Very Brief Blog; CBI Posts Free 40 & 100 page Reports on Digital Healthcare Trends

Many of us probably get advertising emails for $5000 industry reports on healthcare and biotech. 

Currently, research firm CBI is offering two major white paper reports for free.

The first, a 40-page white paper on "How Google Plans To Use AI To Reinvent The $3 Trillion US Healthcare Industry," is available as a very long web page, or a 40-page white paper.  Here.

The second, a 100-page deck on Consumers and Digital Healthcare, is available here.  (It also offers an audio briefing version.)  CBI strategist Nikhil Krishnan also runs his own blog, here.

Kaiser News: Medicare Diabetes Prevention Has Rocky Rollout

April 19, 2018, Kaiser Health News has a long-form story on the "rocky" rollout of a Diabetes Prevention Benefit at Medicare.   CMS doesn't say much, but admits that so far, and since January 1, only 3 DPP programs are enrolled to provide the new preventive benefit to its 40 million beneficiaries nationwide.   Story by Judith Graham here.

Small Policies Generate Colossal Rules and Policies

You can read about the DPP problems at Kaiser.   My take on it, is that it's surprising how often relatively simple concepts at CMS explode into huge rulemaking activities.  I'm not blaming CMS at all; it may be intrinsic to the job.   But let me give a few examples.

Diabetes Prevention Program

The Medicare Center for Innovation can create demo programs at will, evaluate them, and extend them if they improve quality and don't increase costs.  They ran a Diabetes Prevention Program demo - these are CDC-endorsed programs - and found it improved quality and saved costs. 

The program itself is simple: the provider has to be a CDC-certified provider of DPP.   The beneficiary has to meet a couple criteria for weight and blood glucose.   Then they get about 16 sessions over a year in a group-learning format.  Got it?

Rulemaking and policy making involved literally hundreds of pages of fine print in the Federal Register over a period of two years, pages of new regulations, and many other pages of application forms, criteria, rules, and penalties.   This takes hundreds of pages and several years to implement. 

Sepsis Early Intervention Program for Hospitals

In 2015, CMS initiated hospital requirements for sepsis intervention programs.  The core rules are very simple: if a patient is septic, a few key actions like blood tests and starting fluid should occur within 3-6 hours.

This took a several years to implement, has generated a crescendo of complaints in the peer-reviewed policy literature, and has a 160-page manual of rules and definitions plus a legacy of long town hall meetings, question/answer sessions, and transcripts.   See the thorny story in an article by Faust here and one by Rhee here.

CMS NCD for Genomic Testing in Cancer Patients

In March 2018, CMS issued a National Coverage Decision for the use of next generation sequencing tumor panels in patients with advanced cancer.   This document is only a month old, and the "coverage" section of the document is only a few paragraphs long (275 words).

However, it's already foreseeable that interpreting it in the context of the great range of possible cancers, tests, and drugs is going to be very complicated, and there at least several likely dilemmas or unintended consequences caused by the way the short policy is written.   See my own blogs on this topic here and here. 

MIPS/MACRA Physician Quality Reporting

I've listed this one last; you kind of know that Physician Quality Reporting and financial penalties are going to be complicated in any health system.  But of note this week, a new report by the main internists' association - the Amercian College of Physicians - finds that many of the laboriously implemented quality metrics don't even make sense.  Here.   This isn't a new topic; Richard Nixon gave a speech to the AMA in 1971 promising to reduce the government' s "burden of bureaucracy on physicians."

New Letter in SCIENCE Comments on CMS NCD for Cancer NGS Testing

Phillips et al. publish a letter in SCIENCE about the recent CMS NCD on gene panels in oncology.  Phillips published an op ed on this topic in JAMA a few days ago.
_____________

Last November 30, CMS released a proposed NCD requiring coverage with evidence development in a large sector of Medicare cancer patients.  Concurrently on December 1, Rebecca Eisenberg of University of Michigan Law School and Nobel laureate Harold Varmus of Weill-Cornell Medical School published an article in SCIENCE urging insurance coverage for gene panel tests in oncology patients, but under the umbrella of a coverage with evidence development paradigm.*

The NCD was finalized on March 17 without any coverage with evidence development component. 

• Earlier this week, on April 16, JAMA published an op ed on the NCD by Kathryn Phillips of UCSF - here.    
• Today, April 20, SCIENCE publishes a multi-author response to the Eisenberg & Varmus article - here.   This letter is by Phillips and ten co-authors (I am a co-author).   
• Will private payers follow the CMS coverage policy?
• Coverage for Medicaid patients is uncertain.
• CMS didn't consider risk-benefit or economic tradeoffs.
• CMS and other federal agencies need to view this as a beginning of wise oncology policy making, not a conclusion.

* The coincidence of the SCIENCE op ed and the CMS NCD is interesting, if you expect that CMS policies must be secret and not released outside the agency ahead of publication; here.

Medicare LCD for BRCA Testing in Ashkenazi Populations Raises Costs Unnecessarily - While Wasting Time

This morning, I was reviewing the First Coast (FCSO; Florida) Medicare LCD for BRCA testing in women with a personal history of breast cancer.

If the woman has an Ashkenazi history, the LCD requires her to first have founder mutation testing (CPT code 81212, $440) and then if it is negative, to proceed to regular BRCA sequencing (81211+81213, $2948.84). 

Financially this doesn't make sense, and in terms of screening efficiency it obviously doesn't make sense.
_____

The rate of being BRCA-positive in Ashkenazi women with breast cancer has a base rate of 25%.  It is higher (30%) in women < 45, and lower (18%) in women > 45.  Medicare populations are almost entirely > 65, so the rate would be below 18% in a Medicare population.

Running these numbers for FCSO's LCD at current Medicare rates, at the overall population rate of 25% positive, screening testing followed 81211/13 testing would cost $265,163.    Running the numbers with the lower 18% rate for the over-45 population, testing would cost $285,805.   However, if FCSO simply directed patients immediately to BRCA comprehensive sequencing (code 81162), total costs would be $225,293. 

In short, the screen-first policy wastes BOTH money AND time.  Thanks, Medicare.

It's not in the short term interest of a Florida lab to poin tthis out and request a revision, as the lab gets more money under the policy.  (The policy is financially break even for Medicare only if the BRCA-positive rate is higher than 40%.)

Data below.

click to enlarge

___

The LCD requires 81212 testing in all patients then reflex to 81211+81213.  If it reflexed to 81162, it would still lose money in the >45 population, and waste just as much time.

White Paper: Three Rubik's Cube Puzzles of the CMS NGS NCD

We're celebrating one month after the publication of the final NCD from Medicare on the use of NGS testing in advanced cancer patients.   I've written up a 10-page white paper on three puzzles of the NCD.    Find the white paper online here.  Click the "Down" arrow to download.

Based on incoming questions that I hear, andbased on some of my own puzzling, here are the three topics.  The short version is this blog; the long version is the online white paper.

     1.  Why does FMI assert pan cancer coverage under the NCD?
     2.  How does the NCD text limit the scope of the NCD?
     3.  Possible implications of the "advanced cancer" definition in the NCD.

___

1.  Why does FMI assert pan cancer coverage under the NCD?

The NCD provides coverage for tests that are NGS based, FDA-authorized CDx tests, when used in a patient who has an indication for that test.   FMI has two indications.  One indication for use as a CDx in several genes and several cancers (such as EGFR, lung, CDx).   A second indication is for "tumor profiling to inform the physician in solid cancers."   If a Medicare patient has lung cancer, the CDx for EGFR checks both boxes: the test is a CDx test, and, has an indication (EGFR) in that patient.   If a Medicare patient has bladder cancer, or skin cancer, or thyroid cancer, the test is still a CDx test (box 1) and it has "an indication" for that patient, e.g. tumor profiling to aid the physician in a solid cancer.   This analysis concords with FMI statements in several press releases.

click to enlarge

Getting a pan cancer definition isn't easy; FMI provided 80,000 blocks across 43 cancers, and MSK IMPACT provided >10,000 blocks across 17 cancers.

2.  How does the NCD text limit the scope of the NCD?

The proposed NCD had clear cut exclusionary language, which CMS simply deleted from the final version.  In addition, CMS deliberately added several additional sentences for the public, regarding the "scope" of the NCD:

click to enlarge

This language seems initially clear, then a little murky (to exactly what do some terms refer?) on closer reading.  I provide an analysis under Section 2 of this white paper. 

As I've said earlier, trying to read the NCD (including the text above) as broader and more exclusionary collides with common sense and lands in arbitrary and capricious.  For example, applying the NCD to unrelated NGS topics other than cancer therapy would be just like doing an NCD on cardiac devices, limiting the scope to valves, and then concluding that pacemakers aren't covered by implication.


3.  Possible Implications of the CMS "Advanced Cancer" Definition

The title of the NCD is "NGS testing in advanced cancers" and CMS provides coverage for targeted therapy NGS testing only in advanced cancers.  CMS seems to take pains to provide an explicit operational definition of "advanced cancers:" cancers that are recurrent, relapsed, refractory, metastatic, or Stage III or Stage IV. 

Is the NCD Language too restrictive?  For example, a new aggressive leukemia isn't relapsed or recurrent - it's new - and it's not refractory, because we're going to treat it for the first time.  Leukemia isn't usually called "metastatic" or "stage III or IV."  (There are staging systems for leukemia but they're unrelated to the metastatic Stage III/IV in solid tumors).   Could some auditor assert the leukemia doesn't meet any single one of the entrance criteria?

Can the LCD language in some cases too generous?  There's no question that skin cancer is a cancer, and millions of them are "recurrent."  When they recur, patients will certainly want "a therapy" (be it excision, radiation, topical drug, etc).   So here we go: The patient has a "cancer' and it is "recurrent' and he needs "a therapy" (not otherwise specified by NCD), so the skin cancer seems to qualify.  And if he has several (left hand, right hand, forehead) each is a distinct primary eligible for testing.   I'm not recommending this, I'm just reading the NCD.

Very Brief Blog: Behind The Headlines, Ongoing Life for "Tumor of Origin" FDA-Cleared Test

In a Genomeweb article this week, we read that Cancer Genetics (which has been in the news for financial problems) has gotten a special 510(k) clearance for its Tissue of Origin test.   Tissue of Origin tests are useful in cases where a patient presents with metastatic cancer of unknown or uncertain origin.   (Terms include tissue of origin test, tumor of unknown origin TUO, cancer of unknown primary CUP; see one recent review here.)

Genomeweb does not mention the long journey of this test, which classifies 15 tumor types. 

Pathwork Diagnostics developed the test just about ten years ago, and got FDA de novo clearance in June 2010.  (t's not clear yet just what kind of update the "new" FDA clearance represents.)

Pathwork  was folded into LA-based Response Genetics in August 2013.  This was described as "salvaging" the Pathwork test for a mere $1.3M.   According to Crunchbase, during its lifespan, Pathwork Diagnostics had garnered $64M in investments.

Response Genetics, in turn, was in bankruptcy when it was acquired by Cancer Genetics in August 2015, here. 

Currently, in April 2018, Cancer Genetics is the focus of lawsuits and is working to leverage its strategy and access for ongoing business operations.  CGIX stock has slipped from about $18 to about $1.50 over four years, with a current market cap of $40M.  Revenue has been circa $25-30M the past two years against net income of -$15M to -$20M.

JAMA Publishes Rapid Op Ed on CMS NGS NCD (Kathryn Phillips, UCSF)

Noted health economist and expert on precision medicine policy, Kathryn Phillips of UCSF, has published a rapid op-ed in JAMA on the March 2018 Medicare National Coverage Determination for uses of next generation sequencing testing in cancer.

The op ed is open access and online here.

Phillips notes the NCD provides coverage for NGS tests in all Medicare patients with advanced cancer.  However, it doesn't yet come with a "national health policy" for where we are going.  The impact on Medicaid populations or on commercial populations is unknown and remains to be developed.   She notes that a number of studies have asserting that panel NGS testing is "not cost effective," and that the coverage has been rapidly expanded across solid tumors with no kind of tracking or information collection such as could occur under coverage with evidence development.

Phillips will give a talk on the topic at Stanford on April 27.

Phillips directs a longstanding NIH-funded precision medicine policy program at UCSF, TRANSPERS, here.

Very Brief Blog: Harvard Case Study Provides Window into US Health System & Payers

Harvard Business School Case Studies have been famous for decades for providing a concise window into all sorts of businesses and scenarios and the challenges faced by management.   As a med school assistant professor taking MBA classes in 1999, I was struck by my first contact with these and it took a while to "get it."  Typically, a company is presented, some options or dilemmas or even a crisis, and then a portfolio of information and data.   There isn't a right answer; it's like giving you two-thirds of a novel and asking what happens next.

Rebecca Henderson and colleagues at Harvard Business School have published an unusually thorough, 32-page case in February 2018 called, "Aetna and the Transformation of Health Care."  It's available for a few dollars here.    For example, the authors frame a dilemma the CEO faces: 

"Bertolini wondered whether Aetna needed partners to make this strategy work. Should it partner with a firm like Google, Amazon, or Uber—entities that understood digital platforms and had strong consumer brands? Should Aetna double down on its investment in local communities by exploring a relationship with a company that had a major retail presence, such as Walmart, CVS, or Walgreens, or by expanding collaborative relationships with local providers?"  

The case provides background on the CVS-Aetna acquisition that is currently working through the legal channels (e.g. here).

Other interesting Harvard cases recently include the crash and fix of Healthcare.gov (podcast here, case study here) and a case study on precision medicine and the business strategies of cooperative networks for adaptive and basket trials, here.

A Watershed Moment in Healthcare and Reimbursement? FDA Greenlights Auto-Detection of Diabetic Retinopathy

FDA and health tech business IDx announce authorization to market the IDx-DR device, the first artificial-intelligence based diagnostic system.

MedCityNews writes,

    The marketing clearance follows the FDA’s decision to approve a “breakthrough device” designation for IDx-DR, accelerating the review process for the product based on its ability to address an unmet medical need....
    “The FDA’s authorization to market IDx-DR is a historic moment that has the potential to launch a transformation in the way U.S. healthcare is delivered,” said Dr. Michael Abràmoff, founder and president of IDx, in a company news release. “Autonomous AI systems have massive potential to improve healthcare productivity, lower healthcare costs, and improve accessibility and quality. As the first of its kind to be authorized for commercialization, IDx-DR provides a roadmap for the safe and responsible use of AI in medicine.”

MobiHealthnews here.  Medscape here.  FDA press release here. Company website here.

On its website, the company writes:  "IDx is developing algorithms that can detect disease across a number of imaging modalities, with a focus on fundus photography and optical coherence tomography (OCT).  Current prototypes exist for macular degeneration, glaucoma, Alzheimer’s disease, cardiovascular disease, and stroke risk."

CMS Follows FDA.....

Reimbursement?  Coverage?  New Coding?  RVU's?   To be continued....

CEO Gary Seamans and I are both speaking (no relation) at the MedCity INVEST conference in Chicago May 1-2, 2018 - here.

____

IDX is based in Coralville, Iowa, on the other side of Des Moines from the farm where I grew up.

Brief Blog: The NUB Regulation as a Form of CED in Germany

You learn something every day.   Today I heard about the German NUB regulation, which has some features of CED.   

It stands for Neue Untersuchungs- und Behandlungsmethode : New Investigation & Therapies Methods, basically.

NUB was introduced after Germany discovered post 2003 that its DRG inpatient system made it hard for hospitals to innovate.   (USA Medicare has a "New Technology Add-on Payment" or NTAP for our DRG system.)

It's a more hospital-based system in which a hospital can contract for one year or two of special payment for a new device while data is collected.  The device must be CE marked and used less than 3 years in Germany and cost > €500.

NUB Sources

• Sebastian Gaiser, formerly with St Jude, now at J&J, has a 14 slide deck from 2014 about the NUB system available open access on line.  It's not entirely in favor of NUB, and sees flaws, but gives you an introduction. Here.   
• See also a 19 page, 2009 deck about NUB from Henschke et al., here.   
• The group "Assessment in Medicine" has a 24 page PDF from 2017 that discusses NUB.  Here.
• Cloud zip file for the above, here.

Once you've heard of NUB, you start to see it in press releases, e.g. this 2018 press release for the BIOVENTRIX transcatheter device.

There seems to be another German system called "137E" which I've seen cited as supporting rollout of cancer drugs, e.g. here.  See also a 2014 article by Olberg et al. on CED/137E in Germany, here.  NUB is a limited DRG focused program, 137E is a broader CED concept.

___

Hoffman et al. have just an conference abstract on NUB in Value in Health 2016.  Here.

ISPOR has a 2011 detailed webpage on the whole complexity of the German System (includes a bit on NUB).  Here.

The 24-page Assessment in Medicine (AIM) document is on the website of IGES, a technology institute that may be a university spinout, here.

The main German health technology assessment group is called IQWiG, here.

Policy Collaboration with Medicare Agency (Case Study: IDSA, Antibiotic Stewardship)

This month, an active federal prosecution in New York is generating press about how CMS makes decisions, and how information on the decision process might leak.  See e.g. Bloomberg here.

The normal process doesn't make much news or generate big court cases, but it's interesting to see how it can work successfully.

In July 2016, CMS released a special standalone rulemaking specific to a proposed requirement for antibiotic stewardship programs at US hospitals.

 For background see here and here and here.  Remember that antimicrobial resistance and better controlled use of antibiotics are major national and global health priorities.

Google helps us reconstruct some back-story.

June 2014
In June 2014, the Infectious Diseases Society of America (IDSA), along with the Society for Healthcare Epidemiology of America (SHEA), sent a 20 page letter to CMS asking them to create a "condition of participation" for hospitals that they must have formal antibiotic stewardship programs.   See the letter online here. 

Back to 2012
The letter describes multi stakeholder consensus workgroups that go further back, to 2012, and helped support creation of the 2014 request. 

Forward to 2016
Fast forward 24 months, and CMS produced the hospital rule proposal on June 16, 2016 (81 Fed Reg 39448-39480, CMS-3295-P), here.  Comment docket here.  201 comments were submitted.  I've put about a dozen comments in the cloud as a zip file, highlighting "major" ones with XXX in the file name, here.  Interestingly, whereas IDSA and SHEA submitted a joint letter in 2014 requesting the rule, it looks like they commented in 2016 three times: as a joint letter, as a single IDSA letter, and as a single SHEA letter.  Other major groups like AHA commented (see zip file).

And Today: 2017-2018

• CMS hasn't finalized the rule yet; one might expect that at 6-12 months, but the agency has up to several years.   Perhaps it was viewed as excess regulation by the new administration; perhaps it's in somebody's inbox for sign off.
• Meanwhile, Joint Commission has created guidelines for its hospital accreditation process to help fulfill "best practices" for having an antimicrobial stewardship committee.  See JC webpages starting here.   
• There's also a February 2018 article (by Cornell authors Kapadia et al.) in the JC journal, here; open access.    
• A quick search of Pubmed for "antimicrobial stewardship programs" has over 400 hits, here.  
• There's a Presidential Council of advisors here.   (See various work products; the next meeting is May 16).
• CDC is involved here.  
• Lots of stuff at IDSA, enter the webpages here.

June 2019

• Based on SSA 1871, the proposed rule issued in June 2016 will expire unless finalized.

I could go on, but the point is, if you view the center of the wheel as the CMS proposed rule in 2016, you can quickly find several years of predecessor and follow-up events that are part of understanding what CMS is thinking.

click to enlarge

____
Not directly related to CMS rulemaking pathways, but see an April 2018 CDC presentation on the ongoing threat of microbial resistance, here here here.   The public health issues continue even if the bureaucracy stalls.

FDA Releases NGS Review Guidances: NGS Databases and NGS Germline Review

In 2016, FDA released draft guidances on new review practices for NGS, including the use of public variant databases, here.

On April 12, 2018, FDA released final versions of guidances for NGS databases and for review parameters for NGS germline testing.

See FDA press release here.  Trade press at Endpoints here.  Genomeweb here.

Database guidance here.   Note that FDA has set up an elaborate voluntary FDA recognition process for databases; see this new FDA webpage here.  Eventually FDA says it could use third-parties to certify variant databases for FDA recogntion.

Germline guidance here.  Discussion includes choice of indication, expected frequent downclassification from PMA to 510(k) and future use of that (growing library of) 510(k)(s) as a predicate for others, and rolling-on modifications especially under 510(k).

 I've clipped the FDA press release below the break.

Foonote:

You Can't Be Too Rich or Too Thin or Issue Too Many Guidances on IDE Dx Tests

In addition, over the past several years, the FDA has repeatedly revisited the meaning of "investigational use only" and the requirements for registering LDTs or RUOs used in clinical trials as "investigational devices" or IDE's.  Most recently in December 2017, here.  FDA issued another draft guidance in this series on April 12; here.   Generally, FDA has wanted to be sure that IRBs approving local clinical trials ensure that LDTs used in those trials are IDEs.  Perhaps out of IDE exhaustion, this draft guidance is only 7 pages.

Very Brief Blog: Cancer Genetics (CGIX) Describes Payer Write-offs on MoPath Testing

Cancer Genetics (CGIX) is in the news this week due to several investor lawsuits, which follow in short order from CY2017 financial disclosures around April 1. 

CGIX went public with a fairly small IPO ($7M) in 2013.  At the time, its businesses included "a joint venture with Mayo Clinic."  Its business focused on molecular testing both for biopharma R&D and for clinical applications.  It acquired Gentris, a biopharma genomics provider, in 2014.  It acquired LA-based Response Genetics in 2015.   Previously, it had acquired Pathworks, and in April 2018 received 510(k) for the tissue of origin test (here).

According to Yahoo Finance, its stock price has slide from over $11 in mid 2015 to about $1 today.   According to the company's April 2 financial announcement, 2017 revenue was 63% biopharma/discovery, "supporting over 220 clinical trials."   2017 full year revenue was $29M, against a 2017 net loss of $21M.  (E.g. about $50M in costs to generate $29M top line revenue).

Cancer Genetics changed CEO's in February 2018.  On April 9, Genomeweb and other news sources covered a pair of new class actions lawsuits (here).

What's interesting, and familiar to anyone who tracks the messy chase for revenue in clinical molecular laboratories, are the following descriptions:

A major area of concentrated focus during the first quarter of 2018 was the careful evaluation of the Company’s accounts receivables, which had increased to approximately $16 million on the balance sheet prior to any adjustments. 

A significant reason for the increase was disruptions in collections in its Clinical Services business. While the Company continues with its collections efforts on all claims, in the fourth quarter it recorded a bad debt expense of $4.4 million and wrote off $1.8 million of its accounts receivable, with a significant portion of the bad debt expense and write off related to collection issues with respect to the accounts receivable recorded subsequent to the 2015 acquisition of Response Genetics Inc. 

Payors have declined to reimburse the Company on certain performed Clinical services due to: 

• delays in filing its claims, 

• the demands by payors for copies of patient medical records or diagnosis codes which have been difficult to obtain, and 

• reimbursement challenges for certain of our next generation sequencing tests by Medicare and 

• third-party managed care plans, among other reasons. 

• As such, the Company has made a prudent decision to write these off in the fourth quarter. Management believes that its current outstanding accounts receivables are collectible, net of the allowance for doubtful accounts.

[Bullet-points added by me for clarity.   The report also notes that accountants had made a "going concern" warning].

____

See also my blog in January 2018 about an executive leaving Hudson Alpha with the quotation, "The reimburesment [for advanced sequencing] is just not there" - here.

Very Brief Blog: Reasons to Join Personalized Medicine Coalition: Great Speakers (CMS NCD)

One benefit of membership in the Personalized Medicine Coalition in Washington:  great conferences and speakers. 

For members, on April 24 they are hosting Katherine B. Szarama PhD, lead analyst for the CMS National Coverage Determination on Next Generation Sequencing in Advanced Cancer.

According to the PMC website, she's speaking at a members event April 24 in DC.   See the PMC website here.   See the events calendar here. 

Other PMC events coming up include its 14th Annual State of Personalized Medicine luncheon (May 23).  In other news, PMC recently released its 2018 Strategic Plan, online here.

Very Brief Blog: EY Releases 60 Page Report on Life Sciences, Big Data

EY has released a 60-page report on life sciences, capturing value, and big data/digital health. 

Is this real?  When Roche buys Flatiron for $2B - yes.

The EY report is online here and PDF here.

See Q1.2018 reports on digital health investments at RockHealth here [$1.6B], StartUpHealth here [$2.8B].  Trade press at MedCity here, MobiHealth here.    Funding varies with the categorization of companies in genomics; by one tally, Dhealth deals included HeartFlow $240M, Helix $200M, Oscar $100M.   (See my blog on Medicare policy wins for Heartflow, here.)

In March, Bloomberg reported Israel was poised to investment $275M in digital health.

EY Life Science Data White Paper 60pp

___
See a new April 2018 MedCity article on what pharma has learned from efforts to collaborate with Dhealth, here.