Monday, October 31, 2016

Open Access Publications from CMTP: "Barriers to NGS" and "Clinicians and Pragmatic Trials"

The Center for Medical Technology Policy (CMTP) in Baltimore has published two recent, open access publications on genomic test policy and evidence collection.

The first, in Applied & Translational Genomics, is "Barriers to clinical adoption of next generation sequencing: Perspectives of a policy Delphi panel."   By Messner et al., find the open access paper here.  (Abstract clipped below the break).

The second, in Contemporary Clinical Trials & Communications, is "Understanding practice-based research participation: The differing motivations of engaged vs. non-engaged clinicians in pragmatic clinical trials."   Also by Messner et al., find the open access paper here.

Sunday, October 30, 2016

MolDX Updates General Q&A on the MolDX Program (V10, M00086, 10/24/2016)

On October 24, 2016, MolDX updated its general Q&A, which is its Document M00086, to Version 10.   The MolDX link is here.

I don't have an archive of Version 9 to list the most recent updates.  I did have an archive of Version 7, from February 2016.   Updates that occurred at some point from February 2016 to October include:

  • WPS:  WPS MAC added to the program (J5, J8).
  • AMA HAS ITS RULES...:  Continues to affirm, "The effort to publish CPT codes for molecular diagnostic techniques...and the MolDX program are NOT related or interdependent."
  • WHAT CPT CODE?  [Q.34] What resources are available to help me determine the CPT code to bill?   Applications are reviewed for CPT code mapping by the MolDX team....the Team will notify the test provider with final coding, billing, coaverage, and if applicable, fee for the test.
  • G0452 Interpretation:  Molecular interpretation G0452 must be submitted with a Z code for the test.
  • NGS MUST USE SPECIAL CODES:  [Billing Q.10] A test was registered on an NGS platform that is a gene on the MolDX  approved gene list.  Why was the claim denied?  A:  MolDX does NOT consider the T1 and T2 descriptions appropriate for genes interrogated on an NGS platform.  Please reference the coding and billing guidelines in reference article M00130.

Note that the following tests "may not require a TA":

  • NYS certified tests that are industry accepted and have established clinical utility.
  • FDA kit used in unmodified form.
  • PMA-approved tests, and stacking codes, a unique identifier is required, but a TA is not.
  • Tests that use existing or novel technology for targets with established clinical utility.

A cloud copy of V10 is here.

Tuesday, October 25, 2016

Brief Blog: Friends of Cancer Research Holds MoPath/FDA Senate Briefing (October 13, 2016)

On October 13, 2016, the Friends of Cancer Research held a one-hour Senate briefing in the Dirksen Building entitled, "The Future of Precision Medicine and Patient Care: Current Landscape of Genomic Testing."

The session pivoted on a September 2016 publication, "Use of FDA-Approved and Laboratory-Developed Tests in Advanced Non-Small Cell Lung Cancer: Results of a Retrospective Market Analysis" (here) [*].  The survey found  87% use of LDT EGFR tests and 50% use of LDT ALK tests.  Test rates for EGFR and ALK were 95% and 84%, respectively, in adenocarcinomas, somewhat lower among all NSCLC.  (Some recent guidelines suggest that due to ambiguity about diagnosing adenosquamous or some adenocarcinoma cases, all cases should be tested for these markers.)  Testing for ROS1 was circa 30%; other markers were 5% or less.

Open access coverage of the Senate event was available at Bloomberg here or here, and at Medpage Today here.

Hold on to your hats:  Dr. Shuren promoted FDA regulation for uniform quality controls, while adding that FDA regulation would not be excessively burdensome.  Other panelists argued that FDA regulation was costly, burdensome, and unnecessary.  

FOCR raised concerns there may be too much test variability and that holding all tests to the same regulatory standards and quality would lead to more assurance regarding reports.

Panelists include Jonathan Leff, Deerfield Institute, Mike Pellini, Foundation Medicine; Jeff Allen, FOCR; Andrea Ferris, LUNGevity; John Iafrate, MGH; and Jeff Shuren, Director of CDRH, FDA.

Other recent Hill testimony by FOCR is here (Senate HELP committee, September 20, 2016).

[*]  The researchers reached out by mail to about 8000 oncologists, of which about 250 completed an online survey for an honorarium; representing about 750 patients.

Cloud archive of FOCR October 13 webpage, here.

Brief Blog: Cloud Archive of CMS CLFS Pricing Decisions

Someone asked me today about 2012,2013, 2014 era CMS CLFS pricing decisions for crosswalk/gapfill.   Currently, CMS archives preliminary and final decisions back to CLFS-CY2013 (decisions in summer and fall, 2012).

However, at some point they may delete the back archives, such as they are.  I have stored in the cloud in a Zip file what CMS has available as of October 2015.  Here.

Extracting stamp dates inside the PDFs, this is an estimate of when preliminary and final decisions are released.  Note that these data are based only on PDF file dates, and actual public release dates may have been earlier (if the dates we see are slight PDF revisions) or later (if documents were finalized significantly earlier than released).

Update: The CY2017 final was released 11/21/2017.

Note that in genomics, many "final" decisions are for gapfill, which then triggers a twelve month gapfill process.  Preliminary decisions (following the July meetings) are released as early as August 31, and as late as October 14.   This year's Preliminary release was the latest of the group, October 14, but in fact it was only a few days later than the prior two years.  

Final decisions are as early as November 6 and as late as December 23, but often in December.

Monday, October 24, 2016

Brief Blog: FDA and Software as a Medical Device (SaMD)

In mid October, the FDA released a new draft guidance on "software as a medical device," with an unusually robust effort at international harmonization.   The document focuses on definitions and risk stratification.  Note that clinical decision support and other advanced software approaches increasingly blue the traditional lines between a lab report, an EHR, and decision support.  For example, when we read about alliance between Quest and IBM Watson, creating Watson Genomics Powered by Quest Molecular Pathology, the lines between CLIA, FDA, and "none of the above" are being drawn on new territory.

For a trade journal article at MedCity, here.   The FDA's formal Federal Register announcement is here.  The FDA's draft guidance document (46pp) is online here.  Note that the FDA's document is an FDA cover page on top of an international guidance proposed by a working group (45pp, here).

Brief Blog: CMS and FDA make "Parallel Review" Permanent

CMS and FDA have run the "parallel review" program for several years, one example of which was the Exact Sciences Cologuard screening test, which received a draft NCD on the same day as its FDA approval.  (The draft NCD was finalized in a couple months).

CMS and FDA have released policymaking that the program will be permanently available.  For the 8-page Federal Register notice, here.

Thursday, October 20, 2016

Twists and Wrinkles in the OIG's New Report on Medicare and Chiropractic

In October 2016, the OIG issued a new report finding "hundreds of millions of dollars" in Medicare overpayments for chiropractice care (report here, trade coverage here.)

Most of the report finds that chiropractic services were provided for maintenance, not acute problems, and asserts that maintenance care is disallowed by a program manual statement (Benefits Manual Ch. 15:240, here.)

CMS responded it could not write an NCD since there was no data for reasonable and necessary determinations of chiropractic services; but if CMS hasn't found data for an NCD, it seems it would be equally difficult to write an LCD.   Chiropractic services for subluxation of the spine are a covered benefit under Medicare statutes (SSA 1861(r) and (s)).

A nonsensical legacy Medicare policy that "maintenance" services are disallowed (in physical therapy) was thrown out by a court in 2013.  It is an interesting point of policy whether that 2013 judgement should apply to all legacy Medicare bans on maintenance services or only the type of maintenance services in the original case (Jimmo v. Sebelius).  More after the break.

Wednesday, October 19, 2016

Brief Blog: Fortune reports unsealed 66 page investor lawsuit against Theranos.

It's been known there is an investor lawsuit against Theranos, and Theranos has issued a press release denying the charges as unfounded.

On October 14, Fortune reported the 66 page lawsuit was now unsealed (article here).  It does not, however, hotlink the article.  For an insurance industry update collating information on actions in AZ, CA, and DE, published on October 17, here.   Original WSJ coverage on October 10, here.

Tuesday, October 18, 2016

Brief Blog: How did FDA Approve the Roche EGFR Liquid Biopsy Test? (Links)

In June 2016, the FDA approved the Roche cobas EGFR test for the liquid biopsy format.
Take home lesson: The validation and technology review process was very elaborate.  And it's all public record.

 Online, see the FDA's numerous validation and reference documents:
  • FDA PMA HomePage for PMA150047, here.
  • FDA Approval order, here.
  • FDASafety & Effectiveness review, 53p, PMA150047, June 1, 2016, here.
  • FDA/Roche Product Labeling, 71p, here.
For some personal insights from staff both at Roche and at the FDA, see discussion at the FDA Liquid Biopsy conference, July 2016 (here for PPTs and transcripts).

Other links include:
  • FDA's press release, here.
  • FDA home page for test, here and here.
  • Roche 10 page brochure about test, here.
  • Roche product home page, here.
  • Roche product specific web site, here.

Some key FDA documents provided as links above are in one easy to handle Zip file in the cloud, here.


Original FDA EGFR cobas-in-paraffin approval as PMA120019, here.  
FFPE update to include T790M resistance mutation, here.

Brief Blog: Cancer Moonshot Program Publishes Progress Report

Only six weeks ago (September 7) the Cancer Moonshot program released a "blue ribbon report" with 10 strategic goals (here).

On October 17, the program released a 38-page Moonshot Progress Report.   Recall that the Moonshot program was only announced in January (although the President's Precision Medicine Initiative, as a whole, dates back a year earlier.)

See the October Moonshot Progress Report here.   Trade press here.   Washington Post here.

Sunday, October 16, 2016

CMS Hosts Town Hall Call on PAMA Data Reporting - November 2, 230-330 ET

CMS has announced it will hold a town hall call on the new CLFS payment system, data reporting issues.   The call will explain how CMS is implementing PAMA for the CLFS, including the June 2016 final rule and subregulatory policy making that has been released since then.

Hold your calendars for Wednesday, November 2 at 230 ET.   For the CMS webpage and Registration, see here.

Pathology CPT 2017: By the Numbers

Over the last several years, MoPath has grown from a few lines in the AMA CPT codebook to 20 pages, plus another 14 pages of gene table names.   I put this in context of the Laboratory Medicine AMA CPT book for CY2017.

MoPath approaches half of the pathology section now (43 of 100 pages).

Page 479 - Table of Contents
Pages 480-494 - MoPath Gene Name Tables 
    (e.g. international gene abbreviation + full name + corresponding CPT, usually Tier 2)
Page 498 - "Guide"
    (mostly lists the unlisted codes)
Pages 496-497 - Chemistry Panels
     (These are given CMS heartburn re PAMA pricing; see here)
Pages 498-507 - Drug Assays
Pages 509-538 - Molecular Pathology (Tier 1, Tier 2, MAAA, GSP, etc)
Pages 539-567 - Chemistry, Microbiology
Pages 568-576 - Cytopath, Anatomic Path, IHC/ISH
Pages 577-578 - Odds and Ends; Reproductive Path

Wednesday, October 12, 2016

Should We Break "Decision Impact Studies" Into Two Clear Categories?

For decades, we have been told that diagnostic tests are only important if they change care, which begins by changing a decision.   Decision impact studies (the before and after of a physician's recommendation) are important and have been conducted with increasing rigor.  On the other hand, from time to time you hear a payer say, "We don't give any credit to decision impact studies" or "That's only a decision impact study, not clinical utility."

Since decision impact IS so pivotal to a test's clinical utility, we should continue to use the idea, but maybe we should divide it into two parts:
  • Recommendation impact study.
  • Management impact study.
One study can report both, but payers are probably more influenced by management impact studies.  More after the break.

NCI's September 29-30 Workshop on Circulating Tumor DNA: Agenda

While I could not personally attend the September 29-30 NCI workshop on Circulating Tumor DNA in Clinical Cancer Research, feedback on the quality of the meeting has been excellent.  Although decks or streaming media are not available, this is a link to the workshop website (while it is still up) and I have clipped the agenda for the meeting permanently here, below the break.

Monday, October 10, 2016

FDA Issues Long Awaited Final Guidances on Glucose Monitoring

For years, there have been complaints that the FDA was using outdated 1970s standards to clear glucose meters, and not revising the standards to keep up with best of breed newer technologies.  This means the US marketplace could be cluttered with low-performance meters.  (See here and here).  Some studies have modeled the adverse health events that are likely caused by marketable, but low-accuracy, blood glucose meters (here).   In response, the Diabetes Technology Society launched an effort, through surveillance and certification, to endorse high-performance meters that others had no special recognition by FDA (here for homepage, here for article).

For European regulators, ISO announced new, tighter blood glucose meter accuracy standards in 2013 (here).

In October 2016, the FDA has finalized two guidance documents for US blood glucose meter performance, one for devices sold over the counter, and one for use in prescription settings (e.g. hospitals).    The new 2016 guidance documents are here for prescription use, here for OTC use.  (The corresponding brief Federal Register announcements are here and here.)   The guidances are fairly lengthy, at over 40 pages each.

These guidances were issued in draft form, for comment, in 2014 (trade journal articles here and here).  At that time there was some kickback; for example, would tighter standards require more bulky and slow meter equipment (here)?  AdvaMed responded to the hospital device draft guidance with a 47 page critique (here); AACC posted a long article on the same topic here.   An analysis of comments is available here.

For an October 2016 article at RAPS, see here.

For an FDA deck on its BGM review process as of 2014, here.
The FDA pulls down draft guidances and puts up final guidance at the same weblink, but:
    For a cloud copy of the 2014 draft guidance for prescription (institutional) devices, here.
    For a cloud copy of the 2014 draft guidance for OTC (home) devices, here.
In other FDA diabetes news, in August 2016, the agency held a public workshop on the pathway to novel outcome measures; FDA website here, trade news here.  The FDA posts full video and will post a full transcript by November.

Sunday, October 9, 2016

FDA Documents Prove: "Clinical Utility" Was Not Just Invented Recently

Hardly a conference passes without someone saying that today, diagnostic tests require "clinical utility," but this wasn't required until recently.
  • If it's a 1995 conference, clinical utility is recent, being required since 1990.   
  • If it's a 2015 conference, clinical utility is recent, being required since about 2010.   
  • It's a rolling date.   
I was in medical school in the mid 1980s and we were regularly taught, do not order a medical test unless it's medically necessary, and unless there is some action that will pivot on the test.

Lest us human artifacts of the 1980s be criticized for imaginative memory, Richard Ablin provides some extensive documentation in his book, "The Great Prostate Hoax" (St. Martin's Press, 2014.)   Ablin led the discovery and purification of the PSA protein, which was later developed into the clinical PSA immunoassay by others.  Ablin is on faculty at University of Arizona; he is known for a 2010 New York Times op-ed highly critical of PSA testing, and his concerns are documented at length in his 2014 book.  [More about Ablin's book and related citations, at bottom.]   He documents at length the active, even heated, debates over the clinical utility of PSA testing in 1985 and in 1993, based on his access to contemporary transcripts.

(Continues after break.)

ISPOR Releases New Consensus Publication: Health Tech Assessment & Molecular Diagnostics

The International Society for Pharmacoeconomics & Outcomes Research Group has issued an 11-page review of the state of the art of health technology assessments and molecular diagnostics:

The authors conclude that "HTA organizations should provider greater transparency" including "explicit recognition and rationale for differential approaches."   In overview, the authors summarize a wide range of case studies of genomic test technology assessments by different bodies.   Then, the authors summarize ten points, essentially a mix of problems-seen and recommendations for moving forward.
The article Garfield et al. is online here.  It isn't open access; it's $35, or a journal subscription is $332, or a membership in ISPOR including the journal is $150.   There are also three supplemental pages online, however.  A one page summary is open access, here.  A Q&A interview with Garfield is online, here.  The home page for the committee (ISPOR's diagnostics assessment committee) is online, here.

The first author, Susan Garfield, is Principal at the global consultancy EY (formerly Ernst & Young), here.  I also cited her work on pharma strategy a few months ago on this blog (here).  Thanks to Prof. Kathryn Phillips of UCSF for pointing the new ISPOR article out.  Philips heads the Center for Translational and Policy Research on Personalized Medicine.  

Also this fall, Deverka and colleagues produced a position paper on genomics clinical utility, based on interviews and analysis with a large group of consensus-driven stakeholders, here.   Deverka and colleagues provide ten action points to guide better production of clinical utility evaluations.

And also this fall, another recent consensus review article on "clinical utilities in molecular pathology testing procedures" by a collective group of authors at AMP, see Loren Joseph et al., J Molec Dx, here.   Joseph, who is at Harvard/Beth Israel, also published a single-author article on clinical utility of cancer profiling in Exp Rev Molec Dx, here.

Abstracts posted below, after the break.

Thursday, October 6, 2016

Theranos Closing Its Labs and "Wellness Centers"

See an "open letter" on the Theranos website, and an open access post at TechCrunch and at Forbes.

For story at Wall Street Journal, here.

At TechCrunch:
"Theranos is closing its labs and wellness centers, CEO Elizabeth Holmes announced today in a post on the company blog. And this isn’t a temporary closure: the “approximately” 340 employees running them are out of a job.
“After many months spent assessing our strengths and addressing our weaknesses, we have moved to structure our company around the model best aligned with our core values and mission,” Holmes wrote.
The company pivoted away from working on its closely held “nanotainer” technology to a “miniLab” in August. The boxy device — unveiled at the American Association for Clinical Chemistry conference — collects small samples of blood and urine and then uploads them to a centralized system for further analysis.
And it’s a far cry from what the company, once valued at $9 billion, set out to do. According to several experts whom TechCrunch spoke to at the unveiling, it might not be very innovative, either. Although Theranos didn’t want its new device referred to as a “lab on a chip,” that’s essentially what these experts said the miniLab was. And that has been done....."
For an archive of several hundred Theranos articles over ten years, here.

Tuesday, October 4, 2016

Does CMS Calculate Gapfill Medians Wrong? (Revisited)

I have argued before that multiple policy and regulatory statements about the laboratory fee schedule gapfill mathematics process clearly direct CMS to take the "median of MAC" prices.

However, in the recently released CMS spreadsheets for September 2016 gapfill calculations, CMS is clearly taking the median of the 57 fee schedule zones.  Not the 12 MACs.
(By my count, there are 12 MACs (6 in MolDX: Noridian x 2 + WPS x 2 + CGS + Palmetto) and (2 NGS + 3 Novitas + Novitas + FCSO + 1 Cahaba).  So that's 6 + 6 or 12.)  
If you take the median of 12 MACs, so defined, it would be the average of the two middle prices, a MolDX price and a Novitas price.  For example, in a hypothetical where the MolDX price is $3000 and the Novitas price is $2000, the median of 57 fee schedule zones is $2000, but the median of 12 MACs would be the average of the two middle MACs, or in the example, $2500.

Nerd Policy Quotation

42 CFR 414.508(b)(2) states: "In the second year (of the gapfill process) the test code is paid at the national limitation amount, which is the median of the carrier-specific amounts."  In recent rulemaking, CMS updated the word from "carrier" to MAC - same point.  No one ever says there are 57 MACs - so the median of MACs should be the median of 12, not the median of 57.

Monday, October 3, 2016

CMS Posts Final Gapfill Prices for 2016; Most MAAA Tests Rise

On September 30, 2016, CMS posted final gapfill rates for 16 codes that were new CPT codes in January 2016 and in play for gapfill pricing during 2016.    7 of the codes were genomic sequencing procedures (several in pairs, with a sequencing code and a dup-del code).   8 were sole source MAAA tests and one was a trisomy risk test.  Download the file "gapfill 2016 final" at CMS, here.

(A blog discussion of the preliminary June pricing is here.)

MAAA Tests

Prices for three MAAA test were raised by more than $1000 (81525, the Oncotype DX Colon test; 81538, the Biodesix mass spec lung test; 81595, the CareDx Allomap test).  However, the final gapfill prices for some proprietary tests remain below Medicare's local payment rates in recent prior years.  CareDx said in a press release the final pricing was unfair and unwise, here.

The main dynamic was that Novitas/FCSO raised prices a lot (but not to MolDX levels) and NGS raised prices a little, and net-net, the Novitas/FCSO prices set the MAAA medians.   Thus, the MAAA medians are below MolDX prices (further discussion, after the break).

Genetic Tests

Pricing for the HBOC (hereditary breast and ovarian cancer panel) rose from $622 to $925.  (For a press release by Invitae, see here.)   Lab policy nerds will note that MACs also priced the companion dup-del code (at $159); but the MolDX website lists the dup-del code as nonpayable, which would apply in the 25 states in the MolDX system.

Click to enlarge.
Coverage at Genomeweb here.   Nerdy review of changes by MAC, after the break.