The Integrator Blog
Share |
Contact Me, Experience, Mission, Sabbatical in Central America, plus
Editorial Advisory Board
Michael Levin
Taylor Walsh
background resources in PDF
Insurance, Integrative Clinics, Industry Summit Reports, News Files '99-'04
some organization links
Professions, Academia, Research, Policy
some CAM/IM publication links
Electronic, Peer-Reviewed, Blogs, More
Bradly Jacobs, MD, MPH, Revolution Health Blog
supported conference
Institute for Health & Productivity Management - Integrative/Complementary Healthcare
Michael Levin and Stephen Bolles: FICO's Medical Compliance and the Choice of Alternative Medicine PDF Print E-mail
Written by John Weeks   

Michael Levin and Stephen Bolles: FICO's Data Grabbing, Medical Improvement and Consumer Choice of Alternative Medicine

Summary: The stories of patients arriving at doors of integrative practitioner offices with avoidance of prescriptions as their motivation are legion. In an earlier column, Stephen Bolles, DC challenged the move of credit rating firm FICO into drug adherence monitoring in part for its failure to appreciate what may be called positive non-adherence. In this exchange, Integrator columnist Michael Levin notes the substantial estimates of harm from non-adherence. His take is friendlier toward FICO's intent. Bolles responds, and Levin replies. It's a complicated exchange in which paradigm collide. I add a couple comments.  
Send your comments to
to be included in a future Integrator.

Michael Levin
For integrative practitioners, this back-and-forth may be best framed around the well-known practice by patients of seeking alternative therapies or practitioners to avoid or get off a prescription drug they can't stand. As many practitioners know, adversity to taking a prescribed medication is many times a patient's critical step toward self-efficacy and health. If not managed, such a choice of non-compliance can be also be dangerous. 

Last month consumer expert Stephen Bolles, DC opened this subject with a
submission entitled Integrative Practice Issues as Credit Score Firm Begins Monitoring Drug Adherence. The FICO strategy hasn't the subtlety to capture whether or not a patient may in fact have become healthier through abstention from prescribed pharmaceuticals. 

Stephen Bolles, DC
Bolles' provoked an initial response from Integrator columnist Michael Levin. Levin, founder of Health Business Strategies, and knows something about the subject of adherence. He cites evidence of over 100,000 annual deaths associated with non-compliance. In a previous business life, this natural products executive helped lead a medical devices venture focused on adherence. Levin
collaborated with Robert C. Bogash, past-President of the American Society of Health-System Pharmacists on strategies to make what was then called "compliance" easier. They worked in the 1980s and early 1990s on medication dispensing management systems. This included design and intellectual property strategies. "All this work," Levin adds in a note with his column, "was done in an effort to improve outcomes while reducing healthcare costs."

Levin views the FICO gambit as a more severe and evolved example of "using carrots and sticks" to change patient behavior. Levin has written on this theme in previous Integrator columns. Bolles pushes back.

Levin knows Bolles. As he recounts in the lead in, the two met at an Integrative Medicine Industry Leadership Summit a decade ago. Bolles responded to Levin, and Levin replied again. This topic lives at the intersection of the highest technology of data-grabbing, and the collision of conventional and alternative strategies to "improve outcomes while reducing healthcare costs."


Levin-Bolles Dialogue on Consumerism, Alternative Medicine
and the FICO Compliance Scoring Story

Michael Levin: Prelude to a dialogue

I thought Levin's re-introduction to Bolles would be useful in setting the tone. For some of us it will be a reflection on the early days of the integration dialogue at that Leadership Summit.
Stephen - Thanks so much for calling out the FICO scoring story. Medication non-adherence was a passion of mine back in the days when it was called "non-compliance." (Yes, I'm dating myself!) I hope you are doing well. It's been a long time since we visited in Arizona at one of John's "retreats"!  :)

Anyways, attached is my take on the matter. I view this through a slightly different lens than you do. In digging through the FICO info, it appears that patient privacy is preserved (good news) but the probability models are based on certain demographic buckets (bad news - that's just unfair and, if deployed, probably will result in lawsuits). How much weight this will carry in driving premiums or benefit designs remains to be seen. It's simply another profitable product for FICO to sell to a desperate industry needing to drive down costs.

For me, the big news here is that this is just another reaction to economic pain, changing behavior by using economic carrots and sticks. I hope you both enjoy this submission, and, I certainly would welcome your insights. 

Michael Levin
Health Business Strategies
Clackamas, Oregon

FICO Medication Adherence Scores:
A Potentially Dangerous Outcome of Consumer Misbehavior?

Michael D. Levin
Health Business Strategies
Clackamas, OR
The high cost of medication noncompliance is staggering. Thousands of scholarly and not so scholarly articles have been published on this subject over the past 35 years. A research review on the subject published in the Journal of Applied Research in Clinical and Experimental Therapeutics (1) offers these sobering statistics about medication non-compliance:

  • 125,000 preventable deaths/year
  • $8.5 billion in avoidable hospital costs
  • 14% - 21% of patients never fill the original Rx
  • 25% of nursing home admissions are related to improper self-administration of medicine.
We've known for decades that somewhere between 50% and 70% of prescriptions are not taken as prescribed. And we've known for decades that, as a consequence, people stay sick longer or die and money is wasted. And an entire industry emerged 30 years ago to battle this problem.

Will it work as intended? No. Why?
Because it is unfairly based on probability
models - not on an individual's
personal behavior pattern.  

Over the past 25 years, I've had the pleasure of working with brilliant pharmaceutical scientists and engineers working to fight this problem. One of them was Robert C. Bogash, past-President of the American Society of Health-System Pharmacists. Truth be told, he was the man who invented unit-dose drug distribution. (When you open a unit package of salt, or take an individually wrapped tablet, thank Bob - it was his idea back in the 1950s (2).) In the 1990s, Bob developed automated remote drug dispensing technology still in use today. There are scores of patents up to and including a remote-video teleconferencing drug dispenser that records the patient taking his medication once it is dispensed. Think prisons.

Knowing he'd be keenly interested, I called Bob to tell him about Dr. Stephen Bolles' article on the FICO Medication Adherence Score. He granted me permission to quote: "Michael, at age 86, I am staggered by this news and, quite possibly, for the first time in my life, am speechless!"

Bob and I both know that this is clearly an economic attempt to more fairly attribute the costs of consumer misbehavior and, in so doing, improve payer margins.  Will it work as intended? No. Why? Because it is unfairly based on probability models - not on an individual's personal behavior pattern.  While I share Stephen's concern for privacy, and totally agree with him that some medication non-compliance occurs for very valid reasons, I have a different perspective (bias). This FICO model is merely a creative expression of an industry wanting to reduce costs by punishing bad behavior.  As auto insurance premiums are "red-lined" by zip code, the theft frequency of the particular model, and the overall credit score of the insurance applicant, it appears that the FICO model will "red-line" individuals based upon correlating third party (ie, demographic data) with dispensing history for selected base populations, stripped of personal identifiers.

To me, this is simply one more symptom
of economic pain caused by consumers
(that also happens to kill > 125,000
of them each year).

Bottom line: If you have the misfortune of falling into the wrong "bucket of data," you will be scored a medication adherence risk and, therefore, will probably pay a higher premium, co-pay or both than someone who falls into a better "bucket of data". On the 500 point FICO medication adherence scoreboard, over 400 is good; less than 200 is bad. How much influence this score will have on overall premiums and benefit structures remains to be seen. How it will be updated and adjusted is unknown.

If it plays out this way, future lawsuits based on unfair, and possibly discriminating pricing practices will most certainly occur.  Lawyers, start your engines!

But, for me, the real story here is much bigger. This is another creative example of using economic carrots and sticks to alter consumer behavior. Life insurance companies have adjusted premiums to known actuarial risks for decades (think: smoker vs non-smoker rates). This is simply a well-intentioned, but deeply flawed effort to punish consumer misbehavior that wastes billions of health care dollars. It is simply one more symptom of economic pain caused by consumers (that also happens to kill > 125,000 of them each year).

The American Heart Association describes cardiovascular medication non-adherence as "the number one problem in treating illness today"(3). Even in Portugal, this economic dimension of this problem is well documented (4). Indeed, another strategy recently deployed to address problems with antihypertensive medication non-adherence is one in which the insurance company mandates a periodic intervention with a RPh for certain chronic medication prescriptions. Folks who comply have a significantly lower co-pay on the prescription than those who do not comply. Economic carrots and sticks. The days of (as I like to say) the "fishes and loaves" are gone. This is "hardball."

Recently, a colleague challenged me on the arguments I made in the 1990s regarding the economic non-sustainability of our disease-care system. "An economic trainwreck!", I declared back then! After all, he argued, we're still spending, more and more. I reminded my dear friend that we've simply raised the credit limit, just like Congress just did. The bills are growing. The day of reckoning will come. It's simple math. Finite resources, increasing consumption + consumer misbehavior = crisis.

And, when that happens, we may all be forced to take our medicines under the watchful eye of some medication dispensing robot built by Bob.

2.  Stephen Bolles responds to Levin

Michael, a good morning to you. Thanks for your note and very nuanced take on the FICO medication scoring issue.

I appreciate the emphasis on the very real problem of individuals who need medication and don't take them as needed or cannot afford what is prescribed. That problem, as you articulately point out, still is in search of an effective strategy, as those behaviors contribute to poor outcomes. It is a very real tragedy in many respects that does not need to happen.

My concern is that the complexity
of the decision-making context
for these 'non compliant' individuals
is simply not being captured.

The system is engineered, essentially,
to avoid capturing it. That's
another paradigmatic flaw. 

But I am probably even more concerned about this than when I 'flamed on' over the issue when it was first written up. My concerns are certainly in part based on issues of data privacy. The underlying issue for me about these data is not privacy per se, but one of data ownership. These data are either private or not, shared or not, all essentially managed without informed consent.

My primary concern about health data is not about privacy, but about transparency, informed management and ownership decisions, and the value of the data itself. These data have become monetized, and the value of that monetization is not accruing in consumers' favor. People are, essentially, making money on our behaviors without our participation and license. That's a fundamental paradigm difference than what the current system uses. But the current paradigm, in my view, is deeply flawed, and our lack of active participation in these transactions contributes to poor societal health and wellness outcomes.

My second concern, which I probably did not articulate fully enough in my notes to John, is that the complexity of the decision-making context for these 'non compliant' individuals is simply not being captured. The system is engineered, essentially, to avoid capturing it. That's another paradigmatic flaw.

 At UnitedHealth, I vocally argued
that individuals who were 'non complaint'
were making very intentional decisions
for multiple reasons that needed
to be explored and understood.

My third concern--and last for this note--is the issue of the term 'non compliant.' While I was at UnitedHealth, I saw this term in 'fullest flower'--used to describe a basis for partitioning individuals into two convenient but extremely imprecise buckets. I rather vocally argued (totally out of character, I know) that individuals who were 'non complaint' were making very intentional decisions for multiple reasons that needed to be explored and understood, because there were very real consequences of assigning individuals to those categories--consequences those individuals were not being informed about. My insistence, as they say, didn't 'gain any traction.'

For me, all this underscores my concerns that the consumer--as much as many hate that term--does not occupy a position in the data economy where transactions are taking place that use them, but do not engage them.

So I do not disagree with anything you wrote; on the contrary, I see your viewpoint as a necessary vector for a more holistic appreciation of the problem.

Warm regards! I hope our paths cross again.

Stephen Bolles, DC

3.    Levin responds to Bolles'

Good morning to you, Stephen, and thank you for your thoughtful response.

Point One
: You raise an important point concerning data privacy, which, my friend, has become historical artifact, sad to say. Monetizing consumer behavior data is the very foundation for advertising. While one could argue that the monetizing of aggregated data (such as this FICO medication score) is simply an unpleasant marketplace reality, I find myself flaming at targeted digital behavioral advertising which targets the specific consumer in real-time, monetizes his physical whereabouts, and sends a message to his cell-phone telling him to buy a discounted smoothie from Jamba Juice as he strolls past it's front window at the mall. That is an egregious invasion of personal privacy, facilitated through cell phone/satellite technology. Grrrrr.  Ever wonder why those pop-up banner ads seem relevant to you when you visit a website? It's because they are. And folks are making tons of money on targeted behavioral adversting. Just ask Google.

"Not to defend a privacy-busting practice (but)
using behavioral datasets to extinguish unwanted
behaviors which add costs to society, does not
the overall costs to society decline, thus
benefiting others?

Not to defend a privacy-busting practice, I would submit to you that one could argue that this paradigm does not necessarily always NOT accrue in the consumers behavior (usually not, but one could argue that at times it does). How? In using behavioral datasets to extinguish unwanted behaviors which add costs to society, does not the overall costs to society decline, thus benefiting others?  

I'll offer one example I dealt with many years ago. The SIG for a Rx medication was 2 tabs QID. Eight tabs/day, right? It is urinary tract antibacterial. Use 8 tabs, it works great. Use less than 7 and it typically fails. The New Prescription Audits (NPAs) for the pharmacies in my zip codes had an average Rx of 5.5 tablets/day. My job was to educate the docs to get them to prescribe correctly to optimize clinical outcomes. Now, you and I both know that the folks from IMS who sell the monetized data in the form of NPAs to the drug industry profit handsomely from that data product. And you and I both know that the prescribing doctors were generally unaware that their aggregated prescription activity was being monetized. Is this example a bad thing? I'd argue that improving the public health by changing ineffective, costly prescribing habits is a good thing. And in what year was I involved in this example of dataset monetization?  1974.

Point Two
: I agree with you. Capturing only the "what is" and not the "why is it" is a paradigmatic flaw. It prevents us from understanding the "why" behind the "what" of consumer behavior. For me, that's a lost opportunity cost. For businessmen, one could argue "who cares about the why, just change the what, right now!" Capitalism in full bloom.  

Point Three
: Thanks for sharing your real-world experience with the term "non-compliance" at United. Reductionist models are expedient, but, as you point out, imprecise models which unfairly impact consumers (just as this FICO product promises to do!). I salute your efforts at United to challenge the models and, perhaps more importantly, promote transparency and consumer protections. Had our regulators enforced such transparency requirements in the mortgage lending space, we'd not be in the horrific economic mess we are now in.

It's tough being Sisyphus!

I deeply enjoy this thoughtful dialog and again, thank you for your engagement and consumer activism. Keep fighting the good fight!


PS: This "dataset ownership" discussion brings to mind The Little Prince. (I'm going from memory which, at my age, sometimes fails - so please bear with me.) Upon his visit to the planet inhabited by the accountant, he was puzzled why the accountant kept count of all the stars.  The accountant told the Prince that by counting the stars, he owns them. The Prince thought this very odd. How can one own all the stars? In our discussion, do the folks who own all the data own all the people? :)

References from Levin's initial response

(1) Wetheimer, Albert and Santella, T. "Medication Compliance Research: Still So Far to Go" Jrnl App Research in Clinical and Exp Therapeutics
(2) Jorgenson, KM "Robert C. Bogash: Gadgeteer and hospital pharmacist extraordinaire" Am. J. Health Syst Pharm 2010, Jan 15; 67(2):154-7
(4) Morgado, et al "Predictors of uncontrolled hypertension and antihypertensive medication non-adherence" J Cardiovascular Dis Res. 2010 Oct-Dec; 1(4): 192-202


: While I suspect that there may be subject matter in here for author Dan Brown, I was not sure whether this dialogue was proper Integrator fare. I do think it is good for those of us caught in the touch-feel zone of actual human interaction that defines much of the integrative practice space to be reminded of the very different contexts and problem sets that make up much of the US healthcare system.

The stories of patients arriving
at doors of integrative practitioner
office with avoidance of prescriptions
as their motivation are legion. 

Levin's citation of the deaths associated with non-compliance are an interesting counter-point to the deaths and harm associated with compliance with prescriptions. The latter include both adverse effects and drugs that simply worsen conditions over time. This then sets up what Bolles was trying to cajole UnitedHealth to do. He urged them to break down the reasons when there is non-compliance. In the early days of the NIH NCCAM, I innocently thought that this might be the kind of substantive, useful work with which the agency would be engaged.  Which CAM users began their exploration of CAM to avoid prescription drugs? Which are able to back off prescriptions due to chiropractic or yoga or name the alternative treatment? What alternatives did they chose?  That would be useful. The stories of patients arriving at doors of integrative practitioners with avoidance of prescriptions as their motivation are legion.

Ultimately, given the hard numbers Levin offers for deaths and harm from non-compliance, wouldn't it be nice to begin to separate out where non-compliance is health giving? This is truly patient-centered outcomes. Perhaps the new Patient Centered Outcomes Research Institute can start grappling with these kinds of meaty questions.

I wonder if some of the resistance to asking such questions is because we are truly talking, in these patient choices, or in integrative efforts to limit or end undesired pharma, about alternative medicine.

Send your comments to
to be included in a future Integrator.

< Prev   Next >
Integrative Practitioner
The Westreich Foundation
voluntary contributions
Support the work!
All Integrator Round-ups
Integrator Top 10 Lists 2006-2015
Issues #140-#142 Oct-Dec 2015
Issues #137-#139 July-Sept 2015
Issues #134-#136 April-June 2015
Issues #131-#133 Jan-March 2015
Issues #127-#130 Sept-Dec 2014
Issues #123-#126 May-Aug 2014
Issues#119-#122 Jan-April 2014
Issues #116-#118 - Oct-Dec 2013
Issues #113-#115 July-Sept 2013
Issues #110-#112 April-June 2013
Issues #108-#109 Jan-March 2013
Issue #105-#107 Oct-Dec 2012
Issues #102-#104 - July-Sept 2012
Issues #99-#101 - April-June 2012
Issues #96-#98-Jan-March 2012
Issues #94-#95 Nov-Dec 2011
Issues #92-#93 Sept-Oct 2011
Issues #90 and #91 - July-Aug 2011
Issues #88 and #89 - May-June 2011
Issues #86 and #87 - March-April 2011
Issues #84 and #85 - Jan-Feb 2011
Issues #82 and #83 - Nov-Dec 2010
Issues #80 & #81 - Sept Oct 2010
Issues #78 & #79 - July August 2010
Issues #76 & #77 - May June 2010
Issues #74 & #75 - March-April 2010
Issues #73 & #73 - Jan-Feb 2010
Issues #69, #70 & #71 - Nov-Dec 2009
Issues #67 and #68 - Sept-Oct 2009
Issues #65 and #66 - July-August 2009
Issues #63-#64 - May-June 2009
Issues #60-#62 - March-April 2009
Issues #57-#59 - Jan-Feb 2009
Issues #55-#56 - Nov-Dec 2008
Issues #51-#54 - Sept-Oct 2008
Issues #47-#50 - July-August 2008
Issues #46 & -#47 - May-June 2008
Issues #43-#45 Mar-April 2008
Issues #41 & #42 - Feb 2008
Issues #39 & #40 - Dec-Jan '08
Issues #37 & #38 - Nov 2007
Issues #35 & #36 - Oct 2007
Issues #33 & #34 - Sept 2007
Issues #30-#32 - July-Aug 2007
Issues #28 & #29 - June 2007
Issues #26 and #27 - May 2007
Issue #25 - April 2007
Issues # 23 & #24 - March 2007
Issues #21 and #22 - Feb 2007
Issues #19 and & 20 - Jan 2007
Issues #17 and #18 - Dec 2006
Issues #15 and #16 - Nov 2006
Issues #13 and #14 - Oct 2006
Issues #11 and #12- Sept 2006
Issues #9 and #10 - Aug 2006
Issues #7 and #8 - July 2006
Issues #5 and #6 - June 2006
Issues #3 and #4 - May 2006
Issues #1 and #2 - April 2006
All Articles by Subject: 2006
All Articles by Subject: Jan-June 2007
IAYT-Sponsored Series on the Future of Yoga Therapy