CHAVES COUNTY HOME HEALTH SERVICES v. Sullivan, 732 F. Supp. 188 (D.D.C. 1990)

Jerry A. Sternbach and Barbara E. Straub, Pyles & Sutter, Washington, D.C., for plaintiffs.

Henry Goldberg and Donald F. Dickey, Office of General Counsel, Dept. of Health and Human Services, Washington, D.C., for defendant.

JACKSON, District Judge.

Plaintiffs are home health care providers who render medical and related services to Medicare-eligible patients pursuant to agreements with defendant U.S. Department of Health and Human Services ("HHS"). HHS then pays them the reasonable cost of such of those services as are covered by the Medicare Act, 42 U.S.C. *189 §§ 1395 et seq. ("the Act").^[1]

Plaintiffs' claims for their services are processed and paid initially by a fiscal intermediary (usually an insurance carrier), but plaintiffs are (as are all such providers) liable to the government for reimbursement of payments later found by the intermediary or HHS on post-payment review to have been made for non-covered services. See, generally, Mount Sinai Hospital v. Weinberger, 517 F.2d 329, 338 (5th Cir.), reh'g denied en banc with opinion, 522 F.2d 179 (5th Cir. 1975) (holding that provider may look to patient as source of payment), cert. denied, 425 U.S. 935, 96 S. Ct. 1665, 48 L. Ed. 2d 176 (1976).

In this case, HHS made redeterminations with respect to 1,261 claims submitted by and paid to plaintiff Chaves County Home Health Services, Inc., ("Chaves County"), 2,460 paid claims of plaintiff Albuquerque Visiting Nurse Services, Inc., ("Albuquerque VNS") and 10,791 paid claims submitted by plaintiff Bayonne Visiting Nurse Association, Inc. ("Bayonne VNA"). It did so, however, relying entirely upon a review of ostensibly representative samples of 200 claims in the case of Chaves County, 200 claims in the Albuquerque VNS case, and 320 claims in the Bayonne VNA case. Having determined that a certain proportion of each sample reflected payments for non-covered services, HHS demanded repayment of $46,900 from Chaves County, $138,100 from Albuquerque VNS, and $1.5 million from Bayonne VNA. When the repayments were not forthcoming, HHS withheld reimbursement to plaintiffs on subsequent claims by way of offset.

Plaintiffs allege that the use of the statistical sampling method to calculate amounts of overpayment, by which a fraction of the claims submitted by a particular provider is reviewed, and the result projected to the universe of the whole without individual review of each claim, is illegal. They contend that the method operates to deny them due process under the Fifth Amendment; that it is contrary to the terms of the Medicare Act; and that it implements a governmental policy which is not only arbitrary and capricious, but was adopted without the notice-and-comment formalities made obligatory in the case of certain agency rule-makings by the Administrative Procedure Act, 5 U.S.C. § 553 ("APA").

HHS responds by defending the legality and the mathematical validity of the sampling method, insisting that the logistical impossibility of affording an individual review to every Medicare claim submitted justifies its use as a matter of manifest necessity. It also contends that the Court is without jurisdiction to entertain the case, because the plaintiffs have not fully exhausted all possibilities of administrative relief within HHS.^[2]

The case is now before the Court on defendant's motion to dismiss or for summary judgment and plaintiffs' cross-motion for summary judgment. Although the case has been fullyindeed, fulsomelybriefed and argued for many months, the Court has delayed decision in hopes that further guidance might be forthcoming from the U.S. Supreme Court or a circuit court of appeals. No such guidance has materialized, however, and it is necessary that the case be decided now on the basis of such scant authority as is extant.

The clear majority of those few courts having confronted statistical sampling in analogous contexts, while acknowledging its potential for unfairness in the abstract in particular cases, have nevertheless approved its use, primarily as a logistical imperative but also upon the hypothesis that any arbitrariness evens out in the long *190 run.^[3] In what appears to be the only federal appellate case on the subject, the Seventh Circuit upheld statistical sampling in Illinois Physicians Union v. Miller, 675 F.2d 151 (7th Cir.1982), as a means to audit physicians receiving reimbursement for medical services provided under the Medicaid program. The court first considered whether the results of its use were arbitrary:

We find nothing in this procedure, however, to suggest that extrapolation inherently works to the detriment of the physician. Sampling and extrapolation is no more likely to result in a situation where the physician will be required to return monies when there has been no overpayment than in a situation where the Department will not recoup the full amount overpaid. The audit procedures are not arbitrary, capricious or invidiously discriminatory.

Id. at 156. The court then went on to consider the physicians' due process argument:

There is no merit to [appellant's] contention that the Department procedures do not comport with due process. The process due varies with the circumstances and various factors must be considered when evaluation administrative procedures. These factors are: (1) the private interest affected by the official action; (2) the risk of an erroneous deprivation of that interest; and (3) the governmental interest, including the function involved and the fiscal and administrative burdens that other procedures would entail.

We agree with [appellant] that she had a substantial interest in receiving her full statutorily allotted compensation for services actually rendered. However, in balancing the interests of the parties, the balance is heavily weighed in favor of the Department. The Department processes an enormous number of claims and must adopt realistic and practical auditing procedures. We agree with the district court's conclusion that, in view of the enormous logistical problems of Medicaid enforcement, statistical sampling is the only feasible method available.

Id. at 157 (citations omitted).

Several district courts have accepted statistical sampling as an appropriate auditing technique for settling accounts between the government and a private sector care provider under both Medicaid and other social welfare programs. In Georgia v. Califano, 446 F. Supp. 404 (N.D.Ga.1977), a Medicaid case, the court held 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." Id. at 409. It also pointed out that "mathematical and statistical methods are well recognized as reliable and acceptable evidence in determining adjudicative facts." Id. See also Rosado v. Wyman, 322 F. Supp. 1173 (E.D. N.Y.), aff'd, 437 F.2d 619 (2d Cir.1970), aff'd, 402 U.S. 991, 91 S. Ct. 2169, 29 L. Ed. 2d 157 (1971); New Jersey Welfare Rights Organization v. Cahill, 349 F. Supp. 501 (D.N.J.1972) (AFDC), aff'd, 483 F.2d 723 (3d Cir.1973).

Both the Illinois Physicians Union and the State of Georgia cases were relied upon by yet another district court in United States v. DeCosmo, Civil No. 82-631, slip op. (M.D.Fla. Feb. 17, 1984), in a Medicare case. The court found that sampling is "the only practical method" of auditing claims "since a physician may submit literally thousands of claims during the calendar year," id. at 10, and concluded that statistical sampling may properly be used to determine the existence and amount of overpayments under the Act. Id. at 11.

The most recent case, arising in a Medicare Part B context, appears to be Mile *191 High Therapy Centers, Inc. v. Bowen, No. 86-F-1853, slip op., ___ F.Supp. ___ (D.Colo. May 27, 1988), which held that the use of statistical sampling to determine overpayments of Medicare reimbursement, in circumstances where case-by-case review is not administratively feasible, is within the statutory authority of HHS. The court said:

In adopting this method, the agency relied on its authority to use any reasonable means to recover overpayments.... 42 U.S.C. § 1395g(a) authorizes "necessary adjustments on account of previously made overpayments or underpayments." 42 U.S.C. § 1395u(a) gives the Secretary authority to make determinations of the rates and amounts of payments to be made for services and conduct audits to insure proper payments are made. 42 U.S.C. § 1395x(v) (1) (A) (ii) provides for retroactive corrective adjustments in payments when the cost proves inadequate or excessive.

A reasonable interpretation of the statute by the administrator of an agency is entitled to considerable weight.... The above statutory citations give the Secretary considerable discretion and authority to maintain the integrity of the Medicare payment system. The statistical sample method is one way of exercising this power. The agency's policy does not exceed their statutory authority.

Id. at 4 (citations omitted).

Plaintiffs also argue that the HHS' ruling explaining the use of the statistical sampling method (known as HCFA Ruling 86-1 of February 20, 1986) is the sort of "substantive rule" contemplated by 5 U.S.C. § 552(a) (1) (D) of the APA as requiring notice-and-comment rulemaking formalities. Defendant responds that HCFA Ruling 86-1 is merely an "interpretive rule," which is not subject to formal rulemaking requirements. See 5 U.S.C. § 553(b) (3) (A).

The Mile High Therapy Center court again agreed with HHS and held that its various directives counselling upon use of the sampling technique were exempt from the notice-and-comment requirements of the APA, because they represented only an "interpretive rule." The court said:

The Medical Carrier's Manual and the HCFA ruling explaining the statistical sample method are merely interpretations of an existing statute. They do not create new law or depart from prior practice. Agency manuals, guidelines and memoranda are interpretive rules not subject to the APA.... Defendants did not violate the APA in using the Manual as a means of calculating overpayments.

Id. at 3 (citations and footnote omitted).

Thus, each of plaintiffs' contentions here, it appears, have been addressed and rejected by at least one other court. This Court has been shown no reason to believe that those courts were any of them less wise or well-informed, and, in the aggregate, the judicial authority presently available overwhelmingly supports the proposition that statistical sampling in the readjudication of Medicare claims by HHS to determine overpayments and reimbursement liability is lawful.

It is, therefore, this 12th day of February, 1990,

ORDERED, that plaintiff's motion for summary judgment is denied; and it is

FURTHER ORDERED, that defendant's motion for summary judgment is granted, and the complaint is dismissed with prejudice.

[1] The services not covered by the Act are those "which ... are not reasonable and necessary for the diagnosis or treatment of illness or injury" or are for "custodial care." 42 U.S.C. § 1395y.

[2] The Court concludes that it has subject matter jurisdiction under 42 U.S.C. § 405(g). See Bowen v. City of New York, 476 U.S. 467, 106 S. Ct. 2022, 90 L. Ed. 2d 462 (1986). The validity vel non of statistical sampling has been pursued administratively by plaintiff Chaves County to the Appeals Council level within HHS, and the Appeals Council has declined to entertain the issue.

[3] The only case known to the Court to have disallowed a sample-based overpayment determination is Daytona Beach General Hospital, Inc. v. Weinberger, 435 F. Supp. 891 (M.D.Fla. 1977), in which the court took exception to the size of the sample used, but did not categorically reject sampling as a review technique. The court did hold, however, that the extrapolation of the sample results to the entire claims universe constituted a due process violation.