216. United knew that it was problematic that neither it nor the outlier providers 17 reviewed 100 percent of the diagnoses mapping to the problematic HCCs when the 18 providers failed the sample validation test. It knew this because, in those cases where 19 100 percent reviews were conducted for problematic HCCs, more than seventy percent 20 of the HCCs had validation rates below 80 percent. 21 217. Indeed, despite the many flaws of the RACCR Program, United knew from the 22 results of its small sample reviews and the 100 percent reviews that were actually 23 conducted that significant problems existed with diagnoses reported by its financially- 24 incentivized providers. For example, for the sample reviews for 2008, 2009, and 2010 25 encounters, nearly half (49.61 percent) of the HCCs reviewed failed the 80% validation 26 test. Over a third (37.01 percent) of all sampled diagnoses were not supported by the 27 beneficiaries’ medical records. Similarly, for the sample reviews for 2011 encounters, 28 more than half (57.34 percent) of the HCCs reviewed failed the 80 percent validation 70 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 71 of 79 Page ID #:1361 1 test, and over 30 percent of all sampled diagnoses that United reviewed for that year 2 were not supported by the beneficiaries’ medical records. 3 218. Moreover, the problems with incentivized providers’ invalid diagnoses were not 4 isolated incidents, but in many circumstances reflected a clear pattern of miscoding by 5 provider groups. Although United purposefully reviewed only a small sample of 6 medical records each year, it knew that certain provider groups were consistently 7 identified as extreme outliers on certain HCCs and failed the 80 percent sample 8 validation test year after year. 9 219. For example, every year from 2008 through 2011, Edinger Medical Group was an 10 extreme outlier for Spinal Cord Disorders/Injuries (HCC 69). Over those four years, 11 United reviewed medical records for a total of 126 beneficiaries that Edinger diagnosed 12 with Spinal Cord Disorders/Injuries and determined that the medical records of only two 13 of those beneficiaries actually supported those diagnoses. Nevertheless, United never 14 required Edinger to conduct a 100 percent review of its diagnoses that mapped to this 15 HCC and United itself never performed this 100 percent review. 16 220. Similarly, every year from 2008 through 2012, HealthCarePartners was an 17 extreme outlier for Drug/Alcohol Psychosis (HCC 51). During those five years, 18 HealthCarePartners’s highest sample validation rate for this HCC was 57.14 percent. 19 HealthCarePartners conducted 100 percent reviews for 2008, 2009, and 2011, and each 20 time found that less than a third of the diagnoses were supported by its medical records. 21 HealthCarePartners or United should have conducted 100 percent reviews for 2010 and 22 2012 and for subsequent years until it was determined that HealthCarePartners was no 23 longer reporting invalid diagnoses mapping to this HCC, but they did not do so. 24 221. Every year from 2008 through 2012, a provider named CAIPA was an extreme 25 outlier for Chronic Hepatitis (HCC 27). During those five years, CAIPA’s highest 26 sample validation rate for this HCC was 76.92 percent. CAIPA conducted 100 percent 27 reviews for 2008 and 2009 encounters, and it was never able to validate more than 56.73 28 percent of the HCCs that it reviewed. A 100 percent review should have been performed 71 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 72 of 79 Page ID #:1362 1 for 2010 through 2012 encounters and subsequent years until it was determined that 2 CAIPA was no longer reporting invalid diagnoses mapping to this HCC, but neither 3 United nor CAIPA did so. 4 222. Every year from 2008 through 2012, Hemet Community Medical Group was an 5 extreme outlier for Spinal Cord Disorders/Injuries (HCC 69). During those five years, 6 Hemet’s highest sample validation rate for this HCC was 50 percent. Hemet conducted 7 100 percent reviews for 2008 through 2011, and its highest self-audit validation rate also 8 was 50 percent. Hemet or United should have conducted a 100 percent review for 2012 9 and subsequent years until it was determined that Hemet was no longer reporting invalid 10 diagnoses mapping to this HCC, but neither did so. 11 223. Every year from 2008 through 2012, WellMed was an extreme outlier for 12 Disorders of Immunity (HCC 45). During those five years, WellMed’s highest sample 13 validation rate for this HCC was 56.00 percent and its highest 100 percent review 14 validation rate was 63.64 percent. A 100 percent review should have been performed for 15 2010 and 2012 encounters and for subsequent years until it was determined that 16 WellMed was no longer reporting invalid diagnoses mapping to this HCC, but neither 17 United nor WellMed did so. 18 224. Although the sample reviews were limited and the 100 percent reviews were not 19 always performed, United knew that the invalid diagnoses identified through the 20 RACCR Program were significant in value. In late 2012, United estimated that the 21 overpayment based on the RACCR 100 percent reviews for medical encounters in 2008, 22 2009, and 2010 (i.e., with dates of service in those years) was $79 million and, 23 accordingly, took a liability accrual in that amount. By early 2014, United also knew, 24 based on its observation of the results of the RACCR Program, that many of its large 25 incentivized providers had reported unacceptably high rates of invalid diagnoses for 26 numerous conditions. However, rather than redoubling its efforts to address invalid 27 coding by incentivized providers, United restructured RACCR into but another chart 28 review program focused on making ADDS. During the first half of 2014, United made 72 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 73 of 79 Page ID #:1363 1 changes to its RACCR Program that effectively terminated the program as a tool for 2 identifying invalid coding by incentivized providers and refunding Medicare for risk 3 adjustment payments based on those providers’ invalid diagnoses. 4 225. First, United stopped requiring any 100 percent reviews when the sample reviews 5 failed the 80 percent validation test. At the time United made this decision, it was 6 conducting RACCR sample reviews for 2012 medical encounters. Accordingly, no 100 7 percent reviews were conducted for problematic HCCs associated with 2012 medical 8 encounters. 9 226. For 2012, there were 21 outlier provider groups that had failed United’s 80 percent 10 sample validation test. United had determined that the average invalidation rate was 11 34.73 percent for these 21 providers based on the sample reviews of their problematic 12 HCCs. United knew from prior years that 100 percent reviews would result in the 13 identification of more deletes. Nonetheless, United failed to require 100 percent reviews 14 of the problematic HCCs or perform them itself. 15 227. Second, for dates of service years after 2012, United stopped selecting sample 16 beneficiaries for reviews based on whether the providers had submitted diagnoses for 17 them mapping to problematic HCCs. Instead, United began selecting beneficiaries and 18 medical records based on what it believed would yield additional codes and result in 19 increased risk adjustment payments. 20 228. Third, starting with 2013, United reviewed even fewer medical records than in 21 previous years (less than 100 total) for each incentivized provider group. Thus, United 22 reviewed only several thousand medical records as part of its new program. 23 229. By making these changes, United effectively terminated the RACCR Program and 24 deliberately avoided identifying and, thus, deleting invalid diagnoses reported by its 25 financially-incentivized providers and repaying Medicare for risk adjustment payments 26 based on them. The FCA was enacted to prevent and sanction defendants like United for 27 this type of deliberate ignorance, reckless disregard, and, in some cases, actual 28 73 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 74 of 79 Page ID #:1364 1 knowledge of the invalidity of the data they submit to the Government for payments 2 made with taxpayer dollars. 3 VII. United’s False Risk Adjustment Attestations 4 230. United submitted a Risk Adjustment Attestation each year after the final risk 5 adjustment submission deadline. United knew that it was required to submit a truthful 6 Risk Adjustment Attestation to the Medicare Program. United also knew that, if it 7 deleted invalid diagnoses from RAPS prior to the submission of the Attestation, 8 Medicare would not pay for them or would recover any erroneous payments associated 9 with them. However, United failed to do this and knowingly submitted false 10 Attestations. United had actual knowledge that the Attestations were false or acted in 11 deliberate ignorance or reckless disregard of the falsity of the Attestations. 12 231. Starting with the Attestation for payment year 2008 (if not earlier) and continuing 13 forward, United added to its Attestations a footnote which stated that the Attestations 14 were “based on facts reasonably available or made available to” it as of the date of the 15 Attestation. Facts reasonably available or made available to United included the 16 negative results of its medical record reviews as part of its Chart Review, CV, and 17 RACCR Programs. 18 232. For example, on March 9, 2012, United submitted to CMS a Risk Adjustment 19 Attestation attesting to the validity of diagnoses submitted for payment year 2011. The 20 Attestation was signed by UnitedHealth Medicare & Retirement’s Chief Financial 21 Officer Scott Theisen. United added a footnote to his Attestation stating that it was 22 “based on facts reasonably available or made available to [United] as of the date of” the 23 Attestation. In March 2012, facts reasonably available to United included, for instance, 24 the negative results of the blind medical record reviews conducted as part of it Chart 25 Review Program for payment year 2011. In addition, in March 2012, facts reasonably 26 available to United included the negative results of its CV Phase II pilot program 27 showing that there were more DELETES than ADDS. See paragraph 164 above. 28 74 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 75 of 79 Page ID #:1365 1 233. As another example, on April 30, 2014, UnitedHealthcare Medicare & 2 Retirement’s Chief Executive Officer, Steve Nelson, also sent an email to CMS about 3 the Risk Adjustment Attestation that United had just submitted for payment year 2013. 4 The email included that the Attestation was based on facts reasonably available or made 5 available to United as of the date of its submission. Those facts included, for instance, 6 millions of dollars of invalid diagnoses about which United had actual knowledge but 7 never deleted. See paragraph 186 above. 8 FIRST CLAIM FOR RELIEF 9 False Claims Act: Presentation of False or Fraudulent Claims 10 31 U.S.C. § 3729(a)(1)(A) (formerly 31 U.S.C. § 3729(a)(1)) 11 234. The United States repeats and re-alleges the allegations contained in Paragraphs 1 12 to 233 above as though they are fully set forth herein. 13 235. Defendants violated 31 U.S.C. § 3729(a)(1)(A) as follows: Defendants knowingly 14 (as “knowingly” is defined by 31 U.S.C. § 3729(b)(1)) presented or caused to be 15 presented a false or fraudulent claim for payment or approval. Specifically, Defendants 16 knowingly presented or caused to be presented a false or fraudulent Risk Adjustment 17 Attestation to the Government in order to receive and retain risk adjustment payments 18 from the Medicare Program. 19 236. Defendants violated former 31 U.S.C. § 3729(a)(1) as follows: Defendants 20 knowingly presented, or caused to be presented, to the Government a false or fraudulent 21 claim for payment or approval. Specifically, Defendants knowingly presented or caused 22 to be presented a false or fraudulent Risk Adjustment Attestation to the Government in 23 order to receive and retain risk adjustment payments from the Medicare Program. 24 237. By virtue of the said false or fraudulent claim, the United States incurred damages 25 and therefore is entitled to multiple damages under the False Claims Act, plus a civil 26 penalty for each violation of the Act. 27 28 75 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 76 of 79 Page ID #:1366 1 SECOND CLAIM FOR RELIEF 2 False Claims Act: Making or Using False Records or Statements 3 31 U.S.C. § 3729(a)(1)(B) (formerly 31 U.S.C. § 3729(a)(2)) 4 238. The United States repeats and re-alleges the allegations contained in Paragraphs 1 5 to 233 above as though they are fully set forth herein. 6 239. Defendants violated 31 U.S.C. § 3729(a)(1)(B) as follows: Defendants knowingly 7 (as “knowingly” is defined by 31 U.S.C. § 3729(b)(1)) made, used, or caused to be made 8 or used, a false record or statement material to a false or fraudulent claim. Specifically, 9 Defendants knowingly made, used, or caused to be made or used a false Risk 10 Adjustment Attestation material to a false or fraudulent claim for risk adjustment 11 payments from the Medicare Program. 12 240. Defendants violated former 31 U.S.C. § 3729(a)(2) as follows: Defendants 13 knowingly made, used, or caused to be made or used, a false record or statement to get a 14 false or fraudulent claim paid or approved by the Government. Specifically, Defendants 15 knowingly made, used, or caused to be made or used a false Risk Adjustment Attestation 16 to get a false or fraudulent claim for risk adjustment payments paid or approved by the 17 Medicare Program. 18 241. By virtue of the said false record or statement, the United States incurred damages 19 and therefore is entitled to multiple damages under the False Claims Act, plus a civil 20 penalty for each violation of the Act. 21 THIRD CLAIM FOR RELIEF 22 False Claims Act: Reverse False Claims 23 31 U.S.C. § 3729(a)(1)(G) (formerly 31 U.S.C. § 3729(a)(7)) 24 242. The United States repeats and re-alleges the allegations contained in Paragraphs 1 25 to 233 above as though they are fully set forth herein. 26 243. Defendants violated 31 U.S.C. § 3729(a)(1)(G) as follows: Defendants knowingly 27 (as “knowingly” is defined by 31 U.S.C. § 3729(b)(1)) made, used, or caused to be made 28 or used, a false record or statement material to an obligation to pay or transmit money or 76 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 77 of 79 Page ID #:1367 1 property to the Government. Specifically, Defendants knowingly made, used, or caused 2 to be made or used a false Risk Adjustment Attestation material to an obligation to repay 3 risk adjustment payments to which they were not entitled from the Medicare Program. 4 244. Defendants also violated 31 U.S.C. § 3729(a)(1)(G) as follows: Defendants 5 knowingly (as “knowingly” is defined by 31 U.S.C. § 3729(b)(1)) concealed or 6 knowingly and improperly avoided or decreased an obligation to pay or transmit money 7 or property to the Government. Specifically, Defendants knowingly concealed or 8 knowingly and improperly avoided or decreased an obligation to repay risk adjustment 9 payments to which they were not entitled from the Medicare Program. 10 245. Defendants violated former 31 U.S.C. § 3729(a)(7) as follows: Defendants 11 knowingly (as “knowingly” is defined by 31 U.S.C. § 3729(b)(1)) made, used, or caused 12 to be made or used, a false record or statement to conceal, avoid or decrease an 13 obligation to pay or transmit money or property to the Government. Specifically, 14 Defendants knowingly made, used, or caused to be made or used, a false Risk 15 Adjustment Attestation to conceal, avoid or decrease an obligation to repay risk 16 adjustment payments to which they were not entitled from the Medicare Program. 17 246. By virtue of the said false record, statement, and other acts of concealment and 18 improper avoidance, the United States incurred damages and therefore is entitled to 19 multiple damages under the False Claims Act, plus a civil penalty for each violation of 20 the Act. 21 FOURTH CLAIM FOR RELIEF 22 Restitution (Unjust Enrichment) 23 247. The United States repeats and re-alleges the allegations contained in Paragraphs 1 24 to 233 above as though they are fully set forth herein. 25 248. Defendants have received money from the United States to which Defendants 26 were not entitled, which unjustly enriched Defendants, and for which Defendants must 27 make restitution. Defendants received such money by claiming and retaining Medicare 28 77 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 78 of 79 Page ID #:1368 1 risk adjustment payments based on invalid risk adjustment data. In equity and good 2 conscience, such money belongs to the United States and to the Medicare Program. 3 249. The United States is entitled to recover such money from Defendants in an amount 4 to be determined at trial. 5 FIFTH CLAIM FOR RELIEF 6 Payment by Mistake 7 250. The United States repeats and re-alleges the allegations contained in Paragraphs 1 8 to 233 above as though they are fully set forth herein. 9 251. The United States paid money to Defendants as a result of a mistaken 10 understanding. Specifically, the United States paid Defendants claims for risk 11 adjustment payments under the mistaken understanding that such claims were based on 12 valid risk adjustment data. Had the United States known the truth, it would not have 13 paid such claims. Payment was therefore by mistake. 14 252. As a result of such mistaken payments, the United States has sustained damages 15 for which Defendants are liable in the amount to be determined at trial. 16 17 PRAYER WHEREFORE, the United States requests that judgment be entered in its favor and 18 against Defendants as follows: 19 253. On Claims I, II, and III (False Claims Act), against all Defendants jointly and 20 severally, for the amount of the United States’ damages, trebled as required by law, 21 together with the maximum civil penalties allowed by law, costs, post-judgment interest, 22 and such other and further relief as the Court may deem appropriate; 23 254. On Claim IV (Restitution), against all Defendants jointly and severally, for an 24 amount equal to the monies that Defendants obtained from the United States without 25 right and by which Defendants have been unjustly enriched, plus costs, pre- and post- 26 judgment interest, and such other and further relief as the Court may deem appropriate; 27 and 28 78 Case 2:16-cv-08697-MWF-SS Document 114 Filed 05/16/17 Page 79 of 79 Page ID #:1369 1 255. On Claim V (Payment By Mistake), against Defendants for an amount equal to the 2 United States’ damages, plus costs, pre- and post-judgment interest, and such other and 3 further relief as the Court may deem appropriate. 4 5 6 DEMAND FOR JURY TRIAL The United States of America hereby demands a trial by jury. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Dated: May 16, 2017 Respectfully submitted, CHAD A. READLER Acting Assistant Attorney General, Civil Division SANDRA R. BROWN Acting United States Attorney DOROTHY A. SCHOUTEN Chief, Civil Division DAVID K. BARRETT Chief, Civil Fraud Section LINDA A. KONTOS Deputy Chief, Civil Fraud Section Assistant United States Attorneys MICHAEL D. GRANSTON DANIEL R. ANDERSON CAROL L. WALLACK JESSICA KRIEG JUSTIN DRAYCOTT PAUL PERKINS Attorneys, Civil Division United States Department of Justice JAMES P. KENNEDY, JR. Acting United States Attorney KATHLEEN ANN LYNCH Assistant United States Attorney /S/ John E. Lee ______________________________ JOHN E. LEE Assistant United States Attorney Attorneys for the United States of America 27 28 79