United States District Court, District of Columbia
LEE MEMORIAL HEALTH SYSTEM f/b/o LEE MEMORIAL HOSPITAL, et al., Plaintiffs,
SYLVIA M. BURWELL, Secretary of the U.S. Department of Health & Human Services Defendant.
ROSEMARY M. COLLYER United States District Judge
these consolidated cases, Plaintiff hospitals challenge the
methods used by the Department of Health & Human Services
to calculate the “fixed-loss threshold, ” a term
integral to reimbursement under the Medicare program. Because
the Center for Medicare and Medicaid Services, a constituent
agency of HHS, followed the notice and comment rulemaking
process and is entitled to a highly deferential standard of
review, the Court cannot say that it has acted arbitrarily or
capriciously. The regulations will stand.
a group of non-profit organizations that own and operate
acute care hospitals participating in the Medicare program
(Hospitals),  contend that the Center for Medicare and
Medicaid Services (CMS), led by Secretary Sylvia Burwell (the
Secretary), has underpaid them for Medicare services provided
during the fiscal years ending in 2008, 2009, 2010, and 2011.
Plaintiffs challenge CMS's administration of the outlier
payment system, which pays eligible hospitals a percentage of
their costs above the typical threshold for treating a
Medicare patient. Plaintiffs challenge the “fixed loss
threshold” rulemakings promulgated in fiscal years 2008
through 2011, as well as the 2003 amendment to the outlier
before the Court are Defendant's Motion to Dismiss or, in
the alternative, for Summary Judgment, Dkt. 73, and
Plaintiffs' Motion for Summary Judgment, Dkt. 74.
is a federal program that provides health insurance to the
elderly and the disabled. See generally 42 U.S.C.
§§ 1395 et seq. Generally speaking,
hospitals provide care to Medicare beneficiaries and then
seek reimbursement from CMS.
is not a precise exercise. Instead of reimbursing the
providers dollar for dollar, CMS pays fixed rates through the
Inpatient Prospective Payment System (IPPS). Under IPPS,
inpatient services are divided into categories called
“diagnosis related groups” or “DRGs.”
See 42 U.S.C. § 1395ww(d). Each DRG merits a
standard payment rate, intended to reflect the estimated
average cost of treating the service(s) provided. See
id. Because these DRGs correspond to the given
patient's diagnosis upon discharge, the rates may vary
from the costs actually incurred by the provider.
cases, the rates may drastically understate a hospital's
costs. To compensate providers for exceptionally costly
cases, Congress established the “outlier” payment
system. See generally 42 U.S.C. §
1395ww(d)(5)(A). If the cost of health care in a given case
exceeds the DRG payment “plus a fixed dollar amount
determined by the Secretary, ” then the hospital is
eligible for an outlier payment. Id. §
1395ww(d)(5)(A)(ii). Taken together, the DRG plus the
“fixed dollar amount determined by the Secretary”
represents the “outlier threshold.” 42 U.S.C.
§ 1395ww(d)(5)(A)(ii); see also Cnty. of L.A.,
192 F.3d at 1010. If a case qualifies, the provider receives
80% of the costs that exceed the outlier threshold. 42 C.F.R.
§ 412.84(k). This 80% is called the “additional
payment” or “outlier payment.” E.g.,
Id. §§ 412.80(a)(3), (c).
phrase for present purposes is the “fixed dollar
amount, ” which is to be “determined by the
Secretary” and “specified by CMS.” 42
U.S.C. § 1395ww(d)(5)(A)(ii); 42 C.F.R. §
412.80(a)(3). The parties refer to this as the
“fixed-loss threshold” or “FLT.” The
Fixed Loss Threshold functions as an “insurance
deductible” of sorts. Boca Raton Cmty. Hosp., Inc.
v. Tenet Health Care Corp., 582 F.3d 1227, 1229 (11th
Cir. 2009). When the cost of care exceeds the predetermined
DRG payment, the provider must absorb the entire Fixed Loss
Threshold amount before it can recoup any outlier payments
from CMS. The parties' interests are thus diametrically
opposed: CMS benefits from a higher Fixed Loss Threshold and
the Hospitals benefit from a lower Fixed Loss Threshold.
the Medicare Act requires that in any fiscal year
“[t]he total amount of the [outlier] payments . . . may
not be less than 5 percent nor more than 6 percent of the
total payments projected or estimated to be made based on DRG
prospective payment rates for discharges in that year.”
42 U.S.C. § 1395ww(d)(5)(A)(iv). Thus, although the
Fixed Loss Threshold is “determined by the Secretary,
” she must set a Fixed Loss Threshold high enough to
ensure that projected outlier payments do not exceed 6% of
the projected DRG payments, but not so high that projected
outlier payments are less than 5% of the projected
Although the statute's command is unequivocal, Fixed Loss
Threshold rulemakings are predictive. Id. (requiring
outlier payments to be within 5-6 “percent of the total
payments projected or estimated to be made based on
DRG prospective payment rates for discharges”)
(emphasis added); 42 C.F.R. § 412.80(c) (“CMS will
issue threshold criteria for determining outlier payment in
the annual notice of the prospective payment rates published
in accordance with § 412.8(b).”). As a result,
there is no obvious way for CMS to guarantee that
annually prescribed rates and thresholds will yield outlier
payments that are between 5% and 6% of total DRG payments in
the next federal fiscal year. Nor must it take corrective
action if its predictions fall short. See Cnty. of
L.A., 192 F.3d at 1020. The D.C. Circuit has upheld this
practice. See Dist. Hosp. Partners, 786 F.3d at 51.
was a watershed year for the outlier-payment system. The
system had been manipulated in the late 1990s by some
hospitals which exploited certain regulatory vulnerabilities,
arising from “the time lag between the current charges
on a submitted bill and the cost-to-charge ratio taken from
the most recent settled cost report, ” which predated
current charges. Notice of Proposed Rulemaking, 68 Fed. Reg.
10, 420, 10, 423 (Mar. 5, 2003) (3/5/03 NPRM). The outlier
payment system depends on calculating “charges,
adjusted to cost, ” including overhead and
capital costs. 42 U.S.C. § 1395ww(d)(5)(A)(ii) (emphasis
added). That adjustment is made using the
“cost-to-charge ratio” (CCR) mentioned in the
Notice of Proposed Rulemaking. Because hospitals knew that
CCRs were based on past cost reports, some hospitals
increased their charges for patient care between past cost
reports and current reimbursement requests, yielding a CCR
that would “be too high” and thus
“overestimate the hospital's costs.” 3/5/2003
NPRM at 10, 423. Some 123 hospitals were found to have
increased their charges by 70 percent, while only decreasing
their CCRs by two percent. Id. at 10, 424. This
became known as “turbo-charging.” Dist. Hosp.
Partners, 786 F.3d at 51 (describing turbochargers).
The February 2003 Draft Interim Final Rule
Hospitals rely heavily on a Draft Interim Final Rule proposed
in February 2003-before the Notice of Proposed Rulemaking
cited above-and obtained by them through a Freedom of
Information Act (FOIA) request. Hosp. Mot. [Dkt. 74] at 11
(citing AR S3595-S3659) (Draft); see also Joint
Appendix, Ex. 4 [Dkt. 81-4] at 97-161 (same). The 63-page
Draft included a number of findings and proposed various
solutions. The Draft found that turbocharging caused
“nearly all of the increase in the FY 2003 threshold
from FY 2002 ($21, 025 to $33, 560).” AR S3610. It also
described the effect of turbocharging on the Fixed Loss
Threshold: “Because the fixed-loss threshold is
determined based on hospitals' historical charge data,
hospitals that have been inappropriately maximizing their
outlier payments have caused the threshold to increase
dramatically for FY 2003.” AR S3610.
prevent future turbocharging, the Draft said that CMS
“need[ed] to make revisions to [its] outlier payment
methodology, ” primarily by “updating
cost-to-charge ratios [CCRs].” 3/5/2003 NPRM at 10,
421, 10, 423. See also generally AR S3612-15. More
specifically, the Draft proposed to amend CMS's payment
regulations so that “fiscal
intermediaries”-insurance companies who examine
Medicare payment claims under contract with CMS-could
“use either the most recent settled or the most recent
tentative settled cost report, whichever is from the latest
cost reporting period.” AR S3614. But reducing the lag
time alone would not be enough, because some hospitals could
still “increase charges much faster than costs during
the time between the tentative settled cost report period and
the time when the claim is processed. . . . [T]here will
still be a lag of 1 to 2 years.” AR S3614-15. To
counter this possibility, the Draft proposed a new regulatory
provision that would allow CMS to increase a hospital's
CCR if “more recent charge data indicate that a
hospital's charges have been increasing at an excessive
rate (relative to the rate of increase among other
hospitals).” AR S3615. The hospitals could also have
requested a modified CCR if they presented substantial
evidence that the ratios were inaccurate. AR S3615.
the Draft reconsidered CMS's previous policy “that
payment determinations [were] made on the basis of the best
information available at the time a claim is processed and
[were] not revised, upward or downward, based upon updated
data.” AR S3620. Acknowledging that “some
hospitals have taken advantage of the current outlier policy,
” AR S3620, the Draft resolved to reconcile processed
payments with hospital cost reports once they were ultimately
settled. AR S3621; see also AR 3626 (“[W]e
believe the only way to eliminate the potential for such
overpayments is to provide a mechanism for final settlement
of outlier payments using actual cost-to-charge ratios from
final, settled cost reports.”). That proposal would
trigger another problem, however: “in the event of a
decline in the [CCR], some cases would no longer qualify for
any outlier payments while other cases would qualify
for lower outlier payments.” AR S3622 (emphasis added).
In other words, the reconciliation might show that an
instance of patient treatment was never eligible for an
outlier payment to begin with. And because CMS must predict
the “total amount” of outlier payments before the
fiscal year begins to comply with the 5-6% requirement,
“the only way to accurately determine the net effect of
a decrease in [CCRs] on a hospital's total outlier
payments is to assess the impact on a claim-by- claim
basis.” Id. The Draft admitted candidly that
CMS was “still assessing the procedural changes
necessary to implement this change.” Id.
proposed amendments to the outlier payment scheme would have
also made it “necessary, ” according to the
Draft, to lower the Fixed Loss Threshold. AR S3629. After
excluding the 123 offending turbochargers from the CCR pool;
applying actual CCRs (from settled cost reports) to the
hospitals that were previously assigned statewide averages;
extrapolating future CCRs from the national progression over
the previous three years; and reestimating charge inflation
without the 123 turbochargers, the Draft recommended reducing
the Fixed Loss Threshold from $33, 560 to $20, 760.
See AR S3629-33.
FY 2003 Proposed and Final Rules Amending Payment
Draft was never published. Although the Hospitals suggest
that CMS “bow[ed] to pressure from [the Office of
Management and Budget], ” Hosp. Mot. at 11, that
proposition finds no support in the record and may be
inconsequential since both agencies are in the Executive
Branch and headed by presidential appointees exercising their
discretion. Whatever the reason, the Draft was abandoned.
CMS on March 5, 2003 published a Notice of Proposed
Rulemaking, 3/5/03 NPRM, 68 Fed. Reg. 10, 420-29. The NPRM
contained the same modifications listed above to the outlier
payment scheme, but did not propose a corresponding reduction
in Fixed Loss Threshold.
comments, the Final Rule was largely unchanged. See
Final Rule, 68 Fed. Reg. 34, 494 (June 9, 2003) (Final Rule).
“Many commenters recommended that [CMS] lower the
outlier threshold.” Final Rule at 34, 505. CMS
acknowledged having “reestimated the fixed-loss
threshold reflecting the changes implemented in this final
rule that will be in effect during a portion of FY
2003.” Id. Specifically, CMS inflated charges
in the FY 2002 Medical Provider Analysis and Review (MedPAR)
by the two-year average annual rate of change in charges.
Id. Had its analysis stopped there, the Fixed Loss
Threshold would have been $42, 300. Id.
“However, after accounting for the changes
implemented in this final rule, we estimate the
threshold would be only slightly higher than the current
threshold (by approximately $600).” Id.
(emphasis added). Nonetheless, despite concluding that the
Fixed Loss Threshold should be higher, CMS found it
“appropriate not to change the FY 2003 outlier
threshold at this time” because “[c]hanging the
threshold for the remaining few months of the fiscal year
could disrupt the hospitals' budgeting plans and would be
contrary to the overall prospectivity of the [inpatient
prospective payment system].” Id. at 34, 506.
The Fixed Loss Threshold stayed at $33, 560 for the remainder
of FY 2003. Id.
The FY 2004 Regulations
time CMS set the Fixed Loss Threshold for FY 2004, the
changes to the outlier payment regulations were fully in
effect. See generally Final Rule, 68 Fed. Reg. 45,
346 (Aug. 1, 2003) (FY 2004 FLT Reg.). Extrapolating from
2002 MedPAR data, CMS applied “the 2-year average
annual rate of change in charges per case, ” as opposed
to costs per case, “to establish the FY 2004
threshold.” Id. at 45, 476. CMS then took
three steps to update the CCR:
 for each hospital, we matched charges-per-case to
costs-per-case from the most recent cost reporting year; 
we then divided each hospital's costs by its charges to
calculate the cost-to-charge ratio for each hospital; and 
we multiplied charges from each case in the FY 2002 MedPAR
(inflated to FY 2004) by this cost-to-charge ratio to
calculate the cost per case.
2004 Fixed Loss Threshold regulation also reviewed and
evaluated the reconciliation process established by the 2003
amendment to the outlier payment threshold. 68 Fed. Reg. at
45, 476-77. The novel reconciliation process had presented a
roadblock. See Id. (“Without actual experience
with the reconciliation process, it is difficult to predict
the number of hospitals that will be reconciled.”). CMS
resolved to “assess the appropriate number of hospitals
to be reconciled” once “later data bec[a]me
available.” Id. CMS did identify, however, 50
of the turbocharging hospitals as likely subjects of
reconciliation. Id. at 45, 476-77.
on all of this, CMS set an FY 2004 Fixed Loss Threshold of
$31, 000. Id. at 45, 477.
The FY 2005-2007 Fixed Loss Threshold Regulations
pattern largely repeated itself until the years challenged in
this case. See generally 69 Fed. Reg. 48, 916, 49,
276, 49, 278 (Aug. 11, 2004) (FY 2005 FLT Reg.) (lowering the
Fixed Loss Threshold to $25, 800, after initially proposing
$35, 085, in response to comments suggesting that CMS revise
its methodology); 70 Fed. Reg. 47, 278, 47, 493-94 (Aug. 12,
2005) (FY 2006 FLT Reg.) (lowering the Fixed Loss Threshold
to $23, 600, after initially proposing $26, 675, by using the
same methodology but updated data); 71 Fed. Reg. 47, 870, 48,
151 (Aug. 18, 2006) (FY 2007 FLT Reg.) (raising the Fixed
Loss Threshold to $24, 475, after initially proposing $25,
530). To sum up: the Fixed Loss Threshold was set at $25, 800
in FY 2005; $23, 600 in FY 2006; and $24, 475 in FY 2007.
these rulemakings, commenters continually complained that the
Fixed Loss Thresholds were too high, both out of
self-interest and a concern over statutory compliance by CMS.
E.g., FY 2005 FLT Reg. at 49, 276 (“Some
commenters explained that this increase to the threshold
would make it more difficult for hospitals to qualify for
outlier payments and put them at greater risk when treating
high cost cases. . . . The commenters further noted that, in
the proposed rule, [CMS] estimated total outlier payments for
FY 2004 to be 4.4 percent of all inpatient payments.”);
FY 2006 FLT Reg. at 47, 974; FY 2007 FLT Reg. at 48, 149.
cited previous years' outlier payments, which had not
fallen within the 5-6% statutory window. E.g., FY
2007 FLT Reg. at 48, 149 (“The commenters noted that
total estimated outlier payments in FY 2004 and FY 2005 were
well under the 5.1 percent target.”). CMS conceded this
as a factual matter. Id. at 48, 150 (“As the
commenters noted, the outlier thresholds we have projected in
the last several years have resulted in payments below the
5.1 percent target.”). CMS also noted that in earlier
years, payments had been significantly higher than 5.1%
because of turbocharging. Id. (“[I]n the early
years of th[e] decade, outlier payments were significantly
higher than the 5.1 percent target we projected.”).
specifically, commenters decried CMS's failure to (1)
apply an adjustment factor to the CCRs; or (2) account for
the effect of reconciliation. E.g., FY 2006 FLT Reg.
at 47, 494 (“Several commenters suggested an
alternative to the methodology we proposed”: CMS
“should adjust cost-to-charge ratios that will be used
to calculate the FY 2006 outlier threshold.”); FY 2005
FLT Reg. at 49, 277 (“One of the commenters also noted
that none of the calculations above factored in the impact of
reconciliation that would result in an even lower outlier
first point, CMS eventually relented. See FY 2007
FLT Reg. at 48, 150 (“[W]e now agree with the
commenters that it is appropriate to apply an adjustment
factor to the CCRs so that the CCRs we are using in our
simulation more closely reflect the CCRs that will be used in
FY 2007.”). CMS agreed to “apply only a one year
adjustment factor” of 99.73%. Id. The
Hospitals refer to this as a “negative 0.27%.”
Hosp. Mot. at 14.
second point, CMS held firm and did not account for the
potential effect of reconciliation when setting the outlier
threshold. FY 2007 FLT Reg. at 48, 149 (“As we did in
establishing the FY 2006 outlier threshold, in our projection
of FY 2007 outlier payments, we proposed not to make an
adjustment for the possibility that hospitals' CCRs and
outlier payments may be reconciled upon cost report
settlement.”) (citation omitted).
The FY 2008-2011 Fixed Loss Threshold Regulations
now to the Fixed Loss Threshold regulations at issue in this
case. Cf. Hosp. Mot. at 15 (“In Each of FYs
2008-2011 Here at Issue . . . .”). The Hospitals allege
generally that CMS “continued to use the flawed FLT
model that had resulted in substantial underpayment in FY
2007.” Id. With the benefit of hindsight, CMS
has reported that the total outlier payments (as a percentage
of total DRG payments) were 4.8% for FY
2008, see 74 Fed. Reg. at 44, 012 (AR
7084); 5.3% for FY 2009, see 75
Fed. Reg. 50, 042, 50, 431 (Aug. 16, 2010) (AR
10187); 4.7% for FY 2010,
see 76 Fed. Reg. 51, 476, 51, 795-96 (Aug. 18,
2011); and 4.8% for FY 2011, see 77
Fed. Reg. 53, 258, 53, 697 (Aug. 31, 2012).
2008, CMS used the same methodology as it had used for FY
2007 to calculate the outlier threshold. See 72 Fed.
Reg. 47, 130, 47, 417 (Aug. 22, 2007) (FY 2008 FLT Reg.). The
agency applied a one-year CCR adjustment factor (99.12%) to
the CCRs in the October 2006 update to the hospitals'
Provider Specific File, which is a file for each provider
that contains the unique information relevant to that
provider that is used by CMS to compute payments and
repayments for services provided. CMS also artificially
inflated (by 15.04%) the 2006 MedPAR claims by two years.
Id. The result was a proposed Fixed Loss Threshold
of $23, 015. Id. Several commenters thought that
amount was too high. See generally Id. at 47,
417-18. These commenters noted that outlier payments had been
only 4.63% of overall FY 2007 payments; faulted CMS for not
using more recent CCR data; and suggested applying the CCR
adjustment factor over different periods of time (longer or
shorter than one year). Id.
not budge. See Id. at 47, 418 (“Because we are
not making any changes to our methodology for this final rule
with comment period, for FY 2008, we are using the same
methodology we proposed to calculate the outlier
threshold.”). It did use more recent data, however,
which resulted in a lower final Fixed Loss Threshold of $22,
635. Id. at 47, 419. One commenter implored CMS to
use even more recent data. Id. at 47, 418
(“The commenter urged CMS to use the June 2007 update
[to the hospital CCRs] instead of the March 2007 update for
the final rule.”). CMS declined because the June CCR
update would not be ready until the end of July, “which
is beyond the timetable necessary for us to compute the
outlier threshold and publish this final rule with comment
period by August 1st.” Id.
process was the same for FY 2009. See 73 Fed. Reg.
48, 434, 48, 763 (Aug. 19, 2008) (FY 2009 FLT Reg.)
(“For FY 2009, we proposed to continue to use the same
methodology used for FY 2008 to calculate the outlier
threshold.”) (citation omitted). CMS again proposed a
one-year CCR adjustment factor (99.2%) to CCRs calculated
from the previous December's Provider Specific File
update. Id. Once again, the previous year's
MedPAR data were extrapolated out by two years. Id.
The result was a proposed Fixed Loss Threshold of $21, 025.
proposal found a slightly more welcoming reception than its
predecessors. Id. at 48, 764 (“The commenters
commended CMS for making refinements such as applying an
adjustment factor to CCRs when computing the outlier
threshold but noted that, because CMS is still not reaching
the 5.1 percent target, there is still room for
improvement.”). Commenters again called the CCR
adjustment calculation “unnecessarily
complicated”; again urged CMS to use more recent,
historical, and industry-wide rates of change; again asked
CMS to vary the CCR adjustment factor to more or less than
one year; and again asked CMS to apply the June Provider
Specific File update instead of the March version.
again, CMS was implacable. See generally Id.
(providing largely the same reasons as in FY 2008). Applying
the same methodology as in the proposed rule-but with more
recent data-CMS settled on a Fixed Loss Threshold of $20,
185. Id. As it had done previously, CMS refused
to make “any adjustments for the possibility that
hospitals' CCRs and outlier payments may be reconciled
upon cost report settlement.” Id. at 48, 765.
FY 2010, [CMS] proposed to continue to use the same
methodology used for FY 2009 to calculate the outlier
threshold.” 74 Fed. Reg. 43, 754, 44, 007 (Aug. 27,
2009) (FY 2010 FLT Reg.) (citation omitted). The previous
year's MedPAR files were used, and a one-year CCR
adjustment factor (98.4%) was applied to the CCRs as
contained in the previous December's Provider Specific
File update. See generally Id. at 44, 007-08. CMS
proposed a Fixed Loss Threshold of $24, 240, which
represented a 21% increase from the previous fiscal year.
Id. at 44, 008.
2010 proposed increase spurred further protest. See
generally Id. Commenters could not understand why-when
CMS had met its target in FY 2009-there should be any change.
Id. at 44, 009. Others accused CMS of purposefully
erring on the low end of the 5-6% target or below it
altogether. Id. Another asked CMS to make a mid-year
change to the Fixed Loss Threshold if it appeared that the
5-6% target would not be met. Id. Still others
repeated the requests to use June data instead of March data
in the Final Rule, to account for reconciliation.
Id. at 44, 009-10.
insisted in its response that it had “use[d] the most
recent data available to set the outlier threshold.”
Id. at 44, 009. A mid-year course correction was
rejected. Id. (citing 70 Fed. Reg. at 47, 495). All
other suggestions were denied for the same reasons as in
previous years. See generally FY 2010 FLT Reg. at
year 2011-the last at issue in this case-proved to be no
different. See 75 Fed. Reg. 50, 042, 50, 427 (FY
2011 FLT Reg.) (Aug. 16, 2010) (“For FY 2011, [CMS]
proposed to continue to use the same methodology used for FY
2009 to calculate the outlier threshold.”) (citation
omitted). CMS proposed a one-year CCR adjustment factor of
98.9% with a resulting Fixed Loss Threshold of $24, 165, a
4.5% increase from the previous year. Id. at 50,
again pointed out that the previous year had missed the mark
(outlier payments were merely 4.7% of total payments) and
reiterated the previous years' suggestions about how to
fix that. See generally Id. at 50, 428-29.
Commenters also made several discrete suggestions, addressed
in the analysis below. See infra at 38-39.
rejected each of the suggestions. FY 2011 FLT Reg. at 50, 429
(“Because we are not making any changes to our
methodology for this final rule, for FY 2011, we are using
the same methodology we proposed to calculate the outlier
threshold.”). Applying that methodology to the ...