Very Brief Blog: Hospital Charges In Real Life: $15,417 for $446

 I had an MRI in June at a nearby hospital outpatient center.  

When I log onto Medicare.gov/my/claims, I see the MRI code the hospital charged against, and the charge; $15,417.

Medicare allowed $446.50, and Medicare Part B paid 80%, leaving $89.30 to my BCBS Medigap plan.  (Separate paperwork from them shows they paid the $89.30; I owed $0.)

So the hospital charge-to-fee schedule ratio was about 30X.

##

The grammar is a little confusing.  The hospital charged $15417, and the top part of the form says Medicare approved "$15417."  However, further down, the actual payment appears, which is "Total Medicare paid the facility was $350.07" (80% of allowed.)  

Medicare shows a remaining copay of $89.30 while BCBS paperwork shows $89.30 paid (but nothing about the total amounts). 

##

In round numbers, my traditional fee-for-service Medicare costs about $200 a month for Part B and another $200 a month for BCBS Medigap—roughly $4,800 a year. Apart from the annual Part B deductible, there is generally little or nothing more to pay during the year.

A Medicare Advantage plan does not eliminate the Part B premium, but it generally eliminates my $200-a-month Medigap premium, potentially saving about $2,400 a year. It may also throw in modest extras, such as a $50 eye exam and $100 toward glasses.

Less visibly, however, Medicare Advantage can bring a festival of copays throughout the year—for specialists, imaging, outpatient procedures, emergency care, hospital stays, rehabilitation, and other services. In 2026, plans may expose members to as much as $9,250 for in-network medical care, or $13,900 in combined in- and out-of-network spending under a PPO. Few members will experience such a train wreck. But by December, an MA member’s accumulated copays can easily consume much—or all—of the $2,400 initially “saved” by giving up Medigap.  One M.A. copay average was quoted at $1400/year.

 

 

How I Use AI: What Percent of CRC Stool-test Orders Get Done? What Percent of Positives Get Colonoscopy?

Header:  Chat GPT rapidly answers general medical questions, here, slippage in colon cancer screening.

##

We had a screening colonoscopy in the family recently and I wondered:

1) Of stool tests ordered, what percent get completed (e.g. mailed back)?

2) Of positive stool tests, what percent get the indicated colonoscopy?

Below is the direct output from Chat GPT.  I'm not taking it as gospel truth, nor have I fact-checked it.  However, it's probably directonally correct (or better) and there aren't perfect or absolute answers, anyway.  

It's pretty good for, "almost free" and "takes a few seconds."

###

Completion of Stool-Based Colorectal Cancer Screening

How often is an ordered stool test completed?

FIT, United States. FIT completion varies substantially because an “order” may involve anything from handing a kit to the patient to mailing it with reminders and navigation. A reasonable practical estimate is that approximately 35%–50% of ordered or mailed FIT kits are returned in ordinary US outreach programs. In a pragmatic federally qualified health-center program, approximately 46% of ordered FITs were completed in the first year and 41% in the second. Prior behavior strongly predicted adherence: previous completers were much more likely to complete subsequent testing than previous noncompleters (Nielson et al., 2019). Programs with minimal follow-up may achieve only about 20%–35%, whereas organized mailing, reminders, and navigation can raise completion above 50%.

Cologuard/mt-sDNA, United States. Completion of an ordered Cologuard test is generally reported at about 70% within one year. Among 368,494 Medicare beneficiaries with a valid order, 44% completed the test within 30 days, 65% within 60 days, and 71% within one year (Weiser et al., 2021). A later national analysis of approximately 1.56 million first-time users similarly reported 71.3% overall completion, although completion among Medicaid beneficiaries was only 52.0% (Le et al., 2025). These findings are consistent with the likelihood that Exact Sciences’ centralized fulfillment, patient reminders, customer assistance, and tracking infrastructure improve completion relative to low-cost FIT programs. However, these are observational studies, often using Exact Sciences laboratory data, and they do not isolate the independent effect of the company’s outreach system.

Outside the United States. Organized national and regional FIT programs usually measure participation among everyone invited rather than among patients receiving a physician order. Across European programs, average FIT participation was approximately 49.5%, with a wide range of roughly 23%–71% (Senore et al., 2019). Thus, even an organized population program typically reaches only about half of eligible invitees, although the strongest programs approach 70%.

Test and setting	Approximate completion
US FIT, ordinary ordered or mailed programs	35%–50%
US FIT, low-touch or disadvantaged settings	20%–35%
US FIT, intensive organized outreach	50%–60% or higher
US Cologuard, insured populations	About 70%
US Cologuard, Medicaid	About 50%
European organized FIT invitation	About 50% average; approximately 23%–71% range

Colonoscopy After a Positive Stool Test

United States, stool tests combined. In a large study involving nearly 33,000 patients with a positive FIT, guaiac FOBT, or mt-sDNA result across 39 US healthcare organizations, 51.4% underwent colonoscopy within six months and 56.1% within one year (Mohl et al., 2023). The relatively small increase between six and twelve months suggests that the problem is not merely delayed scheduling; many patients with positive stool tests never complete the recommended diagnostic colonoscopy.

Positive Cologuard versus positive FIT. Some US studies have reported substantially higher colonoscopy follow-up after positive mt-sDNA than after positive FIT or FOBT. In one integrated-system analysis, six-month colonoscopy completion was 84.9% after positive mt-sDNA versus 42.6% after positive FIT/FOBT (Finney Rutten et al., 2020). This comparison should be interpreted cautiously because the positive mt-sDNA and FIT groups were small and unequal, and selection, insurance, clinician behavior, and manufacturer-supported communication may have contributed to the difference. Nonetheless, the findings are consistent with the hypothesis that a centralized, trackable Cologuard system produces more persistent follow-up than decentralized FIT testing.

Outside the United States. Organized FIT programs generally achieve better follow-up. An international survey of 35 screening programs found a mean 79% colonoscopy completion rate within six months, although performance varied considerably among programs (Selby et al., 2021). European screening programs have reported average compliance with colonoscopy referral of approximately 81%, with program-level rates ranging from about 64% to 92% (Senore et al., 2019).

A large Italian organized-program study included 111,423 people with positive FIT results; approximately 79% completed colonoscopy. Patients who did not undergo colonoscopy subsequently had about twice the risk of colorectal-cancer death as those who completed the diagnostic examination, although the comparison was observational and residual differences between compliers and noncompliers may remain (Zorzi et al., 2022).

Setting after positive stool test	Colonoscopy completion
United States, stool tests combined	51% by 6 months; 56% by 12 months
United States, positive FIT in routine settings	Approximately 40%–55%
United States, positive Cologuard in selected studies	Approximately 70%–85%
Organized international FIT programs	Approximately 79%–81%
Highest-performing European programs	Approximately 90% or higher

Overall Interpretation

The literature identifies two major points of attrition. First, only about four in ten to one-half of US patients complete an ordinary FIT order, whereas approximately seven in ten complete an ordered Cologuard test. The difference likely reflects not only the test itself but also Exact Sciences’ centralized logistics, reminders, patient support, and ability to track an unreturned kit. 

Second, after a positive stool test, only about half of US patients receive colonoscopy within six to twelve months, compared with roughly four-fifths in organized non-US programs. Positive Cologuard results appear to receive better follow-up than positive FIT results in some studies, but the magnitude of that advantage remains uncertain because the available comparisons are observational and frequently involve manufacturer-generated data.

References

Nielson CM, Vollmer WM, Petrik AF, et al. Factors affecting adherence in a pragmatic trial of annual fecal immunochemical testing for colorectal cancer. Journal of General Internal Medicine. 2019;34:978–985. doi:10.1007/s11606-018-4820-0.

Weiser E, Parks PD, Swartz RK, et al. Cross-sectional adherence with the multi-target stool DNA test for colorectal cancer screening: real-world data from a large cohort of older adults. Journal of Medical Screening. 2021;28:18–24. doi:10.1177/0969141320903756.

Le QA, Greene M, Gohil S, et al. Adherence to multi-target stool DNA testing for colorectal cancer screening in the United States. International Journal of Colorectal Disease. 2025;40:16. doi:10.1007/s00384-025-04805-0.

Mohl JT, Ciemins EL, Miller-Wilson LA, et al. Rates of follow-up colonoscopy after a positive stool-based screening test result for colorectal cancer among health care organizations in the US, 2017–2020. JAMA Network Open. 2023;6:e2251384. doi:10.1001/jamanetworkopen.2022.51384.

Finney Rutten LJ, Jacobson DJ, Jenkins GD, et al. Colorectal cancer screening completion: an examination of differences by screening modality. Preventive Medicine Reports. 2020;20:101202. doi:10.1016/j.pmedr.2020.101202.

Selby K, Senore C, Wong MCS, et al. Interventions to ensure follow-up of positive fecal immunochemical tests: an international survey of screening programs. Journal of Medical Screening. 2021;28:51–53. doi:10.1177/0969141320904977.

Senore C, Basu P, Anttila A, et al. Performance of colorectal cancer screening in the European Union member states: data from the second European screening report. Gut. 2019;68:1232–1244. doi:10.1136/gutjnl-2018-317293.

Zorzi M, Battagello J, Selby K, et al. Non-compliance with colonoscopy after a positive faecal immunochemical test doubles the risk of dying from colorectal cancer. Gut. 2022;71:561–567. doi:10.1136/gutjnl-2020-322192.

More CMS Legal Cases: LCD Challenges

 Header: A dissatisfied beneficiary can appeal his denied claim, OR, challenge the offending NCD or LCD as a whole. Let's see what's current on the CMS website.

##

AI CORNER

##

Medicare beneficiaries may challenge an LCD as unreasonable under 42 C.F.R. Part 426, but the process is obscure and procedurally demanding.

Fifteen recent DAB cases show an active yet largely unsuccessful docket: all were dismissed, eleven for threshold or jurisdictional defects, three after withdrawal, and one after the contractor revised its policy and effectively resolved the dispute. Common failures included unclear standing, failure to identify the precise LCD provision, missing physician documentation, and inadequate scientific evidence.

The cases reveal a mismatch between beneficiary-only standing and the sophisticated evidentiary burden required. NCD challenges remain legally available but appear dormant since 2014.

##

##

 

The Hidden Docket: Medicare LCD Challenges Are Alive—but Rarely Reach the Merits

Medicare beneficiaries have a little-known statutory right to challenge the validity of a Local Coverage Determination, or LCD. These cases still occur: the Departmental Appeals Board published 12 LCD-related decisions in 2024, six in 2025, and two through June 3, 2026. But they are buried in annual lists containing hundreds of unrelated HHS administrative decisions, and most end before the scientific merits are reached. (HHS.gov)

What is an LCD challenge?

An LCD challenge under 42 C.F.R. Part 426 asks whether a Medicare Administrative Contractor’s coverage policy is itself unreasonable. It is not the same as an ordinary claim appeal arguing that a contractor incorrectly applied an LCD to a particular patient.

The distinction matters:

• Claim appeal: “My service should have been covered under the existing rules.” This proceeds through redetermination, reconsideration, OMHA, and potentially the Medicare Appeals Council.

• LCD challenge: “The coverage restriction in the LCD is unreasonable.” This is heard initially by an administrative law judge in the HHS Departmental Appeals Board’s Civil Remedies Division.

Only an “aggrieved party”—generally a Medicare beneficiary who needs or received the service and whose coverage is affected by the LCD—may initiate the challenge. A laboratory, manufacturer, physician society, or other commercial organization cannot ordinarily challenge the LCD in its own name. (eCFR)

The complaint must identify the LCD and exact provision challenged, establish beneficiary standing and timeliness, include relevant treating-physician documentation, explain why the policy is unreasonable, and provide supporting clinical or scientific evidence. The ALJ ordinarily gives the complainant one opportunity to correct an unacceptable filing. Failure to cure the deficiencies produces dismissal and generally a six-month bar on refiling. (eCFR)

How to find the cases

There is no polished “LCD challenge database.” The practical route is:

1. Open the HHS Departmental Appeals Board’s Administrative Law Judge Decisions page.

2. Select a year.

3. Search the page for “LCD Complaint.”

4. Search separately among DAB Board decisions for appeals using “LCD Complaint,” the LCD number, the CR decision number, or “Part 426.”

General DAB ALJ decisions page:

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/index.html

2026 decisions:

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2026/index.html

2025 decisions:

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/index.html

2024 decisions:

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/index.html

DAB Board decisions, including appeals from LCD rulings:

https://www.hhs.gov/about/agencies/dab/decisions/board-decisions/index.html

An appeal from an LCD ALJ decision goes to the DAB Appellate Division, not to the Medicare Appeals Council. The Appeals Council reviews ordinary Medicare claim appeals; the DAB Board reviews Part 426 LCD proceedings. (HHS.gov)

Fifteen Recent LCD Challenges

The following are 15 representative published decisions from February 2024 through June 2026.

1. Gastrointestinal Pathogen Multiplex Panels

Case: In re LCD Complaint: Gastrointestinal Pathogen (GIP) Panels Utilizing Multiplex Nucleic Acid Amplification Techniques (NAATs) (L38229)
Docket: C-26-432
Decision: DAB CR6894
Date: May 13, 2026

Novitas had denied CPT 87507. The beneficiary’s underlying argument was that the LCD had been incorrectly applied to his individual claim, rather than that the LCD provision itself was unreasonable. The ALJ found no jurisdiction over that question and also concluded that the amended complaint lacked the required scientific and clinical support for an LCD challenge. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2026/alj-cr6894/index.html

2. Vitamin D Assay Testing

Case: In re CMS LCD Complaint: Vitamin D Assay Testing
Docket: C-26-368
Decision: DAB CR6909
Date: June 3, 2026

The filing did not adequately establish the complainant’s aggrieved-party status, timeliness, the exact LCD and provision challenged, or the clinical and scientific basis for alleging unreasonableness. The complainant did not respond to the ALJ’s order permitting an amended complaint, and the matter was dismissed. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2026/alj-cr6909/index.html

3. Immune Globulin for Autoimmune Encephalopathy

Case: In re LCD Complaint: Immune Globulins (L34771)
Docket: C-24-215
Decision: DAB CR6782
Date: October 17, 2025

A beneficiary with Hashimoto’s encephalopathy challenged restrictions affecting IVIG coverage. During the proceeding, WPS revised its billing article to add diagnostic codes permitting coverage for autoimmune encephalitis and encephalopathy when documentation requirements were met. The case was formally dismissed, but the contractor’s revision appears to have supplied the practical coverage relief being sought. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/alj-cr6782/index.html

4. Magnesium Testing

Case: In re LCD Complaint: Magnesium (L39400)
Docket: C-25-746
Decision: DAB CR6746
Date: August 11, 2025

The original complaint did not satisfy Part 426’s acceptability requirements. The complainant was offered an opportunity to amend but did not submit a corrected complaint by the deadline, requiring dismissal. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/alj-cr6746/index.html

5. Routine Foot Care and Mycotic Nail Debridement

Case: In re LCD Complaint: Routine Foot Care (L35138) and Debridement of Mycotic Nails (L35013)
Docket: C-25-299
Decision: DAB CR6639
Date: March 13, 2025

The beneficiary sought to reduce the required interval between covered nail-debridement services from nine weeks to six weeks. Although she supplied a podiatrist’s statement, she did not provide qualifying scientific or clinical evidence explaining why the LCD’s interval was unreasonable; a private insurer’s policy was insufficient. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/alj-cr6639/index.html

6. Cataract Surgery and Postoperative Toric Lenses

Case: In re LCD Complaint: Cataract Surgery in Adults (L34203)
Docket: C-24-769
Decision: DAB CR6608
Date: January 22, 2025

The complaint concerned payment for toric contact lenses reportedly needed following cataract surgery. The filing lacked several required elements and supporting evidence, and no amended complaint was submitted after the ALJ identified the deficiencies. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/alj-cr6608/index.html

7. B-Type Natriuretic Peptide Testing

Case: In re LCD Complaint: B-Type Natriuretic Peptide (BNP) (L33573)
Docket: C-25-91
Decision: DAB CR6606
Date: January 16, 2025

The beneficiary adequately demonstrated standing and timeliness, but the complaint did not clearly identify the challenged LCD provision, articulate why it was unreasonable, or supply supporting clinical evidence. No amended complaint followed, and the matter was dismissed. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2025/alj-cr6606/index.html

8. Cologuard Screening

Case: In re LCD Complaint: Cologuard Screening
Docket: C-25-20
Decision: DAB CR6583
Date: December 5, 2024

A nurse practitioner argued that Cologuard was medically necessary for a particular patient. The ALJ explained that beneficiary-specific medical necessity ordinarily belongs in the claim-appeal system; an LCD challenge instead requires identification of an unreasonable policy provision, appropriate authorization, proof of timeliness, physician documentation, and supporting scientific evidence. None was supplied in an acceptable amended complaint. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6583/index.html

9. Hyaluronic Acid Injections for Knee Osteoarthritis

Case: In re LCD Complaint: Hyaluronic Acid Injections for Knee Osteoarthritis
Docket: C-25-44
Decision: DAB CR6580
Date: December 3, 2024

After the ALJ found the initial complaint unacceptable and allowed amendment, the complainant withdrew the challenge. Part 426 required dismissal; a dismissal following withdrawal generally cannot be appealed and prevents refiling for six months. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6580/index.html

10. Trigger-Point Injections

Case: In re LCD Complaint: Trigger Point Injections (L34211)
Docket: C-24-458
Decision: DAB CR6512
Date: July 25, 2024

The beneficiary corrected certain physician-signature and timeliness defects but still did not identify the precise LCD provision alleged to be unreasonable or provide scientific evidence explaining why it should be invalidated. The amended filing therefore remained unacceptable. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6512/index.html

11. Surgical Treatment of Nails

Case: In re LCD Complaint: Surgical Treatment of Nails (L34887)
Docket: C-24-427
Decision: DAB CR6491
Date: June 17, 2024

The filing expressed concerns about nail-avulsion procedures but did not identify a qualifying aggrieved Medicare beneficiary, demonstrate timeliness, or satisfy the substantive complaint requirements. No acceptable amended complaint was presented. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6491/index.html

12. Pressure-Reducing Support Surfaces

Case: In re LCD Complaint: Pressure Reducing Support Surfaces—Group 1 (L33830)
Docket: C-24-371
Decision: DAB CR6463
Date: April 18, 2024

The challenge concerned coverage of a pressure-reducing support surface. The beneficiary’s representative subsequently asked that the proceeding be closed, which the ALJ treated as a withdrawal and dismissed under Part 426. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6463/index.html

13. Vitamin D Assay Testing

Case: In re LCD Complaint: Vitamin D Assay Testing (L36692)
Docket: C-24-135
Decision: DAB CR6441
Date: March 13, 2024

The complaint concerned magnesium and vitamin D testing but lacked an adequate treating-physician statement, identification of the exact LCD provisions, an explanation of their alleged unreasonableness, and supporting clinical evidence. The beneficiary did not submit a corrected complaint. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6441/index.html

14. CT of the Head

Case: In re LCD Complaint: CT of the Head (L34417)
Docket: C-24-204
Decision: DAB CR6438
Date: February 29, 2024

The filing did not adequately identify the LCD provision challenged, establish aggrieved-party status, or provide supporting medical and scientific evidence. The ALJ also questioned whether the beneficiary was actually seeking review of an individual claim denial rather than challenging the validity of the LCD itself. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6438/index.html

15. Therapeutic Shoes for Persons with Diabetes

Case: In re LCD Complaint: Therapeutic Shoes for Persons with Diabetes (L33369)
Docket: C-24-199
Decision: DAB CR6435
Date: February 26, 2024

The initial complaint was incomplete. Before the deadline for correction, the complainant voluntarily withdrew it, resulting in dismissal and the regulatory six-month restriction on refiling the same complaint. (HHS.gov)

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2024/alj-cr6435/index.html

---

Sidebar: The Strange, Nearly Moribund World of NCD Challenges

National Coverage Determination challenges remain authorized under Part 426, but they follow a different route. An acceptable NCD complaint goes directly to the DAB Appellate Division, rather than first being heard by a Civil Remedies Division ALJ. (eCFR)

HHS maintains a dedicated page entitled “Acceptable National Coverage Determination Complaints.” It is less a modern docket than a static historical list:

https://www.hhs.gov/about/agencies/dab/different-appeals-at-dab/appeals-to-board/national-coverage-determination-complaints/acceptable-national-coverage-determination-complaints/index.html

The most recent accepted complaint publicly listed is a vagus-nerve-stimulation proceeding from 2014. The page also includes the celebrated 2013–2014 challenge to the national noncoverage policy for transsexual surgery, which resulted in DAB No. 2576. No accepted complaint dated after 2014 appears on the public list. (HHS.gov)

Thus, NCD review is legally alive but publicly close to moribund. The page does not prove that nobody has attempted a complaint since 2014; it shows that HHS has not publicly listed a newer complaint as acceptable. The likely explanation is structural: standing is confined to affected beneficiaries, the evidentiary burden is substantial, and CMS can reconsider, withdraw, or revise an NCD while a challenge is pending. Those features make an NCD reconsideration request or conventional claim litigation more manageable for sophisticated stakeholders, although that conclusion is an inference from the regulatory design rather than an announced CMS policy. (HHS.gov)

What the Recent LCD Cases Teach

The first conclusion is stark: none of these 15 cases produced a decision holding an LCD either reasonable or unreasonable on the scientific merits. All 15 were formally dismissed.

The dispositions break down as follows:

• 11 of 15 were dismissed at the acceptability or jurisdictional stage.

• 3 of 15 were voluntarily withdrawn.

• 1 of 15, the immune-globulin case, was dismissed after the contractor revised its policy in a way that appears to have resolved the coverage problem.

• 0 of 15 generated a completed evidentiary review of the LCD record and a merits ruling.

The recurring procedural defects were remarkably consistent: failure to identify the exact LCD provision, inadequate evidence of beneficiary standing or timeliness, absence of a treating-physician statement, and—most importantly—failure to provide scientific evidence accompanied by an explanation of why the contractor’s policy was unreasonable.

Several complainants also misunderstood the nature of the remedy. They wanted an adjudicator to decide that a service was medically necessary for one patient or that the contractor had applied the LCD incorrectly. Those are ordinary claim-appeal questions. A Part 426 case requires an attack on the validity of the policy itself, supported by evidence addressing the policy’s clinical logic.

This produces an awkward imbalance. The only parties with standing are individual beneficiaries, but the task resembles sophisticated health-policy litigation: the complainant must identify the operative language, understand the evidentiary record, marshal relevant literature, and explain why the contractor’s synthesis is unreasonable. Manufacturers and medical societies may assist, but they cannot simply substitute themselves as the complainant.

The immune-globulin case is therefore especially instructive. A formal dismissal can disguise a substantive success. Once WPS revised the associated article and permitted payment under additional diagnostic codes, there was no longer a live coverage restriction requiring adjudication. Under Part 426, a contractor revision that removes the challenged provision can carry essentially the same practical consequence for the beneficiary’s claim as a favorable invalidity decision. (eCFR)

The docket is consequently neither hidden nor dead. It is poorly indexed, procedurally unforgiving, and dominated by unsuccessful pro se complaints. The scarcity of merits decisions should not be mistaken for proof that LCDs are universally defensible. It more often shows that very few complainants survive the gateway requirements necessary to place the contractor’s scientific record genuinely at issue.

 

MolDx in Court: Scanning CMS and Federal Cases for "MolDx"

 I'm such a CMS nerd that every year or two I used to scan the CMS Office of Hearings and Appeals (OMHA) for interesting cases.  (Back-when, there were some gems involving HeartFlow).

I'm not sure if OMHA still posts as many cases.  But I asked Chat GPT to search CMS and the web for legal cases involving MolDx.  I show the results below.

##

AI Corner

##

The publicly available  MolDX  appeal record is limited and dominated by Agendia’s disputes over BluePrint and TargetPrint breast-cancer assays. Several OMHA ALJs initially found the tests reasonable and necessary, but the Medicare Appeals Council repeatedly reversed favorable decisions for insufficient deference to MolDX LCDs and technical assessments. Later ALJs generally upheld noncoverage, finding that physician testimony, post-service publications, or evidence of clinical benefit did not adequately establish incremental clinical utility.

Federal courts ultimately sustained the MolDX framework, rejecting challenges to LCD authority and contractor delegation. Overall, the cases show that overcoming MolDX requires a strong, beneficiary-specific evidentiary record rather than generalized scientific disagreement.

##

I found 13 useful decisions: one officially published HHS LCD-challenge decision, eight claim-level ALJ/Council decisions recovered from federal-court appendices or opinions, and four federal judicial-review decisions.

A caution at the outset: HHS now publishes only selected Medicare Appeals Council decisions, and there does not appear to be a comprehensive public database of ordinary OMHA claim decisions. Accordingly, this is a substantial but not necessarily exhaustive MolDX set. (HHS.gov)

I. Published HHS ALJ decision involving a MolDX LCD

1. In re LCD Complaint: MolDX Genetic Testing for Hypercoagulability/Thrombophilia

Case: In re LCD Complaint: MolDX: Genetic Testing for Hypercoagulability/Thrombophilia (Factor V Leiden, Factor II Prothrombin, and MTHFR) (L36159)
Number: DAB CR5997; Docket No. C-22-13
Date: December 9, 2021
Created title: MolDX Thrombophilia LCD Challenge Dismissed at the Threshold

A physician attempted to challenge MolDX LCD L36159 on behalf of a Medicare beneficiary. The ALJ dismissed the complaint without reaching the scientific validity of the LCD because the complainant failed to document timeliness, representative authority, the challenged provisions, supporting evidence, and other mandatory elements after being given an opportunity to amend. (HHS.gov)

Live link:

https://www.hhs.gov/about/agencies/dab/decisions/alj-decisions/2021/alj-cr5997/index.html

From NGS MAC to WELLPOINT FEDERAL

I mentioned several weeks ago that the NGS MAC (New England, New York, and MN-WI-IL) had begun referring to itself as WELLPOINT FEDERAL.  (While using "elevance" emails.)

WELLPOINT FEDERAL is now showing up on CMS.gov websites, as below.

Earlier:

https://www.discoveriesinhealthpolicy.com/2026/06/the-mystery-of-wellpoint-federal-aka.html

CMS and Software-Intensive Lab Services: The Third Rail is "How to Enroll" after Coding and Payment

In both its OPPS proposed rules, and its PFS or Part B proposed rules, CMS has proposed to drop AI-based or software-intensive lab diagnostic tests from the Clinical Laboratory Fee Schedule.  CMS says these services are not "CLIA tests" and do not require CLIA lab status.   In both rules, CMS preserves coding and payment pathways.  In OPPS, coding is the same, payment is shifted to APCs, and in PFS, coding is the same, payment is "contractor priced."

However, CMS neglected to mention a third factor.  How would such labs enroll in Medicare?  You have to enroll as "something" - as a doctor's office, as an ambulance service, as an ASC, and so  on.

I let Chat GPT chew on this dilemma and discuss it.

###

CMS has addressed coding and payment—but not the supplier enrollment pathway

The proposal exposes a third problem distinct from coverage and pricing:

1. Benefit category: Is the service a clinical diagnostic laboratory test or an “other diagnostic test”?

2. Payment system: Is it paid on the CLFS, under the PFS, or through an OPPS APC?

3. Enrollment category: What kind of Medicare supplier is legally permitted to submit the claim?

CMS addresses the first two but largely overlooks the third. It reasons that a downstream algorithmic analysis of previously generated genomic, histologic, or other laboratory data is not itself an examination of human material and therefore need not be performed by a CLIA-regulated entity.  (Some will strongly disagree.)  CMS consequently proposes to treat these analyses as “other diagnostic tests,” contractor-price them under the PFS, and place the corresponding hospital services in new-technology APCs.

But Medicare does not pay an unclassifiable commercial entity simply because its service has acquired a code and price. Under 42 CFR 424.505, a provider or supplier must be enrolled to receive Medicare payment. (eCFR)

The enrollment form is CMS-855B

The classic Part B organizational enrollment application is CMS-855B, Medicare Enrollment Application—Clinics/Group Practices and Other Suppliers. The form requires the applicant to select a supplier type. Among the choices are:

• Independent clinical laboratory

• Independent diagnostic testing facility

• Clinic/group practice

• Pharmacy

• “Other”

But “Other” is not an open-ended invitation to invent a new supplier category. The form expressly says that it may be selected only when the supplier type is already eligible to enroll and bill Medicare but is not shown on the list. An applicant uncertain whether it is eligible is directed back to its MAC. (CMS)

That is precisely where the circularity can begin.

IDTF is the apparent destination—but it is an awkward one

Once CMS calls the service an “other diagnostic test” paid under the PFS, an IDTF becomes the most obvious existing enrollment category for a freestanding, nonphysician software company. Section 410.33 says that PFS diagnostic procedures generally must be performed by a physician, physician group, certain enumerated practitioners or suppliers, or an independent diagnostic testing facility. For a corporate software vendor that is neither a medical practice nor a laboratory, IDTF may be the only plausible item on that list. (eCFR)

There is some precedent for stretching the IDTF concept toward remote services. The regulations recognize IDTFs that perform services remotely and never see beneficiaries at their premises, relieving them of requirements such as handwashing facilities and patient-privacy accommodations. (eCFR)

Nevertheless, the IDTF framework remains heavily designed around imaging and physiologic testing:

• A physical practice location

• Diagnostic equipment maintained at that location

• Equipment inventories, model numbers, and serial numbers

• Named technicians and their credentials

• An interpreting physician, when applicable

• A supervising physician proficient in each procedure

• Code-by-code approval by the MAC

• State licensing and multi-state operational requirements

• Site inspections and record-production obligations

The CMS-855B IDTF attachment requires the company to identify every CPT/HCPCS code and the associated equipment model. More pointedly, it currently instructs applicants point blank that “clinical laboratory and pathology codes should not be reported.” (CMS)  (Conversely, possibly CLIA labs are only allowed to submit CLIA codes - not surgeries or MRIs - and if CMS says the codes are CLIA codes for submission purposes but are not CLIA codes for CLFS purposes, that's a hall of mirrors.)

Therefore, merely moving 0510U, 0512U, 0418U, 0220U, and similar services from the CLFS to PFS contractor pricing does not make them readily enrollable as IDTF services. CMS would have to answer such questions as:

• May these newly reclassified codes now be entered on the IDTF attachment despite their laboratory or pathology origins?

• Is the “equipment” a server, a cloud environment, a software version, or the algorithm itself?

• Who or what is the “technician” performing an automated analysis?

• For IDTFs, physician supervision required, and at what level?  (Is he supervising a computer chip?)

• How does the MAC define the medical specialty?  Must the supervising physician be a pathologist, molecular pathologist, radiologist, oncologist?   Pathologist seems natural but contradicts the CMS position the test is not a laboratory service. 

• Where is the service furnished for place-of-service and multi-state licensing purposes?

• Can one national software center enroll once, or must it enroll separately in multiple MAC jurisdictions?

• Once the code is ejected off the CLFS, is the service technical, professional, or global?

Without national answers, the proposal could produce a service that is covered, coded, and priced—but not billable by the entity that actually performs it.

HeartFlow is the scary precedent

The early HeartFlow experience illustrates this danger almost perfectly. HeartFlow sought IDTF enrollment for its FFRct software service. Noridian denied the application partly because it concluded that the proposed code was not separately payable and partly because the proposed supervising cardiologist did not meet what Noridian viewed as the applicable radiology-supervision requirement. Noridian therefore refused to grant HeartFlow a Medicare supplier number. (HHS.gov)

The Departmental Appeals Board candidly described the circularity: HeartFlow had to meet the requirements applicable to IDTFs, but Noridian primarily denied enrollment because HeartFlow had not identified a service reimbursable under the existing payment structure. The Board observed that HeartFlow could not—and would have no reason to—enroll as an IDTF if it could not bill Medicare for its only service. (HHS.gov)

That is the danger here on a larger scale. A MAC evaluating a SaMS company’s CMS-855B application might ask:

Is this really an IDTF test? Is this code authorized for IDTF billing? Where are the equipment and technician? What supervision applies?

Meanwhile, the payment staff may respond:

The code is payable under the PFS to an eligible enrolled supplier.

Each side can point to the other, leaving the company without a PTAN.

The OPPS route is less troublesome—but only because the hospital can bill

For a hospital outpatient service assigned to an APC, the hospital is already an enrolled Medicare provider and can ordinarily purchase the algorithmic analysis under an arrangement with the vendor. The software company may not need to enroll or submit the Medicare claim itself.

That does not solve the freestanding PFS problem. CMS would effectively be creating two commercial models:

• Hospital: the hospital bills the APC and pays the software vendor contractually.

• Nonhospital: the algorithm company must either find a valid supplier category or arrange for a physician or other enrolled entity to purchase and bill the service.

The latter approach could raise additional purchased-diagnostic-test, reassignment, anti-markup, supervision, place-of-service, and claim-reporting questions. It would also mean that the proprietary test developer no longer necessarily bills Medicare directly.

What CMS should do

CMS should not finalize the payment migration without simultaneously establishing an enrollment pathway. It has at least three options:

1. Preserve laboratory enrollment for this limited purpose.
CMS could permit an entity already enrolled as an independent clinical laboratory to bill designated SaMS codes even though the codes are paid under the PFS rather than the CLFS. Fee schedule and supplier enrollment category do not inherently have to be identical. Pure software entities would still require either a special “dry laboratory” enrollment subtype or another route.

2. Establish a defined remote-SaMS IDTF pathway.
CMS could state nationally that these services are eligible IDTF diagnostic procedures and modify the CMS-855B attachment and PECOS requirements for software-based testing. It would need software-appropriate rules for equipment, personnel, supervision, locations, and code approval rather than allowing each MAC to improvise.

3. Create a separate SaMS supplier category.
A new category on CMS-855B/PECOS would be the cleanest long-term answer. It could include enrollment screening, ownership disclosure, medical-record access, ordering requirements, cybersecurity and software-version controls without forcing cloud software into rules written for imaging machines and technicians.

At minimum, CMS should grandfather existing laboratory enrollments and PTANs during the transition, synchronize enrollment implementation with the January payment change, and prohibit MACs from denying enrollment merely because the affected code was historically categorized as laboratory or pathology.

The central comment point

CMS cannot logically conclude that SaMS analyses require no CLIA-regulated entity, remove them from the CLFS, and then remain silent as to which Medicare-enrolled supplier may furnish and bill them. 

Unless CMS preserves an independent-laboratory enrollment route, expressly designates these entities as remote IDTFs, or creates a SaMS supplier category, the new proposal risks reproducing the HeartFlow problem: a separately payable diagnostic service for which the actual supplier cannot obtain Medicare billing privileges.

SIDEBAR

The CLIA number for ArteraAI (e.g.)  seems to be 10D2258449, Jacksonville, per online Report Form.

The CLIA number for Valar Labs seems to be 45D2278995, per app.Dexzcodes.com.  1200 Binz St, Houston.  NPI 1891554481 as Clin Med Lab, 1200 Binz St, Houston.

Very Brief Blog: FDA Guidance on Payor Communications

Here's one for your horizon scan.  FDA has posted draft guidance updating its policies for communications with payors, such as HEOR information.

Here's my understanding.  FDA already had a guidance that covered marketed product (drug or device) communications with payors about 'health economics' that was not misleading.  

Recall that the starting point for FDA-regulated manufactuerers is what's literally on the FDA label, where HEOR usually isn't.

The old guidance was predicated in part on 502(a) and involved communications on approved products and indications.  There was some FDA "enforcement discretion" for unapproved products (pipelines).  

A 2023 bill adds 502(gg) which specifically extends by law some protections to unapproved (pipeline) products.   

The draft was released June 3 with 60-day comment to August 3.  Think of it this way: 

• Old 2018 final guidance + 2023 new law §3630 creating 502(gg) = this 2026 new draft guidance.

The comments docket is here.  The 24-page draft is here.

_____

Any number of articles about the draft are on Google here.  For extra credit, also check out the 2025 FDA guidance on unapproved uses of approved products here.

Here's the AI White Paper on Paying for AI...Articles from STAT and PETERSON

Do payers pay for AI?   Or do they assume AI-mediated savings, and deflation of prices?  Is one clinical area totally different from another in terms of AI financing and strategies?

STAT just ran a three-part series on AI and reimbursement, with Medicare and other use cases.  Peterson Institute just released a 15pp on AI financing in healthcare.  And there are collateral sources, e.g. a PathAI webpage on the cost strucures of digital pathology.

Here's a consolidated view from Chat GPT...

This essay is also available as a PDF white paper in the cloud - here.

##

Summary.

• Clinical AI is not simply another medical device awaiting a code and payment rate. It can replace human labor, operate continuously, and scale at low marginal cost, creating both opportunities for better care and risks of uncontrolled spending. 
• A useful financial architecture must therefore address six interconnected layers: investment, transactions, workflow and accountability, evidence, market integration, and dynamic governance. 
• The Peterson framework provides the central economic argument. 
• STAT, PathAI, CMS policy, and academic research then show how implementation costs, downstream utilization, platform power, weak evidence, and changing professional roles shape the result. 
• Together, they determine whether AI lowers costs or merely adds another reimbursable layer.

##

AI Guest Author: A Detailed White Paper about Payors and Genomics 2017-2026

Header: An AI-generated 20-page white paper that surveys a decade of articles about payors and genomics.  

Here's another adventure in current AI thinking and writing.  Chat GPT 5.6 surveyed 15 papers on payors and genomics, and for contrast, it also read 5 papers on payors and other medical industries. 

The report provides a general overview and key conclusions.  Trends over the decade are discussed. There's a contrast-and-compare with other medical industries - from continuous glucose meters to radiation therapy.  A side bar reviews two UCSF papers (both Phillips et al...2018, 2026) on payors and genomics.

• Find the white paper in the cloud here.

Summary: 

• Over the past decade, payer–genomics literature has evolved from documenting fragmented coverage policies to examining the broader implementation system. 
• Early studies emphasized inconsistent standards, reliance on guidelines, and uncertainty about clinical utility. 
• Later work showed that coverage alone does not ensure access: prior authorization, counseling gaps, clinician hesitation, patient cost, coding, and claim denials remain substantial barriers.
• Payers increasingly appear not only as gatekeepers, but also as potential implementation partners. 
• The central lesson is that genomic progress depends on more than accurate assays. It remains to be seen how increasingly rapid and advanced genomic testing will fare in an uncertain marketplace of payers expected to reimburse it.

The LabCorp $14M Fine on Tox Testing (Humor)

 You're probably seen press releases about the recent LabCorp $14M DOJ fine related to urine tox testing.

There are some sources that go into numbing detail about the coding situations and the coding definitions and overlap.  First link is DOJ press release; second link is 11p court document.

https://www.justice.gov/opa/pr/labcorp-agrees-pay-145m-resolve-false-claims-act-allegations

https://www.justice.gov/opa/media/1452616/dl

click to enlarge

How I Use AI: "Ten Papers on Barriers to Molecular Test Adoption"

Header: For some questions, AI can search Google or AI can search PubMed more efficiently than I can.

Today I asked Chat GPT to search PubMed and bring back ten papers on barriers to adoption of molecular tests in oncology.  I named some categories like slow adoption of LBx for driver genes [CDx], the slow adoption of Oncotype Dx or similar tests.  

Pro's and Con's

Upsides - The AI does the task quickly, with clean output. I assume it skimmed 100 or more titles to choose 10 as "the best" to display for me.   As shown below, AI can then provide a narrative summary about the group of papers it found.

Downsides - But we don't know what we're missing.  A true academic review might pull 200 abstracts or more and have two human readers each rate to find the most important ones. Many hours. 3 minutes per paper, 20 man-hours.

Can Chat GPT bring me the actual PDFs?   I've tried asking Chat GPT to go and get the actual open access PDF files, where they exist, and deliver to me as a zip file link I can click and download.  

It worked, but very slowly, 30 minutes, once.  It usually failed or retrieved only 1 of 10 open PDFs.   I think you need desktop agentic software and some extra skills to get it to give you PDFs.   Note, the compact freeware citation manager Zotero can easily and reliably pull any number of open access PDFs in a couple seconds, so fancy AI should be able to do the same.

##

Here is the direct output about molecular test adoption.

###

Bruce, I selected these to maximize topic diversity: liquid biopsy, broad tumor profiling, breast prognostic testing, Lynch screening, ovarian HRD/BRCA, prostate germline testing, rural precision oncology, MCED, and a newer companion diagnostic. Only one is specifically about MCED.

INSERT:
Here is an overview essay (written by AI only after it found the 10 barriers papers and 4 extra papers about MRD barriers.)

AI Guest Author: Review of H&E AI Breast Cancer Prognosis, versus Standard Methods (21-gene)

I have seen a scattering of articles that study whether AI interpretations of H&E slides can mirror the prognostic value of molecular tests.   

The linked PDF white paper is entirely machine generated.

Here were the steps:

1. I asked it to seek literature on the topic from PubMed.
2. It got about 8.
3. I asked it to download all the (public access) PDFs in a zip file for me.  (See footnote).
4. I then gave the PDFs back to it, and asked it to write a review article.
1. I gave it about 8 or 10 topics or talking points to include.
5. It produced a 12-page review.
6. I asked it to produce an interesting report cover.

##

Find the white paper in the cloud here.  

(See also footnote below).

##

Here's the executive summary:

From Oncotype DX to “Virtual Recurrence Scores”:

AI Histopathology for  Breast Cancer Risk Stratification

Abstract

Oncotype DX is the best-established multigene assay for guiding adjuvant chemotherapy in hormone receptor-positive, HER2-negative early breast cancer. Its evidence base spans retrospective validation, TAILORx, RxPONDER, and major guidelines, but cost, turnaround time, access, and tissue consumption remain limitations. A rapidly expanding literature now uses routine H&E whole-slide images, often combined with clinicopathologic data, to approximate Oncotype DX or related transcriptomic scores and, increasingly, to predict recurrence and chemotherapy benefit directly.

Recent studies show multicenter generalization, outcome discrimination, and analytical reproducibility. Yet score imitation, prognosis, treatment-benefit prediction, and clinical utility remain distinct claims and they will require separate validation before AI could replace genomic testing.

###

Footnote.

Regarding Chat GPT fetching and downloading PDFs: It did do that that, though after a puzzling half-hour of delay.   For other white papers, Chat GPT just failed to download PDFs (colliding with PubMed Central firewalls, bot detection, etc).   One would have to go further with downloaded AI and true "agentic" modes to reliably get it to find and download PDFs.  On the other hand, freeware reference manager software like Zotero can fetch any number of PMC PDFs nearly instantly.

Regarding US Companies.  A colleague suggested that a future second version could give more direct attention to preciseDx and ArteraAI.  Here are two AI bullets:

PreciseDx is developing PreciseBreast, which combines H&E-derived features with clinicopathologic data to predict six-year recurrence risk.  It has reported performance comparable or superior to Oncotype DX in retrospective cohorts.

ArteraAI has advanced further commercially: its FDA-cleared ArteraAI Breast test uses digitized H&E slides plus clinical variables to stratify distant-metastasis risk, with separate trial evidence suggesting prediction of chemotherapy benefit in selected patients.  (The Oncotype test is not FDA-cleared.)

In the first-version white paper, a third company, the German company Stratif-AI was represented by Boehm et al. 2025 in Nature Communications.

AI Can Do More Than You Think: Downloading the PDFs & Busywork

UPDATE.  I did have one day of success asking Chat GPT in "LLM" mode to download PDF papers and give me a Zip file.  But it took a long time (?hour).   I couldn't make it repeat this successfully on other projects.

BUT...

I understand you can do this reliably if you learn how to use full-strength agentic mode where it "takes over" your desktop browser etc.  I don't know that.

##

Here's an example of AI being able to do more for you than you might guess.

1. Search Pubmed

I asked Chat GPT to find recent papers asserting that computational pathology can predict breast cancer relapse as well as Oncotype Dx.  I mention I have seen several of such papers, including one in the past week.

Chat CPT came back with a number of publications, including flagging for me the most recent of the group.  It wrote, 

• The item you probably encountered “last week” was the Witowski et al. Nature Communications paper. It was initially published May 20, 2026, but its version of record appeared July 6, 2026, accompanied by fresh publicity. That qualifies as the newest directly relevant peer-reviewed paper I found.

Clilck to enlarge:

click to enlarge

2. Downloading PubMed Papers For You

I thought, this list is good, I will now click around on all the links and find the PDF download buttons and download each open access one.

Then I thought, wait a minute.

I asked it:

1. Are you able to retrieve PDFs that are open access and then send to me as a zip file or such?  Or do I have to manually check each link and down load a PDF for each.
2. Yes. I can retrieve the legally open-access PDFs, rename them consistently, and package them into a single ZIP file for download here. You do not need to open and download each one manually.

Here's the ZIP file (the file names style followed my instructions). 

Click to enlarge.  By asking Chat GPT to actually download the papers in a zip, I probably saved 20 minutes.

click to enlarge

AMA CPT: Call for Comments on September 2026 New Codes; also AI in Code Change Application

 AMA has a webpage for the September 17-19 Editorial Panel meeting in Minneapolis.  There are limited in person seats; first come first served; but unlimited virtual participation.  It's at the Hyatt Regency.

https://www.ama-assn.org/membership/events/cpt-editorial-panel-meeting

The public agenda for all codes is now available.  Comment for lab codes is already closed (to allow for early subcommittee meetings).

Click on this link.  Within this PDF you will find instructions to access the AMA CPT Smart App, a web portal that has the dual functions of (1) letting you write and submit new code applications, AND, (2) as a meeting approaches, request to review a packet and submit a comment.

https://www.ama-assn.org/system/files/cpt-panel-september-2026-agenda.pdf

To help  you have a targeted plan to keep updated (codes are withdrawn, etc) AMA says it will update the agenda on July 10, August 14, September 4, September 11.  AUGUST 10 is the last date to submit comments for the "Interested Party" function.  This ensures that all public comments can be reviewed by the voting panelists.

There are 94 agenda items (some numbers may be skipped).  For codes that go into a valuation process (most Category I codes go to AMA RUC), this is the last round for AMA to collect data this winter and get into CMS's public comment process summer 2027 for codebook 2028.

Big Changes for AI for Pathology Codes Only (???)

As we saw earlier in the early release of path-lab codes, tab 94 is extensive revisions to the path-lab application, with many questions on software and its validation.   

• I had expected to see a mirroring tab that would be similar text added to general purpose Category I & III applications, but I don't see that.

See my July 2 discussion of the Path Lab AI CPT changes.

https://www.discoveriesinhealthpolicy.com/2026/07/run-dont-walk-ama-cpt-big-changes-for.html

Assessing the New MolDx LCD for Transplant Rejection; And Our First LCD Report from Agentic Instructions

Header: MolDx releases a long-awaited LCD on transplant rejection testing.  See L40140. Clear improvements in coverage and some explanations.  The actual coverage rules could still be improved for clarity.  

For this blog, this is the first time a report has been entirerly self-generated by AI.  There were 9 steps for creating a report, for example, #10 create a 500-word blog article, and #11 generate a report cover.  The point is, any new LCD could be assessed and reported immediately and automatically by given the AI the same 11 steps.  OK, the result isn't always that great, but it's still a new landmark.

##

See coverage & stock pop at 360Dx here.  Stock pop $30-$38 or as market cap, from ~$1.5B to ~$2.0B.

##

Find the 15 page PDF report online here.  Prompts sidebar that built the LCD review, here.

The 15 PDF is the "deep dive" on the new LCD.   Below is a shorter AI News Story on the new LCD.

##

MolDx Broadens Transplant Testing Coverage—but Keeps a Firm Hand on the Rules

A closely watched final policy

MolDx has finalized LCD L40140, Molecular Testing for Organ Allograft Rejection, replacing the draft issued in summer 2025. The policy governs Medicare coverage for tests such as donor-derived cell-free DNA and gene-expression profiling, used either when rejection is suspected or to monitor an apparently stable transplant recipient for subclinical injury.