Friday, September 21, 2018

Very Brief Blog: CMS Posts Its Preliminary Determinations for Codes under Lab Crosswalk Process

This year, CMS is providing "crosswalk" prices for about 100 laboratory CPT codes, including brand-new CPT codes for 2019 and new PLA codes that met the June 2018 deadline.

On September 21, 2018, CMS posted its proposed prices for public comment.

About 18 codes under the summer gapfill comment and revision process have not been posted yet.

Sometimes in the past CMS has given pretty telegraphic reasons for its choices.  For example, if different stakeholders suggest three codes, CMS might say it has chosen one of the codes "because it is more appropriate."  Well, yes.   This year, CMS gives an introductory explanation of some of its concerns with the range of pricing suggestions.   See the CMS document, but I summarize as:
  1. Multipliers.  Stakeholders suggest multipliers, which is allowed, but rationales may be scanty or inconsistent.  Generally, CMS prefers a crosswalk without any multiplier.
  2. Stacking CDLT codes.  PLA tests may seem to bundle multiple existing codes.  Since economies of scale may be uncertain, CMS will tend to recommend gapfill pricing.
  3. Lots of genes.  Since only one large scale code is available (81455, 51+ genes), CMS prefers to gapfill these codes.
  4. New BRCA codes.  There are a mix of new, revised, or deleted BRCA codes.  Tier 2 MoPath codes were used when possible.
In last year's exercise, in proposed fall pricing I believe CMS rarely recommended gapfill, but switch to gapfill for some codes after fall comments.  Here, CMS seems to take the opposite starting position, to recommend gapfill in September more liberally (I believe on 14 codes, agenda numbers 6, 7, 9, 11, 17, 18, 22, 23, 35, 37, 40, 42-44).

CMS accepts comment until October 22 at an email described in the PDF.

Very Brief Blog: OIG Releases 4th Annual CMS Lab Payments Report

PAMA created a new pricing system for CMS labs in 2014, and also a requirement for annual OIG reports.   OIG has issued its fourth annual report, covering both Part B independent lab payments and also Hospital Outpatient Reference Lab payments, for CY2017, in September 2017.
  • See the OIG 4th Annual Report here.
    • Appendix A provides links to prior OIG reports.
  • See coverage at 360DX here.
1% of labs received 55% of Medicare payments for the top 25 tests.

In this blog, I summarized data from Part B CY2017, and also provided links to the CMS spreadsheets.  My article focuses on the top 10 molecular codes.  That article is here.

By code, the highest paid codes were 84443 (TSH), $484M, blood panel 80053, $473M, CBC 85025, $432M, lipds 80061, $415M, and Vitamin D 82306, $348M.   Next up were 22 or more drug tests, G0483, $307M, and A1c 83036, $257M.  The highest molecular test, Cologuard, 81528, came in #12 overall at $117M.

Comparing data from the 2017 CMS spreadsheets for regular Part B independent labs and the OIG summary data (which includes hospital reference labs) allows the hospital lab billing to be imputed at least for the top 25 codes for which OIG provides data.

The OIG report provides summary data only on the top 25 CPT codes.  Obviously, they could easily provide a cloud link to a table or spreadsheet with all the codes.  

Per the report's Exhibit 2, 1% of labs got 55% of payments; 5% of labs got 79% of payments.  The remaining 95% of labs got 21% of payments. 

Thursday, September 20, 2018

Very Brief Blog: NYT Says that Sloan Kettering Pathology/Venture/Startup Deal "Ignites Uproar"

In a article Friday September 21, 2018, New York Times writes that a deal between Memorial Sloan Kettering, its pathology department, and a spinout venture capital funded company has "ignited a new uproar."

According to the Times, MSK's spinout, Paige.AI, received $25M in new venture funding earlier this year and in return for a 9% institutional equity stake, has exclusive rights to findings from a 25 million-patient slide archive.  According to the times, the archive represents not only patient tissues but "decades of work by its world-renowned pathologists." 

The article describes various facets of the project while stating that: "Experts in nonprofit law and corporate governance questioned whether Memorial Sloan Kettering, which is a charity, acted properly in what is known as a related party transaction..."  The concern seems to be that decision-markers in greenlighting the project were also stakeholders in the project itself.

See the article here.   The Times describes this as "a new uproar" because in the last few days the chief medical officer resigned after concerns about publications that lacked full conflict of interest disclosures - here.

The Times mentioned Paige.AI in a September 14 editorial, but only in passing.  For a February 2018 article about the project launch, TechnoCrunch here.

The Times article today notes that the prior chairman of Pathology (2003-2011) was critical of the deal made by the current chairman (2013-present).

Monday, September 17, 2018

CMS Released CY2017 Claims Data by State

CMS posts annual public Part B data three different ways.

  • CMS posts national Excel files for all CPT codes with utilization and dollars paid (here).   
  • It posts state-level Excel files by MAC subcontract.
    • E.g.. the Noridian subcontract for Utah is one file, the Noridian subcontract for Arizona is another file).  
    • See the state level 2017 data here.  
  • Later - probably around April 2019 - CMS will post physician & lab level claims data for 2017.  
    • That will appear here.   

The state level data appears on the linked webpage as three zip files with a dozen or so Excel spreadsheets each.   Each spreadsheet is named by a Part B MAC contracting zone (usually a state).  You need to read the "Part B Carrier Readme File" pdf on the same webpage to uncode the contractor zones that label the spreadsheets.

I've all put a single zip file of all 50-plus spreadsheets plus the PDF index into the cloud in one zip file here.

State Level MoPath Data

While it's often low utility, tedious, or both, to look at state level data, a few examples.

We know from the national file (link above) that CMS paid $115,586,444 nationally for MoPath unlisted code 81479 in CY2017.   We see from these state files that $44,201,329 was paid in SoCal and $14,533,690 in NCal, totalling $58,735,019 for 81479 or 50.8% of the national spending on 81479.

Utah (03502) has historically led the nation in CMS BRCA payments.  Utah garnered 2017 payments of $30,938,308 for 12,360 uses of BRCA consolidated code (Seq, Dup Del) 81162.   This is relative to nationwide payments of $52,484,618 for 21,018 uses of 81162.    In Utah there were 500 payments for BRCA codes 81211 and 1,244 payments for BRCA code 81213 (respectively $1,044,115 and $728,872).

Saturday, September 15, 2018

CMS Released CY2017 Part B Utilization Data by CPT Code; Including Molecular Codes

A little earlier than some years, CMS released the national Part B fee for service data for CY2017.   See the data website here.

In this blog, I'll focus on molecular CPT code data, with some brief references to data from CY2016 made available last November (here).

As a baseline from the prior year, for code range 81162-81599, I tallied $479,894,633 in "payments allowed" in 2016, which was about the same as in 2015.  (Sharp, $100M-level cutbacks in pharmacogenetic spending was enough to level the annual year-to-year total of all MoPath payments for a while.)

  • Update: Separately, OIG published its fourth annual review of CMS lab payments, which includes hospital outpatient reference labs not included in the Part B data.  Since it includes outpatient reference labs its numbers are always a bit higher.   Here.

Total MoPath Spending Rises, $480M to $644M

For CY2017, the new data, for molecular pathology I tally $644,228,102, a 34% increase year-on-year. 

For all lab spending in this code series, $6.9B.  For example, chem panel code 80053 had 29M uses for $322M dollars at $11 a test.  Net-net, MoPath is circa 10% of lab spending in 2017 data.

10 Codes Get 73% of Spending

As in recent years, MoPath spending is highly concentrated.  Although CMS lists about 175 codes, the top 10 codes get 73% of all spending.  At $118M, Cologuard was 18% of all spending, about tied with the unlisted code 81479.  Oncotype Dx and BRCA 81162 followed with 10% and 8% of spending, respectively.  See table below.

CY2017 Top Ten MoPath Codes (Total Spend $644M)

click to enlarge - Dot Plot of 2017 Allowed Payments by Code

Cologuard Zooms

In CY2016 CMS data, Cologuard was only $62M, on par with BRCA and Oncotype Dx.  In CY2017, it was almost $120M.   Correspondingly:  At end 2016, Exact Science's stock price was about $19, at end 2017, about $50.  (Today it's $76, with a market cap of $9B).

Note that 2016 to 2017 spending rose from $480M to $644M, or +$164M, of which about $60M was accounted for by the growth in Cologuard alone.

BRCA Coding Shifts

BRCA spending totaled $75M, with $52M now being spent under consolidated BRCA code 81162, which pays a few hundred dollars less than the legacy stack coding approach to BRCA (81211+81213), which garnered only about $9M in CY2017.   BRCA panel codes 81432/33, received about $13M for about 12,000 cases.   The shift to 81162 coding impacts CMS spending in 2018-2020, since (legacy 81211+81213) pricing is stable during this period while 81162 pricing drops 10% per year under PAMA.

There is a major revamping of the BRCA code set for CY2019, but coding to services like BRCA1 alone or BRCA2 alone are next to nothing, at least in CMS data.

click to enlarge

Unlisted Code 81479

The second highest code was the Unlisted Code for Mopath, 81479.  At least as of a year ago, nearly all 81479 spending came through MolDx MACs, and almost none elsewhere.  CMS has now asked MACs to report use of unlisted codes by labs (here).   Use of unlisted codes is incompatible with the payment system set up by PAMA.  Use of 81479 rose a little from 2016, $108M, to 2017, $118M. 

Tier 2 Codes

Collectively, the Tier 2 codes (level 1-9) garnered $72M. 

click to enlarge - Mopath Tier 2, Level 1-9
While about 60% of utilization was in Level 1 and 2, these levels had only 28% of payments.  Levels 7,8,9 had 13% of utilization and 41% of payments. 

Tumor Panel Codes

Tumor panel codes for CY2017...81445 had 1,428 uses for $807,000.   81450 had 3,378 uses for $2.2M.   81455 had 4,351 uses for $2.8M.   Together these tumor panel codes were about $6M or just about 1% of genetic/genomic spending at CMS.  Note that 2017 is the same year that CMS proposed its complex NCD on NGS panels in cancer. 

There were no payments for 81415 (exome) or 81425 (genome).


In the data above, I haven't included PLA codes and Admin MAAA codes, which CMS has in a separate spreadsheet.  Impact would be minimal.
It's no masterpiece, but I've included the worksheets used, in the cloud here.

FDA's Gottlieb Gives Major Address on Antibiotics Policy; Cites Need for Payment System Reforms

On this blog, we noted in June 2018 that FDA had used a press release website as an opportunity for Dr. Gottlieb to ventilate some innovative policy ideas about US payments and incentives for antibiotics, here.

On Friday September 14, 2018, the Pew Foundation, which has a major issue in areas like antibiotics and sepsis, hosted FDA Commissioner Gottlieb for a speech on new and innovative FDA initiatives on the major public health issues of antibiotic development and microbial resistance.
  • Dr. Gottlieb's prepared speech is online at FDA, here.
  • FDA rolled out a new consolidated website on "Emergency Preparedness and Response: Antimicrobial Resistance," here.
  • The PEW Foundation website for the event, including Dr. Gottlieb and a panel, is here.
  • The speech and panel stream on YouTube, here (1 hr 38 min).
    • Unofficial transcript here (32 pp).   
    • Trade press on the conference here, here, here.  Coverage at Endpoints here.
    • FDA also released a Request for Information seeking input on setting priorities in antibiotics and microbial diagnostics, here.
While the FDA doesn't directly control payment (or payment rules and systems) at all, the speech and panel touched directly on "payment" 8 different times, "econonomic" 9 times,  "incentives" 10 times, "reimbursement" 10 times and "market" 32 times.  "Diagnostic" appears 25 times.

Extensive direct quotations from the PEW event in regard to payment systems and device innovation follow, after the break.

In addition to efforts at FDA, CDC, and at NIH, the federal government funds BARDA, the bioterrorism research fund.  One major effort is an international public-private partnership called CARB-X, aimed to stimulating the early stages of innovation for new-generation antibiotics despite the market forces that make this difficult.  See BARDA here, CARB-X here.   BARDA also has a new 2018 effort in sepsis innovation and in part, diagnostics... here.  Trade press for BARDA's new Division of Research, Innovation, Ventures (DRIVe), here.  There is also a Presidential Advisory Council on antibiotic resistance, PAC-CARB, here.

Rick Bright, the head of BARDA, attended the conference, getting a call-out from Gottlieb.

Gottlieb left quickly after his speech, after taking one question from Politico and one from Stat.  The following panel was comprised of FDA senior leadership in the four relevant divisions - Edward Cox from Office of Microbial Products; William Flynn, Center for Veterinary Medicine; Steven Gitterman, Microbiology Devices, CDRH, and Carolyn Wilson, CBER.

For an article from Bloomberg on the antibiotic development crisis, here.  There is also an academic literature on antibiotics R&D economic bottlenecks, see e.g. here and here and here.

There's an interesting discussion of something I hadn't heard of, a program called SHIELD to create powerful real time EHR based sharing of antibiotic resistance data from hospital labs.  See comments on that topic, below the break.  It's part of a broader set of important efforts called NESTcc.


In a speech within a week, Janet Woodcock of CDER also took a market-based approach, decrying the impact on generic manufacturers of the monopsony power of drug wholesalers (here).

Brief Blog: What Is the "Society to Improve Diagnosis in Medicine" ??

On September 14, 2018, the trade journal Healthcare Dive carried a story, "Intermountain, Geisinger Form Group to Improve Diagnostic Accuracy."   OK - what?  The tag is, "More than 40 healthcare organizations and patient advocacy programs launched a campaign to cut down on diagnostic errors and ensure timely treatment for patients."
I knew this:  It's a topic the National Academy of Medicine has written about.  See the homepage for their initiatives here.   This workstream inside NAM issued a 2015 450pp report on diagnostic errors in medicine (here) and also held a 2017 follow-up on reducing diagnostic errors (webcast and 37pp briefing book, here). 
Here's the organizational story for SIDM/ACT.

Society to Improve Diagnosis in Medicine (SIDM)

First, the overlying organization is called, the Society to Improve Diagnosis in Medicine or SIDM.   It's homepage is simple:  .   A 501(c)(3), its mission statement is: "The Society to Improve Diagnosis in Medicine catalyzes and leads change to improve diagnosis and eliminate harm from diagnostic error, in partnership with patients, their families, the healthcare community and every interested stakeholder."

SDIM Conferences & Members

They have three conferences in 2018/2019:  In Bern this past August 2018, in New Orleans November 4-6, 2018 (11th Annual), and in Melbourne April 28-30, 2019.   Calendar here

Twitter here.   

Founding members of SIDM were a group of individuals, there are corporate members (including e.g. Abbott), and health provider and  non profit members (including the Gordon & Betty Moore Foundation and Permanente Medicine).  David Newman-Toker of Johns Hopkins is current President.   Paul Epner is the CEO and founder of SIDM.

The New News:  
40 Organizations in SIDM's New "ACT" Coalition

The "new news" is a September 13, 2018 press release from the SIDM, announcing that 40-plus healthcare organizations launching a diagnostic accuracy effort.  Press release here.  This effort is separately branded as, ACT: [ Accurate Communicated Timely ] For Better Diagnosis.  The full press release is worth reading.  There is a generalistic reference to "federal liaisons" including AHRQ, CDC, CMS, VA, here.

Now, ACT has its own website too: .
ACT has this introduction:  More than 40 of the most prominent healthcare and patient advocacy organizations joined together to form the Coalition to Improve Diagnosis and launched ACT for Better Diagnosis(tm).  The goal is to improve the diagnostic process by identifying and spreading practical steps that everyone throughout the healthcare system can take—from patients and physicians to laboratory scientists and health system leaders. By working together, the Accuracy, Communication, and Timeliness of diagnosis can be improved.  ACT for Better Diagnosis is an initiative of the Society to Improve Diagnosis in Medicine and supported by the Gordon & Betty Moore Foundation and The Mont Fund. 


click to enlarge - ACT organizations (9/2018)


Several of the press releases state that the FY2018 Omnibus Appropriations Act has the following language:   The FY 2018 Omnibus Appropriations Act included language emphasizing that improved diagnosis is a “moral, professional, and public health imperative” and requested that “the Agency for Healthcare Research and Quality (AHRQ) convene a cross-agency working group to propose a strategy to enhance scientific research to improve diagnosis in healthcare.”   I couldn't find this wording in the version of Omnibus Appropriations that I had access to (here).  "Moral, professional, public health imperative" is a wording that National Academies have used in regard to diagnostics (e.g. here).

Sunday Reading: Is Your Dog's Genetic Testing Your Best Health Investment? (Would you ask that question?)

Trifecta of recent articles on the value and the marketplace for pet genetic testing.

DTC - Direct To Cat?

  • The original article appeared in Nature, July 26, 2018, by Moses, Niemi, and Karlsson.  It's open access here.
  • At the time, see a story on NPR, by Carey Goldberg, here.
  • In September 2018, a follow-up story with additional reporting by Jessica Hekman, here.
Health economics for hamsters?

Thursday, September 13, 2018

MolDx Proposes Coverage for Pharmacogenetics Panel: A Win for AltheaDx

Palmetto MolDx has posted new LCDs for its October 1 CAC meeting.   Several are duplicates of LCDs that I already recently highlighted because they will appear in the next CGS MAC MoLDx meeting - here.

But also: a new LCD appears. 

This one proposes coverage for a pharmacogenetics test, NeuroIDgenetix® Test for Depression, draft LCD DL37865.   The test was developed by Althea, a San Diego-based genomics laboratory. 

The LCD is quite detailed and will grant coverage of the gene panel test in patients with treatment resistant depression who are -- under age 65.   The company supported a very well done and large randomized controlled trial (685 patients), Bradley et al.  See Pubmed here, journal open access here.  Although a large age range was eligible for the trial protocol, nearly all patients actually enrolled in the Bradley trial were -- under age 65. 

The LCD provides a very detailed clinical setting review, discussion of unmet needs, and lays out a carefully written clinical and technological assessment of the test. 

Public comment is open from October 1 to November 15.


Note that the lab science of PGx is quite complicated.  If you've never delved into this area, see a review in CAP Today, August 2018, here.  See also an August 2018 Genomeweb article here linking to a Pharmacogenomics paper by Caudle et al. here (PMID 29914287).

Screen shot from Bradley et al.   The publication is open-access.

Genetic testing raised remission rates (red bars)
click to enlarge

FDA's Gottlieb Speaks Out Strongly for FDA Diagnostics Law Reform; Plus Third-Party Review & Consensus Standards & Dhealth

On September 13, 2018, FDA Commissioner Gottlieb gave a major speech endorsing substantial reforms, including new law by Congress, to help the FDA be innovating in supporting diagnostics and precision medicine.

Recall that on August 3, 2018, FDA responded to the Hill's lengthy diagnostics reform bill (DAIA) with a 60-page model of its own statutory reform for diagnostics review.  The reform would nix the legacy 510K and PMA systems for diagnostics and set up a new review paradigm.   Entry point here.

Today's speech was at the Blueprint for Breakthroughs / Precision Medicine conference sponsored by the Friends of Cancer Research (FOCR).  See agenda here.

Scott Gottlieb's speech is online at FDA here, and a rapid review by RAPS is here.

There's no question that the FDA ship is steaming to legislative reform.  Gottlieb states in part:
In pursuing a new framework for the appropriate regulation of diagnostics, there will be difficult policy decisions to make and tradeoffs on any path to legislation. We know that the FDA must be flexible and open minded to new approaches that best meet the needs of patients so that patients can have confidence in the results and the treatment that comes from it. That’s the spirit of our proposed reforms. 
FDA’s Center for Devices and Radiological Health has taken exactly this approach in providing feedback to Congress on proposed diagnostics legislation.  We have embraced a more modern, flexible approach to promote the extraordinary innovation that’s already well underway in this space, while ensuring patient protections.
In providing the contours for such a pathway to Congress, we’ve proposed new regulatory concepts that reflect some of the core principles we believe are fundamental to efforts to modernize diagnostics oversight.
He goes on to discuss key points from within the 60-page legislative "brave new model" released by FDA on August 3.


Gottlieb had flagged his interest in statutory reform for diagnostics at his ACLA speech in March 2018, here.

(!!)  Note that the "reform of diagnostics regulation to encourage innovation" as discussed above is also closer interlinked the topics I list below, particularly "third party review" and "use of consensus standards" in applications and approvals review.

Third Party Review
In other important FDA news, this week FDA released a new guideline for 3rd-party review for 510(k), under which organizations like New York State or CAP might do most of a 510(k) test review.   Here (entry point and links from RAPS).

The draft guidance document is here (22 pp) -- and see a quite interesting 16-page white paper proposing to minimize FDA interference after a 3rd party review; here.  (This minimization of FDA re-review is called, "Giving 3rd Party Reviewers the Tools They Need to Succeed."  Nice spin.)

Harmonization / Consensus Standards

Also, FDA released guidance on use of harmonization and consensus standards, here (entry point and link from RAPS).   Recall that FDA released guidances a few months ago on use of NGS clinical genetic databases as part of validation; this is at least a second-cousin to the concept of consensus standards built into FDA applications and approvals.

Digital Health

FDA also released a joint statement from Gottlieb and from the head of Devices, Jeff Shuren, on the FDA's commitment to innovative review and authorizations in digital health, here.  This particular announcement seems to be a restating of their commitment to innovation and not stopping the trains by over-regulation, but they also note they plan to create a new "Center for Digital Health Excellence" inside FDA soon.


For comparison see last year's Friends of Cancer Research conference (September 2017):   agenda here, summary here, and FDA deck here.

House Tweaks, Passes LCD Transparency Bill; Moves to Senate

Local Coverage Determination Clarification Act of 2017

Last year and as recently as mid August, this blog covered the progress of the College of American Pathologist's legislative effort to increase LCD transparency.  See my August 17, 2018, blog here.  See an earlier October 2017 blog here.

On Wednesday, September 12, 2018, the House revised then passed legislation on LCD transparency.   See some links here:
  • CAP press release here.
    • CAP applauds the move, but wishes a few legislative points hadn't been deleted.
    • Statement by CAP president here.
    • CAP's topic web site here.
  • Track HR 3635 text online here.
    • Up to the vote, the bill had garnered 81 co-sponsors.
  • See a YouTube pitch for the bill by Rep. Lynn Jenkins (R-KS), 5 min, here.
    • Transcript here.
    • After one day, the video had 1 hit, zero likes, zero dislikes.
  • See a September 20 article at 360DX, here.
I've clipped the CAP press release below the break.

Note - Notably, CMS has still not officially and fully implemented through instructions to contractors, the smaller legislative reform for LCD clarity passed in 12/2016 in the 21st Century Cures Bill.

Youtube - one hit, zero likes, zero dislikes.

Wednesday, September 12, 2018

Brief Blog: FDA & Payers: Does a New Press Release Make It New?

In February 2016, FDA published a notice in the Federal Register announcing a program through which it invited commercial payers to participate in helping developers understand the differing requirements of payers for coverage decisions and FDA for regulatory decisions.  I covered this and provided links here

Now in September 2018, there are a spate of very similar articles about FDA gearing up its integrated program of communications jointly with commercial payers and FDA.   See an article in Health Payer Intelligence here.   The flurry of "new-news" is stimulated by a September 5, 2018, speech by Commissioner Gottlieb at the Medical Device Innovation Consortium, here.   The full speech highlights other FDA innovations related to devices besides the payer consortium channel.

Does "new" rhyme with "deja-vu?"   It happens. (In fairness to my deja-vu assertion, in a blog Gottlieb does mention the PPP or Private Payer Program dates to 2016). 

73 Inquiries Into CMS-FDA Parallel Review

In a September 2018 blog, here, Dr. Gottlieb notes that the agency has logged 73 inquires about FDA-CMS parallel review (towards NCDs) and received 36 formal applications for the same, although only 2 NCDs have resulted (Exact Sciences Cologuard and FMI F1 CDx).   

CareFirst HealthWorx; CareFirst and FDA

Two payers highlighted by Gottlieb are United Healthcare and CareFirst, the latter known for its entrepreneurial investment program HealthWorx.   Besides UHC and CareFirst, other associations to PPP have included BCBSA, Duke, ECRI, Humana, Kaiser, and NICE.


Company/Payer HEOR communications

A different topic also involves FDA, regulatory guidance, and payers.  In February 2017, FDA issued a draft guidance on communications from companies to payers of data from health economic studies that is not necessarily on the product's FDA labeling; here.  Said guidance was finalized in June 2018, here.   Dr. Gottlieb issued a blog.  For commentary by Peter Neumann and Harry Weissman in Health Affairs, July 2018, here.

Tuesday, September 11, 2018

News Watch: Genomeweb Article on Oregon's Position on Oncology Gene Panel Tests

For those with a Genomeweb subscription, see Turna Ray's deep-dive story on responses from cancer groups to an Oregon policy statement against coverage of somatic mutation gene panel tests in cancer.

The September 11, 2018, Genomeweb article is online here

In brief, the Oregon Health Evidence Review Commission has a busy rolling schedule of policy meetings (calendar here).   Health tech assessment meetings are here.   The June 28 meeting assessed single fraction radiotherapy for palliative care and then assessed "FDA approved next generation sequencing tests."
  • See Agenda here
  • Evidence review materials here (NGS at page 32ff),*
  • Minutes here (minutes on Foundation One are just a few sentences).    
The Genomeweb article discusses the role of OHSU's Vijay Prasad, a noted commentator on evidence for precision medicine.  The responses of various cancer groups are discussed.

According to Genomeweb, one of the issues discussed by stakeholders is the pan-cancer coverage under the NCD and the question of who will be tested under a potential Oregon Medicaid program.   One stakeholder notes the differences in approved drugs for lung vs prostate cancer and said the difference was "apples and oranges." 

Data is publicly available for the distribution of recent tumor types in the tissue archives used for FMI and MSKCC IMPACT test authorizations.   I summarize these below.  MSK reported 17 tissue types, 10,554 blocks, of which the top 5 (58%) were lung, breast, colorectal, prostate, and glioma.  FMI reported 38 tissue types, 80,715 blocks, of which the top 5 (58%) were lung, liver, lymph node, brain, and colon.   Prostate, for example, was 8% of MSK cases and 2% of FMI cases.**  Alternately, prostate volume at MSK was 44% of lung volume, but at FMI, prostate volume was 9% of lung volume.

click to enlarge


Oregon also hosted the OHSU Evidence Based Practice Center, which is supported as part of the national network of practice centers that contract to organizations like USPSTF and AHRQ.

For a negative article on precision medicine in New York Times, appearing the same day as this Genomeweb article, here.

* Page 6 of this document discusses recent review of Oncotype Dx and a "weak" recommendation for coverage in breast cancer patients with 1-3 nodes.

** For 5 tumors found in both top-ten lists, by percent of that lab's cases, the correlation between the FMI and MSK percentages was r = 0.7.

2017 FMI 1130 FDA P170019B Tumor Types b.xlsx

Thursday, September 6, 2018

Very Brief Blog: CMS Updates Public Meeting Video: Streaming Yes, Youtube Archive No

  • CMS has a Youtube channel, CMSHHSgov, available here.
  • Through 2017, CMS both streamed and archived video of major meetings.  
  • The new policy appears to be streaming-only and no archive.


For the last several years, CMS has both streamed live its major medical policy meetings (such as MEDCACs and summer CLFS advisory panel meetings), and also, left them archived on Youtube.

See, for example, a summer 2017 MEDCAC here, and a summer 2017 CLFS meeting here.  An additional feature was that the live subtitling was preserved and could be pulled as a sort of transcript by tapping the "three horizontal dots" symbol below the video and clicking "open transcript."

So far as I can tell, the new policy for 2018 is more limited.  CMS continues to stream the meetings - which means, you could record or screen-grab them* - but CMS doesn't seem to post them as an archive any longer.

For example, unless you were in the office to watch and record the MedCAC on oncology outcomes a few weeks ago, on the day it was being streamed, you'll have to wait (potentially months) for CMS to post a transcript of it (assuming they will).  Other meetings, like CLFS meetings, never get a transcript.


* For example,  I am inferring that MolecularMD captured and posted on Youtube the CMS webstream of its own 18 minute presentation and Q&A on June 25, 2018.  Here.  But the other five hours of presentations are gone.  I'd be happy if anyone can find the whole day's show ion the internet and point it out to me.

Very Brief Blog: Guardant Files for IPO

On September 6, 2018, news releases that Guardant, an oncology liquid biopsy test company in the Bay Area, has filed for an IPO.  Medcity News here.  At Nanylze, here

Medicare LCD coverage for Guardant was finalized over the summer, here.
  • See Guardant Health's SEC web page here.
  • See the S-1 filing here (277 pp).

Underwriters include JP Morgan, B of A Merrill Lynch, Cowen, Leerink, William Blair.

For a July 2018 Bloomberg article about a potential pending Guardant IPO, here.

As noted at Nanylze, Foundation Medicine rolled out a next-generation of its own LDT liquid biopsy test, now called FoundationOne Liquid, September 2018, here.

For a historical comparison, the archived FMI S-1 filing from 2015, here.

Wednesday, September 5, 2018

Very Brief Blog: Friends of Cancer Research Holds 7th Annual "Blueprint" Precision Medicine Forum

On September 13, 2018, the Friends of Cancer Research will hold the 7th Annual "Blueprint for Breakthrough" forum Washington, DC, promoting precision medicine.

See the website here:

It's a half day meeting (8:30-1:00).   It looks to me like registration is required but is free.

Scott Gottlieb MD, head of FDA, gives the opening address.   The first panel features presentations by Katherine Szarama PhD (CMS NCD), and Jonathan Hirsch (Syapse).   Reena Philip PhD, FDA molecular approvals, is also on this panel among others.   The second panel is on "supporting development of high quality diagnostic tests" and includes Timothy Stenzel MD PhD, the new head of diagnostics at FDA, as well as speakers from United Healthcare and Memorial Sloan-Kettering.

See the full agenda at the link above or cut/pasted here below the break (agenda as of 9/5).  The panel is supported and is co sponsored by Alexandria.

See some other fall precision medicine conference dates here.

Monday, September 3, 2018

Weekend Post: Business Models and Med-Tech Innovation

I live in California but frequently work in Washington which gives me a lot of coast-to-coast flights.  In August I had the chance to read two complementary books on business strategy.  These were "Reinvent Your Business Model," a new edition of Mark Johnson's "Seizing the White Space;" and "LEAP, How To Thrive in a World Where Everything Can Be Copied," by Swiss B-school guru Howard Yu. 

Both deal with innovation in a world where the competition comes every closer and the bar moves ever higher. 

Below, I highlight some points from each book and then show a Medtech Take-Away Lesson for each.

Reinvent Your Business Model: White Space and Customer Value Proposition

"Reinvent" describes a world where a company's core business will also be prey to competitors and it needs to constant upgrade its game or move into "adjacent markets," e.g. white spaces that require innovative twists to exploit.  Often, a white space is less competitive but requires altering one's business model.   For example, the iPod moved into a white space when Apple created the ITunes store, which was a wholly new service model for Apple, a desktop and laptop manufacturer at the time.   What Apple learned from ITunes applied forward to its iPad/iPhone App Store, and the rest is history.   Johnson focuses on the challenge of seeing such "white spaces" and exploiting the old core competence while adding the new one.

Johnson also talks about customer value propositions, CVPs, and asks three key questions:
  1. How important is the job-to-be-done to customers?
  2. How satisfied are customers with current solutions?
  3. How well does the new offering get the job done, relative to other solutions?
Medtech Tie-In:  It occurred that most offerings in medtech and diagnostics also try to answer these three questions - we call it an unmet medical need, and improved clinical outcome - and Johnson's three questions capture medtech innovation as well.  (Note that you can't conclude, "if you build it, they will come.")

Medtech Tie-In #2:  In medtech, we have a weird mix of multiple disparate customers to satisfy - physicians, hospital purchasing staff, patients, health technology assessors, payers.  While working within the confines of a legacy and largely fixed payment system. More on that, in the next section.

Find "Reinvent" here and here.

LEAP:  Companies Evolve Onward and Upward

The main point of Howard Yu's book LEAP is that successful companies have to constantly reinvent themselves - clearly not to far afield from Johnson's vision that successful companies reinvent and move into their white space.   

LEAP talks about the history of Proctor and Gamble.  In the 1870s/1880s, it led industrial approaches to making consumer products like soap in giant factories, far above the scale of small business.   By the 1890s, Proctor and Gamble pioneered what we would today call focus groups and effective advertising.    In the 1920s, Proctor and Gamble pioneered bringing more advanced chemistry into the design of household products (something the pharma industry had just discovered in the past decade or two).   This led to Dreft soap in 1935, and Tide in 1946, both landmark product introductions.   Yu's point is that a (1) new type of industrialization, (2) a wholly new school of advertising built on focus groups, and (3) a new approach to consumer soap chemistry were three different core competencies that one company, P&G, was able to roll into over several decades and far ahead of any competitors.

Yu also makes the point that as time rolls on, today's cutting edge becomes tomorrow's commodity as competitors enter and as patents or other thought capital expire or diffuse.  Yu describes three fundamental zones of competition and differentiation:
  • Product differentiation
  •   Services differentiation
  •      Business model differentiation
Product differentiation is obvious; new bells and whistles, faster-lighter, and so on.   Services differentiation can be key: for example, repairing your equipment twice as fast as the competition can repair theirs.   Business model differentiation could be, for example, sell versus lease, or "razor and blades" models or other ways of solving the business interface to your customer.  (Johnson, in Reinvent, presents business model differentiation as the secret sauce that is most often ignored or missed).

Medtech Tie-In:  Yu implies that other industries get faster, better, and cheaper as they roll through multiple cycles of (A) product, (B) services, and (C) business model differentiation.  

In Medtech, what do we always hear?   Don't, for god's sake, try to change physician practice or the way the system finances work (CPT codes, DRGs, and so on).   

If that's true, then medtech is left only with one avenue of differentiation - the first, product differentiation - which is good but only 2/3 of the innovation story, per Yu.   

We make it so difficult to change physician practice, or the way specialties are set up, or the way that forty-year-old codes and payment systems work, that we make many forms of innovation into a forbidden territory, or at least, about as inviting as an icy steep mountainside.  

Find LEAP here and here.


The issue of complexity of our healthcare administration system is discussed in a new 2018 book by Etienne Deffarges, "Untangling the USA," here.

Friday, August 31, 2018

House Messages CMS: Asks CMS to Address Panel Unbundling

Each year, annual budgets prepared at House and Senate come with something called "report language" which describes something of the rationale and purpose of the budget and may make numerous requests of an agency to look into one matter or another. 

This summer, the (health related) report language for House is here and for Senate is here.

CMS and Chemistry Panels

CMS for decades has bundled together tests following CPT codes for common chemistry panels.  A ten-analyte panel might pay $15 as a panel even though each component tallied as a separate test might pay $60 or more.   With PAMA in 2018, CMS dropped the panel bundling rules and right now - if a set of tests fail to meet the exact definition of a panel - CMS pays at the individual line item rates. 

I discussed this for clinical chemistry in a blog in November 2017, here.   As shown in a clipping at bottom, 80053 panel had 29,000,000 uses at CMS.   If paid at the $13 panel cost, this is $377M.  If paid at the stack code analyte rate it's $82.88 or $2.2B.   The same thing can also happen in genetics; as shown in another clipping at bottom, CMS pays for Lynch panels as 81435+81436=$1444, but the genes included at a la carte prices look more like $3,334.

The message from the House to HHS regarding panels is:

"Clinical Laboratory Fee Schedule.  Inconsistencies in panel testing reimbursement in Medicare should be resolved to prevent wasteful government spending.  The Committee encourages the Administrator of CMS to develop and issue a panel pricing policy that ensures the agency is not paying more for a single clinical diagnostic laboratory test, or a group of individual clinical diagnostic laboratory tests, than it would pay for a clinical diagnostic laboratory testing panel that tests for the same analyte(s).  The Committee encourages the Administrator to apply the policy to all types of test panels."  (House report language, page 85.)

OIG Guidance Now Dysfunctional

Labs can follow official OIG guidance, and it leads toward stack coding, not panel pricing.

OIG guidance discourages panel ordering, and encourages physicians and labs to provide only the most precise list of medically necessary tests.  If any one test on a ten analyte panel isn't necessary, then bill the 9 tests, please.   This guidance was written in an era when panels paid a fraction of the single test costs, BUT ALSO, coding and pricing with special rules meant that analyates could only pay LESS than a panel.  (E.g. 9 analytes would pay right about 90% of a ten-analyte-panel.  Today the nine analytes might pay 800% more than the ten-analyte-panel.)

MolDx Releases Set of New LCDs

A Genomeweb article on August 30, 2018, notes that MolDx has released a draft LCD favorable to the Veracyte test Envisia, for pulmonary fibrosis (here).

Four other draft LCDs were also released on the WPS MAC website.  MolDx LCDs may appear first at any MolDx-participating MAC, depending on the various MACs' calendars.

A search of new LCDs this week is available via the CMS LCD database (here).  I've put a bundle of five new MolDx LCDs in the cloud here.

ClonoSEQ (MRD)
In DL37917, coverage is proposed for the ClonoSEQ Assay for minimal residual disease (MRD) in certain lymphoid malignancies.  Analytical validity is cited to "data on file" at Adaptive that is "designed to meet FDA standards" (LCD fn 15).  Guidelines and several validity studies are cited.  MolDx summarizes that "studies have demonstrated the accuracy and clinical utility of clonoSEQ in predicting clinical outcomes in a variety of myeloma settings and timepoints."  Validity in acute lymphoblastic leukemia (ALL) is discussed separately.    While separate coverage is described for myeloma and ALL, criteria are factors such as use in transplant decision-making, as a test for drug therapy response, and testing for sustained MRD "per guidelines at time" as indicated in various guidelines for hematopoeitic MRD.

Veracyte Envisa (Lung disease)
In DL37919, the Veracyte Envisia test will be covered for diagnosing interstitial lung disease in patients who lack a definitive CT pattern and where a bronchoscopic biopsy may be used instead of a surgical lung biopsy.   MolDx notes it consulted specialists from Mayo, Columbia, and USC in addition to its own review.  Criteria for coverage including high resolution CT (hrCT, defined as 1mm reconstructions) which are not definitive.

Invivata InvisionFirst (Liquid Biopsy)
Inivata is a lab dedicated to liquid biopsy testing in oncology.  In DL37921, its LBx test ("InvisionFirst") is covered at diagnosis or at progression in lung cancers that are advanced stage and meet criteria for missing genomic information.   Earlier this year, MolDx covered the Guardant G360 test under similar conditions.

 MyPath Melanoma (Ambiguous Cases)
In DL37923, the MyPath melanoma assay from Myriad is covered for "diagnosis or exclusion of melanoma" in a biopsy when ordered by a board certified dermatologist; the cutaneous neoplasm is ambiguous by histology alone; and the uncertainty impacts the patient such as risk of re-excision and sentinel node biopsy.   According to the LCD, two validation studies assessed concordance compared to histopathological experts.  (This is obviously good, but doesn't directly address the accuracy in cases where histopathology has no diagnosis and experts disagree.)  A third study looked at outcomes at a median of 6 years in 182 prior biopsies.

Decipher Prostate Prognostic
In DL37911, the Decipher prostate prognostic test is covered in low-risk prostate cancer as defined by NCCN.   MolDx has generally granted earlier coverage for prostate risk tests in low risk disease, where the test is most likely to verify a low-risk, lower intervention stance is justified (as opposed to deselecting therapy in higher risk cases).   I suspect this must be a new LCD only for the WPS MAC and has existed in other MolDx MACs.

Several of these assays are NGS assays in cancer, so the LCD should be in compliance with the recent CMS NGS NCD for uses of NGS in cancer.  For example, the Envisia LBx test LCD states, "May be used once per lifetime," following the once-per-patient requirement in the NCD.

Statements of "Promising Data"

Some LCDs include concluding evaluations that data is "promising" and that ongoing coverage is dependent on forthcoming publications.  For example, in DL37919, we read:

"The clinical utility of the Envisia genomic classifier to aid in the diagnosis of patients with an ILD of unknown cause and suspected of IPF, as defined in the intended use above, is quite promising. This contractor believes that forthcoming clinical studies in these patients will demonstrate improved patient clinical outcomes. Continued coverage for Envisia testing is dependent on annual review by this contractor of such data and publications."

This isn't universal, e.g. the MyPath Melanoma LCD  and Inivata LCD don't have this concluding test, and the other three do have it, even though the MolDx evidence ratings shown for ClonoSeq are a bit higher than for MyPath.

I'm not sure how precisely MolDx, or its differently reviewers, classify evidence by "quality, strength, and weight" (see my blog on these MolDx terms here).   With that warning, below is a table with the ratings for the five LCDs discussed today.

Note that the  last row, "Total Citations," is not the citations of the product, but the total citations used throughout the LCD discussion. 

click to enlarge

Thursday, August 30, 2018

Very Brief Blog: The Swing to FDA Approval: FMI Discontinues F1 Test for CDx Test

I may not be the first one to hear it but I heard today:  Foundation Medicine is officially turning off the production line on its F1 test.   According to its website, "On September 28, 2018, FoundationOne will be discontinued.  All [orders] will automatically be converted to FoundationOne CDx."

Website here.  This appears initially as a headline banner.   Web clip below. 

FMI's last SEC report (before its acquisition inside Roche) was in early August; unless I missed something, I didn't see an explicit reference to discontinuing F1.  In that report, it sounded like FMI was billing its CDx test to Medicare via Massachusetts and its LDT test to Medicare via North Carolina.   I don't know if the F1 discontinuation means its CDx supply line in Massachusetts can handle all volume or its North Carolina lab was upgraded to CDx. 

FMI will continue to provide Foundation Heme and Foundation Liquid Biopsy ("ACT") tests as LDTs for the present.

Wednesday, August 29, 2018

Very Brief Blog: MolDx Analysis of Evidence "Quality, Strength, and Weight"

The 21st Century Cures Act requires LCDs to provide rationales for their decisions and a summary of evidence considered.   For an overview, see here.  For the actual 197 words in 21CC about LCDs, see here.

CMS may eventually provide a substantial rewrite of the LCD manual (the current manual is here).  In the meantime, CMS requires LCDs to have one demarcated section which is a "summary of the evidence" and another section which is an "analysis of the evidence."[*]   (This is exactly the same format used in NCDs.   Generally, a "summary" is objective and an "analysis" assigns interpretations, critique, and value judgement.)   

An example of an NGS MAC LCD for a surgical procedure which includes detailed "summary" followed by detailed "analysis" is here.

MolDx Quality, Strength, Weight

MolDx LCDs over the last year have provided a summary assessment of diagnostic tests categorized as "Quality, Strength, Weight."   Frankly, even with a decade of consulting and several publications under my belt, I wasn't sure I could predict exactly what MolDx authors meant by their use of these three words.   Discussion at a California Clinical Laboratory Association meeting this month helped clarify the meaning.

Quality refers to the quality of the trial and study design.  For example, a randomized controlled trial that is double-blinded is higher quality than a retrospective case series.   This is close to a fairly classic "level of evidence" hierarchy (level of evidence 1,2,3,4, etc).  This metric focuses on quality of the trial or study; a 1000-patient RCT may be a very high quality trial, even if it shows the therapy under study is neutral or harmful. 

Strength is the strength of the resulting evidence for Medicare purposes.   For example, a study conducted in children in a country with a wholly different health system and alternatives could be a 1000-patient RCT, but it would have low "strength" for Medicare decision making.   Strong evidence would be large effects accompanied by low p-values, and strength of evidence is additive across multiple trials.   Evidence for Medicare coverage has less strength if it is 95% in patients age 40-55, and not the main Medicare age of >65.[*] 

Weight of evidence is conclusory:  Given the evidence reviewed (perhaps other evidence from other studies, guidelines, or alternatives), how strong is the weight of evidence for Medicare coverage.  Although I am adding my own terms, I suspect the weight of evidence would be strong if it is easy to conclude that, all factors and concerns in the balance, the benefits considerably outweigh the risks or uncertainties. 

Snapshot from a MolDx LCD

[*] The requirement to discuss and summarize evidence considered, and then provide an "analysis" of that evidence, would tend to prevent blanket non coverage LCDs.  For example, an LCD that simply lists 100 Category III codes with a sentence stating they are noncovered would not meet the standard of summarizing and analyzing any evidence for the 100 different services represented by the codes.

[**]  I believe "weight" of the evidence would include the concept, seen in some health technology assessments, is additional evidence likely to materially change the estimate of benefit or risk.  For example, over a large population, there have been very large studies of mammography, and it has a benefit, but/and the benefit is fairly small.   It is unlikely another new study would find that mammograms are harmful or that they have a huge benefit. 

Very Brief Blog: New Medical Director at CMS MolDx Program

The Medicare special contractor program for genomic tests, MolDx, has a new interim medical director, Dr. Paul Gerrard.   Labs who work with the MolDx program have begun receiving emails about the transition in staffing and it was discussed this week at the California Clinical Laboratory Association.

Dr. Gerrard is a Harvard-trained and board-certified physiatrist.  His Linked-In is here.   He holds an MD from Medical University of South Carolina (2009).

Dr. Jim Almas, the senior medical director for MolDx from mid-2017 to August 2018, has retired from that position.   Almas had previously been part of the CMS coverage group in Baltimore (2016/2017), and he had held senior management roles in several large hospital systems.   (Linked-In here.)

According to several sources, MolDx has plans to bring on additional medical director staff this fall.

Wednesday, August 22, 2018

My Short Deck on LCD/NCD and Combination Diagnostics: For NextGenDx Conference in DC

This week is the 1000-person NextGenerationDx conference in Washington, here.

This afternoon I have the chance to present a concise 15-minute PowerPoint deck on changing CMS policy for combination diagnostics - both the LCD and NCD levels.    

The presentation is in the cloud as a PDF here.

Monday, August 20, 2018

Two New Research Studies Propose Updated Approaches to Medicare Dx LCDs

Classically, a new diagnostic test is created and its developer produces a series of studies showing it is clinically valid and has strong clinical utility in patient care.    In two articles I ran across this month, studies are run and published to show potential improvements to existing generic (not product-specific) LCDs.

Lynch Syndrome

The first is a paper in JAMA Oncology by Hampel et al. of the OSU with coauthors from U.W./Seattle, here  Both Ohio and Washington are "MolDx" states and cover Lynch syndrome testing under LCD such as L36161 or equivalents.  See here  This LCD covers Lynch testing through a fairly convoluted system of steps and pathways (see AMP's original comment here.)

Hampel et al. provide data on 419 consecutive CRC cases worked up for Lynch syndrome with NGS panels.  They report that "Up-front tumor sequencing in CRC is simpler and has superior sensitivity to current multi-test approaches to Lynch Syndrome screening."   They also note that Lynch testing (which is recommended now in all colon cancer cases) can thus be performed in parallel with driver oncogene testing for therapy selection. 

HBOC Syndrome

The second is a paper in Annals of Surgical Oncology by Yang et al. which in its title directly addresses Medicare policy criteria (see article here.)    A relevant policy is the MolDx hereditary breast and ovarian cancer LCD (HBOC; e.g. here.)    Medicare requires that a patient already have a personal history of breast cancer, unlike USPSTF guidelines for BRCA testing.   However, even leaving that in place, the authors note that among 4196 Medicare patients, the rate of BRCA positivity was about 9-10%, within and without Medicare's additional criteria.   This "suggests the need for significant expansion and simplification of Medicare criteria." 

Under the suggestion, Medicare patients would still have the entry criteria of personal history of breast cancer.   However, that would become the main requirement.  (For a biomarker test, in a breast cancer patient, does Medicare "have to" require an additional family history besides the cancer itself? No, it doesn't in the case of Lynch syndrome biomarkers in colon cancer nor in the case of e.g. Her2 biomarkers in breast cancer, to give two examples.)  

Levels of Evidence?

The levels of evidence in the articles above would argue to fairly simple standards for policy change.

Genomic Lynch testing overlaps with targeted therapy oncogene drivers (e.g. in a panel test), and simply has superior sensitivity (fewer false negatives).   The current HBOC criteria seem to serve no purpose (segregating patients with 9% vs 10% risk) and are therefore not objectively justified. 


Under recent law for Medicare creation of preventive services, the agency could open an NCD to determine that the USPTF benefit for BRCA testing in women without a personal history of breast cancer should be applied to Medicare patients as well.   Medicare has never opened this particular NCD topic.