Sunday, November 29, 2015

Blackberry, Nokia: Two Business Books

We are well aware of many colossal collapses in the business world, and they can occur in an accelerated way in the digital age - Netscape booms, then loses to IE and Firefox; Myspace is rapidly replaced by Facebook; Yahoo search loses to the Google monolith; Blockbuster caves in with the rise of web-based Netflix.   In the phone world, Motorola and Nokia dominated everything until they abruptly didn't.  Two new books profile the decline of Blackberry and Nokia.

In Losing the Signal: The Untold Story Behind the Extraordinary Rise and Spectacular Fall of BlackBerry (2015), Jacquie McNish and Sean Silcoff tell the story of a small hardscrabble Canadian tech company called RIM that rose to the top of the mobile world.  Blackberries were conceived in the digital stone age when we had (A) PalmPilots that you sync'd in the morning with your PC and there were (B) two-way pagers with very limited text capabilities.  

RIM recognized there could be a huge untapped market for providing full and seamless high security email services on mobile devices.   The early BlackBerries didn't have phones at all.  Management was appalled when iPhones were introduced in 2007: because there wasn't enough bandwidth at carriers for all that internet access (there wasn't), because people wouldn't type on glass, and because the battery was pathetic - 10 hours not 2 days.

BlackBerry avoided the internet like the plague for a while longer.  The company, despite its reputation and success, had always had a head-just-above-water approach to its global servers, and these collapsed a few times, damaging its reputation.  RIM then bungled several attempts at more modern smartphones and had delayed and sank R&D into misguided efforts like its Play tablet.  The "spectacular fall" ensured.  Some of RIM's most tightly held assumptions...that internet and corporate email access on iPhones would be difficult or impossible... proved completely wrong as the bring-your-own-device business world emerged a few years ago.

In The Decline and Fall of Nokia (2015), David J. Cord tells a similar story.  Nokia dominated the global phone market as a miracle Finnish company that could do no wrong - $50B in sales and $7B profit in 2007, the year the iPhone was introduced.  It provided a massive range of devices, for the low, mid and high markets, all rapidly remodeled, and customized to each carrier.  The endless varities of phones ran on a diversely versioned operating system called Symbian.  Android grew rapidly with a completely different model - a one fits all open access operating system.

As Cord points out, iPhone had one phone, one operating system, and was riding on top of Apple's booming success with iPods and iTunes.   Nokia declined rapidly, entering a death spiral by 2011/2012.  Just like RIM, trying to follow the iPhone, it began producing more sophisticated devices that didn't work well.  Attempts to break off Symbian with a proprietary new system, with MS-Windows, and with a Linux-like system all stumbled in turn.  

Lack of a common software platform for Nokia's phones made an iOS like "app and entertainment" ecosystem impossible, and the option of iPads using the same ecosystem made iOS even more attractive, as iPads and iPhones boomed together.

Eventually Nokia was (essentially) acquired by Microsoft, a transaction that soon generated a $7B writeoff.   After a share price peak of $230 around the time the iPhone was introduced, RIM fel to $50 in 2010/2011 and is at $8 today, where it has sat since 2012.

The stories, while different, reproduce at least three of the seven "Billion Dollar Lessons" cataloged by Carroll and Mui in their 2007 book  Billion Dollar Lessons: "Staying the Course [Threat? What Threat?]," "Fumbling Technology: Riding the Wrong Technology," and "Consolidation Blues: Doubling Down on a Bad Hand."

Lessons for healthcare and digital health?  I'm going to think about that....

Decline and Fall of Nokia is paper book only.  Losing the Signal is available in book, eBook and Audible versions, as is Billion Dollar Lessons.

Tuesday, November 24, 2015

My November, 2015 comment on PAMA regarding the definition of ADLTs.

My November, 2015 comment on PAMA regarding the definition of ADLTs.
I disagree with CMS's interpretation of the PAMA definition of ADLT to necessarily include "DNA or RNA."

While the PAMA rule does not give a clear explanation how this was reached, at a CMS advisory board in October, CMS staff explained that the definition of ADLT - "DNA, RNA, or protein" - was read in context of the overarching term "Advanced diagnostic laboratory test" to exclude proteomics-only tests.

This does not make enough sense to hold as a reasonable interpretation of PAMA's ADLT definition, which in contrast should be virtually self implementing.

First, we look to context of a bill when the paragraphs in question are unclear. "A, B, or C" is simple English grammar, and does not require canons of interpretation. If I tell my daughter, "Come home on the bus, subway, or taxi" it does not require her to come home only on the bus or subway. The grammar of ADLTs is the same simple English. There is nothing unclear that requires context and labored interpretation in the concept: "Use A, B, or C."

Second, even though it is inappropriate in the first place, the conceptual outreach to the mere heading, "Advanced diagnostic laboratory test" to define the tests are requiring DNA, RNA, or both, at a minimum, does not pass the sense test either.

(A) The term ADLT has not existed before, therefore it has no pre established meaning.
(B) To the extent the ADLT category close resembles either the IVDMIA (FDA) or the MAAA (AMA CPT) categories, those include RNA, DNA, or protein tests, including protein only MAAAs. The differences between the IVDMIA, MAAA, and ADLT terms are hairsplitting, in reference to multi analyte algorithm tests with single results, AND that group has included MAAAs such as the Crescendo VECTRA test.
(C) There is no bias against proteomics being one of the fundamental modern "Omics" along with DNA and RNA. Proteomics is referred to as the next frontier of molecular medicine, not a backwater.

Thus, the outreach to the  four word category "advanced diagnostic laboratory test", cannot be used to support use of DNA or RNA, as required components.

Third, CMS has built nearly a house of cards (as shown above) in its logic and assertions that ADLT, viewed as a whole, almost magically excludes protein only ADLT tests. It defies belief that such a tortured logical path of assumptions and semantic inferences - not even very well justfied ones, and unexplained in the proposed rule itself - was the main and simple intent of Congress in writing and signing the definition as we have it.

Wednesday, November 18, 2015

My PSA Metric Comment

For the main article, here

"Non recommended PSA screening," as proposed, is a faulty metric for CMS.  First, there is a statutory benefit for PSA screening.  It is unfair to doctors to be placed between legal patient benefits Congress has voted for, and the position of the metric.   Second, there is substantial difference of opinion among major policy bodies on PSA testing, particularly, if a patient and doctor do it appropriately, even if USPSTF does not endorse it.  Third, the text of the metric is confusing and contradicts a CMS National Coverage Determination.  The metric text blocks screening for example "except with history of cancer" and of course that is NOT screening at all.  The text also has rare exceptions like "known dysplasia" but the CMS NCD for PSA  is much broader and endorses PSA for workup of nocturia, hesitancy, and other conditions. Arguably, that is "not screening" but most of the exceptions listed by the metric are not screening either. Hence, it is very confusing.

Instructions for JIRA commenting

For the main article, here

This is the text of the PDF in a zip file for how to comment on a currently proposed CMS performance metric.

You comment via the Office of the National Coordinator (for HHS), Project Tracking Webpage, via a section called "JIRA".  I initially suspected the "IR" stood for Issue Resolution; a management and/or developer based bug tracking system (here) seems to be retrofitted for public policy comments.  Possibly JIRA is simply a shortening of GOJIRA, the Japanese term for Godzilla (here).


If you are a current JIRA user with an active account: (Select “Log In” at the top right corner.)
If you are new to JIRA and do NOT have an account: (Select “Sign up” to set up a new account and log in to the JIRA tool.)
Once logged in, enter your comments by following these instructions:
Select “Projects” at the top middle of the home screen.
Select “View all Projects.”
Under Quality-Measures, select “Comments on eCQMs under development” project.
To enter comments, select “Create issue” (orange button) at the top middle of the screen.
Select the type of issue from the “Issue type” drop-down menu.
Fill out the fields labeled “Summary,” “Contact name,” “Contact email,” and “Contact phone.”
In the “Summary” field, type in one of the following titles, based on the measure you are commenting on:
“Comment—FSA CHF” for the “Functional Status Assessment and Target Setting for Patients with Congestive Heart Failure” measure
“Comment—PSA Screening” for the “Non-Recommended PSA-Based Screening” measure
Enter your comments in the “Description” field.
Select the measure you are commenting on from the “Draft measures” drop-down box:
For the “Functional Status Assessment and Target Setting for Patients with Congestive Heart Failure” measure, select “FSA_CHF”
For the “Non-Recommended PSA-Based Screening” measure, select “PSA_Screening”
Select ‘Create’ at the bottom left to submit your comments. If you would like to enter more comments, select “Create another” and then “Create.”
To be considered, comments must be submitted by November 20, 2015, via the JIRA comment tool.