No. Civ. 98-3075.
May 19, 1999
Charles D. Davidson, Davidson Law Firm, Ltd., Little Rock, AR, for plaintiff.
Matt W. Fleming, Asst. U.S. Attorney, Fort Smith, AR, for defendant.
This case is before the court for review of the decision of the Secretary of Health and Human Services that E. Russell Webb, M. D., had received overpayments under Medicare for certain laboratory services. Review of the Secretary's underlying decision is governed by 42 U.S.C. § 1395oo(f)(1), which incorporates the standard of review of the Administrative Procedure Act (APA). Under the APA, we may set aside agency action only if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law." 5 U.S.C. § 706(2)(A).
"The Medicare program is divided into two parts. 42 U.S.C. § 1395-1395ccc (1994). Part A of the program deals with hospitalization benefits. Part B of the program is a supplementary medical insurance program for the aged and disabled." Clarinda Home Health v. Shalala, 100 F.3d 526, 528 (8th Cir. 1996). This case involves only payment under Part B of Medicare.
To administer the Medicare program, the Secretary is authorized to enter into contracts with carriers. 42 U.S.C. § 1395h(a), 1395u(a). Pursuant to this authority, the Health Care Financing Administration (HCFA) entered into a contract with Blue Cross Blue Shield of Arkansas (BCBSA) to act as the Medicare carrier in this area. "Carriers are authorized to complete several tasks including: determining the rates and proper payment amounts to providers of services; auditing the records of providers; and receiving and accounting for payments made to providers." Clarinda, 100 F.3d at 528.
The Secretary establishes fee schedules for clinical diagnostic laboratory tests for which payment is made under Medicare Part B. 42 U.S.C. § 13951(h)(1)(A). With respect to clinical diagnostic laboratory tests performed by a physician or a laboratory, the type of tests involved in this case, the fee schedules are established on a regional, statewide, or carrier service area basis. 42 U.S.C. § 13951(h)(1)(B). The fee schedule sets out the precise amount of reimbursement Medicare will pay for each particular type of laboratory test. 42 U.S.C. § 13951(h)(2).
Under Medicare, medical payments are made for expenses incurred for items or services which are reasonable and necessary for the diagnosis or treatment or illness or injury or to improve the functioning of a malformed body member. 42 U.S.C. § 1395y(a)(1). To receive payment from Medicare, a physician must submit a claim for the test or tests to his carrier.
Services or items under the Medicare Part B program are identified by a five digit numerical code. Administrative Record at 13 (hereinafter A.R. followed by the page number). "The code numbers are used in the HCFA common procedure coding system (HCPCS) to identify medical services and procedures for standardized billing and payment services." AR. at 13. The "system is based on the American Medical Association's Physicians' Current Procedural Terminology (CPT). The CPT is a systematic listing and coding of procedures and services performed by physicians. The codes are utilized to simplify the reporting and identification process." A.R. at 14. When seeking reimbursement, the physician must indicate, by use of the appropriate CPT code, what tests were performed.
Dr. Webb is a physician engaged in the practice of urology in Mountain Home, Arkansas. On February 27, 1992, BCBSA initiated a post-payment review of Dr. Webb's practice. The claims submitted by Dr. Webb were selected for review because he was a high volume provider overall and his pattern of claims significantly exceeded that of his peers. A.R. at 13. Additionally, Dr. Webb had higher rates of laboratory testing and surgical procedures per 1000 patients than his peers in the region. A.R. at 13.
The period of review was July 1, 1991, to December 31, 1991. A.R. at 10903. BCBSA requested medical records on 24 patients treated during this period. Id.
Following this initial review, it was determined that a full scale investigation covering four years was warranted. A.R. at 10903. The investigation involved Medicare claims filed for specified laboratory services furnished between July 1, 1988, and June 30, 1992. The review was conducted utilizing a random sampling of claims paid during this period. Id. The random sampling method used was "developed by the United States Department of Commerce and refined for the specific purposes of the Medicare Program by the Department of Health and Human Services Inspector General's Office, Division of Program Audit." A.R. at 10919. During this full scale investigation, medical records for 227 Medicare beneficiaries were reviewed. A.R. at 14.
On September 30, 1994, BCBSA notified Dr. Webb of its completion of the post-payment review. A.R. at 10917. BCBSA identified a potential overpayment of $930,976. AR. at 10917. Dr. Webb's total Medicare reimbursement for this period was $4,027,444. A.R. at 11. The review focused on laboratory services billed under the following CPT codes for organ or disease test panels: 53670 (catheterization, urethra; simple); 52281 (cystourethroscopy, with calibration and/or dilation of urethral stricture or stenosis, with or without meatotomy and injection procedure for cystography, male or female); 80050 (general health screen panel); 80058 (hepatic panel); and 80080 (prostatic panel). A.R. at 10920.
With respect to CPT codes 80050, 80058, and 80080, the overpayment letter notified Dr. Webb that the "[m]edical record documentation failed to support the medical necessity of services reviewed for laboratory organ or disease oriented" codes. A.R. at 10921. Further, it was noted that:
the laboratory tests you included in these organ or disease oriented panel codes (80080, 80050, and 80058) failed to include the specific tests required by Medicare as outlined in the February 28, 1990 issue of Providers' News. All individual tests included by you in the organ or disease oriented panel codes (80050, 80050, and 80058), with the exception of prostatic acid phosphatase (84065), are included on the list of tests frequently done as groups and combinations ("profiles") on automated multichannel equipment. Any combination of tests from this list (published in February 28, 1990) and a corrected list published in the April 16, 1990 issues of Providers' News that are medically necessary and are performed for the same patient on the same day should be billed using the appropriate multichannel profile (80002-80019). Medicare does not reimburse for screening diagnostic procedures but does reimburse for services that are reasonable and necessary. The documentation for a diagnostic service must clearly describe the signs, symptoms, conditions and treatment of the patient, establishing the rationale for services rendered.A.R. at 10921.
Dr. Webb was advised that in the event he disagreed with the decision, he could, within six months of the date of the letter, request a hearing before a hearing officer. A.R. at 10924. See 42 C.F.R. § 405.821-836. Dr. Webb requested a hearing. The hearing was held on January 6, 1995, and by letter dated June 28, 1995, Patricia Braswell, a Senior Hearing Officer, rendered her decision. A.R. at 10901 to 10914.
Braswell addressed the following specific issues: (1) "Was Dr. E. Russell [Webb] overpaid for the claims paid July 1, 1988 — June 30, 1992 for procedure codes 52281, 80050, 80058, and 80080?; (2) What is the amount Dr. Webb has been overpaid?; and (3) Was Dr. Webb at fault in causing the overpayment?" A.R. at 10907.
Braswell found as follows: (1) Dr. Webb was overpaid for procedure codes 80050, 80058, and 80080 for claims paid during the period of July 1, 1998 — June 30, 1992; (2) The amount of the overpayment was $708,812.00 after removing the claims involving code 52281; and (3) Dr. Webb was at fault in causing the overpayment. Braswell concluded Dr. Webb's entries in his medical records did not contain adequate information to support medical need for these procedures as billed. A.R. at 10907.
In so doing, Braswell referred to the February 28, 1990, newsletter noting that it included the following list of tests:
80050 (General Health Screening Panel)-82465 (serum cholesterol); 82948 (blood glucose); 83700 (blood lipids), and 85022(CBC).
80058 (Hepatic Function Panel)-82040 (albumin), 82150 (amylase), 82250 (bilirubin), 83705 (lipids), 84080 (phosphatase), 84155 (protein total serum), 84450 (SGOT), and 84460 (SGPT).
80080 (Prostatic Panel)-84060 (acid phosphatase), 84403 (testosterone), 84450 (uric acid).A.R. at 10909. At the end of the article, the following statement appeared: "NOTE: Grouping of specific tests other than the above list for panels 80050-80090 will have to be identified on the claim form." Id. The carrier developed fee schedules for these panels were "based on allowing the fee schedule amount for each test in the panel for one total allowance for the panel." A.R. at 10909.
Documentation submitted by Dr. Webb to the carrier in 1993 showed his panels were as follows:
80050 (General Health Panel)-84132 (potassium), 82948 (blood glucose), 82465 (serum cholesterol).
80058 (Hepatic Panel)-84450 (SGOT), 84460 (SGPT), 82040 (albumin), and 84155 (total protein).
80080 (Prostatic Panel)-84520 (BUN), 82540 (creatinine), 84550 (uric acid), 82310 (calcium), 84080 (alkaline phosphatase), and 84060 (prostatic acid phosphatase).A.R. at 10909.
Braswell concluded, Dr. Webb did not bill the panels for dates of service after February 28, 1990, as stated in the February 28th newsletter. A.R. at 10909. Specifically, she found that "[t]he tests [Dr. Webb] included in his panel codes did not match the tests as stated by the carrier and resulted in an incorrect allowance for his panel codes after February 28, 1990." A.R. at 10909.
Noting that a provider is not responsible for overpayment if he is determined to be without fault in causing the overpayment, Braswell concluded Dr. Webb was at fault in causing the overpayment. Specifically, she stated:
He has billed for services that are not supported in his medical records as being reasonable and necessary for diagnosis or treatment of illness or injury or to improve the functioning of a malformed body member. The medical records submitted as evidence, in the vast majority of claims reviewed for services provided July 1, 1998 — June 30, 1992, do not support medical need for the services billed.A.R. at 10913.
The sampling methodology was reviewed by Dr. Alan Johnson, a statistician and consultant for the carrier and Hearing Officer. Dr. Johnson "determined, through extensive review of the claim file, that the carrier's calculations do represent a statistically valid random sampling methodology and do represent an accurately calculated overpayment amount based on the random sampling methodology." A.R. at 10914.
In the event he disagreed with the decision, Dr. Webb was advised that he had the right to a hearing before an Administrative Law Judge (ALJ) of the Office of Hearings and Appeals of the Social Security Administration. A.R. at 10914. See 42 C.F.R. § 405.855; 20 C.F.R. § 404.967981. On July 25, 1995, Dr. Webb appealed the June 28, 1995, decision. A.R. at 10900.
A hearing was held on December 18 and December 19, 1996. By notice dated April 29, 1997, Dr. Webb was notified of the adverse decision of the ALJ, Garry L. Brewer. A.R. at 8. The ALJ found that "Dr. Webb was overpaid in Medicare benefits in the amount of $708,812.00, that he was not without fault in the overpayment and that he is liable for that amount." A.R. at 12.
The ALJ noted that Medicare grouped tests which are utilized by physicians in specific situations, such as screening purposes or for analysis of a specific disease or organ. A.R. at 14. These grouped tests were coded under the organ or disease oriented panels. Medicare pays for these tests as a group, based on one code, under an established reasonable charge schedule, which takes into consideration each individual test specified for that panel. A.R. at 14.
Laboratory tests that are done as groups or combinations on automated multi-channel equipment result in a reduction in the cost of the tests. A.R. at 15. These tests are coded depending on the number of tests performed, i.e., 80002 for two clinical chemistry tests; 8003 for 3 clinical chemistry tests, and so on. A.R. at 15. Medicare reimburses according to the number of tests performed. A.R. at 15. "If several individual tests are submitted to Medicare for the same patient on the same day, and those tests are the type that can be done on multichannel equipment, reimbursement is prorated as a percentage of the total allowance even for the appropriate multichannel code (80002-80019)." A.R. at 15.
"The organ or disease panel codes contain tests of the automated variety as well as tests that must be done individually. This factor is reflected in the amount of reimbursement a provider receives for a billed panel." A.R. at 15. The ALJ found that when Dr. Webb billed Medicare using a specified code, the carrier assumed the panels were composed according to Medicare guidelines. A.R. at 15.
At the hearing before the ALJ, Barbara McDanel testified that under the 1992 fee schedule medicare would have paid the following amounts for claims submitted using the following panel codes: 80050 — $26.14; 80058-$70.07; and 80080 — $130.50. A.R. at 15. Whereas, she testified that the tests performed by Dr. Webb, if submitted by their appropriate individual codes instead of as a panel, would have been reimbursed as follows: 3 tests in the 80050 panel — $10.17; 4 tests in the 80058 panel — $10.68; 6 tests in the 80080 panel — $44.75. A.R. at 16. As a result of Dr. Webb's panels being comprised of different tests than those specified under HCPCS, the ALJ found Dr. Webb's reimbursement from Medicare was significantly different than it should have been. A.R. at 15.
From this, the ALJ concluded it was clear that Dr. Webb had been "overpaid as a result of the erroneous way he coded his tests and billed Medicare." A.R. at 16. The ALJ further concluded Dr. Webb "had reason to be aware of the potential problem his billing error would make." A.R. at 16.
The ALJ referred to the February 28, 1990, newsletter noting that in the section titled "Laboratory panels — reminder" it stated:
SPECIFIC ORGAN OR DISEASE-ORIENTED PANELS — PANELS WITH SPECIFICALLY ASSIGNED INDIVIDUAL TESTS
The following laboratory panels are considered by Medicare to include the specific tests listed. The listed individual tests will be denied when reported in conjunction with an associated panel test.A.R. at 16. The panel tests at issue herein are listed with the tests specified. Id. Further, the ALJ noted that page three of the newsletter contained the following statement: "NOTE: Grouping of specific tests other than the above list for panels 80050-80090 will have to be identified on the claim form." A.R. at 16.
Donna Dewey, the individual responsible for coding and billing for Dr. Webb, acknowledged that Dr. Webb's office received the newsletter. A.R. at 16. She testified that she relied on a field representative from the carrier who advised her the way the panels were being billed was appropriate. A.R. at 16.
The ALJ concluded the newsletter put Dr. Webb on notice that the billing practices should change in accordance with that notice. A.R. at 17. The ALJ noted that under the Organ and Disease Oriented Panels section of the CPT it states:
These panels were developed for coding purposes only and should not be interpreted as clinical parameters. The tests listed with each panel, identify the defined components of that panel. These panel components are not intended to limit the performance of other tests. If one performs tests in addition to those specifically indicated for a particular panel, those tests should be reported separately in addition to the panel code.A.R. at 17.
The ALJ concluded:
Dr. Webb did not structure his panels according to the guidelines furnished by Medicare. He billed Medicare as if he had. He did not specify in his claims that he was using tests different from the panel. This resulted in Dr. Webb receiving more from Medicare than he was entitled. In effect, Dr. Webb was getting paid for tests he did not do.A.R. at 17. The carrier calculated the overpayment using a date thirty days after the date of the issuance of the newsletter, April 1, 1990, as the reference date. The ALJ found this to be appropriate. A.R. at 17. The carrier did not deny all payment for the tests, instead after April, 1990, the carrier paid Dr. Webb for medically necessary tests based upon the appropriate multichannel test. A.R. at 17-18.
With respect to the determination of medical necessity, the ALJ found both the carrier and the reviewers were generous in their determinations allowing many tests that were of marginal justification. A.R. at 18. The ALJ stated:
A review of the record shows that there is a well defined basis for each of the denied tests. In some cases, tests were repeated at close intervals without documented need. Some tests were done without support shown in the medical history or clinical findings. Many tests appear to have been for screening purposes. Dr. Webb sometimes billed for a CBC separately even though he was paid for it in the general health panel. After review of the record, the Administrative Law Judge finds that substantial evidence supports the findings of the carrier with regard to the lack of medical necessity of the laboratory tests that were denied.A.R. at 18.
Based on Dr. Webb's experience and participation in the Medicare program, the ALJ concluded Dr. Webb should have known that these services were not covered by Medicare. A.R. at 19. Next, the ALJ approved the use of the statistical sampling. Id. The ALJ noted that HCFA Ruling 86-1 provides for the use of statistical sampling to project overpayments when claims are voluminous and reflect a pattern of erroneous billing or overutilization and when a case-by-case review is not administratively feasible. A.R. at 19.
As far as the statistical sampling methodology, the ALJ noted that the carrier first identified a universe of 9131 claims from which a computer program generated a sample selection by use of a random start of 20 and an interval of 36. A.R. at 20. A sample size of 250 was obtained using the five procedure codes. Id. Two codes were later removed but the universe and initial random sample were not changed. Id. The ALJ concluded that the carrier utilized a sufficiently large, objective sample that was fair to the provider and the government. Id.
With respect to the issue of whether Dr. Webb was liable for the overpayment, the ALJ reviewed the applicable principles and concluded:
Dr. Webb has submitted bills which he had reason to know were not correctly coded and he has billed for services which he should have known were not medically necessary. The Administrative Law Judge finds that Dr. Webb cannot be found to be without fault in creation of the overpayment.A.R. at 21. In view of this finding, the ALJ concluded Dr. Webb was liable for the overpayment. A.R. at 23. Dr. Webb requested review of the ALJ's decision. By letter dated September 25, 1998, the Medicare Appeals Council denied his request. A.R. at 5. Accordingly, the ALJ's decision is considered the final decision of the Secretary.
"Judicial review of the Secretary's decision is governed by the Administrative Procedure Act (APA), 5 U.S.C. § 706(2)(A). Shalala v. St. Paul-Ramsey Medical Center, 50 F.3d 522 (8th Cir. 1995). Under the APA, the Secretary's decision shall be set aside if it is arbitrary, capricious, an abuse of discretion, unsupported by substantial evidence, or contrary to law. Federal court review is de novo, id. at 527, but is limited to the administrative record." Hennepin County Medical Center v. Shalala, 81 F.3d 743, 748 (8th Cir. 1996) ( 42 U.S.C. § 1395oo(f), incorporates the standard of review of the APA).
Dr. Webb argues first that the ALJ misinterpreted the February 28, 1990, policy memorandum to physicians, clinical laboratories, and radiation therapy centers. Dr. Webb's position is that the memorandum does not specify an exclusive test panel composition. He contends the newsletter does not require the provider to furnish all of the specifically enumerated tests in order to bill using the corresponding panel code. Instead, he argues the newsletter merely prohibits separate billing for the individually enumerated tests if a claim has also been submitted using the panel code.
If the panel was meant to be exclusive, Dr. Webb submits there would be no purpose in submitting a claim under those codes for a grouping of other tests yet he points out the newsletter specifically directs suppliers to identify alternative groupings of tests when billing under these codes. Dr. Webb contends this language is irreconcilable with the carrier's position that alternative groupings of tests may only be covered under the multi-channel test codes (80002-80019).
Dr. Webb submits his interpretation of the newsletter is consistent with the AMA's description of the relevant organ panel codes during the applicable period. He points out the 1992 CPT Manual provided as follows with respect to the organ panel codes at issue:
The lack of an expanded list of laboratory tests under each number is deliberate. Because no two laboratories utilize the same array of tests in a particular panel, each laboratory should establish its own profile and accompany each reported panel by a listing of the components of that panel performed by the laboratory.
Dr. Webb also points out the AMA in 1993, a change which took effect after the services at issue in this appeal, modified the organ panel codes to specify the precise composition of the test panel. He states the AMA made this quite clear by stating: "This panel must include the following."
Under Chevron U.S.A., Inc. v. Natural Resources Defense Council 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), Dr. Webb contends BCBSA's construction of the policy memorandum cannot be squared with the language of the memorandum and therefore cannot be upheld. Alternatively, even if the court concludes the carrier's interpretation is a permissible construction, Dr. Webb argues the extreme ambiguity of the newsletter provides a basis on which to apply the waiver of liability rules.
The Secretary contends the record clearly shows Dr. Webb improperly billed Medicare for diagnostic laboratory tests he performed on Medicare beneficiaries. With respect to Dr. Webb's application of Chevron to the Secretary's interpretation of the newsletter, the Secretary argues the Chevron analysis is simply inapplicable when the situation being evaluated in not an agency's construction of its governing statute but instead the agency's construction of its rules and regulations. In this latter instance, the Secretary argues substantial deference must be given to an agency's construction of its rules and regulations. In so arguing, the Secretary relies on Thomas Jefferson University v. Shalala, 512 U.S. 504, 114 S.Ct. 2381, 129 L.Ed.2d 405 (1994).
In any event, it is argued that the language of the 1990 policy memorandum plainly supports the Secretary's decision here. Quite apart from the policy memorandum, the Secretary argues the CPT procedures themselves demonstrate that Dr. Webb's billing practices were incorrect. The Secretary also stresses the fact that the CPT states "[t]he tests listed within each panel identify the defined components of that panel." A.R. at 17.
In this case, Dr. Webb is challenging the Secretary's interpretation and application of the February 1990 newsletter. "When, like in this case, the issue is whether the agency has erred in interpreting its own regulations, the Supreme Court has stated that: provided the agency's interpretation `does not violate the Constitution or a federal statute, it must be given controlling weight unless it is plainly erroneous or inconsistent with the regulation.'" Shalala v. St. Paul-Ramsey Medical Center, 50 F.3d 522, 527-28 (8th Cir. 1995) ( citing Stinson v. United States, 508 U.S. 36, 44, 113 S.Ct. 1913, 1919, 123 L.Ed.2d 598 (1993)). The court noted that "as a general rule, where the agency is interpreting its own regulations or it own legislative rules, that interpretation is entitled to substantial deference." St. Paul-Ramsey, 50 F.3d at 528 ( citing Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 510, 114 S.Ct. 2381, 2386, 129 L.Ed.2d 405 (1994)) (agency's interpretation of its own regulation must be given substantial deference); Brewster on Behalf of Keller v. Sullivan, 972 F.2d 898, 901 (8th Cir. 1992) ("as a general rule an agency's interpretation of its own regulation is entitled to great deference"). See also, Friends of Boundary Waters Wilderness, 164 F.3d 1115, 1121 (8th Cir. 1999); HealthEast Bethesda Lutheran Hospital and Rehabilitation Center v. Shalala, 164 F.3d 415, 417 (8th Cir. 1998).
In this case the Secretary is not applying or interpreting a statute or a regulation. Rather, the Secretary is applying a policy or position it has taken with respect to reimbursement for particular tests under Medicare. However, we need not dwell on the issue of whether Chevron or Thomas Jefferson or some other standard applies because we believe the Secretary's position is clearly supported by the plain language of the newsletter.
We find Secretary's interpretation of the February 1990 newsletter to be plainly supported by the language of the newsletter. Certainly we cannot say the Secretary's interpretation is plainly erroneous or inconsistent with the memorandum.
The newsletter clearly informed providers that given panel CPT codes were deemed to include the identified individual tests. Since payment for the panel was based on the cost of the individual tests included therein, the carrier needed to be informed if the grouping of tests comprising the panel was altered. To this end, providers were advised of the need to include on the claim form any changes in the test groupings.
Dr. Webb concedes the memorandum clearly advised providers that they could not bill separately for tests that were identified in the memorandum as being included in a specific panel test. However, he argues the memorandum did not intend to dictate the exclusive panel components. In support, he cites the language in the memorandum which advises providers that if the panel component tests are altered the claim form must so specify.
Even if the memorandum is so interpreted, the flaw in Dr. Webb's reasoning is he did not follow that method. He submitted claims under the panel codes without indicating that the panel composition was other than that indicated in the memorandum. The memorandum clearly advised providers that if the panel components were altered they must indicate the actual panel make-up on the claims submitted.
The method used by Dr. Webb in submitting claims for panels structured differently than the structure specified in the February 28, 1990, newsletter resulted in Dr. Webb being overpaid, or as the ALJ put it, in his being paid for tests he did not perform. We also agree with the ALJ's conclusion that Dr. Webb was not without fault in this matter. Certainly, we cannot say the ALJ's conclusion was arbitrary, capricious, an abuse or discretion, or otherwise not in accordance with the law. Thus, Dr. Webb is not entitled to a waiver of liability on the overpayment.
Second, Dr. Webb argues the statistical analysis used to compute the overpayments was invalid. With a universe of 9131 claims and a selected sample size of only 250 claims, Dr. Webb points out the sampling constituted only 2.74% of the total claims being audited. Dr. Webb objects to the use of the sampling technique and extrapolation of cases to arrive at a huge assessment. He argues such a process deprives him of his federally protected rights to have each individual case examined and his due process rights with regard to assessment and protest of the assessment.
Dr. Webb concedes there are cases, e.g., Chaves County Home Health Service, Inc. v. Sullivan, 931 F.2d 914 (D.C. Cir. 1991), that have held that a sample and extrapolation therefrom are permissible when statistical validity is established; however, he contends he produced expert testimony that the statistical analysis was flawed and that there was no statistical expert who testified in support of the methodology used.
He points out that during the time period in question, he was the only board certified urologist within a 100 mile radius of Mountain Home, making his urologic drawing area very large both in terms of geography and in the number of patients. Due to this extreme skew in his patient population, Dr. Webb contends the statistical survey used in the review was invalid.
With respect to Dr. Webb's arguments regarding the sampling methodology, the Secretary points out that Dr. Webb's expert's opinion was refuted by the fact that the methodology was developed by the U.S. Department of Commerce and refined for the specific purposes of the Medicare Program by the Department of Health and Human Services Inspector General's Office, Division of Program Audit. Further, the Secretary points out HCFA Ruling 86-1 provides for the use of statistical sampling to project overpayments. That experts at these agencies had sanctioned the sampling methodology is sufficient, in the Secretary's view, to reject the opinion of Dr. Webb's expert.
We reject Dr. Webb's arguments that the statistical analysis was flawed. The use of sampling and extrapolation as part of audits to determine overpayments to parties who receive publicly-funded reimbursements has been approved by courts in a number of different settings. It has been noted that "[p]rojection 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. Moreover, mathematical and statistical methods are well recognized as reliable and acceptable evidence in determining adjudicative facts." State of Georgia Dept. of Human Resources v. Califano, 446 F. Supp. 404, 407 (N.D.Ga. 1977). In the Califano case, the court approved the use of statistical sampling in a case involving the Department of Health, Education and Welfare's refusal to reimburse Georgia for $1.5 million the state paid doctors who provided services to Georgia Medicaid recipients.
In the case of Illinois Physicians Union v. Miller, 675 F.2d 151 (7th Cir. 1982), the state audited a sample of 353 records randomly selected from a total of 1,302 records for the audit period and determined a participating doctor had been overpaid $5,018. The doctor contended any formula for sampling and extrapolation was improper per se. Id. at 155. The Seventh Circuit disagreed. It concluded that "the use of sampling and extrapolation is proper provided there is an opportunity to rebut the initial determination of overpayment." Id. at 156. See also Michigan Dept. of Educ. v. United States Dept. of Educ., 875 F.2d 1196 (6th Cir. 1989) (approving audit conducted through random, stratified sampling, and noting that audits of thousands of cases comprising the universe of cases would be impossible).
The use of such sampling procedures was also approved by the court in Ratanasen v. State of California Department of Health Services, 11 F.3d 1467 (9th Cir. 1993) In that case, an audit had been conducted by selecting a sample of 30.0 Medi-Cal beneficiaries out of a total of 8,761 beneficiaries for whom Surabhan Ratanasen, a doctor, had submitted claims. Id at 1469. The doctor failed to appeal the audit findings but later in a bankruptcy proceeding raised the issue of the validity of the sampling methods in connection with an objection to the allowance of a claim.
The court stated:
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. To deny public agencies the use of statistical and mathematical audit methods would be to deny them an effective means of detecting abuses in the use of public funds. Public officials are responsible for overseeing the expenditure of our increasingly scarce public resources and we must give them appropriate tools to carry out that charge.Id. at 1471. See also Yorktown Medical Lab., Inc. v. Perales, 948 F.2d 84 (2d Cir. 1991) (Medicaid provider claimed statistical sampling method violated its due process rights); Mile High Therapy Centers, Inc. v. Bowen, 735 F. Supp. 984 (Colo. 1988) (statistical sampling method is one way the Secretary exercises statutory discretion and authority to maintain the integrity of the Medicare payment system).
In this case, the audit involved Dr. Webb's actual billing records from which the statistical extrapolations of overpayment were made. While Dr. Webb in his brief fails to state why he believes the statistical sampling is flawed, he does mention, without a single citation to the record, that his expert testified the sampling was flawed.
The record does contain a transcript of the testimony presented to the ALJ. At that hearing, Dr. Webb presented the testimony of Robert C. Walls, a statistician and teacher at the University of Arkansas for Medical Sciences, College of Medicine. A.R. at 11276.
Walls testified he had two concerns with the sample. A.R. at 112-79. First, he was concerned with the size of the sample. Id. Second, he was concerned about the validity of the calculations of the overpayment and of the precision of that overpayment. A.R. at 11279-11280.
With respect to size, Walls testified that the original sample size of 250 was reduced in number as procedure codes were removed from consideration. A.R. at 11280-11286. With respect to the allowance of claims for certain codes prior to the April 1, 1990, date, Walls testified these claims were not dropped from the universe but instead considered to be an "allowed amount equal to the claim amount, so that the difference was zero and it did not affect the summation of the total disallowed on the sample." A.R. at 11286. He believed this procedure was inconsistent with the handling of the codes dropped from consideration and it appeared these claims were "left in seemingly to inflate the count of the sample size." A.R. at 11286.
He testified 188 claims were listed by the Secretary on the document showing the final overpayment calculation. A.R. at 11285. If the claims that were allowed because they were made before April 1, 1990, were dropped from the universe, Walls testified the universe would have been reduced to 176 which would have been in violation of HCFA guidelines which require a minimum sampling size of 200 claims/services. AR. at 11286-87.
Another concern Walls had with the sampling methodology was the program used by Medicare was set up to analyze stratified samples. A.R. at 11290. According to Walls, part of the basic definition of a stratified sample is that the sample in each strata is taken independently of the sample in any other strata. Id. As Medicare in this case sampled claims and not procedure codes, Walls testified its methodology ignored "in the calculation of variance" the "code variance." A.R. at 11291. Walls stated this meant "the calculation for the precision of the estimated overpayment" was wrong. A.R. at 11291.
However, Walls went on to testify it was possible the sample might have turned out better than is stated but that he didn't really know because he didn't have the data to recalculate it. A.R. at 11292. In fact, he testified he could not tell "precisely how much difference it would have made, or even in what direction in the final estimate of the overpayment and in the estimate of the precision of that overpayment, except I believe that they were calculated the wrong way, not corresponding to the way the sample was designed and executed." A.R. at 11293-11294. He therefore concluded the sampling was flawed. A.R. at 11294.
We believe the ALJ did not err in rejecting this testimony and concluding substantial evidence supported the carrier's use of sampling method. The audit was conducted on the basis of random statistical samples of claims paid during the four-year period being examined. Once the claims were randomly selected, if more than one procedure was involved with a particular claim, each individual procedure was examined and a decision was made as to that particular procedure. A.R. at 11297. This individual examination by procedure code resulted in some services being found to be covered services for a particular service date while others were found not to be covered for that date. A.R. at 11297-11298. When a particular procedure code was removed, the samples were stratified for the remaining codes. The universe and initial random sample were not changed. The results, the percentages of procedures that were allowable and the percentage of procedures that were not allowable, were then projected across the entire universe of cases resulting in the overpayment amount. A.R. at 11298.
Although Dr. Webb makes the bare assertion that the sample size was so small that its use to arrive at a huge assessment deprives him of his federally protected rights to have each individual case examined and his due process rights with regard to assessment and protest of the assessment, it is clear that Dr. Webb had the opportunity to challenge the audit method and has had access to an appeals procedure. See e.g., Ratanasen, 11 F.3d at 1472. We do not believe that there is a "statistical `floor' that auditors must exceed in order to guarantee providers due process." Ratanasen, 11 F.3d at 1472 (sample of 3.4 percent); Michigan Department of Education, 875 F.2d at 1199 (random, stratified sample of .4 percent (259 out of a universe of 66,368) used as a starting point for determining improper expenditures). So long as the audit methods used are valid and reliable, they are sufficient to satisfy due process. See Ratanasen, 11 F.3d at 1472; Michigan Dept. of Education, 875 F.2d at 1206 ("There is no case law that states how large a percentage of the entire universe must be sampled. However, when, as here, the state is given every opportunity to challenge each disallowance as well as the audit technique itself, it appears that the state has been treated as fairly as is practicable under the circumstances."). Clearly an individual audit of each Medicare claim would be impossible. Dr. Webb has had ample opportunity to challenge the Secretary's findings. We find that his due process rights were not violated.
For the reasons stated, the judgment of the Secretary is affirmed. A separate judgment will be entered in the defendant's favor.