Saturday, December 30, 2023

Xifin Offers Advanced Payor Analytics on Lab Pricing

Xifin, whose services include a large lab billing consultancy, offers an elaborate analytics system for payor contracting and pricing data.   Find it at:

https://www.xifin.com/resources/payor-rate-transparency-monitor/

The online web page provides a sampler; for data contracts and strategic services contact Xifin.

###

For example, in the screen shot below, set to code 80048 (metabolic panel), United has 590 distinct rates, Aetna has 578.   United's weighted average is $8.39, Aetna's is $8.57.   

click to enlarge

For 88305, another sample available to try, a common biopsy surg path code, United has 3,718 distinct rates, Aetna 661.   United's weighted average is $95.38, Aetna's is $73.95.   

click to enlarge

The bubble charts show volume per price (Dollar-cent), for example, Aetna has concentrations at $58, $56, $61 and $65.

(Somewhat similar data was seen in 2017 CMS PAMA pricing data, including popular price bands and including very high and very low outliers). 

Hospitals & HHS

Federal laws require a level of hospital price transparency (at HHS here) but this doesn't display payor & independent lab adjudication.  

 

###

See an article at Politico on whether AI can reduce payor administrative burdens.

https://www.politico.com/news/2023/12/31/ai-medical-expenses-00132557





Friday, December 29, 2023

Medicare's Festival of Codes for Care Management, Update 2024

Is it a new era in Medicare?   Are we rocketing into a new Medicare that has extensive resources for care navigation and care coordination?   This article unravels some new 2024 policies, and shows a surprising connection between "new codes" and White House cancer policy.  

For a quick overview, see an AI summary of this blog here.

###

A Bit of History

Traditionally, Medicare Part B paid for direct physician services (and services incident-to a physician, like the nurse giving the flu shot).   Getting anything other than a direct and face-to-face physician service was pretty difficult.  

  • For example, take-home, home-use allergens from an immunologist had to receive their own statutory benefit category (found at SSA 1861(s)(2)(G)) since they were not a face-to-face service at the time of use, at home.   

Then some bona fide non-face-to-face services were introduced, like physician monthly care management plans for home health or hospice (G0181, '82).  These codes go back decades and pay about $100/month (2022, G0181, 400,000 uses for $42M; G0182, 34,000 uses for $3.5M).  The key feature of the home health and hospice management code pair is, that they are for not-face-to-face services, which at the time was quite uncommon.

Care Management Services

There are now a lot of care management services, and CMS provides a 15 page PDF guide vintage 2021, which will have to be a few pages longer when updated for 2024.  Be aware that I'm only cherry-picking a few of the long list of care management codes.

In this blog, enjoy a snapshot of part of what's recent or even brand new in the last several weeks.    I find this area of care management and navigation services to be extremely interesting, because many genomics labs (see my recent blog about YouScript and Tempus AI here) are featuring their integration into care pathways and their role at decisional timepoints in care management.   Navigating cancer care would certainly include navigating the access to, and flow of funds for, genomic tests.   Is it a "you snooze, you lose" proposition for genomics labs?

[Separately, re AI and care management, see the extremely proactive recent HHS rulemaking on any kind of AI (any algorithm with data), whenever incorporated into EHR services - here).    

AMA, White House on Navigation

See a special AMA update on new care management codes here and here.   

Remarkably, see a White House announcement about the new care navigation codes, their relevance for oncology patient navigators, and their role in the larger Cancer Moonshot project.  From the WH Office of Science and Technology Policy, OSTP, they also cite to an auxiliary document on standards for high quality care navigators/navigation here (in, The Journal of Cancer Navigation, update, here).  Bills to require patient navigators in Medicaid have been floated (here).    The White House document was written by Anabella Aspiras, who worked on Cancer Moonshot under Obama (2016) and rejoined in January 2023. (In between she was with Mendel.AI).



Code Set #1

"Principal Care Management"

See codes 99424, 99426, which are physician or qualified health care professional [PA, NP] care management services, first hour, and add-on hour.  (2022, 99424, 75,000 uses, $6.5M, $87; 99425, 23,000 uses, $1.M, $64.)   There are also a pair of codes for "clinical staff" directed by a physician, 99426, 99247, paying about 2/3 as much and used in roughly the same volumes.  

The 75,000 uses of 99424 are pretty tiny compared to the 40m Medicare Part B patients.   Checking the Medicare Coverage Database (MCD), I saw no LCDs or articles about usage guides or audit risks or record-keeping tips for these codes.

Code Set #2

"E&M Add on Code G2211" ($17)

CMS has been trying to add code G2211 for several years and an Act of Congress actually delayed it until January 2024.   (Discussed, for example, by Urology Times.)  

The code is an add-on code to office visits that adds $17 (prorated for more expensive areas).  This was generally opposed by procedural specialties who either don't bill E&M (e.g. radiology, pathology) or who have bundled postsurgical visits (e.g. surgeons) rather than standalone E&M codes.   CMS estimated it could be added to as many as a third of primary care visits.  

CMS fielded much debate on this code in fall final rulemaking (88 FR 78969ff, November 16, 2023).   On 78980, CMS bluntly stated "The median compensation for surgical specialities was $441,000, well aove the $264,000 that primary care earned...This large compensation gap makes careers as primary care providers less financially attractive...the supply of primary care physicians in the U.S. has been declining."

See code text at bottom.  See CAP on G2211 here.  See AAFP here.  

Code Set #3

Brand New: Principal Illness Navigation, PIN, G0023/G0024 and G0140/G0146

Unlike G2211, which has been vollied back and forth across the net for a couple years, CMS introduced PIN - Principal Illness Navigation - in Summer 2023 and finalized it November 1.   These are for staff, who are under the "general supervision" of a physician but considered an "incident to" benefit.   G0023 and G0024 are for 60 minutes of time of lower-level staff, ad add-on 30 minutes, respectively.   Since the lower-level navigator staff are under the "general supervision" of a physician, the physician must have an initiating E&M visit with the patient and while not necessarily seeing the patient monthly, provide ongoing supervision and care.   There were debates about what could be delivered in person or remotely and even a few paragraphs about the navigator work being outsourced (e.g. subcontracted) by the physician to an outside entity.   

See 88 FR 778937ff (November 16, 2023); regarding outsourcing, "PIN services provided by auxiliary personnel who are external to, and under contract with, the practitioner or their practice..." see p. 78947-8.   I am not an attorney and I refer companies to Medicare policy attorney experts for interpretation of the rule and formation of such outsourcing contracts (e.g. creating a close financial relationship but without kickback concerns).  

In another care area, the creation of AMA's "remote physiological monitoring" or the RPM and related codes, a whole sub-industry of service providers appeared quickly who could provide the home monitoring equipment to the patient as subcontractors to the physician receiving the CMS RPM code fee.   [Find just one of them here.]  I suspect both RPM and PIN fields will development their own case studies, compliance best practices, and legal experts, and perhaps OIG advisory opinions.   

Codes G0023 (+G0024) and G0140 (+G0146) are billable by physicians;/PA/NP  and by clinical psychologists, respectively.   Pricing is circa $80 and $65 for the primary and the add-on codes.   

See code text at bottom (for G0023).

###

Comment and Analysis

  • Boom, Here's New Policy.  Major new areas of policy - G2211 and the G0023/G0140 series - are being introduced in November, and will be active in January, probably before MACs have time to write any guidance articles or even LCDs on the topic.  
    • CMS transmittal November 22, 2023, to MACs, CR13452, has just a few vague sentences in comparison to the lengthy pages of CMS policy in the November 16 Fed Reg.
      • Nerdy CR. Though there's already one nerdy CMS followup, CR13272, re use of -25 modifier.
      • AAFP tries to fill the gap with "coding tips" for G2211.
      • Who Says What "Trained" Means?  For example, G0023 refers to staff who are "credentialed and trained."  What constitutes credentialing?  Bob's Credentialing?  What constitutes training?  15 minutes?  15 hours?   15 months?   CMS isn't saying, nor are MACs.  So nobody knows the answers for sure.  Will some of the answers and enforcement vary MAC to MAC - undercutting the even distribution (the social equity) of the new federal benefit?
      • Doctor Hires Incident-To Staff Out of State?  If a doctor hires [contracts] what Medicare legally calls healthcare incident-to staff, out of state, are their any implications?  The benefit doesn't seem like "medical practice," but CMS classifies it as a service "reasonable and necessary to diagnose or treat illlness," which does.    
    • That said, the older codes 99424,'26, have been around 2 years and seem to have no guidance yet on the Medicare Coverage Database.
  • Outsourcing Compliance TBD.  How the new codes G0023 etc for PIN will interface with an outsourcing industry while remaing "under general supervision and incident to the physician" remains to be seen.  
    • It seems like physicians who get paid for the G0023 service for PIN, who use outsourced navigators, would be vulnerable to the navigators' medical records quality if the physician gets audited.
    • It seems like navigators could use a lot of time in 60 and 30 minute periods.  It takes time to be acquainted with the highly complex patient before starting to coordinate, many of the tasks in the code text seem time-consuming, and don't care coordinators spend a lot of time on hold?
  • "Creation of a Benefit by Proxy." If service providers with no statutory benefit- e.g. "care navigators" or "peer care navigators" can be paid via an incident to and removte general supervision benefit, I'm at a loss to know why genetic counselors might not be paid this way.
    • (Historically, I think genetic counselors have been considered as taking on parts of the healthcare that would otherwise be a doctor's E&M benefit, but E&M can only be billed by MD/DO/PA/NP).
  • "It's a Code."  It's interesting that what look like a new form of healthcare benefit requires no action in statue, or even action in regulation, or even subregulatory policy printed by MACs.  Instead, all the annual rule acomplishes for G0023/G0140 is the creation of massively long HCPCS G-codes that themselves look and read almost like C.F.R. regulations.
    • It's unlikely doctors or coders or MACs would read the dozens of pages of Fed Reg rulemaking.
    • The $100-plus payments ofor PIN have some similarities to the $160/mo payments in the CCMI Oncology Care Model ver1 (here), which wasn't all that successful.


###
CODES

99424 (25, 26, 27)

99424
Principal care management services, for a single high-risk disease, with the following
required elements: 
  • one complex chronic condition expected to last at least 3 months,
  • and that places the patient at significant risk of hospitalization, acute exacerbation/ decompensation, functional decline, or death, 
  • the condition requires development, monitoring, or revision of disease-specific care plan, the condition requires frequent adjustments in the medication regimen, 
  • and/or the management of the condition is unusually complex due to comorbidities, ongoing communication and care coordination between relevant practitioners furnishing care; first 30 minutes provided personally by a physician or other qualified health care professional [PA NP],
  • per calendar month
  • 99425: Additional half-hours
G2211

G2211
  • Visit complexity inherent to evaluation and management associated with medical care services that serve as the continuing focal point for all needed health care services 
  • and/or with medical care services that are part of ongoing care related to a patient’s single, serious condition 
  • or a complex condition. 
  • {Add on to E&M]

G0023 (etc)


G0023
Principal illness navigation services by certified or trained auxiliary personnel under the direction of a physician or other practitioner
  • including a patient navigator
  • 60 minutes per calendar month, 
  • in the following activities:  
    • person-centered assessment, performed to better understand the individual context of the serious, high-risk condition. 
    • ++ conducting a person-centered assessment to understand the patient's life story, strengths, needs, goals, preferences, and desired outcomes, including understanding cultural and linguistic factors and including unmet sdoh needs (that are not separately billed). 
    • ++ facilitating patient-driven goal setting and establishing an action plan. 
    • ++ providing tailored support as needed to accomplish the practitioner's treatment plan.  identifying or referring patient (and caregiver or family, if applicable) to appropriate supportive services.  practitioner, home, and community-based care coordination.
    •  ++ coordinating receipt of needed services from healthcare practitioners, providers, and facilities; home- and community-based service providers; and caregiver (if applicable). 
    • ++ communication with practitioners, home-, and community-based service providers, hospitals, and skilled nursing facilities (or other health care facilities) regarding the patient's psychosocial strengths and needs, functional deficits, goals, preferences, and desired outcomes, including cultural and linguistic factors. 
    • ++ coordination of care transitions between and among health care practitioners and settings, including transitions involving referral to other clinicians; follow-up after an emergency department visit; or follow-up after discharges from hospitals, skilled nursing facilities or other health care facilities. 
    • ++ facilitating access to community-based social services (e.g., housing, utilities, transportation, food assistance) as needed to address sdoh need(s).  [social determinant of health] 
    • Health education- helping the patient contextualize health education provided by the patient's treatment team with the patient's individual needs, goals, preferences, and sdoh need(s), and educating the patient (and caregiver if applicable) on how to best participate in medical decision-making.  building patient self-advocacy skills, so that the patient can interact with members of the health care team and related community-based services (as needed), in ways that are more likely to promote personalized and effective treatment of their condition.  health care access / health system navigation. 
    • ++ helping the patient access healthcare, including identifying appropriate practitioners or providers for clinical care,  and helping secure appointments with them. 
    • ++ providing the patient with information/resources to consider participation in clinical trials or clinical research as applicable.  facilitating behavioral change as necessary for meeting diagnosis and treatment goals, including promoting patient motivation to participate in care and reach person-centered diagnosis or treatment goals.  
    • facilitating and providing social and emotional support to help the patient cope with the condition, sdoh need(s), and adjust daily routines to better meet diagnosis and treatment goals.  
    • leverage knowledge of the serious, high-risk condition and/or lived experience when applicable to provide support, mentorship, or inspiration to meet treatment goals.


 ##
AI illustration, Dalle 3, vis Chat GPT4, created by "dumping" the blog into the AI and asking for an illustration suitable for the article.  

Thursday, December 28, 2023

Natera Wins Injunction re MRD Testing (versus NeoGenomics)

MolDx has a broad policy allowing use of minimal disease detection (MRD) for cancer management, but MolDx issues coverage step by step, one test and condition at a time.  In July 2023, Neogenomics announced it had won MolDx coverage for its RaDaR molecular test (here).  The coverage is complex and fairy narrow, a signal for other companies seeking MolDx MRD coverage:

[Neogenomics:] Effective as of March 24, 2023, the RaDaR assay is now covered for fee-for-service Medicare patients within the United States with hormone receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2-) breast cancer. The coverage includes patients with a personal history of high-risk stage II/III HR+/HER2- breast cancer, five or more years from diagnosis who presently do not have evidence of disease.  NeoGenomics' tumor informed RaDaR assay has been designed to detect extremely low levels of circulating tumor DNA.

In a new December 28 headline, Natera announces it "has won a preliminary injunction" against Neogenomics for clinical sales of the MRD assay.   I believe the article at Genomeweb here is open access; it contains several links to prior stages and decisions regarding other products.  Neogenomics will appeal.  The case is related to certain tumor-informed approaches to MRD.  Some labs are offering non-tumor-informed technologies for MRD.  The Natera press release, dated 12/27, is here.  A 12/28 headline states, "NeoGenomics Stock Plunges" - here.  The "tumble" was from $20 to $17, circa $250M of market cap.

Additional Details

According to Fierce Biotech, Inivata's owners sold its assets such as RaDaR to NeoGenomics in 2021 for $390M.

Some of the NC case links are here.  One version of the complaint (26pp) is here.  Focus is on Natera '454 and '035 patents.   Details of exactly how the alledged infringement occurs are presented in attachments to the filing at para 38,  "Exhibit 3-4, Exhibits 5-16").  








Wednesday, December 27, 2023

Tempus and YouScript: Emphasizing Integration of Diagnostics, Data, Healthcare

Things have quieted done for the Christmas-New Years holiday week, but I ran across two interesting posts on Linked In.  What they have  in common is that each company is summarizing its products or clinical research to show the important of integration of diagnostic results into the care pathway.

From YouScript, a post by Kristine Ashcraft talks about a decade of YouScript research into pharmacogenetics and clinical outcomes.  She highlights the emphases in the post text, but see also a 7 page embedded PDF download.  Here:

https://www.linkedin.com/feed/update/urn:li:activity:7140501837056516097/

Similarly from Tempus AI.   See their 9 page embedded PDF about "providing physicians with the data needed to optimize patient care, seamlessly integrating with existing workflows."

https://www.linkedin.com/feed/update/urn:li:activity:7142573730865590272/


I always find the Linked In download feature a bit of a trick.   You hover over the embedded PDF, and at bottom you'll see a broken-up square.  Click on that, and you'll then find a download arrow at the top edge of the new page.

##
See a later blog about Medicare's new [oncology & other severe illness] care navigation codes, here.  I would think part of cancer care navigation is, navigating to cancer genomic tests.

Thursday, December 21, 2023

Foundation Medicine Ad Provides Consumer Portal to Test Coverage

 Noted online:  Foundation Medicine banner ad, "See who covers our CGP tests..."



Landing page:

https://info.foundationmedicine.com/coverage-counts


and leads to




FDA Approves Opioid PGx Test - But With Significant Conditions

UPDATE - PUSHBACK - APRIL 2024 - https://www.discoveriesinhealthpolicy.com/2024/04/rare-event-experts-warn-cms-and-fda.html 

##

FDA has approved the first test to help identify elevated risk of opioid addiction, the AvertD test from AutoGenomics.

That's good news.   But there's more - take a look at the FDA press release and the news reports, and there are a number of conditions.

The test requires physician education, is only to be used in the context of a clinical evaluation (e.g. nowhere near DTC), and the FDA wants postmarketing studies as well.   

Find more here:

  • FDA press release here.
  • Company press here.
    • I assume several paragraphs here, likely reflect the official indication and use labeling.
  • Reuters here.
  • Precision Medicine Online here.
  • The Hill here.
  • Update your Google search here (I think).

The project might tie in to a 2018 FDA initiative for new tests and technologies to fight the opioid crisis (here).  FDA has a 2023 guidance devices [not drugs] to treat opioid use disorder, here.

FDA PROCESS - ADVISORY BOARD 2022

The FDA notes that "an earlier version of AvertD" was discussed at a October 2022 Ad Comm.  The Ad Comm didn't seem to generate much press at the time, but is archived on YoutTube and the meeting documentation is online here.  The FDA's questions to panel are here.  The company's 79 page  document on AvertD is here; it proposed a new version of a de novo classification.  

The 18 page draft package insert is here.  I wasn't sure what to expect from the FDA's meeting minutes, but they're pretty detailed at 26 pages here.  (At some point, a PMA page will be created with a full final FDA safety and effectiveness review of 2023 for the final PMA.  As of December 21, the last posted PMA device was September 29, so the S&E write-up may take 90 days.)

The  FDA panel voted 11-2 against approval; a group that had published a negative article made public comment.  For more detail see 26-page FDA summary (or see ten-point AI summary of it here.) See an AI comparison of the Sponsor vs FDA presentations here. See a primary 2021 publication about the test here.  

See most of the FDA files in one cloud zip file here.

Noteworthy: the Ad Comm was set up against a de novo 510k; the final 2023 decision upgraded to a complicated PMA approval plan with conditions.

FDA and PMA Uncertainty

Per FDA, post-ad-comm revisions were collaborative between the company and the FDA.  The FDA discusses the pathways that would occur if there was a false negative or false positive.   The FDA specifically points to uncertainty about risks and benefits (e.g. the gap between what we know about clinical validity and what we'd like to know about clinical utility or real world evidence) - FDA on uncertainty here.

###

AI CORNER

I hadn't heard of the 22 page Uncertainty Guide, so I asked Claude.AI for a 150 word summary:

Here is a 150 word summary of the key points from the FDA guidance document on considering uncertainty in benefit-risk determinations for medical devices:

The FDA guidance discusses how the agency considers uncertainty in making benefit-risk determinations to support premarket approval (PMA), De Novo classifications, and humanitarian device exemptions (HDEs) for medical devices. It notes there is inherent uncertainty in medical product safety and effectiveness. 

The appropriate extent of uncertainty depends on factors like device type, intended use, available alternatives, and regulatory standards. More uncertainty may be acceptable for breakthrough devices addressing unmet needs or devices for small patient populations, with robust postmarket data collection to address uncertainty. 

The guidance provides examples of how greater uncertainty could allow smaller premarket studies but require more postmarket evidence. Overall, marketing authorization depends on meeting statutory standards based on total valid evidence, considering uncertainty and using controls like labeling and postmarket surveillance when appropriate.

##

Business Holdings.  Fairly complicated; Autogenomics in Carlsbad CA was acquired in 2019 by Prescient Medicine Holdings, whereas the former company called Prescient Medicine Inc was renamed Solvd Health.  Here.  The various entities (Autogenomics, Prescient Holdings, and Solvd) all appear to be privately held. One public source available on Google suggests at one point Solvd had raised $120M.

AI Corner 2

You've heard of "prompt engineering" - the art of "talking to AI."  
Here's the dialog with AI that led to the picture above.

Wednesday, December 20, 2023

CAP TODAY: Continues Rising Coverage of Digital Pathology

CAP TODAY, which is open access, has published a series of detailed articles on digital pathology and AI and lab medicine in 2024.  They're closing out the year with an interview panel on Digital Pathology.

See a Linked In entry point by Katie Maloney here, and the CAP TODAY article here.  (Note that CAP TODAY splits the article in 2 parts, both of which are accessed by the "create pdf" button.)

 The interview is led by publisher Bob McGonnagle, talking with Katie Gillette of Decibio, M.E. de Baca (VP, Sysmex), Michael Quick (Dig Pathol Assoc, DPA), Lisa-Jean Clifford (Gestalt Dx)and Eric Glassy (Affil Pathol Med Group).   

AI Corner

I asked ChatGPT4 what it saw as ten top takeaways from the article.

ChatGPT: "The document, "2023 CAP TODAY 1201 Decibio Gillette Dig Pathology," contains a series of expert interviews discussing the interplay of digital pathology and AI in healthcare. The key takeaways from these interviews are:

  1. Digital Pathology Adoption: Adoption varies widely across institutions, from minimal use to fully digitized workflows.
  2. Role of Academic Centers: Leading centers are keen on adopting state-of-the-art digital pathology tools and contributing to their development.
  3. Perceptions of Digital Pathology: There's a disparity in how digital pathology is perceived across different pathology meetings and associations.
  4. Drivers of Adoption: Early drivers included needs for tumor board, education, and research. Now, pathologist shortages and remote sign-out capabilities are key factors.
  5. Pathologist's Job Security: Concerns about job loss due to digital pathology and AI are diminishing as the field recognizes AI's role in augmenting, not replacing, pathologists.
  6. Impact on Workforce: Similar to the evolution in cytology, digital pathology and AI are expected to transform, not diminish, the roles of pathologists and cytotechnologists.
  7. Budgeting and Financial Constraints: Budgeting remains a challenge, especially in determining the financial equation for adopting digital pathology.
  8. Regulatory and Reimbursement Landscape: The environment is evolving, with new CPT codes and an increasing focus on data interoperability and patient safety.
  9. Patient Involvement and AI: The role of patients in driving technology adoption and the necessity of pathologist involvement in AI development are highlighted.
  10. Future Trends and Concerns: Anticipating future needs and challenges, such as regulatory approvals and the impact of AI on clinical practice, is crucial.

###

What are the five main points made about Reimbursement and Dig Path?

ChatGPT: The five main points made about the reimbursement topic in the "2023 CAP TODAY 1201 Decibio Gillette Dig Pathology" interviews are:

  1. Implementation of Category Three CPT Codes: The CAP and American Medical Association have added over 30 category three CPT codes for digital pathology, and they are now working on AI codes. These codes are essential for CMS (Centers for Medicare & Medicaid Services) to gather data to inform the need for reimbursement for new services​​.
  2. Challenges in Budgeting: Laboratories struggle with budgeting for digital pathology due to the complexity of financial equations. Large labs can leverage economies of scale, while smaller hospitals find value in remote consultations and addressing pathologist shortages. Academic centers focus more on research opportunities than the direct financial implications, as direct reimbursement is not yet fully established​​.
  3. Need for Effective Use of Current Billing Codes: There's an emphasis on the importance of properly utilizing existing Category Three codes for digital pathology. This ensures that their intended purpose for tracking and reimbursement is realized. Without effective use, these efforts might not be fully recognized or compensated​​.
  4. Shift in Attitude Towards Technology Adoption: Initially, proposals for lab automation and digital pathology were rejected due to cost concerns and lack of direct reimbursement. However, as the demand and workload increased, the same institutions that initially rejected these technologies sought to implement them, recognizing their value in improving efficiency and managing increasing volumes​​.
  5. Reimbursement as a Key Factor for Patient Involvement: For patients to be involved in digital pathology, three conditions need to be met: physician access to information, availability of technology in the lab, and insurance or payer reimbursement. The success of direct-to-consumer approaches in areas where these conditions are met indicates the significance of reimbursement in advancing digital pathology​​.

These points highlight the complexity and importance of the reimbursement landscape in the adoption and implementation of digital pathology.

###

BQ: There are a couple problems with the Category III codes.  (1) They are structured as add-on (+) codes, so they must be on the same claim same day as the original surg path (or staining) codes.  (2) A huge proportion of relevant specimens originate in hospital outpatient (or inpatient) environments, where CMS has made the codes nonpayable (bundled or packaged).

###

BQ: See also a 2023 article, open access, by Lennerz et al, on diagnostic quality and a framework for adding AI-ML to it.  

###

click to enlarge

###

BQ: On a humorous note, I recently asked ChatGPT to write a health policy news article about a digital pathology conference that will occur in the future, in March 2024 - here.

Very Brief Blog: Transitions: Dr Joe Lennerz Joins BostonGene

 Harvard molecular pathologist, policy expert, and prolific author Dr Joe Lennerz has left his position at MGH Brigham to join BostonGene as its Chief Scientific Officer.   See a recent open-access policy article from his group, here.

For more, see his postings on Linked In here:

https://www.linkedin.com/in/joe-lennerz-90029219/

Find the BostonGene website here:

https://bostongene.com/about-us/

BostonGene announced $150M in new funding eighteen months ago - here.


Tuesday, December 19, 2023

FDA Issues Updated Real World Evidence Guidance [Also 3rd Party Review] (Drafts for Comment)

FDA has an existing guidance document for real world evidence (RWE) but now they've issued an update for public comment.  It's directed at medical device (or potentially diagnostics) studies.

Find the FDA web page, draft document, and Fed Reg link here:

https://www.fda.gov/regulatory-information/search-fda-guidance-documents/draft-use-real-world-evidence-support-regulatory-decision-making-medical-devices

Comments to 2/20/2024.

  • Early coverage at MedTechDive here.
  • In September 2023, we saw final new RWE guidance for biopharma - here.
  • Also in September 2023, we saw draft guidance for "clinical data in 510(k) applications," here.
  • A subscription-only article at BioWorld here.
WHEN IT RAINS IT POURS
FDA also issues draft guidance on third party review of 510k and third party review of emergency (EUA) submissions. Here.

click to enlarge


Monday, December 18, 2023

HHS OIG Releases Consolidated Guide to Compliance (Nov 2023)

 In November 2023, OIG at HHS released a "general compliance program guidance," attempting to knit together the helter-skelter rules that govern compliance.

Read about it at PolicyMed.com here:

https://www.policymed.com/2023/12/hhs-oig-releases-general-compliance-program-guidance.htm

Find the OIG Home Page for the project here:

https://oig.hhs.gov/compliance/general-compliance-program-guidance/

If you download "complete guidance," you get a 91 page PDF:

https://oig.hhs.gov/documents/compliance-guidance/1135/HHS-OIG-GCPG-2023.pdf


Here's an AI summary (GPT4):

The 2023 OIG HHS General Compliance Program Guidance document is a comprehensive resource for healthcare stakeholders, outlining compliance practices and legal standards. It covers federal laws like the Anti-Kickback Statute, False Claims Act, and HIPAA, emphasizing the importance of compliance infrastructure including written policies, leadership, training, and effective communication. 

The guide addresses issues for both small and large entities, offers insights into quality and patient safety, financial incentives, and financial arrangement tracking. It also details OIG resources and processes, providing a valuable toolkit for navigating health care compliance and legal landscapes.

Key Points:
  1. Offers comprehensive compliance guidelines for healthcare stakeholders.
  2. Covers key federal laws like Anti-Kickback Statute and False Claims Act.
  3. Emphasizes the importance of an effective compliance program infrastructure.
  4. Provides specific guidance for both small and large healthcare entities.
  5. Addresses quality and patient safety, and financial incentives in healthcare.
  6. Highlights the significance of tracking financial arrangements.
  7. Includes detailed sections on OIG resources and compliance processes.
  8. Serves as a toolkit for understanding healthcare compliance and legal requirements.
  9. Updated to reflect modern healthcare industry challenges and practices.
  10. Aims to aid in preventing fraud, waste, and abuse in healthcare systems.

The document doesn't have anything specific to genetics/genomics, and mentions the word "laboratories" only once or twice.  For example, guidance on how a lab can avoid OIG penalties around the 14 day rule or colon screening gift cards, are not here.

OIG Releases 48-Page Guide to Saving Govt Money in Medicare Advantage Plans

The OIG has released a 48-page guide to "improper payment reduction" for Medicare Advantage plans.   The OIG summarizes the project this way:

This toolkit offers Medicare Advantage (MA) organizations information that will enable them to replicate Office of Inspector General's (OIG's) techniques to identify and evaluate high-risk diagnosis codes to ensure proper payments and provide better care for enrollees. This toolkit is meant to be a practical, hands-on device that will help MA organizations improve the accuracy of their submitted diagnoses that are at high risk for being miscoded.

Find the project home page here:

https://oig.hhs.gov/oas/reports/region7/72301213.asp

Find the 48 page PDF here:

https://oig.hhs.gov/oas/reports/region7/72301213.pdf


Here's some further background:


  • The 2023 OIG Toolkit aims to decrease improper payments in Medicare Advantage by identifying and addressing high-risk diagnosis codes. 
  • It offers practical techniques for Medicare Advantage organizations to ensure accurate payments and improve enrollee care. 
  • The toolkit is based on OIG's audits, which found significant miscoding of diagnosis codes, impacting around 70% of submissions. It includes programming codes and methodologies used in audits, enabling organizations to replicate OIG's techniques for identifying miscoded diagnoses and enhancing compliance and accuracy in Medicare billing.

The tool kit doesn't have sections that are lab (or genetics-genomics) specific.  However, the OIG does state this:

"Although MA organizations make their own payment arrangements with providers for theseservices, CMS requires MA organizations to submit copies of all claims to CMS. These claims include services that can be used for risk adjustment purposes (physician, outpatient, and inpatient) as well as claims whose services are not used for risk adjustment purposes (such as home health services, skilled nursing facility services, durable medical equipment, laboratory services, and x-rays, among other services). We have developed computerized programs to access and analyze these claims data. " (p.5)

Sunday, December 17, 2023

HHS Has Produced TWO Very Bad Cost-Benefit Analyses: (1) For FDA LDT Rule, (2) for HHS ONC EHR-AI Rule

This year, HHS has put out two major and broad rules that impact innovative medical technology.  In both cases, the published cost-benefit analysis was a mess.   The two rules are an FDA proposal to regulate lab developed tests (proposed 9/2023), and an HHS ONC rule on health system information systems and AI (proposed 4/2023, final 12/2023).  Both rules divided stakeholders.  The lack of an ability to propose credible cost-benefit data in favor of either rule, which is the duty of proponents and advocates, is noteworthy.

FDA - Terrible Cost-Benefit Discussion

Most familiar to readers here is the FDA LDT rule, in proposal, commens closed December 4.   Among other problems, the rule was extremely vague in some key aspects, and had a god-awful cost benefit analysis.   I wrote about that early in the fall in a white paper.   Among many commenters, ACLA emphasized the awful cost-benefit analysis with a 25-page economist's appendix to their comment letter. (See an AI tutorial on Prof. Carrigan's report, here.)

HHS ONC - EHR AI Rule Is Impactful,
Suffers From Another Terrible Cost-Benefit Mess


HHS ONC releases EHR-AI Rule ("HTI-1")

I regret I didn't stop and pay enough attention to an HHS (ONC) proposal to upgrade their regulation of EHR's and AI, issued in proposal form last April.   See an outstanding white paper on the ONC proposal issued in June by McDermott Will & Emery, here.    Turning to the new final rule, see Fierce Healthcare here, ONC home page for the rule HTI-1 here.

Others were paying attention - see descriptions of major comment letters from the likes of Amazon, Google (here at Becker's Health It, here at Stat).  Each was just 5pp.  (For an AI analysis of the 2 letters, here.)  (UHG had a vehement, longer letter here.)

The final rule appeared in Fed Reg on 1/9/2024 - 89FR1192 (247pp).  Here.

HHS ONC Cost-Benefit Discussion- A Tissue of Confusions

Having been appalled from A to Z by the cost-benefit analysis in the FDA LDT proposal, I looked at the cost-benefit analysis in the HHS ONC proposal.   

The ONC proposal matches the final, because, in contrast to the barrage of criticism on this topic at FDA-LDT comment period, there was NO COMMENT to most of the equally awful cost-benefit analysis served up by the writers at HHS ONC.  My head was spinning.

They make predictions of benefits for sepsis management of $432M (over ten years when there are $200B of sepsis admissions) and $600M of saved general admissions (in a ten year period with a base cost of $240B of a type of admission) (proposal table 19).   (See discussion at 88 FR 23889 ff).  Software costs range up to $335M (proposal, table 18).  When they get to summary figures (proposal 88 FR 23903) it's $742M for total annual IT costs and net annual benefits the suspiciously round number, "1 billion," allowing the whole thing to pass the test for "unfunded mandates" (88 FR 23905).  

Of course, if hospitals (with passed-on IT developer costs) incur $0.7B in expenses and have $1B less in DRG revenue, they are $1.7B down (in the red), not $300M ahead (in the black).      

What Staff?

ONC also makes unusual assumptions about the labor require, as either management analysts (about $45 per hour before overhead) and software analysts (about $55 per hour). But the rules involve new and extremely complex conceptual medical risk-benefit reports and other reports on every piece of software (see McDermott white paper).  The work of doing this is unlikely to be within the grasp of an entry-level programmer or analyst.  (Plus, AI software experts are hardly $55 an hour!)

Overall, I was surprised that the important cost-benefit section of the ONC EHR AI rule attracted few to no comments.  If you don't know the benefits exceed the costs...?

Another Kinship: Statutory Overlap is Underdeveloped:  FDA-CLIA vs ONC-FDA

Another topic in itself.  The FDA LDT proposed rule collides with CLIA in many ways that the FDA left wholly for future "filling in the details."   

Same deal here, but in a final rule.  The ONC EHR AI rule, appears to have many points of collision with FDA regulation of medical software.   Writers complain about this (Final Rule, Inspection Copy, 200ff).  However, ONC handles these concerns and some vague inline sentences and a couple uncommented crosslinks to FDA documents.

Updates Sexual Orientation and Gender Identity

The final rules update on sexual orientation and gender identify in several contexts (45 CFR 170.207).

(o) Sexual orientation and gender information

(1) Standard. Sexual orientation must be coded in accordance with, at a minimum, the version of SNOMED–CT® codes specified in paragraph (a)(4) of this section for paragraphs (o)(1)(i) through (iii) of this section and HL7 Version 3 Standard, Value Sets for Administrative Gender and NullFlavor (incorporated by reference, see § 170.299), up until the adoption of this standard expires on January 1, 2026, for paragraphs (o)(1)(iv) through (vi) of this section, attributed as follows: 

(i) Lesbian, gay or homosexual. 38628009
(ii) Straight or heterosexual. 20430005
(iii) Bisexual. 42035005
(iv) Something else, please describe. nullFlavor OTH
(v) Don’t know. nullFlavor UNK
(vi) Choose not to disclose. nullFlavor ASKU 

(2) Standard. Gender identity must be coded in accordance with, at a minimum, the version of SNOMED–CT® codes specified in paragraph (a)(4) of this section for paragraphs (o)(2)(i) through (v) of this section and HL7 Version 3 Standard, Value Sets for AdministrativeGender and NullFlavor (incorporated by reference in § 170.299), up until the adoption of this standard expires January 1, 2026, for paragraphs (o)(2)(vi) and (vii) of this section, attributed as follows: 

(i) Male. 446151000124109
(ii) Female. 446141000124107
(iii) Female-to-Male (FTM)/Transgender Male/Trans Man.
(iv) Male-to-Female (MTF)/ Transgender Female/Trans Woman. 407376001
(v) Genderqueer, neither exclusively male nor female. 446131000124102
(vi) Additional gender category or other, please specify. nullFlavor OTH
(vii) Choose not to disclose. nullFlavor ASKU 

(3) Standard. Sexual Orientation and Gender Identity must be coded in accordance with, at a minimum, the version of SNOMED CT® codes specified in § 170.207(a)(1). 

(4) Standard. Pronouns must be coded in accordance with, at a minimum, the version of LOINC codes specified in 170.207(c)(1). 


##

I posted in a sidebar, an initial back-and-forth AI discussion of the ONC EHR AI rule - here.

28 companies signed a general commitment to develop AI responsibley; Reuters Dec 14 here

 

Excellent Article about Alz Diagnostics - And Some Additional Points

Mara Aspinall, a Partner at Illumina Ventures and a Professor in Biomedical Diagnostics at ASU, edits and writes "Sensitive and Specific: The Testing Newsletter," available at Substack.

[By the way, Substack has an interesting business model.]

Over at S&S, Aspinall and Ruark have just published, first, a special report on Alzheimer diagnostics, and second, a broader special reporton neurologic diagnostics.  Find them here:

https://sensitiveandspecific.substack.com/p/special-report-alzheimers-disease

https://sensitiveandspecific.substack.com/p/neurologic-disease-diagnostics-special

Not mentioned, newly clinically (commercially) available diagnostics for alpha-synuclein by special protein aggregation assays - see Siderow in Lancet here, 1100 patients, and see the lab Amprion here.

On the side of Alzheimer diagnostics, they quote Mattke e al. that the vast majority of minimal cognitive impairment MCI cases are undiagnosed:



The first Alzheimer tests, a decade ago (Amyvid), were PET scan tests not paid for by Medicare (until very recently).  More recently, FDA CSF tests have been approved.   However, patients show a drop-off in willingness to take CSF tests (and CMS bundles the costs of proteomic testing of CSF tests and the costs of special PET tracers, when they are taken in a hospital outpatient setting, where they often are).  


Aspinall and Ruark remark, "The first potentially curative treatments for Alzheimer’s disease were approved this year. These developments have dramatically increased interest in developing diagnostics that are less invasive than a CSF spinal tap, cheaper than imaging, and more accurate than either."

The Accuracy of Alzheimer Tests - A Fierce Problem

Alzheimer tests are one of the worst examples of spectrum problems in diagnostics, a topic that's been discussed for decades.   Since the 1990s, researchers have regularly touted accurate new Alzheimer tests, but by testing the "tails" of the population.  This means 25 perfect healthy patients (75 years old, working at a bank, doing their own taxes, and playing tennis every weekend) with 25 severe Alzheimer patients (several years in a nursing home, profound dementia.)  These two populations are easy to tell apart.   But they aren't real test candidates, who would be the 72 year old who has mild memory concerns and a little confusion when driving.  Not only are those patients less studied in pilot tests, proof of concept tests, but there's no gold standard for their real diagnosis, either.

So  "spectrum effect" is a problem, and often makes a test look better than it is.

The second problem is summary statistics.  If we're told Text X is 90% sensitive and 90% specific, that's misleadingly useless unless you know the population tested, which you probably don't.

The third problem is a test is not a test.   Let's say you have a test for phospho-Tau at position 217 (p217tau).   You can't say "Such tests are 90% sensitive and 90% specific."  Different tests for p27tau could have quite different results, depending on the accuracy of the monoclonal used in that test.  And it's probably two monoclonals in an immunoassay, a binding one and a detecting one, both of which may different from one company to another.  These changes can push the data around quite a bit, so it's a fallacy to conclude something like ("p200tau is better than p300tau by 8%.")  

So, in conclusion, if you hear a particular Alzheimer biomarker is summarized in a review as "90% sensitive and 90% specific" you know nothing til you know something about the population tested and the particular monoclonals used.





Thursday, December 14, 2023

Very Brief Blog: CMS Publishes 2024 CLFS Transmittal; New Crosswalks All Listed

 Just before Thanksgiving, CMS published 84 final prices (or gapfill status) for new 2024 lab CPT codes.

Now, CMS has published the official transmittal with all the CY2024 crosswalks in one place.  CR13467.  This is worth keeping track of, because it's the last time CMS publishes this data.  

  •  It's important to know because if code "B" is crosswalked now to code "A," code B will move up and down for several years tied to the price of Code "A".  (For example, if Code "A" becomes an ADLT code, or is changed and repriced, or is deleted, if A goes through regular PAMA triennial repricing that B isn't involved in because B is too new.)
  • There is no single rolling listing of all the currently crosswalked codes, so you have to look through all of these annual transmittals - as PDFs - one by one to find out whether a code of your interest is crosswalked to anything.
    • It would be nice if CMS published a spreadsheet of all currently crosswalked codes, rolling up each year's additions into one spreadsheet.

https://www.cms.gov/files/document/r12389cp.pdf



Monday, December 11, 2023

AI Corner: AI Reports on Digital Pathology Conference That Hasn't Happened Yet

Next March, I'm excited to be chairing a session and appearing on a panel at the annual TRI-CON conference, which will be held in San Diego March 26-28, 2024.  Find it here:

https://www.triconference.com/


I noticed one of the tracks is, "ARTIFICIAL INTELLIGENCE IN DIAGNOSTICS."  I fed the agenda and presentation summaries into Chat GPT4 and asked for a journalist-style discussion of the conference, as if the conference had already occured.

###

Artificial Intelligence in Precision Oncology: 
A Transformative Leap in Cancer Care

By [Name], Health Technology Journalist

The recent conference on "Artificial Intelligence in Precision Oncology" has concluded, leaving attendees, including myself, buzzing with excitement about the potential AI holds in transforming cancer care. The event showcased groundbreaking innovations, thoughtful discussions, and a shared vision for a future where AI and healthcare converge to offer unprecedented precision in oncology.

Innovations at the Forefront

Douglas Flora, MD, set the tone with an insightful opening, emphasizing the need for AI in overcoming the challenges faced in oncology. His vision was echoed in Ben Freeberg's presentation, which focused on using AI to innovate and commercialize in oncology. Freeberg highlighted the unsustainable rise in cancer care costs and how AI could be a game-changer in this landscape.

One of the most compelling talks was by Eric Stahlberg, PhD, who delved into the world of digital twins and virtual models. His presentation painted a future where diagnostics, virtual models, and AI algorithms merge, offering new avenues for optimal treatment in precision medicine.

Paving the Way for Integration

The next day kicked off with Stephen T. C. Wong, PhD, who brought to light the challenges of integrating AI into the complex healthcare ecosystem. This was further explored by Sonya Makhni, MD, who emphasized the necessity of a novel approach to solution evaluation and bias assessment in AI models.

An intriguing approach was presented by Yuan Luo, PhD, who talked about using deep reinforcement learning for cost-effective medical diagnosis. Luo's work in applying reinforcement learning to optimize lab test panel selection could lead to significant cost reductions in clinical tasks.

The conference also shined a spotlight on cognitive automation, with Wong demonstrating how AI could surpass clinicians' performance in tasks like breast cancer overdiagnosis and stroke diagnosis in emergency settings.

Future Directions and Regulatory Considerations

The final day, chaired by Jithesh Veetil, PhD, focused on digital pathology. Veetil's presentation on the role of public-private partnerships highlighted how collaboration could fast-track the adoption of AI in healthcare.

The panel discussion on digital pathology and AI, moderated by Veetil, was particularly enlightening. It brought industry and regulatory leaders together to discuss the progress and hurdles in implementing digital pathology and AI/machine learning. The session underscored the importance of recent regulatory and legislative developments in shaping the future of diagnostics.

Final Thoughts

This conference was not just about the technicalities of AI in oncology but also about envisioning a future where cancer treatment is more personalized, cost-effective, and efficient. It's a future where AI doesn't replace clinicians but works alongside them to enhance precision in cancer care. The enthusiastic participation of leaders in the field and the innovative ideas presented make it clear that this is not a distant dream but an imminent reality. As a health technology journalist, I am excited to see how these advancements unfold in the coming years, bringing hope and improved care to cancer patients worldwide.

###

Here is an "article" written by Claude.AI.

Artificial Intelligence -
Poised to Revolutionize Cancer Care

I attended the lively AI in Precision Oncology conference this week, which showcased some of the most promising innovations in leveraging AI to improve cancer diagnosis, treatment, and outcomes. As cancer rates and costs continue rising at an alarming rate, there is growing urgency around finding new solutions. Based on the enthusiasm and optimism of presenters and attendees, AI-enabled technologies seem poised to drive a sea change in oncology care.

Several presenters offered compelling examples of how AI tools can support clinicians, reduce costs, and improve patient outcomes all through better use of data. Eric Stahlberg’s presentation on “digital twins” particularly piqued my interest. These virtual computer models of patients, built by aggregating their clinical, molecular, and imaging data, can enable researchers to simulate personalized treatments and predict outcomes. This has enormous potential to get patients on optimal therapies faster while sparing them from unnecessary or overly toxic treatments.

I was also intrigued by the presentation on using reinforcement learning algorithms to optimize diagnostic testing. By continually evaluating which tests provide the best balance of cost and added diagnostic value for a patient, the AI system learned to diagnose conditions accurately while greatly reducing costs. The ability to curb excessive testing offers hope for alleviating economic strains on patients and health systems.

However, as Sonya Makhni cautioned in her talk on responsible AI deployment, unleashing the power of these innovations requires much more than just developing accurate algorithms. Thoughtfully integrating them into complex clinical workflows poses major challenges that could undermine their success if not adequately addressed. Her proposals around novel frameworks for risk assessment and monitoring offer wise guidance for those striving to implement AI tools.

Collaboration is Key

Of course, the 18 experts on the panel discussions repeatedly drove home that partnerships, collaboration, and open channels between stakeholders across specialties will be instrumental in realizing AI’s potential in oncology. Public-private initiatives like those led by MDIC, intended to align interests and amalgamate precious data resources, represent the kinds of joint efforts that will clear the way for AI adoption.

Concluding Notes

If the enthusiasm I witnessed at this conference is any indication, all signs point to AI rapidly permeating virtually every facet of cancer care in the near future. This technological revolution cannot unfold quickly enough for the patients and providers hungry for better solutions. Judging by the fruitful exchanges and ambitious visions on display at the event, many innovators are already racing to unleash AI’s radiant possibilities. I came away confident that this extrahuman intelligence will soon be saving countless lives.



See a Lewis Black version here.

Scott Gottlieb: FDA Anecdotes from Trump Administration

Dr. Scott Gottlieb was the 23rd FDA Commissioner, from May 2017 to April 2019.   He was the first keynote speaker at the Personalized Medicine Coalition's annual meeting on November 30, 2023.   His discussion recounted two anecotes about the FDA under President Trump.

CAR-T in the Oval Office

Gottlieb's tenure included the first CAR-T approval in August 2017.   He and then-Secretary Dr. Tom Price were debriefing the President on health policy issues a few weeks ahead of that.   Dr. Price gave Dr. Gottlieb the floor, and Gottlieb described the imminent approval and Trump listened closely and with enthusiasm.   ("He like wins.")   Leaving the Oval Office, Gottlieb grew concerned that the President might prematurely tweet the news, and asked Jared Kushner, passing by, to be sure and tell him not to do so.   As he was leaving, Gottlieb caught a glimpse of Trump making a motion "I know, I know."

A few weeks later, Gottlieb got a couriered envelope from the White House, hand-delivered to the FDA, which was a favorable Washington Post op ed. And a handwritten note from the President, "Great job!"

##

VALID and More: An Agile HHS

Gottlieb also discussed the recent FDA LDT regulatory proposal, which he felt was not well crafted.  Gottlieb talked about the several years of work between FDA and Hill to  craft and revise the VALID Act (currently stalled on the Hill.)   Gottlieb noted that in the early Trump years there may have been more degrees of flexibility and pivoting in negotiating things like that, with less tiers of burocracy (OMB, levels within HHS, etc) for every sign-off.   Gottlieb surmised that the layers of controls and gates and sign-offs are probably back in place now, making it harder to craft creative solutions (like VALID) into regulations or statutes.

##

Note 1
The video of Gottlieb's whole session is here:

Note 2
For 2024, Gottlieb is finishing a book on CAR-T and cellular therapies.

Note 3
See my long blog in August 2019 about the Trump adminsistration's roll-out of new kidney transplant initiatives (includes citation to Adam Boehler at CMMI and Abe Sutton in WH.)