Thursday, October 31, 2019

Very Brief Blog: CDx Forum; Illumina Forum; PMC in Boston; Many More Meetings

The flurry of conferences and events marches on.  These crossed my inbox in the last few days.  (One or two of them, unfortunately, seem to cross my inbox daily.  Ok, I get it already.)

On October 29-30, see the Fourth Annual Companion Diagnostics Forum in Princeton. Here.

On November 6-8,  see the California Clinical Lab Association in Orange CountyHere.

On November 7-9, see the Association for Molecular Pathology in BaltimoreHere.

On November 13-14, see the 15th Annual Personalized Medicine Coalition at Harvard Medical School in BostonHere.

My Linked-In feed brought advertisements for an Illumina "Market Access Symposium" in NYCNovember 15 - right after PMC in Boston.  Here.  The agenda is an interesting read.  Illumina brings you lots about payers.

On December 3-4, see the Diagnostic Coverage and Reimbursement Conference from Q1 in BostonHere.  (Or see the West Coast version, February 11-12, San Diego).

January 21-24, see the Precision Medicine World Conference in Silicon Valley.  (I'm a panelist).  Here.

On February 11-12, see the West Coast edition of the Diagnostic Coverage and Reimbursement Conference from Q1. San DiegoHere.

On February 27, the Foley Lardner Business of Personalized Medicine Summit in South San Francisco. (I'm a panelist.)  Here.

On March 1-4, the 27th International Molecular Medicine Tri-Con in San Francisco.  (I'm running a workshop).  Here.

Very Brief Blog: Center for Innovation Role Open at CMS; New Leader Likely Soon

Recently, Adam Boehler stepped down as head of the Center for Innovation (CMMI) at CMS.  He spoke this week at the HLTH conference (here), and remarked that his successor was likely to be named soon.

According to an article in Politico a few weeks ago, Brad Smith was a favored contender (here).  Smith, not to be confused with the Brad Smith who is President at Microsoft, is a 36-year-old former Rhodes scholar and McKinsey consultant.  (The first head of CMMI, Patrick Conway, also served a stint at McKinsey between college and med school).  Smith served as staff under Sen. Bob Corker and developed the palliative care firm Aspire, sold to Anthem for several hundred million dollars in 2018. 

CMMI recently rolled out new payment models that include capitated payments for primary care (here) and direct-contracting models (here, here).  (It takes some expertise, though, to recognize the difference between "total care capitation" and regular Medicare Advantage.)

Tuesday, October 29, 2019

Very Brief Blog: Snapshot of CY2018 MoPath Payments at CMS

Each fall, CMS releases CPT code payments for Part B in the prior year (for CY2018, here).

I've done a quick review of MoPath spending in CY2018.  Cloud excel here.

By my tally,  MoPath payments (81162-81599, plus U and M codes) was about $1,151,475,000.  This compares to $480M in the same data for CY2017, so PAMA 2018 did not kill MoPath.

As in prior years, the billing is highly concentrated.

Out of 204 codes, the top 5 gathered 50% of payments.   The top three were Cologuard (81528, 335,455 uses, $171M; see also here), Unlisted Code 81479 (used almost exclusively in MolDx states, 89,721 uses, $136M) and MoPath Tier 9 (62,280, $123M). 

click to enlarge  -MoPath CMS Pt B CY2018

Corus CAD
The 9th largest test was Corus CAD, at $30M - at the end of 2018, coverage was discontinued and the company quickly folded, here.

PLA Codes
The largest PLA code, 0037U, FMI F1 CDx, gathered $37M.  It was active beginning April 2018 and under NCD coverage (the CMS NGS NCD) for three quarters of 2018.  The $37M represents circa 10,000 patients.  (PLA Excel here; See additional PLA chart at bottom).

GSP Codes
Among GSP codes, 81432/81433 (HBOC panels) were used in 19,448 and 16,492 patients, respectively, for $16M and $9M (total $25M).   5-50 gene tumor panel 81445 was used 3,244 times for about $2M.  81455 (50+ tumor genes) was used 7,494 times for $21M.  While MolDx generally insists "we don't pay for 81455," 7,006 of the 7,494 payments (94%) were in Noridian Jurisdiction E under MolDx, suggesting the Z-code and master edit system is somewhat leaky.

Whole exome 81415 was paid 1 time; whole genome 81425 paid 0 times.

CMS data shows claims paid and gives no insight into denial rates.  Based on 2018 data, XIFIN showed that payor denial rates for 81445, 81450, 81455 ranged from 82%-96% (here).

BRCA Codes

CY2018 was the first year in which CPT codes 81211/81213, legacy BRCA codes, were deleted and replaced by less expensive code 81162.  Although BRCA remains a top-ten revenue item in MoPath, revenues trended downward due to this coding/pricing change.  See article on Myriad revenue changes due to coding, Genomeweb, 11/5/2019, here.

Tier 2 Watch

Tier 2 codes altogether tallied $190M uses, bigger than Cologuard.  Most of this was 81408, Tier 2, Level 9 (big genes).

click to enlarge - MoPath Tier 2

Despite the profusion of CPT codes and PLA codes, 25-30% of payments still flow through an unlisted code.

Each fall, OIG releases a report on CLFS spending in the context of PAMA implementation.  The 2017 reported appeared in 9/2018, but the 2018 report has not appeared yet.  That report includes hospital outpatient labs, which my Part B data does not.


PLA Code Concentration

2 codes were 98% of PLA code usage (CMS listed about 40 codes in CY2018 data).  In round numbers, FMI F1 CDx was 90% of PLA usage and Oncotype Prostate was the next 10%.

Most PLA codes had no usage followed by some that had a few dozen uses or less. PLA Excel here.



Unlisted code 81479 is used almost exclusively in MolDx states.  In the MolDx master edit file of all Z codes (received through FOIA), there are 13,000 lines, of which 5000 lines track to 81479 (probably meaning that 5000 Z codes are assigned to 81479).

Exact Sciences is working on an improved Cologuard test; here.

Opko 4KScore, code 81539, had 25,572 uses for $19.4M.  Early press here.  Parent company Opko was recently down 30% (here).

Category I MAAA Codes (excluding PLA  or Admin MAAA codes, the U and M codes) tallied $443M, though a large share was Cologuard 81528 and no other tests approached its volume or revenue. Leaving out Cologuard (which has a huge volume at about $500), the average MAAA paid $2500.  The average 81479 payment was $1500, while the average 81599 payment was $4000.

Category I MAAA Codes

Very Brief Blog: CMS Releases Part B Utilization Data at National and State Levels

Update: In January 5, 2020, the State level data is now posted by CMS only up to 2017, although the page for 2018 still exists, no data is there.  I have the 40mb of 2018 state data posted in October 2019; email me.

Each fall, CMS released data on the prior year's Part B CPT code utilization and payments both at the National and the State levels.   CY2018 data has been released.

See national data here.

See state level data here[missing as of 1/2020]

The state data is actually by historical Medicare Part B carrier zones, so some states have two parts (e.g. NCalif and SCalif).  Also, the Excel files are given cryptic contractor numbers and you need to decode with a provided PDF.  (Noridian Contract 03102 = CMS Excel 03102 = Arizona).

Separately, and on a longer delay, CMS releases physician-level CPT code payments.  These appear around June for the year 18 months earlier.  Here.

Example: BRCA

For example, the CY2018 National data file #04 contains 80,000-series codes, and in CY2018 CMS Part B paid for 22,615 uses of 81162 (BRCA Seq + Dup Del) totaling $50,896,967.

If we turn to the CY2018 state data, we find that Utah is File #03502, where there are 12,459 uses of 81162 paying $28,069,254. (In 2018 in Utah, older BRCA code 81212 had only 496 uses for $28,240.)   Research can be done at national and state levels for any code.

Example: FMI CDx Test

In national data, Code 0037U for Foundation Medicine FDA test was paid $33,396,800 (used 9,870 times).   However, I believe this code was effective in April 2018, so this may reflect only FMI payments in three quarters of 2018.   I believe there are 100,000-200,000 incident Medicare patients with advanced cancer (and multiples of that number as prevalent patients), so the FMI tests was not used in too many of them in CY2018, despite the NCD becoming effective in March 2018.

0047U is the Genomic Health prostate panel, which tallied 1,277 uses for about $4M in 2018.

Payments for all the other PLA codes 0U-70U were only about 4M more than payments for 0037U and 0047U.  Many had 0 CMS payments.

CMS Releases Revision for Comment of NCD for Next Gen Sequencing in Cancer

On October 29, 2019, CMS released a revision of its NCD for next generation sequencing in cancer patients.

The original NCD, released in March 2018, was criticized for being liable to ambiguity but potentially adverse to NGS testing in Stage 1/Stage 2 cancer patients or those who require repeat testing.   This criticism accelerated after some restrictive additional interpretative statements released by CMS in November 2018.   CMS held a comment period in May 2019, and received uniformly negative comments (here).
  • The tracking sheet for the NCD revision is here.
  • The revised NCD, for new cycle of public comment, is here.
  • I've put a PDF version in the cloud, here.
Scroll down to Appendix B for a redline proposal to revise the NCD decision.  I've also clipped the redline proposal as the bottom section of this blog.

Comments are during on Thanksgiving, Thursday, November 28, so many stakeholders may want to view Friday November 22 as an effective corporate pseudo-deadline.  The final NCD is due about 60 days from November 28, e.g. January 28.

Coverage at Genomeweb here.

What Happened in Brief

Warning - these are my first-pass interpretations; it could be that other readers will find textual ambiguities or problems I haven't noticed.
#   #   #

click to enlarge

NCD Coverage.  Previously, CMS nationally guaranteed coverage for NGS testing for NGS tests approved as CDx, if the patient had advanced cancer (metastatic, stage 3/4, etc).   For example, the FMI F1 CDx test was covered per its FDA labeling in such patients.

   >>> Now, CMS also proposes to nationally guarantee coverage for FDA approved or cleared tests (NOT necessarily CDx) in patients with ovarian or breast cancer, indications for germline testing, risk factors for inherited cancer.   Such patients have no stage restrictions like Stage 3/4.

In short, the areas of the NCD which nationally guarantee coverage pivot on being an FDA CDx test, and now it adds coverage for FDA tests that are not CDx for any indicated genes (per FDA) in breast and ovarian cancer.

LCD Flexibility.  Previously, CMS gave MACs discretion to cover non-FDA-approved (LDT) NGS tests, or FDA tests off-label, as long as it was ONLY for one test and ONLY in advanced cancer patients.

   >>> Now, CMS will let MACs allow coverage for NGS testing with LDT tests in patients with diagnoses other than breast and ovarian cancer (e.g. colon cancer), as long as they have inherited cancer risk factors and not prior tested with NGS.  There are no stage restrictions. 

#   #   # 

I see three issues right away.

  1. HBOC patients get only FDA tests that don't exist.  Breast and ovarian cancer patients can only get germline (inherited) testing for FDA cleared or approved tesst.  The only ones are for BRCA, and in fact, the Myriad germline BRCA test is not NGS (it's Sanger) and the Foundation BRCA test and Myriad MyChoice BRCA test are NGS but are somatic tumor tests.  Oops.
    1. The standard of care is almost certainly more than BRCA gene testing alone, it's panel testing, while only BRCA has FDA approvals (albeit defective ones relative to the NCD, as just stated).
    2. Myriad CEO Marc Capone also remarked that the NCD text here was very important and that the NCD points to FDA tests that don't exist, in a November 4 investor call (here).
  2. Still a one test per patient rule.  Only one test per patient, so not obviously helpful yet to recurrent disease detection.   
  3. Glitch re types and classes of testing.  The new passages about germline testing preclude any prior NGS method testing, but, for example, you might have to use one somatic FDA test for genes like EGFR and KRAS, and a different germline FDA test for BRCA.  The NCD doesn't seem to contemplate this problem.   For sure, there isn't any one FDA test today that covers both somatic CDx mutations and germline management mutations.   The NCD would cover both, as long as you don't do more than one.  Oops.
    1. The new parts of the NCD allow germline testing as long as "patient has not had a prior NGS test."  Really? Any prior NGS test?  FDA has approved an NGS HIV test - if you've had that, it's a prior NGS test, so you can't have BRCA testing?  Here.
NCD has confusing different sections applying to stage.  The new germline sections don't refer to stage.  And it's clear from the body of the NCD that CMS staff want to cover germline testing in breast and ovarian regardless of stage.  But CMS leaves in place a "section C" which is a blanket ban on NGS test coverage for patients who don't meet rules in "section B.1" which gives coverage to only metastatic/advanced patients.  (Eye-crossing.)

CMS distinguishes breast and ovarian cancer patients because they found the best germline literature in these groups.  Throughout the NCD, they are cautious to refer only to published NGS literature for outcomes, which is a little daffy, since a particular mutation sequenced by Sanger or by NGS gives the same outcomes.

There's also still a barrier to technology modernization that I dislike.  Many women with breast cancer get the Oncotype Dx or similar tests, which aren't NGS so they aren't under this NCD.  But if they were migrated to NGS platforms (and Mammaprint had been able to do this), they would be under the NCD but in very limiting or counterproductive ways. 

click to enlarge

Nerd Points.

The prior NCD covered certain tests, if that test hadn't been used in that patient before.   The new language covers certain germline testing, if the patient hasn't recent [any] NGS test before.  This is confusing.   Although it's now rare, someday, a patient might have e.g. NGS microbiology testing, which would be "a prior NGS test."  Who would track that?  Obviously an unintended issue.   Or a patient might have an NGS epilepsy panel as a child, but need NGS testing for cancer as an adult.  The NCD doesn't seem to contemplate that, since it blocks testing if the patient had "a prior NGS test."   This will hopefully be fixed in the final revision.


Medicare National Coverage Determinations Manual
We are seeking public comments on the proposed language that we would include in the Medicare National Coverage Determinations Manual. This proposed language does not reflect public comments that will be received on the proposed decision memorandum, and which may be revised in response to those comments.
Table of Contents
(Rev. in Redline)
90.2 Next Generation Sequencing (NGS) for Patients with Advanced Cancer
A.  General
Clinical laboratory diagnostic tests can include tests that, for example, predict the risk associated with one or more genetic variations. In addition, in vitro companion diagnostic laboratory tests provide a report of test results of genetic variations and are essential for the safe and effective use of a corresponding therapeutic product. Next Generation Sequencing (NGS) is one technique that can measure one or more genetic variations as a laboratory diagnostic test, such as when used as a companion in vitro diagnostic test.
Patients with cancer can have recurrent, relapsed, refractory, metastatic, and/or advanced stages III or IV of cancer. Clinical studies show that genetic variations in a patient’s cancer can, in concert with clinical factors, predict how each individual responds to specific treatments.
In application, a report of results of a diagnostic laboratory test using NGS (i.e., information on the cancer’s genetic variations) can contribute to predicting a patient’s response to a given drug: good, bad, or none at all. Applications of NGS to predict a patient’s response to treatment occurs ideally prior to initiation of such treatment.
B. Nationally Covered Indications
Effective for services performed on or after March 16, 2018, the Centers for Medicare & Medicaid Services (CMS) has determined that Next Generation Sequencing (NGS) as a diagnostic laboratory test is reasonable and necessary and covered nationally, when performed in a Clinical Laboratory Improvement Amendments (CLIA)-certified laboratory, when ordered by a treating physician, and when all of the following requirements are met:
1. Patient has:
  • either recurrent, relapsed, refractory, metastatic, or advanced stage III or IV cancer; and,
  • either not been previously tested using the same NGS test for the same primary diagnosis of cancer, or repeat testing using the same NGS test only when a new primary cancer diagnosis is made by the treating physician; and,
  • decided to seek further cancer treatment (e.g., therapeutic chemotherapy).
2. The diagnostic laboratory test using NGS must have:
  • Food & Drug Administration (FDA) approval or clearance as a companion in vitro diagnostic; and,
  • an FDA-approved or -cleared indication for use in that patient’s cancer; and,
  • results provided to the treating physician for management of the patient using a report template to specify treatment options.
^ Effective for services performed on or after [Month/XX] [Day/XX], [20XX], the CMS, proposes that NGS as a diagnostic laboratory test when performed in a CLIA-certified laboratory, when ordered by a treating physician and when all of the following requirements are met:
The patient has:
  • ovarian or breast cancer;
  • clinical indications for germline (inherited) testing,
  • risk factors for germline (inherited) cancer breast or ovarian cancer; and
  • not been previously tested using NGS.
The diagnostic laboratory test using NGS must have all of the following:
  • FDA approval or clearance;
  • an FDA approved or cleared indication for use in that patient’s cancer; and
  • results provided to the treating physician for management of the patient using a report template to specify treatment options.
C. Nationally Non-Covered
Effective for services performed on or after March 16, 2018, NGS as a diagnostic laboratory test for patients with cancer are non-covered if the cancer patient does not meet the criteria noted in section B.1. above.
D. Other [aka local coverage section]
Effective for services performed on or after March 16, 2018, Medicare Administrative Contractors (MACs) may determine coverage of other NGS as a diagnostic laboratory test for patients with cancer only when the test is performed in a CLIA-certified laboratory, ordered by a treating physician, and the patient has:
  • either recurrent, relapsed, refractory, metastatic, or advanced stages III or IV cancer; and,
  • either not been previously tested using the same NGS test for the same primary diagnosis of cancer or repeat testing using the same NGS test was performed only when a new primary cancer diagnosis is made by the treating physician; and,
  • decided to seek further cancer treatment (e.g., therapeutic chemotherapy).
^ Effective for services performed on or after [Month/XX] [Day/XX], [20XX], Medicare Administrative Contractors (MACs) may determine coverage of other Next Generation Sequencing (NGS) as a diagnostic laboratory test when performed in a CLIA-certified laboratory, when ordered by a treating physician, when results are provided to the treating physician for management of the patient and when all the following conditions are met:

The patient has:
  • a cancer diagnosis other than breast or ovarian cancer,
  • clinical indications for germline (inherited) testing,
  • risk factors for germline (inherited) cancer other than inherited breast or ovarian cancer, and
  • not been previously tested using NGS.

Monday, October 28, 2019

Very Brief Blog: AMA PLA Code Watch: 9 New Proposals

AMA takes applications for PLA codes (proprietary lab codes) quarterly, and processes them very rapidly.   For codes accepted October 10, they were posted for comment October 17 and comment closed October 22.   They will come to a panel vote in early November, be published January 1, and effective April 1.

Ten New Applications

There were ten new code applications in the current cycle.   There is also one code deletion ("detection of 120 medications, substances, supplements, foods") and two code revisions (therascreen FGFR kit, therascreen  PIC3CA).

The AMA calendar for 2020 is not published yet, but applications are about the 10th of each quarter (about January 10, April 10, July 10, October 10).

  • PLA home page here.
  • PLA codes for October 2019 here.
  • PLA calendar home page here.
The next publicly available PLA call is Tuesday, November 12, 7 ET 4 PT.   However, this calls are often quite short (the codes may be stated to be a "consent agenda" with rapid "aye" votes and the meeting closes.)

Friday, October 25, 2019

Very Brief Blog: CMS releases Panel Actions from September 2019 AMA CPT Seattle Meeting

About a month after each AMA CPT public meeting, AMA releases summary actions of the panel.  They've just released the summary actions for the Seattle meeting in September.

Find them online here.

There were about 80 main agenda items and about 10 executive committee actions (merely editorial).  Using a word search function, I tallied 46 accepted proposals, 5 postponed, 18 withdrawn, and 13 rejected.   Applicants get feedback regarding association positions on their proposals, and typically most applicants withdraw if it is clear that major associations will vote down the item.  If we tally the "withdrawns" as "rejected," they would total 31.

On the lab side, a number of genes were lifted from CPT Tier 2 to Tier 1.  A gene panel code for epilepsy was withdrawn.  Several infectious agent applications were also withdrawn.  A new MAAA code was created for cutaneous melanoma (815XO).


Some of the Category I codes are revised for January 2020 instead of January 2021, even though this September meeting occurred after the 2020 book had gone to press.  New to me, and potentially a little confusing.

Since it was revised, my attention was drawn to process code 3170F, used when flow cytometry baseline studies are performed prior to initiating hem-onc therapy.

Newswatch: German commission dings LBx cancer researcher re molecular screening tests

On October 25, 2019, Science published an open access news item with the provocative title, "German university finds ‘severe’ misconduct by researcher who promoted questionable cancer blood test."  I recall news items about this test over the past year; it concerns a new LBx test sensitive and specific enough to be used for cancer screening.   Now it's being labeled "questionable" in some news reports.  (For early favorable press, here.  For skepticism as far back as April 2019, here.)
  • See the new article by Hinwerk Feldwisch-Drentrup in Science, online here.
In fact, the article describes a flip-flop series of events.  In February 2019, a press release described the screening cancer liquid biopsy test as "revolutionary."  In past months, however, "the hospital has been rocked by scandal" including "criminal proceedings."   According to Science, "in July, an external panel established by the hospital board presented the results of a preliminary investigation, criticizing the press releases as premature and 'yellow press instead of serious science communication.' "

Then there is another twist, the accusatory report was pulled from its public release, and a court ordered a press conference on the alleged wrongdoing to be cancelled.

For full details, see Science link above.  See website of the company HEISCREEN, here.

Some Heiscreen webtext after the break.

Wednesday, October 23, 2019

Very Brief Blog: Clinical Labs Competing on Technology's Cutting Edge; Spotlight on Caris

In recent months, we've seen announcements from labs like Ambry and Invitae that they are adding services related to RNA sequencing to improve their accuracy in germline genetic mutations (here, herehere).  See a publication from Ambry, Karam et al., here.

I ran across an interesting article on marketing molecular oncology, featuring Michael Sullivan, SVP and Chief Commercial Officer at Caris.  It's at trade journal MM&M, here.  

Along the way, he mentions some of the new technologies that Caris is working on and launching.  These include:
  1. Whole Transcriptome Sequencing (WTS), looking at RNA sequence of 22,000 genes
  2. Modifications to Caris's current tumor DNA panel
  3. Germline-hereditary panel to be introduced
  4. NGS profiling tool that "applies our DEAN artificial intelligence tool to our library of cancer cells"
  5. A liquid biopsy assay
Caris currently runs 4000 cases per month, about 50,000 per year.  Sullivan comments Caris has tripled its US sales force in "the last six to nine months" and has transitioned onto  One year ago, in October 2018, Caris announced $150M in growth capital (here).  This spring, Caris received FDA breakthrough status for its MI Transcriptome CDx test (here).

Regarding DEAN, Sullivan comments, "We’ve developed next-generation profiling (NGP), which uses a deliberation analytics tool (referred to as DEAN) in an AI-powered platform to power analysis of outcomes and biomarker data. We use all of this to generate potential results that will help diversify our offering and improve the overall amount of information we’re able to provide to physicians and clinics."


MM&M is a medical marketing journal with a lot of open access content.  There is a partner journal Exits&Outcomes which focuses on the digital health industry ($200/year), here.

It's Official: Launch of CMMI Diabetes Prevention Program Was A Hot Mess

With much fanfare, CMS announced an Innovation Center program for diabetes several years ago.  The program underwent review by CMMI, and literally two annual cycles of rulemaking (hundreds of pages of rulemaking) before it was launched.

I wrote an analysis of potential flaws in the program and data in 2016, here.

The program started in CY2018, and by late spring, only 3 providers had been able to enroll (Kaiser Health News, April 2018, here.)   In providing a link to the KHN article, I noted that the Diabetes Prevention Program was just one of several CMS efforts that got bogged down in dizzying complexity (here.)    About the same time, Spring 2018, GAO decried the low productivity of the Center for Innovation (here).

New News: Bad News

This week, Politico (open access) runs an article by Darius Tahir, "Languishing Medicare Diabetes Program Frustrates Provider."  (With a picture that makes the Humphrey Building, the DC home of HHS, look like a prison designed to keep diabetes providers from getting in.)

Here's the key passage:
A flagship Medicare program that HHS expected to engage up to 110,000 people annually each year in measures to help them avert Type 2 diabetes only managed to enroll about 200 people last year, according to an analysis of CMS data.
He continues, "People familiar with the program said a heavy regulatory burden had kept many providers from offering the diabetes prevention service. CMS’s rollout of the program, these sources said, could presage difficulties for any future programs..." with similar goals.

Do The Math

HHS under Obama announced the program with great fanfare in 2016 (see original press release here, fact sheet here, NPR article here.)  They estimated the program would save $2600 in the first year for every enrollee.  (I decimated this phony math in my 2016 blog, above).  With 200 enrollees, let's hope the program saved the $600B Medicare program $520,000.   At that rate, it would take 23 years until 2040 just to pay off the $11.8M pilot study that kick-started the protocol.*

Providers complained that the proposed reimbursements were intolerably low (here).  Some providers told Politico they still haven't been paid for 2018 claims.


Politico notes that about 400 patients were enrolled by 1Q2019.


Track Darius Tahir at Twitter, here.   He recently retweeted (rt'd, in the lingo) a 2019 article in Politico about Seema Verma's alleged high costs for media consultants (here).


For a 2016 article on overuse of the concept "pre-diabetes," here.

Philips and Deverka Publish Major Review of Lab Benefit Managers in Health Affairs

On this blog a few weeks ago, I highlighted new articles in CAP TODAY and at 360DX on lab benefit management in its different forms (here). These range from (A) software that adapts guideline-based care or appropriate use criteria into EHR's at the physician's point of care to "traditional" lab benefit management systems which are often consultancies to (B) payers which implement denials or prior authorization.   Concurrently, there were Hill hearings on the burden of prior authorization (links in the blog cited).

October 23, 2019

At Health Affairs, UCSF's Phillips and Deverka publish a major review of the LBM industry for genetic testing.   It's open access here.  

In its August 2019 quarterly call, Myriad noted it had lost as much as $50M in recent quarters to more aggressive use of LBMs (here, here).  It's an impactful business and policy area.


Concurrently, the CRO-based consultancy Evidera publishes a nice suite of open-access articles on real world evidence, pragmatic randomized trials, and use of RWE for health technology assessment and payers.  See index here.


For an Oct. 24 article on payers and ICD10 claims denials, here.


One thing that LBMs do is manage prior authorization.   There are two articles on prior authorization published online in February 2020 in JAMA (see Resneck; see Gaines).  While these two op ed's focus on drug prior authorization, there is a logical crossover with diagnostics prior authorization.   

Monday, October 21, 2019

Medicare Trivia: MACs Can Now Have LCD Public Meetings as Rarely As They Want

For many years, until January 2019, Medicare's instruction manual for LCDs required them to be presented in a cycle of no longer than every 3-4 months.   The presentation cycle included a public comment meeting, a CAC (contractor advisory committee) meeting, and a 45 day public comment.

Among the numerous changes to the LCD manual in January 2019, any references to the frequency of LCD cycles have been dropped.   Want to release a batch of new LCDs every two months?  That would be OK, but releasing new LCDs only once a year would also be OK.

I've heard one MAC suggest it was downgrading LCD releases to no more than twice a year, such as an April and a September cycle. 

Nerd Note: Details

I have a copy of the Internet Only Manual - instructions for MACs - Chapter 13, vintage 2015, v. 608, which describes frequency rules in section  Section F there used to state that LCD meetings must be held a minimum of 3 times a year, with no more than 4 months before meetings.

The currently effective version is v. 863, issued 2-22/2019.  It's been completely reformatted, but Section 13.2 describes the LCD release and comment process.

While there are clearly rules for submitting a New LCD Request and clear rules for an LCD Reconsideration Request, both of which the MAC must acknowledge in writing, there is simply no remark anywhere as to how often, or at what delay, the MAC has to do anything about the submitted request other than acknowledge it as received.

Notwithstanding, an LCD Timeline Is Easily One Year

Here's a potential timeline:
  • December 2019: You have your publications, request letter, overview PPT deck all prepared and request a MAC meeting.
  • January 2020:  You get a telecon or fly-in meeting with the MAC medical director(s)
  • February 2020: MAC considers what to do, does its own literature search, calls its own experts
  • March 2020:  MAC writes up LCD text, circulates internally, finalizes
  • April 2020:  Most sources say CMS requests 4 weeks to review draft LCD "to check for conflicts with NCDs."
  • May 15, 2020:  MAC posts draft LCD online. MAC sets public meeting date (June 15) and a deadline for submission of public meeting comments (June 1).
  • June 15, 2020: Public meeting.
  • June-July 2020:  Public 45 day written comment period.
    • This may start variably with the LCD release date (May 15) or the actual public meeting date (June 15)
  • August, September, October: MAC reviews those public comments
    • 100+ days is typical based on real world data
  • November 15: MAC releases new final LCD and its responses to comments
  • January 1, 2021:  After 45 days, LCD is active.
Regarding the timeline for comments review, MolDx (for example) released 8 LCDs in May, comment closing June 20, and as of late October, only 1 of 8 has been released-to-final.  That's 120 days or 4.0 months and counting for 7 of 8 LCDs.  Hence my remark that >100 days for review is typical.  (Here).

While each cycle of LCD releases must be accompanied by a public comment meeting, holding CAC meetings (ever) is now optional, and could occur before, after, or never, relative to the date of an LCD release.


I've put copies of documents referenced into a zip file in the cloud, here.

For the New Yorker cartoon, "How about never? Is never good for you?" - here.

Saturday, October 19, 2019

Brief Blog: CMS Solicits New MEDCAC Advisory Panel Members; November 18 deadline

SEE CMS Correction of Email Address for Submissions:

For many years, CMS has had an official standing advisory committee for coverage decisions and evidence called the MEDCAC - Medicare Evidence Development and Coverage Advisory Committee. 

It has its own webpage, here.  See the Index of Meetings, here.

On October 21, 2019, CMS published a Federal Register announcement that it is seeking new members of the committee.  See 84 Fed Reg 56193-4, here.  The deadline is 30 days (November 18).

MEDCAC consists of a "pool" of 100 members, including 90 standing members (10 of this 90 are patient advocates) and 10 industry members.  Membership is designed to span medical specialties as well as public health, epidemiology, biostatistics, economics and ethics.  25 positions expire in 6/2020, the open slots being 2 industry representatives, 5 patient advocates, and 18 at-large professionals.   Candidates serve 2-year terms.


Unlike FDA advisory panels, MEDCAC panels at CMS are rare.  Recent year tallies include:

2019: Zero so far.

2018:  Two (TAVR and CAR-T)

2017:  Two (Heart failure treatment; Bariatric surgery)

2016:  Two (Treatment resistant depression; Lower extremity venous disease)

2015:  Two (Molecular tests for cancer prognosis; Lower extremity arterial disease)

In earlier years, meetings were more frequent, e.g. 5 in 2010 and 5 in 2009.

Friday, October 18, 2019

Hold On To Your Hats: MolDx Publishes Strict New Panel Test Coding Article

Hold on to your hats. Test panel coding has been a hot topic in the past 12 months.  In November 2018, OIG published a report about potential overbilling of clinical chemistry panels (here).  Also in November 2018, CMS published strict new rules against stack coding of NGS tests (the national Correct Coding handbook; here). 

There's even been a court case settled in which documents were filed pivoting on panel coding issues (here).

Finally, MolDx continues to have an older article against code stacking and about panel billing (of either PCR, Sanger, or NGS tests), dating from 2014 (here).  This is called "Test Panel Alert M00101."

What's New! A57503!

Flash forward to October 2019 and MolDx drops a new article that stringently defines single gene versus multiple gene test orders (whether PCR, NGS, etc) and how they should be coding.  MolDx appears set on enforcing the Correct Coding handbook cited above. 

The brand new article is called "Billing and Coding: MolDx: Testing of Multiple Genes" (A57503).  Find it on the CMS website here and see a cloud PDF copy here.

A Little Dizzying

The instructions are fairly lengthy, but seem to include using appropriate CPT coding for known panels (e.g. 81432, a ten-gene breast cancer panel), but also covers other situations such as billing for "two or more genes on the same day of service."   Whereas the Correct Coding edits referred above only apply to NGS testing (and require 2 or more genes to be billed with 81479, which most payers and MACs would find manual and cumbersome), this new article A57503 is like the Test Panel Alert of 2014, because it applies to any method, whether NGS or not. 

It's unclear if this article supersedes the Test Panel Alert for MolDx, that is, if the T.P.A. will be withdrawn from the website.  If both MolDx articles are in effect, and the CCI edits, it may be a little confusing.  I've put a cloud copy of the T.P.A. article here.

Three Simultaneous Test Panel Instructions: MolDx, CMS, MolDx


I've clipped the core text of A57503 below the break.

Noted Life Sciences Investor Brook Byers Gives Stanford's Annual Fogarty Lecture

On October 21, 2019, a leading Silicon Valley life sciences investor, Brook Byers, gave the 21st annual Fogarty Lecture at the Biodesign Center at Stanford.  The talk streams online at YouTube.

  • One-hour talk at YouTube, here.
  • Byers' page at the Kleiner Perkins Caufield Byers website, here.
    • Wikipedia here.  Lists the long line of top-tier companies where Byers has been a board member.
Stanford (Stanford-Byers) Biodesign center here.  There's also a Byers Hall at UCSF.
Fogarty Institute for Innovation here. (Dr. Fogarty at Wikipedia, here.)

Byers highlights four influential books, John Doerr's Measure What Matters, Bill Campbell's Trillion Dollar Coach, Thomas Byers et al. Technology Ventures, and the Stanford Biodesign textbook, "Biodesign," by Paul Yock and coauthors.

Companies highlighted in his talk include Genentech, IDEC Pharmaceuticals, Oculeve, Optimedica, CareDx, and Verana. 

Thursday, October 17, 2019

Vegas Conference Offers Who's Who of Trump Administration Health Leaders

I was skimming trade journal headlines and the annual HLTH (health innovation) conference in Las Vegas has posted a who's who of health policy leaders on its agenda.

The conference is October 27-30, 2019, at the MGM Grand in Las Vegas.

The agenda right now includes:
  • Seema Verma, head of CMS
  • Amy Abernethy, deputy director of FDA
  • Joe Grogan, head of National Policy Council
  • Eric Hargan, HHS Deputy Secretary
  • Tomas Philipson, Chairman of Council of Economic Advisors
  • Adam Boehler, until very recently head of Center for Innovation at CMS
Also on board is Sen. Michael Bennet (D-CO), a presidential candidate.  (He didn't make the cut for this week's Dem debate, though).

See:  Trade press at MedCity News here.   Conference website here.  Registration is $2600.


If you get bored at the conference, there are two WeWork locations in Las Vegas.

Tuesday, October 15, 2019

Very Brief Blog: Myriad MyPath Melanoma Is the Newest ADLT

Myriad Test Is the Sixth ADLT

While the ADLT process for new lab test pricing has been in place for several years, only a handful of tests have been granted ADLT payment status.  This list now totals six tests, with the addition of the Myriad myPath Melanoma test in September 2019, "initial period" pricing running from October 1, 2019 to June 30, 2020.

Two Types of ADLTs

ADLT is a special annually revised payment status for tests that are "sole source" tests (not IVD kits) which must be either (A) MAAA type tests covered by Medicare, or (B) sole source lab tests with 510(k) approval.   Five of the six ADLTs are MAAA ADLTs; only one is an FDA-pathway ADLT.  The abbreviation stands for Advanced Diagnostic Laboratory Tests.

ADLTs Usually Have PLA Codes

Of the six ADLT tests, 4 are represented by PLA codes ("U codes"), one has a Category I code (81538, Biodesix Veristrat test), and one has a TBD code (Castle DecisionDx-Melanoma). 

ADLTs usually don't go through the gapfill/crosswalk process, as they are priced for the first 9 months at list price and thereafter at an annually-reset market price.  However, an ADLT can qualify as an ADLT late in its life cycle, when it already has a code and CLFS price.  In that case, I believe the annual mark-to-market pricing policy starts, but there is no initial "list price" phase.

Seeing ADLT Prices

On the approved ADLT spreadsheet, CMS lists only the "New ADLT Initial Period" price (the list price) and not the later, annually reset price which appears on the Clinical Laboratory Fee Schedule on a calendar year basis.   ADLT prices range from $1950 (Myriad myPath Melanoma) to $7193 (Castle DecisionDx Melanoma). 

See screenshot below (click to enlarge).

click to enlarge

Where To Learn More

See the ADLT home page here.   It includes complicated directions about how to apply for ADLT status and details about the ADLT rules. 

The PDF showing current ADLTs (as clipped above) is here.

Very Brief Blog: FDA Releases Suite of Digital Guidances, including Clinical Decision Support

In late September 2019, FDA released a suite of guidance documents aimed to encourage the development of digital health tools and to continually modernize the agency's approach.

They were released with a statement from Amy Abernethy MD PhD, an oncologist with industry experience at Flatiron who is currently Principal Deputy Commissioner.

The detailed statement, with links to multiple new and existing agency documents, is online here.
  • One of the most important is a draft guidance on Clinical Decision Support - here.
    • Comments open for 90 days.
    • Summary at Ropes & Gray here.
    • Genomics tie-in:  CDS is considered important in genetics, precision medicine and oncology due to the tsunami of new facts.
  • See the 2017 FDA digital health plan white paper here.
  • FDA also finalizes changes to existing guidances based on 21st Century Cures Section 3060 - here.
  • FDA aims to harmonize where possible to the IMDRF - the International Medical Device Regulators Forum - here.
  • McDermott Will & Emery on the multiple FDA documents here.
Several existing guidances were also updated; see Abernethy's statement here.  These include the guidance for mobile applications, wellness devices, off-the-shelf software used in medical devices, and medical imaging software.


In December 2019, SCIENCE had an interesting article on FDA, artificial intelligence, algorithms, and "regulatory lockdown" by Babic et al. - here.

Wednesday, October 9, 2019

Very Brief Blog: ICER Launches First Annual Report on "Unsupported Drug Price Increases"

Ouch!  The Boston-based Institute for Clinical and Economic Review (ICER) has published a new 125 page report on "unsupported price increases" for drugs in the US.

  • See ICER's home page here.
  • See Press Release on new report here.
  • See 125 page PDF report here. (*)
  • See trade press at BioPharmaDive here.


(*) In Windows, I had to right click on the VIEW button and select "save link as."

Brief Blog: CMS Announces New Anti Kickback Proposals; Make Value Based Care Easier

Separate but related:

See a Seema Verma blog on wider program integrity efforts from September here, and a CMS Program Integrity "Request for Information" (due Nov. 20, but worth reading even if you missed the deadline) here.

On October 9, 2019, CMS and OIG announced several hundred pages of pending proposed rulemaking intended to lighten coordination of care barriers for value-based care, such as ACO's.  

The rules were published in a website version, which is prior to both the "public inspection" copy at and the official Federal Register publication, which typically follows in 5-15 days.

Associated Press summarizes as follows:
Azar said the idea is to encourage hospitals, doctors and other service providers to enter into formal “value-based arrangements,” in which they collaborate to improve care for patients and commit to delivering measurable results. 
For example, a hospital may send a kidney patient home with technology to monitor critical health indicators and automatically transmit back any signs of problems. Under existing rules, such an arrangement could be interpreted as the hospital providing the patient an illegal “inducement” to continue using its services.
The original rulemaking for ACO's faced some conflicts with similar kickback laws.  Also on the program integrity front, a few weeks ago CMS released final rulemaking for program integrity, generally making it easy to kick abusive providers out and keep them out (here).

See links below:
  • CMS press release, including additional links about anti kickback statute (AKS) and Stark Law, here.
  • OIG proposed rulemaking:
    • OIG Federal register version, October 17, here.
    • 84 Fed Reg 55694-55765 (72pp), 10/17/2019.
    • Comment to 12/31/2019.
    • Focus on "anti-kickback"
  • CMS proposed rulemaking:
    • CMS Federal Register version, October 17, here.
    • 84 Fed Reg 55766-55847 (82pp), 10/17/2019.
    • Comment to 12/31/2019.
    • Focus on "self-referral"
  • CMS Fact Sheet on Self-Referral changes, here.
  • OIG Fact Sheet on Safe Harbor changes, here.
  • Coverage at Associated Press here.
  • Coverage at Reuters, here.
  • Coverage at Lexology, here.  
  • National Law Review here (McDermott), here (Drinker Biddle).
  • Wilson Sonsini here, Greenberg Traurig here
  • Wynne et al. at Health Affairs, here.
  • Mintz produced a three-part series of the proposes rules:
Note that in September 2019, CMS released new regulations making it hard for fraudulent providers to use CMS as a revolving door through a facade of name changes or corporate shells; here.

Trump Name Watch

Some CMS press releases very prominently feature the President's name, including in headlines. (E.g., this week, see "Trump Administration Empowers Nursing Home Residents," here.) 

In contrast, this lengthy press release only mentions the President's name in passing ("Regulatory reform has been a key piece of President Trump's agenda.")  

Press Release Extract

These are examples of problems that the new amendments are intended to fix:

"Below are examples involving coordinated care, value-based care, data sharing, and patient engagement activities that, depending on the facts, could currently be difficult to fit under existing protections and could potentially be protected by the Stark Law, Anti-Kickback Statute, or Civil Monetary Penalties Law proposals if all applicable conditions are met:
Hospitals and physicians could work together in new ways to coordinate care for patients being discharged from the hospital. The hospital might provide the discharged patients’ physicians with care coordinators to ensure patients receive appropriate follow up care, data analytics systems to help physicians ensure that their patients are achieving better health outcomes, and remote monitoring technology to alert physicians or caregivers when a patient needs healthcare intervention to prevent unnecessary ER visits and readmissions. 
A physician practice could provide smart pillboxes to patients without charge to help them remember to take their medications on time.  The practice could also provide a home health aide to teach the patient and the patient’s caregiver how to use the pillbox.  The pillbox could automatically alert the physician practice and caregiver when a patient misses a dose so they could follow up promptly with the patient.   
A local hospital could improve its cybersecurity and the cybersecurity of nearby providers that it works with frequently.  To do so, it could donate, for free, cybersecurity software to each physician that refers patients to its hospital.  The hospital and the physicians often share information about their patients, so it is important that there are no weak links that might compromise everyone else.  The software would help ensure that hackers cannot attack the physician’s computers.  Improving each physician’s cybersecurity would help prevent hackers from spreading the attack to other physicians and the hospital. 
To improve health outcomes for patients with end-stage kidney disease, a nephrologist, dialysis facility, or other provider could furnish the patients with technology that is capable of monitoring the patient’s health and two-way, real-time interactive communication between the patient, facility, and physician.  In addition, the facility could equip the physicians with data analytics software to help them monitor patients’ health outcomes. 
In an effort to coordinate care and better manage the care of their shared patients, a specialty physician practice could share data analytics services with a primary care physician practice."

Sunday, October 6, 2019

Very Brief Blog: Journal Watch: Flurry of Colon Screening Articles

Last winter, I noticed that Medicare has a special policy for frequent colonoscopy benefits, and it covers everything except genetics. 

If you have a family history of colon cancer or polyps, or a family history of familial polyposis, or family history of non hereditary colorectal cancer, Medicare will be happy to give you annual rather than per-ten-years colonoscopy.

But wait.  If you simply don't have the gene that runs in your family, you still get the Medicare benefit because you have a "family history;" as far as genetics, Medicare doesn't care. 

If you are gene-positive and therefore at high risk, but an orphan, you don't get a benefit, unless you can get an adoption agency to release records that your mom died of colon cancer.  CMS chugs on exactly as if Mendel never lived.  (Blog here).

Six Current Publications Related to CRC Prevention Policy

In the last few weeks, a flurry of research articles potentially relevant to colon screening policy and practices.

I'll rattle them off here and readers with further interests can use the links to go further.

1) Outcomes With Higher or Lower Intensity Screening

In September 2019, in Annals of Internal Medicine, Meester et al. studied the benefits of high-intensity screening, here.  Higher intensity screening had benefits, though small, and was cost-effective, though on the high side (<$100,000/QALY).  Comes with an op ed by Weinberg & Schoen on advising patients (here), and a "patient page" here.

Weinberg cites recent studies that up to 50-60% of patients have at least one tiny polyp on screening, thus qualifying for more intense screening (Weinberg citing Rex; and see Pilonis).

2) Outcomes With FIT, Colonoscopy, Sigmoidoscopy: Similar

In October 2019, in BMJ, Buskermolen et al. studied the differential outcomes expected, by modeling, assuming patients used either FIT, or sigmoidoscopy, or colonoscopy.   The conclusion: all were OK, and it didn't matter much.   Here.

3) European Guidelines: "Risk" More Important than "Age"  

In October 2019, a European guideline appeared in BMJ, Helsingen et al., recommending that CRC screening decisions be based on risk, not on age.

This would be quite different than US, where USPSTF and AHRQ recommendations are primarily based on age and definitely don't consider genomics (in or out) for either mammography or colon screening.  Article here, trade press at 360Dx here.

4) Towards Using Genetics in Population Health Recommendations for CRC Screening

In October 2019, in Cancer Epidemiology, Biomarkers, Prevention, McGeoch et al. published a systematic review of risk prediction models for CRC incorporating genetic variants.

They conclude that, "Public health modeling studies suggest that, if determined by risk models, the range of starting ages for screening would be several years greater than using family history alone."

See similarly Jeon 2018, here.

5) Cologuard: CMS Funds Cost-effectiveness Study

In September 2019, there was a media splash when a CMS-funded study appeared ascribing low cost-effectiveness to the Cologuard test.   See PLOS One, Naber et al., here.   Trade press here.  A STAT Op Ed on Cologuard here.   (My blog here.  I agree there are some errors, like lowballing the CMS cost of colonscopy.)

If this had appeared, say, in JAMA, there would be some letters pointing out errors, easily pegged to the article and easily found, co-listed in PubMed, etc.  Not so at PLOS One.

6) 2018 News Meets 2019 News: CRC Screening at Age 45 not 50

In late 2018, American Cancer Society new guidelines that CRC screening should start at 45, not 50.  Here

New tie-in this fall:  In September 2019, FDA approved Cologuard for patients 45-50,  here.


How CMS Makes New Preventive Benefits 

CMS can make new preventive benefits through an NCD if they are first endorsed by USPSTF.

Separately, for exactly two cancers, prostate and CRC, CMS can make screening benefits by acting on its own (without USPSTF) if it wants to.   For a discussion of the underlying patchwork of laws, here.



Essay comparing making new laws to making sausages.  Here.  Probably could adapt to "making new Medicare CRC preventive screening benefits."

But that's in keeping with the spirit of the original sausage metaphor.  The idea "Don't watch how sausage is made," from Bismarck, stated in full: "Don't watch how either law or sausages are made." Or colorectal cancer preventive screening benefits.

Original online here.

Essay, on the original Bismarck quote,"The less the people know about how either laws or sausage are made, the better they will sleep."  ("Je weniger die Leute wissen, wie W├╝rste und Gesetze gemacht werden, desto besser schlafen sie!“)