Friday, February 23, 2018

Very Very Brief Blog: MEDPAC gets course in Cost Effectiveness

There is a cliche' that "Medicare doesn't consider cost" but each year it has hundreds of pages of policymaking on value based purchasing, lowering costs of certain services, bundled purchasing with no extra payment for extra stuff, and so on. 

The MEDPAC advisory body for CMS gets a "primer on cost effectiveness analysis" or CEA at its next meeting, on Friday March 2.  Pic:

click to enlarge

CAP Filing with Court Supports ACLA Lawsuit against CMS PAMA Lab Rates

In mid December, 2017, ACLA filed legal action against CMS for for having defined "laboratory," under PAMA, in a way that excluded almost all hospital labs.

CAP has announced that it filed an amicus brief  supporting the ACLA action, with a press release here and the CAP legal filing here.

More than ever, the case shows that the underlying law is tortuous and difficult to implement by CMS. 

The law requires that CMS collect data from "laboratories" that have more than half their revenue from Medicare from Part B (CLFS or PFS). 

This means they get less than half derived from Part A (inpatient) or other Medicare sources.   I believe CMS staff faced what they saw as a tortuous or impossible mission to figure out how much of an embedded hospital lab's "revenue" filters in from Part A, Part A overhead, Part A continuing graduate education, Part A's complicated cost buildup with real estate and facility amortization, and so on.   Fleeing this rats-nest, CMS looked to the NPI, making it trivial to calculate how the NPI was paid by Medicare, because an NPI is a thing that CMS cuts checks to and can add up the checks.   However, embedded hospital labs roll up into a hospital NPI with all of the inpatient and outpatient services; the NPI result is no longer conceivable as a "laboratory" income statement in any sense.  Thus, PAMA law is violated, says CAP, because "laboratory" must be the unit of analysis.

While CMS can argue that it interpreted the law acceptably, and ACLA/CAP can argue that its interpretation was illegal under the law, neither party probably feels free to say that the law is a Rubik's cube with no good solution.   (There's an old puzzle, "The barber shaves everyone who doesn't shave himself; so who shaves the barber?" If it's himself, he's shaving someone who shaves himself.  If it's someone else shaves him, it violates the rule that "the barber" shaves everyone who doesn't shave himself.  There's no answer.)

CAP makes an argument that definitely does, in some ways, usefully expand on ACLA's original brief.  However, CMS may respond that it must have a solution it can implement, and that no one has provided one.  The nature of these CAP/ACLA legal documents is to argument CMS is wrong, and stop.  But implementation would require somebody somewhere (at CAP, ACLA, CMS, HHS, DOJ, Congress, or heaven) to provide a solution that would work for an embedded hospital lab that is all interwoven with the hospital accounting and finances in ways that weren't designed to be disentangled neatly for PAMA.

Thursday, February 22, 2018

Very Brief Blog: Waiting for CMS Oncology Genomics NCD: How Will CMS Handle Huge Expansion of Results?

In November 2017, CMS released a proposed new national coverage position on the use of next generation sequencing in oncology.   The final version will be released by February 28, and it will be interesting to see how CMS has triaged the diverse, incoming public comments.

One certain effect is that far more Medicare patients will receive genomic results for uses not currently labeled by FDA or endorsed by NCCN.   CMS proposed that if tests like the several hundred gene Foundation One CDx test are used off-label (for example, in kidney cancer), it would require and trigger coverage with evidence development, studying the genes assayed, genetic results, drugs chosen, and clinical drug responses. 

However, what might easily be overlooked is that large numbers of Medicare patients with nominally on-label initial testing will get off-label drug suggestions for the first time. 

Right now, most MACs only pay for genetic tests (like EGFR or KRAS) in those cancers where that gene, if positive, leads to an on-label drug.   This means that labs, MACs, and clinicians currently are  "coloring within the lines," getting results for BRAF drugs in melanoma, ALK and EGFR genes in lung cancer, and so on. 

However, under the proposed NCD, even when the test is initially nominally on-label (like the FMI test in lung, colorectal, and melanoma), many patients will NOT have a positive gene for an on-label drug, but they and their clinicians will have a range of genetic results for off-label drugs.   Since the original NGS panel test was on-label (like an FMI CDx test in a lung cancer patient), the NCD doesn't require any CED or suggest what will happen at the MAC level with the panel-based or "passenger" genes that are being reported to the clinician and patient. 

But a lot of patients may fall here, especially if they had already screened negative for the common mutations in their cancer (e.g. EGFR/ALK in lung) and the FMI test is used as a fall back test.  If MACs provide the drugs, but no clinical data is gathered, we're missing out.  If MACs don't provide the drugs, after they have paid for the tests and have supported delivery of the accompanying off label genetic results to the cancer patient, who knows what will happen.

Relevant is a new paper in Nature by Human et al. of MSKCC, also profiled at Genomeweb.  (The paper is open access to view, but not download).   In the SUMMIT trial, patients with mutations in HER2 or HER3 were studied across a wide range of cancers - NSCLC, breast, biliary, salivary, and so on.   The main conclusion is that only a complex and clinically empirical study of mutations, genes, and tumor type gives a good predictor of drug response.   For example, no responses were seen in 16 patients with bladder cancer or in 12 patients with colorectal cancer.   The SUMMIT trial, and similar trials underway, are important amplifications of the early lessons learned or early truisms in precision oncology (e.g. early lessons that ALK gene-drug pairs seem to work in most cancers, while in contrast BRAF gene-drug pairs work great in melanoma but not in colorectal cancer, etc). 

OHSU's Vinay Prasad also wrote about theory and outcomes in basket trials in the February 2018 issue of JAMA Oncology.

Very Brief Blog: New York City Tallies $175M in Recent Healthcare Tech Investments

Will New York City be Silicon Valley East?   One feature of digital health, and to some degree next gene sequencing, is the relatively light required footprint and space.   Thus, significant companies can locate in places like New York and San Francisco and central urban Cambridge, MA (the latter being the home base of Foundation Medicine). 

New York City has released a 2018 report on venture capital investments in healthcare industries.  Most are on the artificial intelligence/innovative data management side or predictive analytics.   Report here, 16 pp.

Digital health can move quickly and at scale.  Roche just announced its $1.9B acquisition of Flatiron, while COTA (cancer health analytics) raised a new $40M, which would top out the NYC list above.  West coast in the last few months, Illumina announced a $230M venture fund and 23andMe raised $250M.   In the genomics digital space, DNANexus very recently pulled in a new $58M which would also top out the NYC list above.

There are also fizzles in Dhealth.  News reports that Nokia may be backing out of digital health after several hundred-million dollar investments (here).  For a detailed review of trends in digital health in 2018, see a 2 part series at MobiHealthNews here and here.  See also a nice article on "5 trends that will matter," here.

Wednesday, February 21, 2018

Very Brief Blog: HIMSS Las Vegas 2018: Precision Medicine Symposium (Monday March 5)

The giant annual HIMSS conference comes to Las Vegas March 5-9.   This year it features a special all day symposium ($350) on "Precision Medicine: Journey to the Summit Using Clinical and Business Intelligence." 

Find the webpage here.   On the webpage, note that each of the six hours has a pop-up that opens its own webpage and detailed description.

High level topics are:

  1. Keynote: PM at the Inflection Point: Damon Hostin, Precision Medicine Alliance
  2. Preparing for the Journey: Tools of the Trailblazers
  3. Data Interoperability and Sharing
  4. Steps to the Summit (Panel with Aetna, Quest, Syapse)
  5. Ethics in PM
  6. Closing Keynote: Steven Kalkanis, Henry Ford

What is Precision Medicine Alliance?  Well, there's an entity with that named based in Berlin, here (located just north of Mitte between Rosenthalerplatz and Bernauerstrasse).  Umm, but in THIS context, with Damon Hostin speaking, it represents an alliance between the 150 hospitals of SF-based Dignity Health and Colorado-based Catholic Health, see here.

Very Brief Blog: Checkpoint Inhibitors in Hematological Malignancies (Open Access Review)

If you're like me, you think of checkpoint inhibitors first in solid tumors - melanoma, lung cancer, and so on.   Knowing less about it, I'm caught a bit off guard in the shift to leukemia/lymphoma therapy.

Chi Young Ok and Ken H Young provide a 16-page, open access, 2017 review of checkpoint inhibitors in J Hematol Oncol 10:103, 2017.   They're at MD Anderson.   Open access here.

There's cool stuff.  For example, PD-L1, PD-L2, and well known oncogene JAK2 are all at 9p24.  Who knew?  Errors as diverse as JAK2 upregulation, copy number and fusion events involving PD-L1, and 3'-untranslated region PD-L1 aberrations can all disrupt immuno surveillance in lymphomas.

For a non open access article with a quite different perspective based on B-cell malignancies, Muschen of City of Hope provides a perspective in Nature Reviews Cancer 18:103-116, 2018 (here).


Not related directly to heme tumors, but a NYT article today discusses rare variants of ovarian cancer that respond well to nivolumab where the majority of ovarian cancers don't.  The emerging rules about immunotherapy biomarkers (PDL1, TMB, etc) have strong exceptions and it's still being worked out (a point also made by Ok & Young).  {The NYT patient example is SSCOHT ovarian cancer driven by SMARCA4 mutations}.

Another "who knew?" interesting  Nature Reviews article recently pointed out the multiple ways that "regular" driver somatic mutations (EGFR, etc) also interplay with turning on immune evasion mechanisms (Seton-Rogers, Nat Rev Ca 18:67, 2018, here).

Very Brief Blog: The Messy State of Tech Assessments and Economics in Genomics (two papers)

Two current articles, both subscription, cast a light on the still-debated state of health economics and technology evaluations in genomic medicine.

For readers of Genomeweb....NICE is currently reviewing the whole field of breast cancer prognostic tests (such as Oncotype Dx, Mammaprint, Prosigna, Endopredict, and others).   A few weeks ago, the news was that NICE had provisionally found none of these to be cost-effective.   Public comment was requested.

On February 7, 2018, Genomeweb's Monika Ashford provides a very detailed, test by test rundown of public comments submitted and argumentations used (here).   Much discussion of where the NICE report could be wrong, data misinterpreted, conclusions incorrect, important points overlooked, and so on.

Also under subscription, see an article with the promising title, "Cost effectiveness analyses of genetic and genomic diagnostic tests," by Payne et al., as an Advance Epub in Nature Reviews Genetics (here).   While the title is promising, the 12-page review actually walks the reader through a wide range of what the authors describe as uncertainties, dilemmas, puzzles, missing data, and contentious issues... All of which would be faced in producing anything that is a strong statement on the "cost effectiveness of genetics." 


Proposed legislation would instruct the US HHS to review and report on the effectiveness of genomics.

Very Brief Blog: Intel's Hospital-System-Level Survey of the Challenges to Precision Medicine

Organizations like the Personalized Medicine Coalition regularly discuss challenges to personalized medicine, and so to studies by major government bodies. 

This winter, Intel published the results of a large interview- and survey-based study of challenges to personalized medicine at the health system level.   Results include the major drivers of moving forward with investments, education, and personnel, as well as challenges (lack of money for investments, lack of education, and lack of personnel.)   The interplay between insourced and outsourced or mixed solutions is also discussed.

See the commissioned review article at HealthcareITNews here and the actual Intel white paper report here.

This and many additional figures in Intel (2017)

The Intel white paper discusses choices between insourcing and outsourcing.  In December, I published a short article on the growing specialized companies in "Digital Genomics" that allow closer-to-turnkey outsourcing of IT aspects after sequencing is done, e.g. Fabric Genomics, DNANexus, Edico, and many others.  Here.

I recently mentioned, a brilliant article on how real world health systems and CEOs view new initiatives was produced by Spellberg in 2017, here

The HIMSS 2018 conference, March 5-9 in Las Vegas, has a number of precision medicine tracks and side conferences.

Tuesday, February 20, 2018

Brief Blog: Health Policy was 140 Pages of the 250 Page 2018 Budget Bill

Avoiding a long term government shutdown, the Hill passed a national budget bill of 250 pages on February 9, 2018.    The legislation is 250 pages(HR 1892).  There have been numerous summaries of various healthcare provisions.

Notable, I would say, is the raw fact that health care policy occupies 140 of the 250 pages (from 105 to 245).    

I've put the health care pages in the cloud here.

Many parts are nearly unreadable, because they amend existing law.  For example, the heavily debated and contentious funding of Children's Health (CHIP) looks like this:

Some tidbits and highlights after the break.

BCBS CareFirst and HealthWorx for Payer/Innovator Collaborations

For some years, CMS has developed a long list of variably successful programs for "coverage with evidence development." In November, CMS proposed one of the largest and most complex CED programs, for molecular oncology, which is expected to be finalized by February 28.   

Meanwhile, it is a cliche' that commercial payers don't do much in the CED space.[*]  However, CareFirst has a formal program called HealthWorx that is gathering increased attention.  I've pulled together a few links.

CareFirst is a BCBS plan serving 3.2M members in Maryland, D.C., and northern Virginia.  The current CEO is Chet Burrell, who has served since 2007.  He was previously an EVP for Anthem and before that held several state government positions in New York. 

A January 2018 article in Baltimore Business Journal describes the HealthWorx program as "working with biotech companies to validate innovative, cheaper health care options" and bring them to its members.   The insurer "searches across the country for innovations and new technologies that fit the criteria" and then the company and CareFirst collaborate in programs that test the innovation in CareFirst members.  After time, a plan-wide coverage decision can be made.  According to this article, it's working actively with 5 companies, talking to a dozen, and could grow in scale to work with 25 companies at a time.

A pilot program with Exosome Diagnostics was announced in August 2017, for tests that would make prostate cancer care pathways more efficient and accurate (also here and here and here).  A pilot program with Myriad Genetics was announced in November 2017, looking at pilot coverage for the Vectra DA test in rheumatoid arthritis. 

Key point:  The Exosome Diagnostics pilot program led rapidly to full coverage for appropriate patients of the ExoDx Prostate IntelliScore (EPI) test in January 2018.   The $500 molecular test could displace more costly $2000 biopsies.  The CEO told the Baltimore Business News, ""This is something we could offer to our members that is less expensive, less invasive and more accurate than current options...It illustrates exactly the kind of thing we are looking to do with HealthWorx."

 An advertisement for HealthWorx program director, reporting to Chief Medical Officer, gives some insight into how CareFirst views its program.


Cross-reference:  Illumina announced a collaborative coverage plan with Partners Health Plan in January 2018.
Tidbit:  Not directly related to CareFirst, but there is a Boston-based consultance ALVA10 which aims to help diagnostic innovators collaborate early and rapidly with payers. 
Tidbit: The HealthWorx trademark dates to 4/2017
Tidbit: A San Mateo, California, regional Medicaid plan also uses the term "HealthWorx."
Tidbit:  The Chief Medical Officer at Carefirst was Dr. Rahul Rajkumar (as of 2016), here, but he moved to BCBS in early 2018, according to a press release
Tidbit:  Also focused on cost saving innovation, the West Health Cost Innovation Summit was held in DC on 2/21/2018, here.   

CareFirst is looking for innovation that reduces costs ("really.")   This is sometimes termed, "reverse innovation" which seems like a terrible term to me.  See an article by Govindarajan in Lancet 2017 here and a forthcoming Amazon book by him, published by Harvard Business School, here (6/2018).


[*]  While I've quoted the cliche' there is not much commercial CED, there is a fairly lively dialog on at risk contracts in biopharma.  See a 2017 "Forum" (open access) at J Managed Care & Specialty Pharmacy here, and 2017 blog at Deloitte here, and an article on Value Based Contracts at National Pharmaceutical Council here and a June 2017 article in Managed Care Executive here.  A 2017 survey by the Academy of Managed Care Pharmacy is here and a 2017 survey of both US and European stakeholders is here.

Bill Promoting Genomic Testing: Advancing Access to Precision Medicine Act

This week, California's Eric Swallwell (D) introduced a bill with four other congressional supporters, the "Advancing Access to Precision Medicine Act."

The AAPMA bill has two parts.

The first part requests HHS and National Academy of Science to issue a comprehensive report on genomic medicine and overcoming barriers to precision medicine, including preventive uses of genomics. 

The second part provides federal funding for several years to any state Medicaid plans which opt in, and would provide 90% federal payment for genome sequencing in children with intractable undiagnosed disorders that have also required ICU admission.
  • Coverage at Genomeweb here.
  • Press release by Swalwell here.
  • The bill is H.R. 5026 and the status is "referred to the House Committee on Energy and Commerce."  Here.
    • Cosponsors are Shimkus (R-IL), Peters (D-CA), Paulsen (R-MN), and Vargas (D-CA).
      • Note it is a very different bill than the similarly titled Advancing Precision Medicine Act back in 2016 by Lamar Alexander (here).
      • Swallwell and other legislators urged HHS to study and promote precision medicine in a letter to HHS in March 2017, here.
  • Supporters include PMC, BIOCOM, American Medical Technology Association, and AACR, as well as the EveryLife Foundation for Rare Diseases.
    • Via a July 2017 letter to Swallwell, PMC supported the legislation in draft form, here.
The bill probably wouldn't score too expensively as one part is just a written federal report and one part is a very narrowly defined pediatric genomic benefit (e.g. 1000 children a year at $10,000 each would be $10M in a federal budget around $3T).

Personal note - 37-year-old Swalwell was born about ten miles from where I grew up in rural Iowa (his town was the county seat and had two stoplights, mine zero.  We had 17 people average per square mile; LA where I live has 7000 per square mile).  Swalwell eventually went to law school in Maryland and worked on the Hill before becoming a Democratic congressman for the SF East Bay/Oakland area.   

Sunday, February 18, 2018

Brief Blog: Berlin Health Entrepreneurship and Innovation Conference, May 7-8, 2018

Information is up for the 11th annual Health Entrepreneurship Summit in Berlin, sponsored by the Charite' hospital and the Berlin Institute of Health.   This is a large international event (language English) with about 400 participants. 

Main topics this year include:  AI, Digital health; Blockchain; Innovation Ecosystems; Microbiome; Regenerative and Cell Therapies; and Novel Funding and Affordability.   Registration is 360E/470E, before/after March 15.


For Berlinophiles there is also an Aging conference earlier, March 15-18, 2018.  There will also be a "World Biotech Congress" July 16-17, 2018.

The Charite' Summit is held in a rebuilt historical 1880s building in downtown Berlin (facing Gendarmenmarkt.)  There are plenty of downtown business district hotels, but I would recommend going a bit north, to Hackescher Market (just NE of Museum Island), for example Adina Apartment Hotel Hackescher Markt or other area hotels,   The pleasant riverside walk is 1 mile, 20 minutes or a correspondingly short taxi ride.

Alternative Tips.

Brief Blog: Health Europe 2.0 Conference near Barcelona, May 27-29, 2018

Co-sponsored with HIMSS Europe, see the Health 2.0 Conference in Sitges, a beach city about 20 miles south of Barcelona, on May 27-29, 2018.   The theme is "Working Together to Advance Digital Health in Europe."   The conference website is here.  Speakers include the Minister of Health for the Netherlands, the Global Head of Digital Health at Bayer, and the CEO of the NHS (UK) Digital Academy.   Full program here.   The commercial registration rate looks like $2099 to me, here (e.g. they state that a consultant or industry employee is "a non sponsoring vender.") 

Brief Blog: Health Innovation Conference in Vegas, May 6-9, 2018

In SF this week, several different people mentioned a large health innovation conference to be held May 6-9, 2018, at the Aria in Las Vegas.   It's abbreviated HLTH and the full title is THE FUTURE OF HEALTHCARE.   The speakers page already lists 240+ speakers - David Feinberg, CEO of Geisinger, Kieran Murphy, CEO of GE Healthcare, Ann Wojcicki, CEO of 23andMe, Andrew Hayek, CEO of Optum Health, Amy Abernethy, CMO/SVP of Flatiron, and so on.  (Speakers here).  The agenda runs five different tracks on each of four days (e.g. "Paying for Value, ROI Strategies," and "Digital Innovations," and  "Robotics and AI," and "Strategic Investors," and "Identity Crisis:Shifting Roles."  Don't miss the page-flipping brochure version with sound effects. The current registration rate is $1850.

The Aria is on the West Side of the strip, a block south of the Bellagio and in the CityCenter complex with Vdara and Cosmopolitan and the older Monte Carlo. 

Liquid Biopsy Panel from Guardant Health Awarded "EXPEDITED PATHWAY" at FDA

On February 15, 2018, Guardant Health announced that its multi-gene liquid biopsy somatic mutation panel has received the prestigious "Expedited Access Pathway" designation from the FDA. 

According to co-founder Helmy Eltouky, "Our FDA submission for Guardant360 is Guardant Health's top priority for 2018."  Press release here.

Liquid biopsy is a high visibility space, with 1060 hits currently on Pubmed.    FDA and AACR held a conference on the topic a few months ago in October 2017, decks and transcripts here.

Within the field, Guardant Health has been a leading player, having raised an additional $360M in May 2017 (here).    According to this week's press release, the value of the Guardant G360 has been supported by  >30 clinical outcome studies.

Wednesday, February 14, 2018

Brief Blog: Pitching Innovative Programs to the Health System C-Suite: Spellberg et al (2016)

Everyone working in healthcare talks about innovation, and this month Health Affairs features a dozen articles on diffusion of innovation in healthcare.

In the diagnostics industry, there have been articles and conferences about the need to move labs from "Lab 1.0" (specimen in, report out) to "Lab 2.0," a lab that is closely integrated with the goals of its health system, its ACO, best practices management, utilization management, high quality bundled care episodes, etc.   This perspective is not brand-new, but it's a growing emphasis.  For example, see Lab 2.0 articles in CAP TODAY here and here, an academic article by Crawford here, and the record of the November 2017 Santa Fe Lab 2.0 conference here.  I would also include Lewandroski's (MGH) book on modern lab goals and management here

Anyone interested in this topic, especially if you are based outside a health system, should read the excellent article by Spellberg and colleagues on pitching large scale innovation programs to the health system C-suite. 

The article uses the case study of funding and implementing new antibiotic stewardship programs.   However, I urge any reader to see it more broadly and thus applicable to bringing a "precision medicine program" to a health system, a pharmacogenomics program to the clinic, etc.   There may been many articles in the personalized medicine sphere over 10-15 years in which we are always on the cusp of some adoption revolution that is always two years away.  Spellberg's article provides a helpful perspective on what adoption and diffusion really means, in practical terms, in the nation's busy and overstresssed health care systems.

The article is open access here.

Spellberg B, Bartlett JG, Dilbert DN (2016)  How to pitch an antibiotic stewardship program to the hospital C-suite.  Open Forum Infect Dis 3(4):ofw210.

On a tangent to Spellberg, new articles on Alternate Payment Models and on ACOs in the 2/20/2018 JAMA, here and here.

Tuesday, February 6, 2018

How Long Might It Take CMS To Implement Rules for a Massive NGS NCD? (Scary)

On November 30, 2017, CMS released a massive draft NCD which would cover the whole domain of next generation sequencing technologies in oncology patients, both those with advanced cancer (where many new controls would be implemented) and those with non-advanced cancer (where NGS technologies would be banned outright).   Comment period closed in January (see comments here) and the final decision will be released on February 28 and effective on the same day.

How long would it take to figure out and implement all the associated rules?   One recent NCD example took a nearly ten month timeline.
  • On May 25, 2017, CMS released an NCD covering supervised exercise therapy (e.g. by physcial therapists) for patients with peripheral artery disease.  
  • According to some fairly obscure program materials, three months later, they sent an email to Medicare Advantage plans on August 12, 2017, stating that costs under the NCD would be paid by CMS FFS MACs rather than billed to MA plans through the end of CY2018.
  • On February 2, 2018, eight months after the NCD was released as final, CMS released "Change Request 10295," aka NCD Transmittal 204, describing the new coverage policy to contractors, with an "effective date" seven months in the past (May 25, 2017) and an "implementation date" of....wait for it...two months in the future at April 3, 2018.
    • The corresponding "claims processing" transmittal is here (also Change Request 10295, but a separate file as Claims Processing Transmittal 3969).

Hopefully for the hugely impactful NGS Oncology NCD, CMS could get the rules to medical centers and laboratories and MACs in faster than 8 months -- but that was the actual timeline for this 2017 NCD example

Very Brief Blog: Intelligent Digital Pathology & AI? Three Media Sources This Week

We've talked for years now about "digital pathology" as in storing pathology slide images as readily as we do MRI and CT scans. 

Three articles on advanced digital pathology and the intersection with AI in the last several days.
  • In MedCityNews, an article about a $25M spinoff for digital pathology oncology decision support coming out of Memorial Sloan Kettering,  here.

  • In FierceHealthcare, an article on whether "computer algorithms may soon outperform pathologists," here.
  • The FierceHealthcare trade press is tied to an original opinion piece in JAMA Oncology, by Acsa and Rimm at Yale, "Not Just Digital Pathology, Intelligent Digital Pathology," here.   
    • This in turn is an op ed on a JAMA research article by Bejnordi et al on "deep learning algorithms for the detection of lymph node metastases in breast cancer," here.


For an NEJM 2016 article on machine learning in radiology and medicine, here; trade press on radiology and AI here and here.   For much more in that direction, just google the key words    machine learning radiology.   

Friday, February 2, 2018

Speaking on Medtech & Investors; MEDCITY Tuesday May 1 in Chicago

I'm speaking on a panel on medtech reimbursement at the MEDCITY INVEST 2018 conference in Chicago, May 1-2, 2018.

See the full agenda here.

May 1, 2018, 10:15-11:00 AM

  • Bruce Quinn MD PhD, Bruce Quinn Associates
  • Everett Crossland JD MSC, Market Access & Reimbursement, Pear Therapeutics
  • "Show Me the Money:  How Investors Are Seeking Robust Reimbursement"

More after the break.

Very Brief Blog: Would Costs under Proposed NGS NCD Be Paid by Medicare Advantage Plans or by CMS FFS?

This blog provides an entry point to some CMS policy information that can be hard to nail down in its obscure resting places.

Medicare Advantage plans are generally obligated to pay for all services covered under Fee for Service Medicare, including the provision of immediate coverage for new NCDs.   This means a lot of new costs could occur under the proposed umbrella NCD for all next-generation sequencing services for cancer patients.   Medicare Advantage plans are not, however, responsible for costs that CMS has deemed exceed a regulatory "significant cost threshold."  These costs would be billable directly to the local Medicare FFS MAC even though the patient, otherwise, is in a Medicare Advantage plan.
  • An explanation of the rules is provided in Section E, page 23, of a February 1, 2018, CMS Medicare Advantage document (here).   
    • That's page 23 of a 231-page policy document. 
Rule 1:  CMS states that costs for "routine costs of clinical trials" under NCD 310.1 are billable to FFS Medicare, even for Medicare Advantage beneficiaries (the relevant regulation is 42 CFR 422.109).

Rule 2:  However, CMS states that costs for all other NCDs, including "coverage with evidence development" NCDs, are generally the responsibility of the Medicare Advantage plan from its capitated budget. 

Rule 3.  If and when CMS determines that there are special billing rules under the significant cost law for a particular NCD, instructions will be released by CMS as part of the billing instructions for that NCD.  So there's another Easter egg hunt for you.  For example, this document on page 182 states that "supervised exercise therapy NCD for symptomatic peripheral artery disease" is a MAC FFS benefit during CY2018 but not in CY2019.[*][**]

CMS will have to determine whether the NGS NCD, whatever its final form, meets the significant cost threshold that "holds harmless" the Medicare Advantage plans for unanticipated costs.   That accounting activity may be especially difficult if the prices of the tests are based on PLA codes or G codes and don't have fee schedule prices yet.   Additionally, CMS would have to estimate whether maximal costs are likely (many hundreds of thousands of patients at $3000? per test is many hundreds of millions of dollars), or whether clinical uptake of actual testing might be smaller than the NCD allows and only reach a fraction of the eligible patients, thus becoming a much smaller cost burden on Medicare Advantage plans.

For real nerds, Section E of the cited document also gives some complicated rules that Medicare Advantage plans must follow for tracking of patient copays for services that are shifted to FFS MAC payment.

[*] And they indicated this was conveyed in a MA program email on 8/12/2017.  See the NCD here and some trade press here.  CMS wrote, ""Trials showed that SET (supervised exercise therapy) decreases mortality, reduces cardiovascular risk factors, increases exercise capacity, and increases quality of life in older adults.  While the increase in general exercise capacity alone, which was an endpoint in a number of studies, would not be an appropriate outcome, it is a meaningful outcome for IC (intermittent claudication) due to symptomatic PAD since there is a well-established evidence link to all-cause mortality."
[**]  While the MA documentation in this blog (its page 182) says the exercise therapy NCD leads to MAC-covered benefits, not MA-benefits, that's not obvious to me from CR10295, which provides instructions to MACs.  Note also that while the NCD was released in May 2017, the CR to implement it was not released until February 2018 with an implementation date at MACs of April 2018 (effective date retroactive to May 2017).

The rule also contains some interesting text but the agency voluntarily broadening the definition of "health benefits" that may be supplemental benefits (rather than being rejected as outside the possible scope of supplemental benefits.).   However, MA plans must still go through arduous paperwork for each "supplemental benefit," which is becoming increasingly arcane as the whole fabric of the health system and modern delivery system evolve.  See page 182ff.

Very Brief Blog: Precision for Value's Map (Literally) of Value Propositions and Payers

Someone asked me this morning about biopharma pricing consultancies, and I ran across the homepage of the firm Precision For Value.   They have an interesting homepage aerial map of the product and payer world:  From Strategy to Health Outcomes Research to Promotional Effectiveness to Payer Insights to, finally, PayerLand.    Homepage here.  Pic below.

click to enlarge
A different consultancy, Evidera, has a less colorful home page but is currently promoting two interesting white papers.   The first is about payer sensitivities to surrogate endpoints (here), the second is about parallel scientific advice from European payers and life sciences regulators (here).

Brief Blog: What Were EXCITE and PACER (accelerated device coverage) and What Happened To Them?

At the early January 2018 JP Morgan conference, there was a fireside chat between Jeff Shuren (FDA), Tamara Syrek Jensen (CMS), and the CEOs of Exact Sciences and Foundation Medicine about parallel review.    Shuren made a strong reference to a program in preparation that would provide accelerated device coverage at CMS which would be different from parallel review. 

In late January, Modern Healthcare and other sources reported that the rapid payment program, EXCITE, had been pulled from the regulatory process.   If you missed this news, here are some links.

According to Modern Healthcare, here, there had been an industry proposal presented to CMS in 2H2016 called PACER - Provisional Accelerated Coverage to Encourage Research.   Early in 2017, CMS pushed a similar idea forward in the regulatory pathway called EXCITE - Expedited Review of Innovation Technology.   Modern Healthcare quotes Elizabeth Halpern of Hogan Lovells that nobody outside the agency knows how close EXCITE was to the industry idea, PACER.    The EXCITE proposed regulation was canceled without being released by about January 20, 2018.

  • Modern Healthcare, January 22, here.
  • Mass Device, January 23, here.
  • Fierce Biotech, January 24, here.
  • For more on PACER, here and here.
Separately of interest, see a detailed overview of CMS FDA Parallel Review online at Modern Healthcare, here.   See also a RAND publication, "Medical Device Innovation in the Era of Affordable Care Act: The End of Sexy," 2016, by Mattke et al., here or here.

For a recent general National Academy of Sciences review of expedited innovation efforts, here.   In February 2018, Health Affairs features a special issues on diffusion of innovation and clinical translation, here.

Thursday, February 1, 2018

NGS MAC Proposes LCD: PLA Codes and Admin_MAAA Codes Not Covered

On February 1, 2018, the NGS MAC (several midwestern states, New York, and New England) issued a draft LCD for public comment that states all administrative MAAA codes (00XXM) and PLA codes (00XXU) are non covered until an explicit coverage decision is made.  LCD DL37600 is online here.

This would mean the MAC does not need to do claim-by-claim review for each instance of each MAAA or PLA code as they come in.   Rather, they could be denied up front until a future date when a coverage decision was made.   The policy wouldn't apply to Category I MAAA codes (815XX).

The comment date runs through April 11, 2018.   The policy echoes a master MolDx LCD, used in all 28 states where MolDx is effective, that non-covers a wide range of molecular CPT codes until such time as a coverage article or policy is written for each CPT code in the non-covered range (here).

The proposal may be one new step in an ongoing tug-of-war between efforts to move regulatory processes faster (including accelerated approval at FDA, here, and rapidly issued codes, here) and the amount of evidence required before coverage decisions by payers.

Another consultant pointed out there is a small silver lining.  If a test has no LCD, and the manufacturer requests an LCD, there is actually no official path for that.  The MAC doesn't have to provide a meaningful response or make an LCD.   They can all but "circular file" a request for a new LCD.  But, here, a new PLA test would formally request a "reconsideration" of the PLA LCD, and the MAC would have to log it, process it, and respond formally under a set of rules and a fixed timeline.