Genesis Hospice Care, LLC v. Mississippi Division of Medicaid

IN THE SUPREME COURT OF MISSISSIPPI NO. 2018-SA-00284-SCT GENESIS HOSPICE CARE, LLC v. MISSISSIPPI DIVISION OF MEDICAID AND DREW SNYDER, IN HIS OFFICIAL CAPACITY AS INTERIM EXECUTIVE DIRECTOR OF MISSISSIPPI DIVISION OF MEDICAID DATE OF JUDGMENT: 01/23/2018 TRIAL JUDGE: HON. WILLIAM H. SINGLETARY TRIAL COURT ATTORNEYS: LAURA L. GIBBES WILLIAM CLARK PURDIE JANET McMURTRAY RANDALL ELLIOTT DAY, III PHILIP JOSEPH CHAPMAN COURT FROM WHICH APPEALED: HINDS COUNTY CHANCERY COURT ATTORNEYS FOR APPELLANT: PHILIP JOSEPH CHAPMAN RANDALL ELLIOTT DAY, III JULIE BOWMAN MITCHELL ATTORNEYS FOR APPELLEES: JANET McMURTRAY LAURA L. GIBBES DION JEFFERY SHANLEY NATURE OF THE CASE: CIVIL - STATE BOARDS AND AGENCIES DISPOSITION: AFFIRMED - 04/18/2019 MOTION FOR REHEARING FILED: MANDATE ISSUED: BEFORE KITCHENS, P.J., MAXWELL AND CHAMBERLIN, JJ. MAXWELL, JUSTICE, FOR THE COURT: ¶1. Genesis Hospice LLC provided outpatient hospice care to Medicaid beneficiaries in the Mississippi Delta.1 As a Medicaid provider, Genesis was required to “maintain auditable records that will substantiate the claim[s] submitted to Medicaid.”2 Because the hospice claims Genesis submitted were outside the norm, the Mississippi Division of Medicaid audited a statistical sample of 75 of the 808 billed claims. And what the multilevel audit revealed was that, of the 75 claims audited, 68 were not substantiated by the patients’ records and thus were not eligible for payment. ¶2. To be eligible for hospice care, a patient must be certified by a physician as having a “terminal illness,” defined as an expectation of death within six months if the disease follows its normal course. The auditing physicians specifically found that the patient records for the 68 rejected claims lacked sufficient documentation to support the given terminal- illness diagnosis and/or lacked documentation of disease progression.3 Medicaid’s statistician extrapolated that 68 of 75 unsupported claims represented a total overpayment of $1,941,285 for the 808 claims Genesis billed during the relevant time period. And Medicaid demanded Genesis repay this amount. 1 Medicaid is a federal and state program of medical assistance to qualified individuals. The Mississippi Division of Medicaid is the state agency designated to administer the Medicaid program in Mississippi. Miss. Div. of Medicaid Provider Policy Manual (PPM) § 14.01 (June 1, 2005). 2 PPM § 7.03 (July 1, 2000). Though now replaced by the Mississippi Administrative Code, the PPM controlled during the relevant time period. 3 Two of these rejected claims had no documentation at all. 2 ¶3. Medicaid’s decision has been affirmed in an administrative appeal before Medicaid and by the Hinds County Chancery Court, sitting as an appellate court. On further appeal to this Court, Genesis essentially argues Medicaid unfairly imposed documentation requirements not found in the federal or state Medicaid regulations. Genesis instead insists the only requirement was a physician’s certification that in his or her subjective clinical judgment the patient was terminally ill, which Genesis provided. ¶4. But the regulations are clear. A physician’s certification of terminal illness is indeed required, but so is documentation that substantiates the physician’s certification. Substantiating documentation is necessary when, as in this case, Medicaid conducts an audit. Without substantiation—or, to use the auditing physicians’ terminology, without “objective clinical data supporting a terminal diagnosis and/or . . . supporting a progression of a terminal disease”—the auditors may reasonably conclude, as they did here, that hospice care was not in fact medically necessary. ¶5. Because Genesis’ records failed to support 90 percent of its hospice claims, Medicaid had the administrative discretion to demand these unsupported claims be repaid. Therefore, we affirm. Background Facts and Procedural History I. Audit ¶6. In 2009, Medicaid ran a data analysis of all hospice claims paid during January 1, 2006, through December 31, 2008, for beneficiaries who lived longer than six months.4 4 Excluded from its analysis were beneficiaries diagnosed with cancer or acute liver failure. 3 Medicaid was looking for billing aberrations evidencing overpayment. And Medicaid found many aberrations among the 808 line-item claims Genesis billed during this time frame. So Medicaid initiated an audit of a statistical sample of 75 claims billed for 30 patients. ¶7. These 75 claims were scrutinized at three levels—an internal audit by Medicaid, conducted by one of its staff nurses; a third-party audit by a reviewing physician; and a third- party audit by a three-physician peer-review panel.5 What the panel ultimately concluded—after giving Genesis the opportunity to rebut its findings—was that the patient records did not support the medical necessity of the hospice care billed for 68 of the 75 claims. ¶8. Medicaid defines hospice as palliative care for terminally ill patients. PPM § 14.02 (Sept. 1, 2007). To be eligible for hospice, “the beneficiary must be certified as being terminally ill with a life expectancy of six (6) months or less, and there must be a documented diagnosis consistent with a terminal stage of six (6) months or less.” Id. Moreover, there must be a plan of care and documentation of the beneficiary’s terminal illness. PPM § 14.03 (Sept. 1, 2007). The panel’s unanimous finding hinged on Genesis’ failure to document “objective clinical information that corroborated the signed certification of terminal illness.” (Emphasis added.) For two claims, both for Patient Doe,6 no records were submitted. And for the 66 other uncorroborated claims, “the information presented in the records did not support a reasonable clinical expectation of death within 6 months time 5 Medicaid contracts with consulting firm Health Systems of Mississippi to conduct further audits when Medicaid’s initial audit reveals potential billing violations. 6 For healthcare privacy reasons, this patient’s name is withheld. 4 either because of insufficient objective clinical data supporting a terminal diagnosis and/or [because of] insufficient objective clinical data supporting progression of a terminal disease.” In other words, for 68 of 75 of the claims, the patients’ records did not substantiate the physician certification that the beneficiary was terminally ill, leading Medicaid to conclude hospice care was not medically necessary and thus not compensable. ¶9. Hospice providers “must maintain auditable records that will substantiate the claim submitted to Medicaid.” PPM § 14.12 (Oct. 1, 2000). “If a hospice’s records do not substantiate [hospice] services paid,” the provider “will be asked to refund to [Medicaid] any money received for such non-substantiated services.” Id. That is exactly what Medicaid did here. Based on the panel’s conclusion that 68 of the 75 audited claims were not hospice appropriate, Medicaid’s statistician estimated that, for the 808 line-item claims billed, Genesis was overpaid $1,941,285.7 By letter on March 10, 2013, Medicaid demanded Genesis repay this amount. IV. Appeal ¶10. Genesis responded by requesting an administrative appeal. While that appeal was pending, Genesis ceased operating July 31, 2013. Following a hearing, the hearing officer recommended to affirm, which Medicaid did in October 2014. ¶11. In November 2014, Genesis further appealed to the Chancery Court of Hinds County. See Miss. Code Ann. § 43-13-121(1)(j) (Supp. 2014) (statutorily authorizing, after July 1, 2014, appeals of Medicaid’s recoupment decisions to the Hinds County Chancery Court). 7 Genesis had been paid $2,096,053 total, leaving $154,768 in approved claims. 5 While its appeal was pending before that court, Genesis was administratively dissolved in December 2014. Following the chancery court’s January 2018 ruling to affirm, Genesis appealed to this Court.8 Issues on Appeal ¶12. On appeal, Genesis argues Medicaid acted arbitrarily, capriciously, and contrary to the substantial weight of the evidence: (1) by denying claims based on the to failure to document “disease progression”; (2) by denying claims based on a physician certification of terminal illness; and (3) by denying Patient Doe’s two claims, for which no records had been submitted to the auditor. Genesis also asserts: (4) the Administrative Hearing Officer improperly upheld Medicaid’s use of extrapolation. ¶13. This Court applies the same discretionary standard of review as the chancellor who first reviewed the administrative decision—reversing only if the decision (1) was not supported by substantial evidence, (2) was arbitrary or capricious, (3) was beyond Medicaid’s power to adopt, or (4) violates a constitutional or statutory provision. Crossgates River Oaks Hosp. v. Miss. Div. of Medicaid, 240 So. 3d 385, 387 (Miss. 2018). Analysis I. Motion to Dismiss 8 While this appeal was pending, Genesis’ sole member and administrator, Charlene Brandon, pleaded guilty to one count of conspiracy to commit federal healthcare fraud. 6 ¶14. As a preliminary matter, Medicaid has filed a motion to dismiss. This motion was passed for consideration with the merits of the appeal. ¶15. Because Genesis was administratively dissolved in December 2014—more than three years before it filed its notice of appeal with this Court—Medicaid asserts Genesis had no authority to file the notice of appeal, because it no longer legally exists. ¶16. As support, Medicaid cites Mississippi Code Section 79-29-831, the statute that spells out the effects of an administrative dissolution.9 Medicaid selectively hammers on one provision—“A limited liability company that has been administratively dissolved may not maintain any action, suit or proceeding in any court of this state until such limited liability company is reinstated.” Miss. Code Ann. § 79-29-831(4) (Rev. 2013). Medicaid argues this provision precluded Genesis from continuing to pursue its action against Medicaid without first being reinstated. ¶17. Genesis has responded by citing another provision of Section 79-29-831—“The administrative dissolution of a limited liability company shall not . . . prevent such limited liability company from defending any action, suit or proceeding with any court of this state.” Miss. Code Ann. § 79-29-831(2) (Rev. 2013) (emphasis added). Genesis takes the position that, by further appealing Medicaid’s decision, it is not “maintaining an action” against Medicaid. Rather, Genesis continues to defend itself against Medicaid’s administrative decision. 9 Unlike a voluntary dissolution, which is initiated by a company itself, an administrative dissolution is initiated by the Secretary of State when a company makes one of the enumerated missteps in Mississippi Code Section 79-29-821 (Rev. 2013). 7 ¶18. We agree with Genesis. While Medicaid tries to characterize this matter as an original action against Medicaid, both the administrative regulation and the statute Medicaid cites clearly refer to Genesis’ requests, first, for an administrative hearing and, next, for chancery- court review as appeals. Miss. Code Ann. § 41-13-121(1)(j) (Supp. 2014); 23 Miss. Admin. Code Pt. 300. To follow Medicaid’s logic, an administratively dissolved limited liability company loses its right to appeal—even in a proceeding in which the company began in a defensive posture. Since this position contradicts Section 79-29-831(2), which allows Genesis’ appeal to proceed, Medicaid’s motion to dismiss is not well taken and is therefore denied. II. Medical Necessity of Hospice Care ¶19. Turning to the merits of Genesis’ appeal, the now-dissolved hospice provider first attacks Medicaid’s reasoning for denying 66 of the 68 rejected sample claims. The peer- review panel concluded hospice care for these 66 claims was not medically necessary “because of the failure of [Genesis] to document objective clinical information that corroborated the signed certifications of terminal illness.” ¶20. While presented as two separate issues, Genesis’ first two points of error are intertwined. Basically, as Genesis reads the applicable regulations, the medical necessity, and thus eligibility, for hospice care hinges “strictly” on the physician’s certification that, in his or her subjective clinical opinion, the patient has a terminal illness. This certification, Genesis asserts, is entitled to the “utmost deference,” especially here, where 20 different physicians certified the terminal illnesses for 30 patient records audited. So as Genesis sees 8 it, by requiring objective documentation—especially of “disease progression”—to substantiate the certification, Medicaid has retroactively imposed a new substantive requirement in violation of its due-process rights. ¶21. But Genesis’ claim that hospice eligibility is “strictly” based on a physician’s certification of terminal illness is clearly contradicted by the regulations. Genesis is correct that, “[t]o be eligible to elect hospice care under Medicaid, the beneficiary must be certified as being terminally ill with a life expectancy of six (6) month or less[.]” PPM § 14.02 (Sept. 1, 2007). However, there are other express requirements. “[T]here must be a documented diagnosis consistent with a terminal state of six (6) months or less.” Id. (emphasis added). The regulations also required a “written plan of care established and reviewed by the hospice’s interdisciplinary team at each enrollment period10 and updated as required by the beneficiary’s condition.” PPM § 14.02 (June 1, 2005). And “[a] new or updated plan of care, along with supporting documents that explains the beneficiary’s condition (i.e., progress notes from MD, nursing notes that explain beneficiary’s physical condition), must be retained in the patient’s medical record[.]” Id. (emphasis added). ¶22. Thus, Medicaid regulations very clearly required additional documentation to substantiate a physician’s certification of terminal illness. See PPM § 14.12 (Oct. 1, 2000) (requiring the beneficiary’s record to include, in addition to a physician’s certification, a plan of care that supports each hospice service rendered, treatment rendered, documentation to 10 The hospice benefit is divided into distinct enrollment periods. The first period lasts up to 90 days. The second period is also for 90 days. After that, each benefit period is for 60 days. The beneficiary must be recertified as terminally ill for each period. PPM § 14.02 (June 1, 2005). 9 show relationship of the treatment plan and medications to the terminal illness). And so did federal regulations. See 42 C.F.R. § 418.74 (2008) (requiring each beneficiary’s record to contain, among other things, “[c]omplete documentation of all services and events (including the evaluation of treatments, progress notes, etc.)”); 42 C.F.R. § 418.104 (2008) (requiring “correct clinical information” for each patient, including “[r]esponses to medications, symptom management, treatments and services[,] . . . [o]utcome measured data elements, . . . [and] [p]hysician certification and recertification of terminal illness”). ¶23. Though none of provisions contained the term “disease progression,” we agree with the administrative hearing officer that, “[t]aken together, these regulations require that the hospice provider document disease progression or lack of disease progression or improvement.” (Emphasis added.) That the term “disease progression does not appear in the regulation does not change the fact that the regulations require documentation from which an auditor can determine whether or not the treatment is necessary and appropriate.” So “[i]t was entirely appropriate for the physicians auditing hospice records to note an absence of evidence of disease progression, since the patient was initially certified as having an illness, which, if it followed its normal course, would result in death within 6 months.” ¶24. This documentation requirement clearly distinguishes this case from the one Genesis cites in its brief, Bethany Hospice Services of Western Pennsylvania v. Department of Public Welfare, 88 A.3d 250 (Penn. 2013). In Bethany Hospice Services, Pennsylvania’s guidelines for hospice eligibility only required a physician to certify the beneficiary as terminally ill. There was no documentation requirement. Id. at 252. So, by later rejecting 10 a hospice claim for lack of documentation that the patient was terminally ill and reasonably expected to die within six months, the Pennsylvania Department of Welfare retroactively imposed an additional, extra-regulatory requirement, supported only by its expert’s testimony. Id. at 255. ¶25. By contrast, Mississippi’s Medicaid policies clearly require more than just a physician certification of terminal illness. So by basing its decision, in large part, on the lack of “objective clinical data supporting progression of a terminal disease,” Medicaid did not retroactively create a new substantive requirement in violation of Genesis’ due-process rights. Instead, it was acting within already-existing, clear policy that patient records must contain documentation to substantiate each hospice claim. See PPM § 14.12 (Oct. 1, 2000) (“If a hospice’s records do not substantiate services paid for under the Mississippi Medicaid program, as previously noted, the [provider] will be asked to refund to the Mississippi Medicaid Program any money received for such non-substantiated services.”). ¶26. Hospice provides palliative care, which Medicaid defines “as the relief of severe pain or other physical symptoms and supportive care to meet the special needs arising out of physical, psychological, spiritual, social, and economic stresses which are experienced during the final stage of illness and during dying and bereavement.” PPM § 14.02 (Sept. 1, 2007). Here, the three physicians reviewing the records were unanimous in their findings. Each found the vast majority of the patients’ records contained insufficient objective data showing the patient was experiencing an illness, which in its normal course would result in death 11 within six months. Thus, it was reasonable for the three-physician panel to conclude Genesis had failed to substantiate the medical necessity of hospice care. ¶27. At the administrative hearing, “Genesis submitted no evidence, medical or otherwise,” to rebut the panel’s finding. So, by adopting the panel’s decision, Medicaid was not acting contrary to the weight of the evidence. Thus, there is no basis to overturn this decision. III. Patient Doe ¶28. Next, Genesis challenges the denial of the two claims for Patient Doe. The peer- review panel rejected these two claims because Genesis failed to submit any records until after the panel had deliberated. ¶29. At the administrative hearing, Genesis challenged the inclusion of these two claims in the sample size because Medicaid records indicated Medicaid had already been reimbursed for these two claims. The hearing officer rejected this specific challenge because Genesis did not repay the claims. Instead, the claims were reimbursed by a third party as part of a personal-injury settlement with Patient Doe. Consequently, the hearing officer found Genesis was not entitled to keep the payments for these claims. The chancellor similarly rejected Genesis’ argument, holding that the two claims for Patient Doe were properly included within the sample, because they fell within the criteria and relevant time period. The chancellor further found the two claims were properly denied, because Genesis failed to provide any supporting records until after the peer-review panel had already made its final determination. 12 ¶30. In its appeal to this Court, Genesis abandons its argument that Patient Doe’s claims should not have been included in the statistical sample. Instead, it accuses Medicaid of arbitrarily refusing to review Patient Doe’s records, which Genesis insists it submitted. Further, Genesis argues it never received timely notice these records were missing. So it should have been allowed to resubmit these records, even after the audit had been completed. ¶31. This issue is procedurally barred. Genesis’ argument to the hearing officer focused on the appropriateness of including the two claims in the sample audit based on the third- party reimbursement. Genesis did not ask—or at least does not point to where in the record it asked—the hearing officer to fault Medicaid for the records not being submitted. See Fowler v. White, 85 So. 3d 287, 293 (Miss. 2012) (holding that an issue raise for the first time on appeal is procedurally barred). Still, even if not barred, Genesis did not develop this issue. We have scoured the record and find no evidence that supports Genesis’ claim it initially provided Medicaid with Patient Doe’s records and was never timely notified the records went missing. In other words, there is no definitive record evidence showing Medicaid’s disapproval of these claims for lack of records was arbitrary, capricious, or unsupported by the substantial weight the evidence. Thus, we find no basis in the record to overturn Medicaid’s decision to include Patient Doe’s two claims in the audited sample. IV. Extrapolation ¶32. Finally, Genesis challenges Medicaid’s use of extrapolation—using the audit of the statistical sample of 75 claims to estimate Medicaid’s total overpayment for all 808 claims. 13 ¶33. Like Genesis’ medical-necessity argument, Genesis points to the physician certifications of terminal illness to justify why Medicaid could not use extrapolation. Because 20 different physicians certified the 30 patients in the sample as terminally ill—and even more physicians certified terminal illness in the entire universe of claims—Genesis argues each hospice claim billed was too fact-dependent and unrelated to apply statistical sampling. ¶34. As support, Genesis quotes an unreported Texas federal district court case. It pitches this case for the proposition that, when “each and every claim at issue was fact-dependent and wholly unrelated to each and every other claim, and determining eligibility for each of the patients involved a highly fact-intensive inquiry involving medical testimony after a thorough review of the detailed medical chart of each individual patient, . . . the case [is] not suited for statistical sampling.” United States ex rel. Wall v. Vista Hospice Care, Inc., No. 3:07-CV-00604-M, 2016 WL 3449833, at *12 (N.D. Tex. June 20, 2016). But Wall posed a different scenario than this one. It concerned a qui tam action on behalf of the United States for violations of the federal False Claims Act (FCA) in connection to hospice claims. Id. at *1 (citing 31 U.S.C. § 3729 (2012)). And the question the district court faced was “whether statistical sampling and extrapolation can be used to establish liability in an FCA case where falsity depends on individual physicians’ judgment regarding individual patients.” Id. at *12 (emphasis added). That is not the question presented in this case. So Genesis’ reliance on Wall and that case’s focus on individual physicians’ judgment is totally misplaced. 14 ¶35. The specific question here is whether Medicaid could use statistical sampling to calculate overpayment to Genesis. And the majority of courts that have addressed this question have approved of sampling. See, e.g., Ratanasen v. California, 11 F.3d 1467, 1471 (9th Cir. 1993) (“We now join other circuits in approving the use of sampling and extrapolation as part of audits in connection with Medicare and other similar programs, provided the aggrieved party has an opportunity to rebut such evidence.”); Yorktown Med. Lab., Inc. v. Perales, 948 F.2d 84, 89-90 (2d Cir.1991) (rejecting the claim that extrapolation violates due-process rights); Ill. Physicians Union v. Miller, 675 F.2d 151, 155 (7th Cir. 1982) (holding that “extrapolation based on review of a relatively small sample is a valid audit technique in cases arising under the Social Security Act”); United States v. Cabrera- Diaz, 106 F. Supp. 2d 234, 240 (D.P.R. 2000); Chaves Cty. Home Health Serv. v. Sullivan, 931 F.2d 914 (D.C. Cir. 1991), cert. denied, 502 U.S. 1091, 112 S. Ct. 1160, 117 L. Ed. 2d 408 (1992); Mile High Therapy Ctrs., Inc. v. Bowen, 735 F. Supp. 984 (D. Colo. 1988); Georgia v. Califano, 446 F. Supp. 404, 409-10 (N.D. Ga. 1977) (“Projection of the nature of a large population through review of a relatively small number of its components has been recognized as a valid audit technique and approved by federal courts in cases arising under Title IV of the Social Security Act.”). See also Goldstar Med. Servs., Inc. v. Dept. of Soc. Servs., 955 A.2d 15, 31 (Conn. 2008) (“It is well established that proof of damages through the use of statistics and statistical sampling has been endorsed in numerous cases involving medicare and medicaid overpayments.”). Indeed, the Seventh Circuit has held that, “in view 15 of the enormous logistical problem of Medicaid enforcement, statistical sampling is the only feasible method available.” Miller, 675 F.2d at 157. ¶36. That is exactly what the administrative hearing officer concluded here, noting the Genesis matter was “a prime example of why extrapolation is necessary.” The 75-claim audit still required countless man hours to copy records, conduct an initial review, hire a consultant for a second review, send the records to the peer-review panel for a third review, and then make a demand for repayment and allow for an administrative appeal. To individually audit all 808 claims, in the hearing officer’s view, “would be worse than impractical; . . . it would be impossible.” ¶37. We find no abuse of discretion. Sampling was the only feasible method to determine how much Medicaid overpaid Genesis for unsubstantiated hospice claims. And apart from one irrelevant, unreported district court case, Genesis cites no authority supporting its argument that Medicaid’s use of extrapolation was improper. ¶38. Alternatively, Genesis argues the particular extrapolation performed by Medicaid’s statistical expert, Dr. Mark McComb, was problematic and should have been rejected by the administrative hearing officer. ¶39. As support, Genesis points to the expert testimony of Dr. Harold Haller, Genesis’ sole witness at the administrative hearing. At one point in his testimony, Dr. Haller challenged Dr. McComb’s claims-based sampling unit. According to Dr. Haller, because most beneficiaries had multiple claims, the auditor would not have been able to independently judge each claim for a single beneficiary independently of the beneficiary’s other claims. 16 Instead, Dr. Haller believed the sampling unit should have been based on beneficiaries, not claims. The administrative hearing officer rejected Dr. Haller’s opinion noting the auditor still “would be in the same position of judging multiple claims for a single beneficiary.” On appeal, Genesis points to Dr. Haller’s rejected solution as proof that Dr. McComb’s sampling methodology was fatally flawed because independence was impossible. But the administrative hearing officer expressly found, based on subsequent testimony, that each claim had been audited independently of other claims. Because the administrative hearing officer supported his finding of statistical independence with substantial evidence, we find no abuse of discretion in accepting Medicaid’s extrapolation as statistically proper. ¶40. Therefore, we affirm Medicaid’s administrative decision that Genesis must repay $1,941,285 for unsubstantiated hospice claims. ¶41. AFFIRMED. RANDOLPH, C.J., KITCHENS AND KING, P.JJ., COLEMAN, BEAM, CHAMBERLIN, ISHEE AND GRIFFIS, JJ., CONCUR. 17