UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

BLUE CROSS AND BLUE SHIELD OF MASSACHUSETTS, INC. et al.,

Plaintiffs, Case No. 1:25-cv-693 (TNM) v.

ROBERT F. KENNEDY, Jr., Secretary, U.S. Department of Health & Human Services et al.,

Defendants.

MEMORANDUM OPINION

Every year, the Centers for Medicare and Medicaid Services (“CMS”) rate health

insurance plans on a one-to-five scale to reflect those plans’ quality of care and services. The

promise of better advertising to enrollees and more money from CMS creates a strong incentive

for plans to seek a high rating. So when Blue Cross and Blue Shield of Massachusetts Inc. and

its subsidiary, Blue Cross and Blue Shield of Massachusetts HMO Blue, Inc. (together, “Blue

Cross”) received 2025 ratings lower than expected for two of its plans, they sued.

Blue Cross claims that two parts of CMS’s rating-calculation process contravene agency

regulations and are therefore arbitrary and capricious under the Administrative Procedure Act

(“APA”). First, Blue Cross questions CMS’s adjustments to raw data that account for

demographic characteristics outside a health insurer’s control. Next, it challenges how CMS

compares one plan’s scores to national average scores. Because both processes satisfy

governing regulations, the Court disagrees with Blue Cross. It thus denies Blue Cross’s motion

for summary judgment and grants Defendants’ cross motion.

1 I.

The Court starts by summarizing CMS’s rating process and then overviews its regulatory

footing. How Blue Cross’s dispute fits in that picture follows.

A.

Title XVIII of the Social Security Act establishes the Medicare program, which provides

the elderly and disabled with health insurance from the federal government. See generally

et seq. CMS, a component of the Department of Health and Human Services

(“HHS”), runs the Medicare program.

Medicare offers four types of coverage plans, labeled Parts A–D. See

§ 1395c–1395i-

5 (Part A); id. § 1395j–1395w-4 (Part B); id. § 1395w-21–1395w-29 (Part C); id. § 1395w-101–

1395w-152 (Part D). This case concerns Part C, which details what is called the Medicare

Advantage Program. Under that program, insurers provide coverage that individuals would

otherwise receive through traditional Medicare (in Parts A and B). See id. § 1395w-22(a). Part

C providers—called Medicare Advantage Organizations (“Advantage Organizations”)—contract

with CMS and agree to offer coverage for a price lower than CMS’s “benchmark” rate (the per-

capita cost of covering traditional Medicare beneficiaries in a given geographic area). Id.

§ 1395w-23(n);

. An Advantage Organization makes that offer through a

“bid” to CMS that indicates what payment it would accept to cover a beneficiary.

. In exchange for offering a lower cost of coverage, CMS pays Advantage

Organizations back a certain amount. The lower a provider’s bid compared to CMS’s

benchmark rate, the more CMS will pay back a provider. 42 U.S.C. § 1395w-24(b)(1)(C);

. How much Advantage Organizations receive also depends in part on its “Star

Rating.” 42 U.S.C. §§ 1395w-23(o)(4), 1395w-24(b)(1)(C)(v).

2 Star Ratings require a bit of explanation. CMS assigns a Star Rating to each Advantage

Organization contract based on the “plan’s quality.” See Policy and Technical Changes to

Medicare Programs,

, 16,520 (Apr. 16, 2018) (codified in scattered sections

of 42 C.F.R.). Star Ratings appear as a number between 1 and 5 stars (in half-star increments).

(c)(3). The higher the Star Rating, the more money an Advantage

Organization receives annually from CMS. Medicare laws explain in great detail how Star

Ratings affect that amount. For instance, Advantage Organizations with 4-Star contracts qualify

for bonus payments in the form of more bidding power. See 42 U.S.C. § 1395w-23(o)(1).

When those Organizations contract with CMS for a new year, they can propose a higher bid

(which reflects a higher cost of coverage) while keeping the CMS rebate amount that lower-

rated Advantage Organizations would receive only with a lower bid. Id. As another example,

4.5-Star contracts receive back seventy percent of the gap between their bid and CMS’s

benchmark rate, while a 3.5-Star contract earns only sixty percent of that amount, and lower

rated contracts only fifty percent. 42 U.S.C. § 1395w-24(b)(1)(C)(v) (listing the “final

applicable rebate percentage[s]” by rating);

(a)(2)(ii), 423.186(a)(2)(ii)

(same).

Rebates are not the only benefit. Star Ratings affect plan enrollment too. By looking at

various providers’ Star Ratings, potential beneficiaries can more easily compare options and

choose the best provider. Indeed, the Star Rating system “is designed to provide information to

the beneficiary that is a true reflection of the plan’s quality and encompasses multiple

dimensions of high-quality care.” See Policy and Technical Changes to Medicare Programs, 83

Fed. Reg. at 16,520; 42 U.S.C. § 1395w-23(a), (o). Understandably, contracts with higher Star

Ratings prove more attractive to beneficiaries. All of this creates a strong incentive to aim for a

3 high Star Rating.

The regulations detail at length how a Star Rating comes to be. Each overall Star Rating

derives from several “measure-level” Star Ratings.

(c)(1); 422.162(a).

Measure-level ratings refer to contract-wide scores on a plan’s specific features that reflect part

of that plan’s overall quality. Over thirty criteria serve as measures, each receiving its own

rating.

(a)(4); Joint Appendix (“J.A.”) 7, ECF No. 22. Those criteria cover,

for example, patient outcomes, access to care, complaints about plans, and provider processes.

J.A. 6. 1

How does CMS gather data to calculate measure-level ratings? Data comes from

multiple sources. CMS, Advantage Organizations themselves, multiple surveys, and CMS

contractors all gather data relevant to the calculation. Ex. A. (“Abernathy Decl.”) ¶ 6, ECF No.

15-2. The data relevant to this dispute stems from surveys by the Consumer Assessment of

Healthcare Providers and Systems (“Consumer Assessment”). See

(a); J.A.

125–40 (survey data). In 2025, around a quarter of quality measures used Consumer

Assessment survey answers as their main data source. J.A. 105. These surveys ask patients to

evaluate features of their healthcare that are otherwise difficult to measure. See

(a); J.A. 115–16. For example, to measure “Ease of Getting Prescriptions” with a

plan, the survey asked enrollees: “In the last six months, how often was it easy to use your

prescription drug plan to get the medicines your doctor prescribed?” J.A. at 85. Or to measure

how quickly patients can seek care, the survey asked: “In the last 6 months, when you needed

care right away, how often did you get care as soon as you needed?” J.A. at 60.

Turning that raw survey data into contract-wide measure-level ratings, and then a final

1 All page citations refer to the page numbers that the CM/ECF system generates.

4 Star Rating involves a few more steps. Two of them prompted this dispute.

Case-mix adjustments. Before converting raw Consumer Assessment survey data into a

contract-wide measure-level rating, CMS sometimes applies what is called a “case-mix

adjustment” to raw survey answer scores. A CMS guidance document called the Technical

Notes details when CMS should do so. See e.g., J.A. 64. The case-mix adjustment lowers or

raises certain measures’ scores to take “into account the mix of enrollees.” J.A. 103. Doing so

prevents certain traits of a contract’s enrollees from excessively skewing that contract’s data.

If, for example, one Advantage Organization contract serves an especially large number of folks

aged 75–79, it may be at an unfair advantage without the case-mix adjustment. That is because

individuals in this age range tend to provide more positive ratings than their peers in the age

group below them.

Who knew? The case-mix adjustment controls for differences like these

and thus removes characteristics outside the control of health-care providers from the Star-

Rating inquiry.

National average comparison. Next, CMS transforms the case-mix adjusted Consumer

Assessment data into a contract-wide measure-level rating for each criterion through a method

called “[r]elative distribution and significance testing.”

(a)(3). This

method compares one contract’s average survey score for a measure to the national average

survey score for that measure.

Depending on how highly that contract’s score ranks

compared to the national average, the relevant measure-level rating rises or falls. Id.; J.A. 8. To

calculate the national average, CMS weights contract scores based on how many individuals a

contract covers and then takes the average of those contracts. CMS, Summary of Analyses for

Reporting, MA & PDP CAHPS 2 (Aug. 2024) https://perma.cc/E626-FQ2N; J.A. 9.

Once CMS has an Advantage Organization contract’s measure-level ratings for all the

5 relevant criteria, it assigns each a specific weight based on that characteristic’s importance and

averages those values. See

(e); J.A. 20. After rounding this raw score, a

contract’s final Star Rating is born.

(c).

B.

Three regulatory sections promulgated by HHS govern the Star Rating process.

To start,

provides (among other things) several definitions relevant

to the Medicare Rating System, including the “case-mix adjustment,” which, recall, refers to

adjustments to measure scores to prevent enrollee composition from skewing the data. Here is

how § 422.166(a) defines that term:

“Case-mix adjustment means an adjustment to the measure score made prior to the score being converted into a Star Rating to take into account certain enrollee characteristics that are not under the control of the plan. For example age, education, chronic medical conditions, and functional health status that may be related to the enrollee’s survey responses.”

This section also outlines the Consumer Assessment Surveys, or “CAHPS” as the regulations

call them. Id.

Next,

provides the operative framework for how CMS “adds,

updates, and removes measures” that it uses to assess various providers’ plans and ultimately

“calculate the Star Ratings.”

§ 422.164(a). Besides changing measures, this section also

explains that each year, “CMS lists the measures” it will use for Star Ratings “in the Technical

Notes or similar guidance document.” Id.

Last,

marches through a few adjustments to measure-level scores as

part of the Star Rating calculation process. This is where the “[r]elative distribution and

significance testing” methodology lives,

§ 422.166(a)(3), and calls for the use of a “national

average” in turning Consumer Assessment Survey answers into contract-wide measure scores

6 that eventually form a Star Rating, id. § 422.166(a)(3)(i)(B). Section 422.166 also explains how

to calculate the final Star Rating by “using the weighted mean of measure-level” ratings.

Id. § 422.166(c).

C.

This case arises from that regulatory landscape. Blue Cross is a non-profit medical

service corporation. See Compl. ¶¶ 7–9, ECF No. 1. It operates several health-care plans that

provide medical and prescription drug coverage to Medicare beneficiaries in Massachusetts

under Medicare Part C. Id.

Blue Cross Massachusetts executed two Medicare Advantage contracts—H2230 and

H2261—with CMS. J.A. 117–24. Both received 2025 final Staring Ratings of 3.5. Abernathy

Decl. ¶ 19. Unhappy with that rating, Blue Cross turned to the courts to challenge CMS’s Star

Rating calculation as unlawful and arbitrary and capricious under the Administrative Procedure

Act (“APA”). Compl. ¶ 41. It objects to two portions of CMS’s rating calculation process.

Blue Cross contends that CMS circumvented HHS’s regulations when it used (1) the case-mix

adjustment provided for in the Technical Notes and (2) a weighted national average of survey

answer scores rather than a simple average. Had CMS not done so, Blue Cross claims its Star

Ratings would have been half of a point higher. Abernathy Decl. ¶ 20. And this missing half-

star cost Blue Cross at least $35 million. Compl. ¶ 45. After Blue Cross moved for summary

judgment on this issue, Blue Cross Mot. for Summ. J., ECF No. 15, CMS cross-moved. CMS

Mot. for Summ. J., ECF No. 17. The Court considers the arbitrary-and-capricious challenge

now, taking in turn the case-mix adjustment and national average. 2

2 Technically, Blue Cross sued Robert F. Kennedy Jr., the HHS Secretary, in addition to CMS. But the Court refers to Defendants as “CMS” for simplicity’s sake. This Court has subject matter jurisdiction under

.

7 II.

The Court reviews agency actions under the APA’s standards of review. See

; Nursing Ctr. of Buckingham & Hampden, Inc. v. Shalala,

(D.C. Cir.

1993). Normally, a court will grant summary judgment when there “is no genuine dispute as to

any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P.

56(a); see also Anderson v. Liberty Lobby, Inc.,

(1986). But Rule 56’s

standards do not apply to a court’s review of a final agency action under the APA. See

Cherokee Nation v. Dep’t of Interior,

, 117 (D.D.C. 2024). In these cases,

summary judgment “serves as the mechanism for deciding, as a matter of law, whether the

agency action is supported by the administrative record and otherwise consistent with the APA

standard of review.”

at 90 (citing Richards v. INS,

, 1177 & n.28 (D.C. Cir.

1977)).

When reviewing an interpretation of an agency’s own regulations,

(2)(A)

governs and requires courts “to set aside agency action that is arbitrary, capricious, an abuse of

discretion, or otherwise not in accordance with law.” Nat’l Envtl. Dev. Ass’n’s Clear Air

Project v. EPA,

(D.C. Cir. 2014) (cleaned up). Because it “is

axiomatic . . . that an agency is bound by its own regulations,”

(cleaned up), courts

set aside agency actions as arbitrary and capricious if they flout the agency’s own regulations.

Erie Boulevard Hydropower, LP v. FERC,

(D.C. Cir. 2017). So unlike

arbitrary-and-capricious challenges focused on whether an agency adequately explained a

decision, see Seven Cnty. Infrastructure Coal. v. Eagle Cnty.,

, 1511 (2025),

even a well-reasoned agency action may contravene its own regulations and thus fail APA

review, see Panhandle E. Pipe Line Co. v. FERC,

, 1135–36 (D.C. Cir. 1979)

8 (setting aside an agency order that “ignore[d]” agency regulations despite the agency’s

justification for its order).

III.

First up, the Court considers whether the Department’s regulations permitted the use of

the case-mix adjustment as outlined by CMS’s Technical Notes in calculating Blue Cross’s

measure-level ratings. They did. Multiple aspects of the regulatory scheme show why.

Start with § 422.164, which, recall, creates the operative framework for how CMS

handles the measures it uses to assess plan quality. This section, titled “[a]dding, updating, and

removing measures,” contemplates (before anything else) that CMS will use the Technical Notes

as part of the rating calculation’s scheme.

(a). Specifically, it notes that

“CMS lists the measures used for a particular Star Rating each year in the Technical Notes or

similar guidance document with publication of the Star Ratings.”

No further detail limiting

what Technical Notes or other guidance may cover appears.

But the rest of § 422.164 provides a clue. There, one finds the methods for how CMS

adds new measures and makes “substantive” changes to old ones. Id. § 422.164(c), (d)(2). It

may do so only after an announcement and two-year waiting period. Id. Section 422.164 also

notes specific instances when CMS will remove certain measures in entirety. Id. § 422.164(e).

It then outlines how to determine the most heavily weighted (and thus most important)

measure—what reflects a contract’s improvement over multiple years. See id. §§ 422.164(f);

422.166(e)(1). The last portions overview data integrity and the CMS review process.

§ 422.164(g)–(h). Together, these topics cover general, bigger-ticket items that either require

rulemaking (significant changes to measures) or describe more clerical aspects of the program

(data accuracy and review processes).

9 Meanwhile, the Technical Notes detail frequently changing aspects of the scoring process

including the numerical value of each case-mix adjustments, and the demographics and measure

to which those adjustments apply. Compare CMS, Medicare 2023 Part C & D Star Rating

Technical Notes, 105–107, https://perma.cc/53PE-4PMJ, with CMS, Medicare 2024 Part C & D

Star Rating Technical Notes, 111–118, https://perma.cc/EB9Z-MP4A, and CMS, Medicare 2025

Part C & D Star Rating Technical Notes, 106–113, https://perma.cc/XJ8V-N63N. Nowhere in

§ 422.164’s broad mechanisms are steps that preclude or otherwise conflict with the Technical

Note’s explanation of those case-mix adjustment details. To the contrary, when put together

with its reference to the Technical Notes, § 422.164’s broad focus suggests that regulators left

certain details—details like the regularly changing “more specific identification of a measure’s .

. . case-mix adjustment”—up to those Technical Notes. See Medicare Program; Changes to the

Medicare Advantage and the Medicare Prescription Drug Benefit Program for Contract Year

2024,

, 30,636-37 (Apr. 23, 2024). In short, CMS’s use of the Technical

Notes to set out the case-mix adjustment follows § 422.164’s operative scheme.

Nearby provisions track that conclusion. Recall that

(a) defines

case-mix adjustments as any “adjustment to the measure score made prior to the score being

converted into a Star Rating.” Meanwhile, its neighbor, § 422.166(f)(2)(ii)(A), excludes

measures that have “already” been “case-mix adjusted for socioeconomic status” from a

different part of the Star Rating calculation process called the categorical adjustment index.

More, § 422.166(f)(3)(i)(A) explains how to account for measures “that are case-mix adjusted”

in calculating what’s called the health equity index, a summary of a contract’s performance

among those with certain social risk factors. Id. § 422.162(a). 3

3 To be sure, § 422.166(f)(3)(i) does not go into effect until 2027, and thus has only limited relevance here.

10 What do these provisions have in common? None dictates when CMS may or may not

apply case-mix adjustments. But all refer to case-mix adjustments in a way that assumes CMS

already uses the feature for at least some measures. The Court hesitates to conclude from this

pattern that regulations cabined the use of case-mix adjustments to specific, unwritten

circumstances. The better view favors a reading that permits CMS’s use of Technical Notes to

address those details. This is especially true given § 422.164’s authorization of extra-regulatory

Technical Notes for details like the case-mix adjustment.

Examples of conflict between guidance and regulations reinforce the absence of it here.

Consider National Environmental Development Association Clean Air Project v. Environmental

Protection Agency,

(D.C. Cir. 2014). In that case, an EPA regulation called for

“Regional Consistency” when implementing the Clean Air Act.

. Following defeat

in a Sixth Circuit lawsuit, however, the EPA issued a directive explaining that it would change

its practice only in that jurisdiction while maintaining its established practice elsewhere.

. The D.C. Circuit rejected that approach. Because the regulations annunciated “in clear

terms” the EPA’s “firm commitment to national uniformity in the application” of its rule, a

circuit-specific approach could not stand.

at 1009–10.

Panhandle Eastern Pipe Line Co. v. Federal Energy Regulatory Commission provides

another example of conflict. There, FERC adopted regulations allowing pipeline companies to

protect themselves from supplier rate increases by adjusting their rates to reflect changes in

certain gas costs. Panhandle,

. In a separate order, FERC required pipeline

companies to pass through to customers the revenue they gained from transporting other

companies’ natural gas.

at 1122–23. Under the order, they would do so in part by placing

that revenue in a “purchase gas account.”

11 A pipeline company objected to that requirement and the D.C. Circuit agreed. The

trouble for FERC was that its other regulations limited purchase gas accounts to include only

“purchase gas costs,” not transportation costs.

. Indeed, the regulations defined

“purchased gas cost” specifically. Those costs included: “the cost of wellhead purchases, field

line purchases, plant outlet purchases, transmission line purchases, and pipeline production from

leases.”

(quoting

(d)(4) n.1 (1979)). In other words, these accounts

covered “costs, not revenues,” and “purchased gas items, not transportation items,” so FERC’s

order stood in direct conflict with existing regulations.

(cleaned up).

CMS has not similarly tried “to play fast and loose with its own regulations.”

Nothing about § 422.164 calls for a specific practice that the Technical Notes’ case-mix

adjustments ignore. Cf. Clean Air Project, 752 F.3d at 1009–10. And nothing about § 422.164

sets up measures in a way that excludes case-mix adjustments from that picture. Cf. Panhandle,

613 F.2d at 1134–35. In contrast, § 422.164 points to the Technical Notes where case-mix

adjustments live and provides nothing limiting how CMS must treat raw survey data, while

§ 422.162 defines case-mix adjustments in manner that assumes CMS will use them. See supra

10. Unlike other examples, then, CMS “adhere[d] to its own rules.” Reuters Ltd. v. FCC,

, 950 (D.C. Cir. 1986).

One more note. Even if the regulatory text and structure left “uncertainty,” the Court

would defer to CMS’s reading as reasonable. Kisor v. Wilkie,

, 574–75 (2019).

This scheme’s “interpretive issue[s] arise[] in the context of a complex and highly technical

regulatory program.”

; Cmty. Care Found. v. Thompson,

(D.C. Cir.

2003) (noting the Medicare program’s “tremendous complexity”) (cleaned up). And when that’s

the case, courts’ “preference” for deferring to the agency interpretation is “strongest.”

12 Given that reality, the “character and context” of CMS’s ruling “entitles it to control weight.”

Id. at 576. 4

None of Blue Cross’s contrary arguments persuade. Blue Cross trains most of its fire on

§ 422.166(a)(3), which outlines the calculation process for measure-level ratings based on

Consumer Assessment survey data and makes no mention of the case-mix adjustment. Blue

Cross Mot. Summ. J. at 20–21, 23–24. Had the Department meant to include the case-mix

adjustment in that process, the argument goes, this section would have said so. Id. But that

view ignores the full picture. Section 422.166 covers specific adjustments CMS makes to

measure-level scores, id. § 422.166(a), and then explains how to transform those scores into the

final Star Rating, id. § 422.166(c). It leaves out how CMS should transform raw data into

measure-level score themselves, and that is where the case-mix adjustments come in.

More, adopting Blue Cross’s argument focused on § 422.166 would require the Court to

ignore the neighboring provision, § 422.164. And, again, this section contemplates that CMS

will use “the Technical Notes or similar guidance document” to fill in details. Id. The Court

must view the full picture. Cf. Davis v. Michigan Dep’t of Treasury,

(1989)

(“It is a fundamental canon of statutory construction that the words of a statute must be read in

their context and with a view to their place in the overall statutory scheme.”).

Blue Cross also questions CMS’s invocation of the 2018 rulemaking preamble

4 While the D.C. Circuit has not directly addressed whether Loper Bright Enterprises v. Raimondo,

(2024) preserves Kisor deference, it has cited Kisor’s principle since Loper Bright. See Duke Energy Progress, LLC v. FERC,

(D.C. Cir. 2024). Other circuits, too, have applied Kisor post-Loper Bright. See, e.g., United States v. Boler,

, 322 n.4 (4th Cir. 2024) (“Since Loper Bright dealt specifically with ambiguities in statutory directives to agencies and did not address the issue of agency interpretations of their own regulations, we will apply the Supreme Court's recent guidance in Kisor to address the issue before us today.”); United States v. Trumbull,

, 1118 n.2 (9th Cir. 2024) (“The Supreme Court did not call Kisor into question in Loper Bright . . . .”). The Court thus applies its framework.

13 commentary, which more extensively discussed case-mix adjustments than did the codified

regulations. See 83 Fed. Reg. at 16,537–38 (explaining how the Technical Notes include “more

specific identification of a measure’s . . . case-mix adjustment”); Blue Cross Mot. Summ. J. 24–

25. According to Blue Cross, the Court must ignore those references and look solely to the

codified regulation, which, it contends, demands the opposite result. True, preambles do not

themselves establish law when the regulation’s text demands another outcome. See Elevance

Health, Inc. v. Becerra,

, 23-24 (D.D.C. 2024) (finding a conflict where

regulatory text required the application of a score modification to certain “actual” points and the

preamble required application of the modification to different, “hypothetical” points). But here,

the regulatory text and the final rule’s commentary point in the same direction. Indeed, even

without the preamble, as the Court has explained, the regulations permit CMS’s case-mix

adjustment. The preamble confirms that view.

Last, Blue Cross points to Congress’s express reference to the case-mix adjustment in

other Medicare statutes. Blue Cross Mot. Summ. J. at 14. For instance, 42 U.S.C.

§ 1395rr(b)(12) states that “[t]he Secretary shall establish a basic case-mix adjusted prospective

payment system for dialysis services” in a statute about “End stage renal disease program.” Id.

As its second source, Blue Cross cites 42 U.S.C. § 1395fff(b)(3)(B)(iv), which allows for “case

mix changes” for the cost of certain home health services.

These uses of “case mix” do not move the needle. In general, language “in one statute

usually sheds little light upon the meaning of different language in another statute.” Russello v.

United States,

(1983). So the Court gleans little about the Medicare Advantage

statute from those unrelated to it. More, an “agency’s exercise of its regulatory authority”—the

crux of this case—is “distinct from an agency’s interpretation of a statute it administers. See

14 Reckitt Benckiser Inc. v. E.P.A.,

(D.C. Cir. 2010); Coeur Alaska, Inc. v. Se.

Alaska Cons. Counc.,

, 283–84 (2009) (distinguishing between agency

interpretations of regulations and agency interpretations of statutes). However little unrelated

statutory provisions say about the Medicare Advantage statute, then, they say even less about

calculating Star Rating, measure-levels, and measure-scores, all which derive from agency

regulations, not from congressional statutes.

Finally, note what Blue Cross did not argue. Blue Cross argues only that CMS deviated

from existing regulations in applying the case-mix adjustment. It never develops an argument

that case-mix adjustments constitute a rule that should have been promulgated through notice-

and-comment rulemaking. See Blue Cross Mot. at 24; Appalachian Power Co. v. E.P.A.,

(D.C. Cir. 2000). Because the Court does not consider arguments made only in

passing, cf. Hutchins v. District of Columbia,

(D.C. Cir. 1999), it expresses

no view on the distinct questions such an argument involves.

IV.

Next, consider whether CMS acted reasonably and in accordance with law in calculating

the national average for Consumer Assessment survey score answers. Once more, the regulatory

text confirms that it did.

Recall that when CMS relies on Consumer Assessment survey data, § 422.166(a)(3)

instructs CMS to discern a contract’s measure-level for each criterion in part by comparing that

contract’s “average” survey answer score against the “national average” survey answer score.

Contracts with average survey scores “statistically significantly higher than the national

average” survey score for a measure have a chance at the highest rating for that measure.

(a)(3)(v)(B).

15 CMS operates within those parameters. Its guidance tracks § 422.166(a)(3) by requiring

CMS to consider the “statistical significance of the difference of the contract mean from the

national mean” in its calculations. J.A. 108. To do that, CMS takes each contract’s average

score for a certain measure, weights that score according to the number of beneficiaries enrolled

in that contract, and then averages those scores. CMS, Summary of Analyses for Reporting.

This sequence makes sense. The size of Advantage Organizations’ contract enrollment varies

substantially. Contracts with as few as 600 survey-eligible enrollees must participate in

Consumer Assessment surveys, see CMS, Medicare CAHPS Fact Sheet (Mar. 2025),

https://perma.cc/Y89X-X7SV, alongside contracts with over 140,000 enrollees, see CMS

Monthly Enrollment by Plan (Sept. 2025), https://perma.cc/HU9P-G56P. Weighting contracts

by enrollment size prevents outlier contracts’ averages from skewing results and instead

produces one that better reflects a national trend.

Blue Cross’s alternative confuses matters. As it reads § 422.166, CMS must add up

every contracts’ measure-level score for a criterion and then divide that value by the number of

contracts without weighting for their size. Blue Cross Mot. Summ. J. at 28–29. Anything else,

it contends, strays from the plain meaning of “national average” in § 422.166. But that method,

not CMS’s, diverges from the regulatory scheme. While CMS’s guidance instructs it to compare

the “the difference of the contract mean [to] the national mean,” Blue Cross proposes that CMS

instead compare one contract’s mean to the national contract mean. J.A. 108. Courts cannot

insert words into the regulatory scheme in that fashion. Cf. Jawad v. Gates,

(D.C. Cir. 2016) (“We will not read a phrase into the statute when Congress has left it out.”).

Blue Cross approach would require that much.

CMS’s method also tracks common sense. The Consumer Assessment survey answers

16 come from enrollees across the country. One could imagine those scores’ average stemming

from the sum of all individual answers divided by the number of survey-takers from across the

country. CMS’s method follows that logic while Blue Cross’s strays from it. When CMS

weights each contract’s average score by the number of enrollees that contract covers, it evens

out how much an individual’s answers factors in. Without weighting for contract size—as Blue

Cross’s proposed method entails—answers from an enrollee who belongs to a large plan will

factor in less than answers from small-plan enrollees. To put in another way, Blue Cross’s view

is like “taking the average of all fifty states’ average heights in order to get the national average

height, rather than adjusting for each state’s population.” Elevance Health, Inc. v. Kennedy, ---

F. Supp. 3d ---,

, at *7 (N.D. Tex. 2025). Blue Cross’s arithmetic does not

add up.

V.

For all these reasons, Blue Cross’s Motion for Summary Judgment will be denied. And

CMS’s Cross-Motion for Summary Judgment will be granted. A separate order will issue today.

2025.11.03 15:34:26 -05'00' ____________________________ Dated: November 3, 2025 TREVOR N. McFADDEN, U.S.D.J.

17