Lee Memorial Hospital v. Sebelius

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA

)
LEE MEMORIAL HOSPITAL, et al., )
)
Plaintiffs, )
)
v. ) Civil Action No. 13-643 (RMC)
)
SYLVIA M. BURWELL, Secretary, )
U.S. Department of Health and Human )
Services, )
)
Defendant. )
)

MEMORANDUM OPINION

Plaintiffs, a group of non-profit organizations that own and operate acute care

hospitals participating in the Medicare program, 1 contend that the Department of Health and

Human Services has underpaid them for Medicare services provided during the fiscal years

ending in 2008-2010. The dispute requires a huge leap into Medicare and its regulations, but, in

essence, Plaintiffs allege that the Secretary knew that her basic approach and formulas produced

the wrong results but continued to underpay them for years, notwithstanding.

At issue here is Plaintiffs’ Motion to Compel: the Secretary has produced an

Administrative Record, which Plaintiffs complain is incomplete. The Secretary of Health and

Human Services repeatedly insists the record is more than sufficient for judicial review. For the

reasons set forth below, Plaintiffs’ motion will be granted in part and denied in part.

1
While Plaintiffs’ Memorandum in Support of its Motion to Compel the Administrative Record
[Dkt. 51] states that Plaintiffs “are thirty-four acute care hospitals participating in the Medicare
program,” id. at 1, the Third Amended Complaint [Dkt. 58] states that Plaintiffs are a group of
non-profit organizations that own and operate thirty-three acute care hospitals, id. ¶ 1.
1
 I. BACKGROUND

It is not necessary to take a reader through the underlying dispute in this case.

There are, however, a few fundamental points. Under Medicare, certain hospitals may be

reimbursed in part for their operating costs per patient. “Because different illnesses entail

varying costs of treatment, the Secretary uses diagnosis-related groups (DRGs) to ‘modif[y]’ the

average rate.’” Dist. Hosp. Partners, L.P. v. Burwell, No. 14-5061, 2015 WL 2365718, at *1

(D.C. Cir. May 19, 2015) (quoting Cape Cod Hosp. v. Sebelius, 630 F.3d 203, 205-06 (D.C. Cir.

2011)). Hospitals are paid at fixed rates determined by the Secretary based on DRG prospective

payment rates, which are intended to reflect the estimated average cost of treating a patient

whose condition falls within that DRG. See 42 U.S.C. § 1395ww(d). When patient costs

become extraordinarily high, hospitals may request an “outlier payment” in any case “where

charges, adjusted to cost, exceed . . . the sum of the applicable DRG prospective payment rate . .

. plus a fixed dollar amount determined by the Secretary.” 42 U.S.C. § 1395ww(d)(5)(A)(ii).

“[T]hree particular numbers are important” in calculating outlier payments:

“(1) the cost-to-charge ratio, (2) the fixed loss threshold, and (3) the outlier threshold.” Dist.

Hosp. Partners, 2015 WL 2365718, at *2. First, a hospital’s cost-to-charge ratio “represents a

hospital’s average markup” and “is calculated from data in its most recent cost report.” Id.

(internal quotations and citations omitted); see also Def. Mem. in Opp. to Pl. Mot. to Compel

(Def. Opp.) [Dkt. 53] at 4 (“The Secretary estimates a hospital’s costs for a case by multiplying

the hospital’s charges by a cost-to-charge ratio, which is a fraction that represents the estimated

amount that the hospital incurs in costs for every dollar that the hospital bills in charges.”). A

hospital’s cost-to-charge ratio is generally calculated specifically for that hospital based on data

contained in its prior cost reports. Def. Opp. at 4 (citing 42 C.F.R. § 412.84(i)).

2
 Second, as noted above, a hospital can request an outlier payment if its charges

exceed the sum of the DRG payment rate and a “fixed dollar amount.” 42 U.S.C.

§ 1395ww(d)(5)(A)(ii). The “fixed dollar amount” is otherwise known as the “fixed loss

threshold.” The fixed loss threshold “‘acts like an insurance deductible because the hospital is

responsible for that portion of the treatment’s excessive cost’ above the applicable DRG rate.”

Dist. Hosp. Partners, 2015 WL 2365718, at *2 (quoting Boca Raton Cmty. Hosp., Inc. v. Tenet

Health Care Corp., 582 F.3d 1227, 1229 (11th Cir. 2009)); see also Def. Opp. at 5 (“The fixed

loss threshold essentially represents the loss that a hospital must absorb before it is eligible to

receive an outlier payment.”). The fixed loss threshold is set annually in advance of each fiscal

year based on projections about aggregate payments to hospitals and a consideration of past

charges. Def. Opp. at 5. The Secretary determines the figure in part by looking at historical data

on charges actually submitted by hospitals and then applies an inflation adjustment factor to the

data to produce an approximation of what hospital charges might look like in the future. Id. The

Department of Health and Human Services (HHS) attempts to set the fixed loss threshold at a

level such that total outlier payments for the upcoming year will represent 5.1% of projected total

DRG payments. Pl. Mem. in Support of Mot. to Compel Complete Admin. Record (Pl. Mem.)

[Dkt. 51] at 2 (citing 72 Fed. Reg. 47,130 at 47,419).

The third relevant number—the outlier threshold—is the sum of the fixed loss

threshold and the DRG rate. Dist. Hosp. Partners, 2015 WL 2365718, at *2. “Any cost-adjusted

charges imposed above the outlier threshold are eligible for reimbursement under the outlier

payment provision.” Id. (citing 42 U.S.C. § 1395ww(d)(5)(A)(ii)). 2

2
An example is instructive: “Assume that the Secretary sets the fixed loss threshold at $10,000.
Assume also that a hospital treats a Medicare patient for a broken bone and that the DRG rate for
the treatment is $3,000. The Medicare patient required unusually extensive treatment which
3
 These figures are set by HHS each fiscal year. In this case, Plaintiffs challenge

HHS administrative regulations governing outlier payments and the fixed loss thresholds,

asserting that those regulations led to an incorrect determination of their outlier payment

amounts for 2008-2011. Specifically, they allege that HHS improperly applied two sets of

regulations: (1) “Payment Regulations,” which establish a model for determining whether

individual hospital cases qualify for outlier payments; and (2) “Threshold Regulations,” which

set the annual fixed loss threshold. Pl. Mem. at 5. They also take issue with HHS amendments

to the rules governing outlier payments made in 2003, maintaining that those amendments form

the basis for how the fixed loss thresholds were set in 2008-2011. Id.

In their Motion to Compel, Plaintiffs argue that HHS failed to produce

information used by the agency in determining the fixed loss threshold. They seek the following

materials: (1) the draft Interim Final Rule from the 2003 amendments to the payment regulations;

(2) the formulas used to calculate the fixed loss thresholds; (3) data used to calculate a cost-to-

charge adjustment factor and an inflation factor, which were then used to calculate the fixed loss

thresholds; (4) the formulas and data that HHS used to calculate estimated outlier payments,

made during previous fiscal years, which HHS considered in determining the fixed loss

thresholds for the relevant years; (5) the supporting data which HHS used to determine certain

key assumptions for projected outlier payment calculations as set forth in HHS’s Impact Files;

(6) materials supporting HHS’s regulatory impact analysis considered in each of the fixed loss

threshold regulations; and (7) materials supporting HHS’s statements in the fixed loss threshold

caused the hospital to impose $23,000 in cost-adjusted charges. If no other statutory factor is
triggered . . . the hospital is eligible for an outlier payment of $8,000, which is 80% of the
difference between its cost-adjusted charges ($23,000) and the outlier threshold ($13,000).”
Dist. Hosp. Partners, 2015 WL 2365718, at *3.
4
regulations that it would not consider the mandatory reconciliation of outlier payments in setting

the fixed loss thresholds. 3 Pl. Mem. at 3. The parties conferred extensively but were unable to

resolve this dispute. HHS contends that the materials sought were properly excluded from the

Administrative Record and avers that “HHS has provided certified administrative records of the

rulemaking proceedings for regulations concerning the establishment of fixed loss thresholds for

2008 through 2011 which contain: the agency’s proposed rule and final rule; the comments

received; and the data that the agency considered in developing the outlier payment amount and

rule and the fixed-loss threshold.” Def. Opp. at 8 (citing Ex. A, Decl. of Ing-Jye Cheng,

Director, Division of Acute Care, Hospital and Ambulatory Care Group, Centers for Medicare

and Medicaid Services (CMS), HHS [Dkt. 53-1] (Cheng Decl.) ¶¶ 2–3).

II. LEGAL STANDARDS

A. Jurisdiction and Venue

This Court has jurisdiction to review Plaintiffs’ challenge to the agency

regulations under the Medicare Act, which incorporates the Administrative Procedure Act

(APA). See 42 U.S.C. § 1395oo(f)(1); 5 U.S.C. § 706. Venue is proper under 42 U.S.C.

§ 1395oo(f) and 28 U.S.C. § 1391(c).

3
Plaintiffs originally also sought the administrative record for the 2003 amendments to the
payment regulations as well as the Medicare Claims Processing Manual in effect for FYs 2008-
2011. The Secretary has now agreed to supplement the record with those materials, with the
exception of the Interim Rule and the Impact File for the 2003 rulemaking. See Def. Opp. at 9,
11. Defendants also note that they have supplemented the record with MedPAR data for FYs
2006 and 2007, which had been inadvertently omitted from the rulemaking records for FYs 2008
and 2009, respectively. Id. Plaintiffs also requested a comment to the FY 2009 fixed loss
threshold regulations, but subsequently withdrew their request upon receipt of HHS’s
representation that it did not consider the comment for outlier policy purposes. See Pl. Reply in
Support of Motion to Compel Administrative Record (Pl. Reply) at 21.
5
 B. Standard of Review for Supplementation of Administrative Record

The Administrative Procedure Act requires reviewing courts to “set aside agency

action, findings, and conclusions found to be . . . arbitrary, capricious, abuse of discretion, or

otherwise not in accordance with law.” 5 U.S.C. § 706. In reviewing agency rulemakings, the

APA requires courts to “review the whole record or those parts of it cited by a party.” Id. “If a

court is to review an agency’s action fairly, it should have before it neither more nor less

information than did the agency when it made its decision.” Walter O. Boswell Mem’l Hosp. v.

Heckler, 749 F.2d 788, 792 (D.C. Cir. 1984); see also Citizens to Preserve Overton Park v.

Volpe, 401 U.S. 402, 420 (1971), abrogated on other grounds by Califano v. Sanders, 430 U.S.

99 (1977) (APA requires courts to review “the full administrative record that was before the

Secretary at the time he made his decision”).

“The ‘whole’ administrative record . . . consists of all documents and materials

directly or indirectly considered by agency decision-makers and includes evidence contrary to

the agency’s position.” Stainback v. Sec’y of Navy, 520 F. Supp. 2d 181, 185 (D.D.C. 2007)

(internal quotation omitted); see also Banner Health v. Sebelius, 945 F. Supp. 2d 1, 15 (D.D.C.

2013) (“Courts in this Circuit have interpreted the whole record to include all documents and

materials that the agency directly or indirectly considered . . . [and nothing] more nor less.”)

(internal quotations omitted). The record must include “all materials that might have influenced

the agency’s decision, [ ] not merely those on which the agency relied in its final decision.”

Stainback, 520 F. Supp. 2d at 186 (internal quotation omitted). An “agency may not skew the

record by excluding unfavorable information but must produce the full record that was before the

agency at the time the decision was made.” Blue Ocean Inst. v. Gutierrez, 503 F. Supp. 2d 366,

369 (D.D.C. 2007). “[A]n agency may exclude arguably relevant information that is not

6
contained in the agency’s files but that may be available from third parties” and “generally may

exclude material that reflects internal deliberations.” Fund for Animals v. Williams, 391 F. Supp.

2d 191, 197 (D.D.C. 2005).

“Although an agency may not unilaterally determine what constitutes the

administrative record, the agency enjoys a presumption that it properly designated the

administrative record absent clear evidence to the contrary.” Id.; Pac. Shores Subdivision, Cal.

Water Dist. v. U.S. Army Corps of Eng’rs, 448 F. Supp. 2d 1, 5 (D.D.C. 2006) (“[A]n agency is

entitled to a strong presumption of regularity that it properly designated the administrative

record.”). Accordingly, “[s]upplementation of the administrative record is the exception, not the

rule.” Pac. Shores, 448 F. Supp. 2d at 5.

The D.C. Circuit has held that supplementation of the record is only permitted in

one of three “unusual circumstances”: “‘(1) the agency deliberately or negligently excluded

documents that may have been adverse to its decision; (2) the district court needed to supplement

the record with background information in order to determine whether the agency considered all

of the relevant factors; or (3) the agency failed to explain administrative action so as to frustrate

judicial review.’” Dist. Hosp. Partners, 2015 WL 2365718, at *7 (quoting American Wildlands

v. Kempthorne, 530 F.3d 991, 1002 (D.C. Cir. 2008)); see also City of Dania Beach v. FAA, 628

F.3d 581, 590 (D.C. Cir. 2010). “To rebut the presumption of regularity, the party seeking

supplementation must introduce ‘concrete evidence’ to ‘prove’ that the specific documents

allegedly missing from the record were ‘before the actual decisionmakers’ involved in the

challenged agency action.” Banner Health, 945 F. Supp. at 16-17 (quoting Pac. Shores, 448 F.

Supp. 2d at 6). In making this showing, the party seeking to supplement the record “must

identify the materials allegedly omitted from the record with sufficient specificity, as opposed to

7
merely proffering broad categories of documents and data that are ‘likely’ to exist as a result of

other documents that are included in the administrative record.” Id. at 17. A district court’s

refusal to supplement the administrative record is reviewed for abuse of discretion. Kempthorne,

530 F.3d at 1002.

III. ANALYSIS

Plaintiffs seek to compel several types of information from HHS, arguing that

the materials should be supplemented to the Administrative Record because they meet one or

more of the Kempthorne criteria.

A. 2003 Draft Interim Rule

In 2003, HHS initiated a rulemaking for Medicare payment regulations in order to

more accurately compensate hospitals for their costs exceeding the fixed loss threshold. In

February 2003, then-HHS Secretary Tommy H. Thompson executed a draft Interim Rule and

sent it to the Office of Management and Budget (OMB) for its review. The Interim Rule

recognized that a small group of hospitals had gamed the system by rapidly inflating charges,

making it appear that they had incurred greater costs, so that they would obtain greater outlier

payments. As a result, HHS set falsely high fixed loss thresholds, thereby causing insufficient

payments to be made to other hospitals that had not inflated their charges but that still provided

care that was more expensive than the set Medicare rate. 4 According to Plaintiffs, the Interim

Rule concluded that HHS should immediately lower the 2003 fixed loss threshold. However, the

Interim Rule was never implemented and the Rule that was ultimately proposed did not lower the

threshold or include any of the analysis or data that had underscored the Interim Rule.

4
The higher the fixed loss threshold, the more expensive patient care must be to qualify for an
outlier payment.
8
 HHS has failed repeatedly to include the Interim Rule in the Administrative

Record on challenges to outlier payments. See Banner Health, 945 F. Supp. 2d at 24-26. Just as

in Banner Health, Plaintiffs here correctly argue that “this document goes to the heart of

establishing the Secretary’s promulgation of and continued application of invalid Fixed Loss

Threshold Regulations as arbitrary and capricious, because it demonstrates that the agency knew

that lowering the threshold would correct the problems engendered by its earlier regulations and

believed it was obligated to do so immediately, but did not.” Id. at 26. Moreover, Plaintiffs

provide evidence that the Interim Rule was considered by the agency in the rulemaking process,

as it was signed by Secretary Thompson and both the Interim Rule and the proposed rule bear the

same Regulatory Identification Number and are mostly identical in content. See Pl. Mem. at 23-

24. Accordingly, the Hospitals have met their burden of showing that agency decisionmakers

considered the material at issue and that the agency failed to include in the record documents that

may have been adverse to its decision. See City of Dania Beach, 628 F.3d at 590.

HHS argues that inter-agency documents should be granted the same

“predecisional” and deliberative status as internal agency documents. Def. Opp. at 11-16. While

it is settled law that “materials reflecting an agency’s internal deliberations should not be part of

an administrative record unless there is a strong showing of bad faith or improper behavior,” see

id. at 13 (citing San Luis Obispo Mothers for Peace v. U.S. Nuclear Regulatory Comm’n, 789

F.2d 26, 44 (D.C. Cir. 1986)), that argument carries no weight with respect to proposed drafts of

agency rules that are submitted to OMB and then publicly posted on OMB’s website. See

Executive Order (E.O.) 12,866, 58 Fed. Reg. 51,735 (Sept. 30, 1993) (requiring that after a

regulation becomes final, OMB make available to the public all documents exchanged between it

and agency during the inter-agency review). The deliberative process privilege is intended to

9
“ensure open communication between subordinates and superiors, prevent premature disclosure

of policies before final adoption, and to avoid public confusion if grounds for policies that were

not part of the final adopted agency policy happened to be exposed to the public.” Ctr. for

Medicare Advocacy, Inc. v. Dep’t of Health and Human Servs., 577 F. Supp. 2d 221, 234

(D.D.C. 2008) (citations omitted). A document is properly withheld if “disclosure of [the]

materials would expose an agency’s decisionmaking process in such a way as to discourage

candid discussion within the agency and thereby undermine the agency’s ability to perform its

functions.” Formaldehyde Inst. v. Dep’t of Health and Human Servs., 889 F.2d 1118, 1122

(D.C. Cir. 1989) (internal quotation omitted). “[A] document protected pursuant to the

deliberative process privilege loses protection if the agency used the document in its dealing with

the public.” Banner Health, 945 F. Supp. 2d at 22 (internal quotation omitted).

Here, neither HHS nor Secretary Thompson would have anticipated that the

Department’s Interim Final Rule would remain confidential. Indeed, protecting the document as

privileged would not further the purpose behind the rule because the draft would be seen by the

public. This Court concurs with Banner Health that there is no way that “inclusion of the

Interim Final Rule from the administrative record would in any way ‘expose’ HHS’s

decisionmaking process in a manner that would discourage candid discussion within the agency

or otherwise contravene the purpose of the deliberative process privilege.” Health v. Sebelius,

Civil Case No. 10-1638 (CKK), 2013 U.S. Dist. LEXIS 147713 (D.D.C. July 30, 2013); see also

Dist. Hosp. Partners, 2015 WL 2365718, at *7 n.3 (affirming district court’s decision to

supplement record with the OMB draft outlier correction rule). Accordingly, the draft Interim

Final Rule must be produced as part of the Administrative Record.

10
 B. Impact File for 2003 Rulemaking

Plaintiffs also argue that HHS improperly excluded the Impact File for 2003

Rulemaking. Pl. Reply at 7. Impact Files are Microsoft Excel spreadsheets that contain “‘data

used to estimate payments under Medicare’s hospital inpatient prospective payment systems for

operating and capital-related costs,’ including the outlier fixed loss thresholds.” Id. at 16

(quoting 72 Fed. Reg. 24,680 at 24,828 (AR at 150)). HHS has included in the record the Impact

Files for FY 2008-2011 and acknowledges that they, along with other data, “are the bases for

HHS’s determination of the fixed loss thresholds.” Def. Mem. at 16 (citing Cheng Decl. ¶¶ 2–3,

18) (“These are the materials that the Secretary has determined properly comprise the

administrative record for each of the fixed loss threshold rulemakings at issue, and they are

sufficient for meaningful judicial review.”). Indeed, Impact Files are a fundamental part of the

HHS process as they contain the data used to estimate outlier payments and fixed loss thresholds.

Cheng Decl. ¶¶ 2, 11. Accordingly, now that HHS has also now supplemented the record with

the 2003 rulemaking, see infra n.3, so too should it provide the 2003 Impact File as Plaintiffs

have made a non-speculative showing that the Secretary considered such information. Further,

the Court finds that the material is necessary “to determine whether the agency considered all the

relevant factors” in taking the challenged actions, see City of Dania Beach, 628 F.3d at 590,

based on Plaintiffs’ contention that the analysis therein “served both as justification for the

wholesale revision to the Payment Regulations and for HHS’s decision not to lower the

turbocharged, hyper-inflated fixed loss threshold.” Pl. Reply at 12 (“For example, HHS

estimated that implementing the 2003 Rulemaking would ‘reduce outlier payments for the

remainder [2 months] of FY 2003 by $150 million.’” (quoting 68 Fed. Reg. 34,494, 34,514 (June

9, 2003) (AR 3217))). See also Banner Health, 945 F. Supp. 2d at 32-33 (ordering

11
supplementation of record with Impact File relating to 2003 amendments), aff’d on

reconsideration, 2013 U.S. Dist. LEXIS 147713 at *28-29.

C. Raw Data, Formulas, Other Materials Underlying Fixed Loss Thresholds and
Outlier Payments

1. Formulas

HHS describes its process for determining fixed loss thresholds as follows:

HHS (i) simulate[s] payments to hospitals at different possible fixed
loss threshold amounts by applying the particular fiscal year’s rates
and policies to actual MedPAR files from two years prior, then
inflating the charges on the MedPAR claims by two years, and
estimating costs based on cost-to-charge ratios, and then
(ii) select[s] the fixed loss threshold at which projected total outlier
payments would equal 5.1 percent of total DRG payments.

Def. Opp. at 19. MedPAR files are “data files, which contain records, by FY, relating to claims

paid on each Medicare beneficiary inpatient hospital encounter; they establish the universe of

hospital inpatient claims which HHS uses to model claims for the upcoming FY.” Pl. Mem. at 8-

9.

Plaintiffs allege that HHS failed to file the relevant formulas that it “necessarily”

relied on in setting the fixed loss thresholds and outlier payments. Pl. Mem. at 26-27. They

argue that HHS “superficially” described its process, id. at 26, and that without the formulas

“neither the hospitals nor the Court can test . . . whether HHS considered all relevant factors.”

Pl. Reply at 14.

HHS insists that the information provided in the Administrative Record is

sufficient for judicial review of its annual Fixed Loss Thresholds without supplementation of the

record. It relies on the “presumption of agency regularity and the general rule against

supplementation,” see Banner Health, 945 F. Supp. 2d at 29–30, and other decisions in this

Court refusing to order the production of underlying raw data and formulas. Moreover, HHS

12
asserts, Plaintiffs’ motion should be denied in this respect because they have failed to identify the

specific documents they want added to the Administrative Record.

While affording HHS the presumption of regularity, it is also true that “[o]f

course the Hospitals cannot identify, by file name, the specific documents containing the

formulas, because only HHS knows those file names.” Pl. Reply at 13. However, Plaintiffs here

have met their burden of showing an “unusual circumstance” justifying supplementation of the

record. Kempthorne, 530 F.3d at 1002. Plaintiffs explain:

[Data from HHS’s published rulemakings] shows a consistent trend
of HHS assuming positive (upward) inflation in hospital costs
during all the FYs at issue, but a consistent downward trend (in all
FYs but one) of deflation in the fixed loss threshold. The fact that
the FLT shrunk each year means that some unknown factor was
applied during HHS’s described step two (“select[ing] the fixed loss
threshold at which projected total outlier payments would equal 5.1
percent of total DRG payments”), to counter the effect of the charge
inflation factor. Without any explanation as to how HHS projected
increased estimated charges and costs but nevertheless lowered the
fixed loss threshold, HHS’s path in setting the threshold remains
hidden.

Pl. Reply at 14 (quoting Def. Opp. at 19). Moreover, Plaintiffs point to “concrete proof,” noting

that in setting the formulas and running “simulations,” “HHS used computer algorithms

embedded in one or more software applications.” Pl. Reply at 13 (citing 72 Fed. Reg. 47,417

(AR 1145) (FY 2008)).

Because Plaintiffs contest the annual fixed loss thresholds as arbitrary and

capricious, the Court agrees that the Administrative Record must include formulas used to derive

those numbers, if such formulas exist. As indicated above, the basic contention here is that HHS

knew or had reason to know that its annual fixed loss threshold calculations were inaccurate,

thereby depriving Plaintiffs (and other hospitals) of legitimate outlier payments, but HHS

repeatedly failed to make necessary adjustments. Without these calculations, this dispute cannot

13
be decided. Plaintiffs have provided a reasonably specific showing that the Agency relied on

such formulas in making decisions about the fixed loss thresholds and outlier payments which

are directly challenged in this case and have thus met their burden to support supplementation of

the record because review of the formulas is necessary for the Court “to determine whether the

agency considered all the relevant factors” in taking the challenged actions. City of Dania

Beach, 628 F.3d at 590.

2. Data Trims

In setting fixed loss thresholds, HHS engages in a process known as “trimming,”

which “‘refers to the practice of disregarding data records that are invalid or otherwise may

unduly distort the analysis.’” Def. Opp. at 19 (quoting Cheng Decl. ¶ 19). HHS argues that

Plaintiffs are not entitled to supplementation of such materials because “[d]ata trims do not

modify data files or produce new data, and instead exclude certain data points from analysis, e.g.,

Medicare Advantage managed care enrollee data and critical access hospital (‘CAH’) data,

without expunging them from the data set.” Id. at 19-20.

Plaintiffs recognize that the data trims do not modify or produce new data. They

argue that HHS “misapprehended” their supplementation request because they do not seek “new

data,” but instead seek “the systematic instructions which HHS applied to ‘exclude certain data

points from analysis.’” Pl. Reply at 13 (quoting Def. Opp. at 19). While HHS provided two

examples of certain excluded data—“Medicare Advantage managed care enrollee data and

critical access hospital data,” Def. Opp. at 19—Plaintiffs request “a complete list of such

exclusionary principles” that HHS used in setting fixed loss thresholds. Pl. Reply at 13.

Plaintiffs also point out that HHS, in responding to a commenter who discovered a discrepancy

14
when attempting to simulate outlier payments for FY 2008, stated that the discrepancy could be

because the commenter used different data trims than those used by HHS. Pl. Mem. at 27.

Plaintiffs have thus pointed to “concrete evidence” that HHS relied upon various

“exclusionary principles” in determining fixed loss thresholds. Id. Understanding how HHS

applied its methodology is necessary for the Court to weigh the merits of Plaintiffs’ case and to

understand whether HHS considered all the relevant factors in making its decisions. 5

Accordingly, HHS will be required to supplement the record with all instructions applied by

HHS to exclude data from its analysis in conducting data trims.

3. Formulas and Data used to Calculate Estimated Outlier Payments

Plaintiffs further argue that HHS failed to include formulas and data used to

calculate the estimated outlier payments made during previous fiscal years. Pl. Mem. at 28.

Specifically, Plaintiffs quote an HHS statement explaining that, in estimating FY 2006 outlier

payments, the Agency originally relied on “simulations using the FY 2005 MedPAR file,” that

is, the Agency applied FY 2006 rates and policies to available FY 2005 bills. Id. (quoting 72

Fed. Reg. at 47,420 (AR at 1148)). However, HHS subsequently updated its estimate using FY

5
The Circuit’s decision in District Hospitals Partners, 2015 WL 2365718, is not to the contrary.
There, plaintiff hospitals sought inclusion of “trimmed” versions of MedPAR data files that were
already included in the administrative record. Dist. Hosp. Partners, L. P. v. Sebelius, 971 F.
Supp. 2d 15, 25 (D.D.C. 2013). Plaintiffs argued that the produced “untrimmed” files included
data that the Secretary drew from to calculate the thresholds at issue, but were not the actual data
subsets that the Secretary used in her calculations. Id. The district court found that plaintiffs
“fail[ed] to provide any non-speculative basis for a finding that distinct, smaller ‘trimmed’
versions of MedPAR files in fact exist.” Id. at 26. The Circuit affirmed, holding that “the
process of ‘trimming’ involved neither the modification of the [data] files currently in the
administrative record, nor the creation of new [data] files not in the record.” Dist. Hosp.
Partners, 2015 WL 2365718, at *8 (internal quotations omitted). The Circuit also held that the
trimmed files were “neither background information nor material that is needed because the
agency failed to explain itself” and there was “no showing that the exclusion of the trimmed files
was done in bad faith.” Id. Here, however, as set forth above, Plaintiffs do not seek actual data,
but rather request a full list of what data points were excluded from the analysis.
15
2006 bills. Id. Plaintiffs contend that “[d]espite the fact that HHS expressly relied on data and

formulas (‘simulations’) to estimate past payments, the agency has refused to provide any of

these data and formulas with the administrative records.” Id. Plaintiffs also maintain that HHS

relies on estimated prior overpayments when calculating the annual fixed loss threshold. Id. at

29.

HHS argues that its notices of final rulemaking adequately explain that in

calculating outlier payments, it applied the past fiscal year rates and policies to the bill

information in that year’s MedPAR files. Def. Opp. at 21. With respect to the formulas used to

calculate estimated outlier payments, the Court cannot adequately determine if HHS applied the

relevant factors in determining past outlier payments without knowing how it came up with those

calculations. Plaintiffs point to the fact that HHS relied on specific documents to make such

estimates. Pl. Reply at 16 (“HHS stated that its estimates were based on ‘simulations’ (i.e.,

formulas) using the FYs 2005-2006 MedPAR files and FYs 2006-07 ‘rates and policies.’ In the

same rulemaking, HHS also described using ‘the [Provider Specific File] for this final rule’ in

combination with the FY 2006 MedPAR file to estimate FY 2007 outlier payments.”) (internal

citations omitted). This concrete data demonstrates HHS’s consideration of such formulas and

such information is needed to determine whether HHS considered relevant factors in taking the

challenged actions. Therefore, just as the formulas underlying the calculation of fixed loss

thresholds, HHS must supplement the record with formulas or documentation of HHS’s

calculations of estimated outlier payments in previous fiscal years.

However, as for the data underlying those calculations, “requiring an agency to

produce source data upon source data so that its analysis can be replicated in minute detail would

appear, in most instances, to exceed the bounds of arbitrary and capricious review.” Banner

16
Health, 945 F. Supp. 2d at 28. Indeed “[P]laintiffs’ bare desire to replicate each calculation

contained within the Secretary’s analysis – without more – will not suffice to justify

supplementation, as [t]here is no general requirement that the [Secretary] include in the record

the data underlying each factor considered in its decision.” Dist. Hosp. Partners, L.P. v.

Sebelius, 971 F. Supp. 2d 15, 24 (D.D.C. 2013) (internal quotation omitted). Plaintiffs here have

not shown that raw data underlying the calculations and formulas are necessary for judicial

review. Thus, the request for data used in calculated estimated outlier payments will be denied.

D. Data Used to Calculate Annual Cost-to-Charge Adjustment Factors

Plaintiffs also seek data used to calculate a cost-to-charge adjustment factor (used

to determine the fixed loss threshold) that was implemented in FY 2007 to account for a

consistent decline in cost-to-charge ratios that happened after each projection of the threshold

and payment during upcoming fiscal years. Pl. Mem. at 30. Plaintiffs contend that the

adjustment factor is small as compared to the actual rate of decline in ratios and that they need

the underlying data to understand HHS’s methodology. Id. at 30-31. Plaintiffs quote HHS

explaining how it calculated the FY 2008 cost-to-charge adjustment factor:

For FY 2008, we calculated the [cost-to-charge] adjustment [factor]
by using the operating cost per discharge increase in combination
with the final market basket increase determined by Global Insight,
Inc., as well as the charge inflation factor described above to
estimate the adjustment to the [cost-to-charge ratios]. We
determined the operating [cost-to-charge] adjustment by taking the
percentage increase in the operating costs per discharge from FY
2004 to FY 2005 (1.0564) from the cost report and dividing it by the
final market basket increase from FY 2005 (1.043). We repeated
this calculation for 2 prior years to determine the 3 year average of
the rate of adjusted change in costs between the market basket rate
of increase and the increase in cost per case from the cost report (FY
2002 to FY 2003 percentage increase of operating costs per
discharge of 1.0715 divided by FY 2003 final market basket
increase of 1.041, FY 2003 to FY 2004 percentage increase of
operating costs per discharge of 1.0617 divided by FY 2004 final

17
 market basket increase of 1.04). For FY 2008, we averaged the
differentials calculated for FY 2003, FY 2004, and FY 2005 which
resulted in a mean ratio of 1.0210. We multiplied the 3 year average
of 1.0210 by the 2006 market basket percentage increase of 1.0430,
which resulted in an operating cost inflation factor of 6.49 percent
or 1.0649. We then divided the operating cost inflation factor by the
1 year average change in charges (1.062) and applied an adjustment
factor of 1.0027 to the operating CCRs from the PSF.

Pl. Mem. at 31 (quoting 72 Fed. Reg. at 47,418 (AR at 1146)). Plaintiffs argue they are entitled

to the data from contractor Global Insight, Inc., as well as several years of market basket and cost

report data.

HHS argues that the quoted explanation above clearly delineates how it

determines the cost-to-charge adjustment factor. The rulemaking records for the fiscal loss

thresholds include the final market basket increase and the costs per discharge based on cost

reports, which were used to adjust the cost-to-charge ratios. Def. Opp. at 23 (citing Cheng Decl.

¶ 20). What Plaintiffs seek, HHS asserts, is underlying source data that is unnecessary for

judicial review. Further, HHS argues, there is no evidence that unspecified data were before

HHS decisionmakers: the Federal Register indicates only that HHS considered the final market

basket increase and cost reports in setting the fixed loss thresholds.

HHS is correct. While Plaintiffs argue that HHS considered Global Insight data,

HHS’s explanation clearly states that it only relied upon the “final market basket increase” and

relevant cost reports; those have already been provided to Plaintiffs in the Administrative

Record. Def. Opp. at 23. Ms. Cheng further confirms that “[f]or purposes of outlier policy,

[HHS] does not review raw data used in setting market baskets” but instead “applies the final

adjustment factors to adjust the CCRs”—“the raw data used in deriving the market basket

inflation factor is not considered in setting the outlier threshold.” Cheng Decl. ¶ 20. Thus,

Plaintiffs fail to overcome the presumption of regularity afforded to HHS in this respect.

18
Plaintiffs do not present evidence sufficient to show that such documents “were ‘before the

actual decisionmakers’ involved in the challenged agency action.” Banner Health, 945 F. Supp.

2d at 17 (quoting Pac. Shores, 448 F. Supp. 2d at 6). Moreover, courts generally do “not need to

examine the raw data in order to determine whether or not the [Secretary’s] decision was

arbitrary and capricious or otherwise not in accordance with law.” Dist. Hosp. Partners, 971 F.

Supp. 2d at 24 (internal quotation omitted). HHS cogently explained how it calculates cost-to-

charge adjustment factors and raw underlying source data are not needed to determine whether

the agency considered the relevant factors. See Kempthorne, 550 F.3d at 1002; see also Dist.

Hosp. Partners, 2015 WL 2365718, at *7 (affirming district court’s refusal to supplement record

with source data used to approximate cost-to-charge ratios because Secretary explained how

ratios were calculated, data were not “critical background information,” and there was no

evidence that data were deliberately or negligently excluded from record). Therefore, the Court

will deny Plaintiffs’ request to supplement the Administrative Record with underlying data ssed

to calculate cost-to-charge adjustment factors.

E. Data Used to Calculate Inflation Factors

Plaintiffs also seek to supplement the Administrative Record with all MedPAR

data relied upon by HHS in calculating annual inflation factors used for setting the fixed loss

thresholds. According to HHS, in calculating the proposed FY 2008 outlier threshold, it

“simulated payments by applying FY 2008 rates and policies using cases from the FY 2006

MedPAR files” thereby “inflat[ing] the charges on the MedPAR claims by 2 years, from FY

2006 to FY 2008.” Pl. Mem. at 32 (quoting 72 Fed. Reg. at 47,417 (AR at 1145)). HHS

described its process of inflating the charges in the final FY 2008 stating that “[u]sing the most

recent data available, [HHS] calculated the 1 year average annualized rate of change in charges

19
per case from the first quarter of FY 2006 in combination with the second quarter of FY 2006

(October 1, 2005 through March 31, 2006) to the first quarter of FY 2007 in combination with

the second quarter of FY 2007 (October 1, 2006 through March 31, 2007).” Id. (quoting 72 Fed.

Reg. at 47,418 (AR at 1146)).

Plaintiffs contend that HHS has refused to produce MedPAR data that was used to

calculate “the 1 year average annualized rate of change in charges per case,” as described above,

and that HHS should be required to supplement the record with such data as it was used to

calculate annual inflation factors. Id. HHS responds that it has included in rulemaking records

for the fixed loss thresholds the MedPAR data for each of the fiscal years between FY 2006 and

FY 2011, which is the data used by the agency. Pl. Opp. at 24 (citing Cheng Decl. ¶ 21). HHS

further explains “the MedPAR data that [HHS] used for the charge inflation calculation is from

an early update of MedPAR that is not publicly available, and that the MedPAR data that is used

in the final rules (which is the MedPAR data that can be made available publicly for limited uses

and is included in the rulemaking records produced to Plaintiffs) could be used to closely

approximate the inflation factor that [HHS] calculated.” Id. Plaintiffs counter that they are

entitled to the data that was before the agency.

Again, this Court agrees with District Hospitals, 971 F. Supp. 2d at 25, and

Banner Health, 945 F. Supp. 2d at 29, and concludes that Plaintiffs are not entitled to raw data

underlying HHS’s calculations, despite the fact that HHS may have used that data as part of its

rulemaking process. Plaintiffs broadly claim that the “data established a key variable (the

inflation factor) that was used twice in the setting the fixed loss thresholds,” Pl. Reply at 20, but

they provide no further explanation as to why the raw data is necessary for this Court’s review of

the outlier threshold determination. See Dist. Hosp. Partners, 971 F. Supp. 2d at 25. Because

20
“there is no general requirement that the agency include in the record the data underlying each

factor considered in its decision,” id. (internal quotation omitted), the Court will deny Plaintiffs’

request to supplement the record with data used to calculate inflation factors.

F. Data Underlying Cost-to-Charge Ratios in Impact Files

While the Administrative Record contains the Impact Files for the fiscal years at

issue here, it does not include the “underlying assumptions and associated data used to compute

the conclusory data contained in the Impact Files.” Pl. Mem. at 34. According to Plaintiffs, the

cost-to-charge ratios in the Impact Files were drawn from data in the “Provider Specific Files,”

which are available to the public on the Centers for Medicare and Medicaid Services (CMS)

website, but there are material discrepancies between the cost-to-charge ratios set forth in the

Impact Files and those in the Provider Specific Files. Id.

HHS argues that there is no evidence that the Provider Specific Files requested by

Plaintiffs, or any other data from which the impact files were derived, were relied on by HHS

decisionmakers. To the contrary, Ms. Cheng’s Declaration states that the Provider Specific Files

were used to derive cost-to-charge ratios contained in the impact files. Cheng Decl. ¶¶ 12-13.

However, Plaintiffs fail to establish how such underlying source data is necessary to assist the

Court in determining whether the Agency considered all the relevant factors. Plaintiffs contend

that the material discrepancies between the cost-to-charge ratios in the Impact Files and the

Provider Specific Files constitute “unusual circumstances” warranting supplementation of the

record. The Court disagrees. Ms. Cheng explains that Provider Specific Files are updated

quarterly and may also be subject to data trims. Cheng Decl. ¶¶ 10, 15, 19. As the Court in

Banner Health explained on reconsideration, in denying Plaintiffs’ request to supplement the

record with source data underlying the Impact Files:

21
 the Provider Specific File data on the CMS website is updated (and
may be retroactively corrected) by fiscal intermediaries and
therefore cannot be relied upon to mirror the data that was used to
generate the Impact Files. Accordingly, because any alleged
inconsistences between the Provider Specific File data on the CMS
website and the Impact Files do not undermine the Secretary’s
account, as stated in the Federal Register, of how the Impact Files
were created, the ostensibly “unusual circumstances” on which the
Court relied are non-existent.

Health v. Sebelius, Civil Case No. 10-1638 (CKK) 2013 U.S. Dist. LEXIS 147713, at *35

(D.D.C. July 30, 2013) (emphasis added). Here too, the Court will deny Plaintiffs’ motion to

compel with respect to this data.

G. Regulatory Impact Analyses

Plaintiffs also request that the record be supplemented with the regulatory impact

analyses for each of the rulemakings at issue here, as well as all underlying data. For every

major rule, agencies must assess all costs and benefits of available regulatory alternatives and, if

regulation is necessary, select regulatory approaches that maximize net benefits (including

potential economic, environmental, public health and safety effects, distributive impacts, and

equity), after which they must prepare a regulatory impact analysis (RIA) detailing their

findings. Pl. Mem. at 38. Plaintiffs cite to a statement in the RIA section of the FY 2008

Threshold Regulation stating that FY 2007 outlier payments were lower than projected, which

they argue shows that HHS directly considered the RIA in implementing outlier payment

regulations. Id. at 39.

HHS explains that it already “included with each fixed loss threshold notice of

final rule a robust RIA as an appendix to the notice” and argues that what Plaintiffs really seek is

underlying source data that is “beyond the scope of what the Court needs for meaningful judicial

review.” Def. Opp. at 29-30. The Court agrees. There is no suggestion that data underlying each

22
RIA would help the Court determine whether HHS acted arbitrarily and capriciously, and thus

Plaintiffs’ motion to compel with respect to such data will be denied.

H. Documents Relating to Reconciliation

When HHS implemented the 2003 amendment to the outlier payment regulations,

it also required the reconciliation of outlier payments made to providers upon the settlement of

cost reports. This process was intended to have the excess outlier payments of those hospitals

that had engaged in “turbocharging” reconciled and recouped, with interest, by HHS. Pl. Mem.

at 12. HHS ultimately stated that it would not be “‘making any adjustments [to its fixed loss

threshold] for the possibility that hospitals’ CCRs and outlier payments may be reconciled upon

cost report settlement.’” Id. at 40 (quoting 72 Fed. Reg. at 24,837 (AR at 159)). HHS explained

that it believed that

due to the policy implemented in the outlier final rule (68 FR 34494,
June 9, 2003), CCRs will no longer fluctuate significantly and,
therefore, few hospitals will actually have these ratios reconciled
upon cost report settlement. In addition, it is difficult to predict the
specific hospitals that will have CCRs and outlier payments
reconciled in any given year. We also noted that reconciliation
occurs because hospitals’ actual CCRs for the cost reporting period
are different than the interim CCRs used to calculate outlier
payments when a bill is processed. Our simulations assume that
CCRs accurately measure hospital costs based on information
available to us at the time we set the outlier threshold. For these
reasons, we are not making any assumptions about the effects of
reconciliation on the outlier threshold calculation.

Id. at 13 (quoting 72 Fed. Reg. at 47,419 (AR at 1147)).

Plaintiffs maintain that HHS has failed to file any documents underlying its

assertion that few hospitals had their CCR ratios reconciled after cost settlement. This position,

Plaintiffs contend, is contradicted by a report from the HHS Office of Inspector General (OIG)

identifying $664 million in outlier payments made in FYs 2004-2009 which were not reconciled

23
in accordance with HHS regulations. Pl. Mem. at 41. Plaintiffs seek either documents adverse

to HHS’s original statement or documents showing why it refused to account for the impact of

reconciliation when setting the fixed loss thresholds at issue here.

HHS argues that the above-quoted language explains why HHS did not factor the

effects of cost-to-charge reconciliation into its calculation of the fixed loss thresholds:

as a result of the policy implemented through the outlier payment
regulation [in 2003], (i) HHS expected that “cost-to-charge ratios
[would] no longer fluctuate significantly,” (ii) consequently it
expected that “few hospitals [would] actually have these ratios
reconciled upon cost report settlement,” (iii) predicting the specific
hospitals that would undergo reconciliation in any given year would
be difficult, and (iv) the rationale for reconciliation (which is based
on the time interval between interim cost-to-charge ratios and actual
cost-to-charge ratios) did not apply to the fixed loss thresholds
because HHS’s simulations assumed accurate measures of hospital
costs.

Def. Opp. at 31 (quoting 72 Fed. Reg. at 47,419 (Admin. R. 1147)).

Plaintiffs have not met their burden of demonstrating why supplementation is

warranted with respect to their vague request for missing documents regarding reconciliation—

they have pointed to no specific documents and fail to “identify the materials allegedly omitted

from the record with sufficient specificity, as opposed to merely proffering broad categories of

documents and data that are ‘likely’ to exist as a result of other documents that are included in

the administrative record.” Banner Health, 945 F. Supp. 2d at 17. HHS has explained that its

expectation that “few hospitals” would have their CCR ratios reconciled upon cost report

settlement is based on the implementation and effect of its amended policy for outlier payments.

Def. Opp. at 31 (quoting 72 Fed. Reg. at 47,419). Whether HHS’s decision may be deemed

unreasonable in light of the OIG report is a question to be addressed upon the Court’s review of

the merits.

24
 IV. CONCLUSION

For the reasons set forth above, Plaintiffs’ Motion to Compel [Dkt. 51] will be

granted in part and denied in part. No later than July 2, 2015, HHS will be required to

supplement the record with the following materials: (1) the draft Interim Final Rule from the

2003 amendments to the payment regulations; (2) the Impact File for the 2003 Rulemaking;

(3) the formulas used to calculate the fixed loss thresholds; (4) all instructions applied by HHS to

exclude data from its analysis in conducting data trims; and (5) the formulas used to calculate

estimated outlier payments, made during previous FYs, which HHS considered in determining

the fixed loss thresholds for the relevant years.

A memorializing Order accompanies this Memorandum Opinion.

Date: June 11, 2015 /s/
ROSEMARY M. COLLYER
United States District Judge

25