How the Search for Perfect Markets has Damaged Health Policy

Sometimes ideas in healthcare are so powerful that they haunt us for generations even though their link to the real world we all live in is tenuous. The idea of “moral hazard” is one of these ideas. In 1963, future Nobel Laureate economist Kenneth Arrow wrote an influential essay about the applicability of market principles to medicine entitled “Uncertainty and the Welfare Economics of Medical Care”.    

One problem Arrow mentioned in this essay was “moral hazard”- the enhancement of demand for something people use to buy for themselves that is financed through third party insurance. Arrow described two varieties of moral hazard: the patient version, where insurance lowers the final cost and inhibitions, raising the demand for a product, and the physician version–what happens when insurance pays for something the physician controls by virtue of a steep asymmetry of knowledge between them and the patient and more care is provided than actually needed. The physician-patient relationship is “ground zero” in the health system.

Moral hazard was only one of several factors Arrow felt would made it difficult to apply rational economic principles to medicine. The highly variable and uniquely threatening character of illness was a more important factor, as was the limited scope of market forces, because government provision of care for large numbers of poor folk was required.  

One key to the durability of Arrow’s thesis was timing: it was published just two years before the enactment of Medicare and Medicaid in 1965, which dramatically expanded the government’s role in financing healthcare for the elderly and the categorically needy. In 1960, US health spending was just 5% of GDP, and a remarkable 48% of health spending was out of pocket by individual patients. 

After 1966, when the laws were enacted, health spending took off like the proverbial scalded dog. For the next seven years, Medicare spending rose nearly 29% per year and explosive growth in health spending rose to the top of the federal policy stack. By 2003, health spending had reached 15% of GDP! Arrow’s  moral hazard thesis quickly morphed into a “blame the patient” narrative that became a central tenet of an emerging field of health economics, as well as in the conservative critique of the US health cost problem.  

Fuel was added to the fire by Joseph Newhouse’s RAND Health Insurance Experiment in the 1980s,  which found that patients that bore a significant portion of the cost of care used less care and were apparently no sicker at the end of the eight-year study period. An important and widely ignored coda to the RAND study was that patients with higher cost shares were incapable of distinguishing between useful and useless medical care, and thus stinted on life-saving medications that diminished their longer term health prospects. A substantial body of consumer research has since demonstrated that patients are in fact terrible at making “rational” economic choices regarding their health benefits. 

The RAND study provided justification for ending so-called first dollar health  coverage and, later, high-deductible health plans. Today more than half of all Americans have high deductible health coverage. Not surprisingly, half of all Americans also report foregoing care because they do not have the money to pay their share of the cost!   

However, a different moral hazard narrative took hold in liberal/progressive circles, which blamed the physician, rather than the patient, for the health cost crisis.  

The Somers (Anne and Herman) argued that physicians had target incomes, and would exploit their power over patients to increase clinical volume regardless of actual patient needs to meet their target income. John Wennberg and colleagues at Dartmouth later indicted  excessive supply of specialty physicians for high health costs. Wennberg’s classic analysis of New Haven vs. Boston’s healthcare use was later shredded by Buz Cooper for ignoring the role of poverty in Boston’s much higher use rates.

The durable “blame the physician” moral hazard thesis has led American health policy on a  futile five-decade long quest for the perfect payment framework that would damp down health cost growth–first capitation and HMOs, then, during the Obama years, “value-based care”–a muddy term for incentives to providers that will eliminate waste and unnecessary care. Value-based care advocates assume that  physicians are helpless pawns of whatever schedule of financial rewards is offered them, like rats in a Skinner box. If policymakers can just get the “operant condition schedule” right, waste will come tumbling out of the system. 

The end result of this narrative: thanks in no small part to the festival of technocratic enthusiasm that accompanied ObamaCare, (HiTECH, MACRA, etc), physicians and nurses now spend as much time typing and fiddling with their electronic health records to justify their decisions as they do caring for us. Controlling physician moral hazard thru AI driven claims management algorithms has become a multi-billion business. The biggest “moral hazard mitigation” company, UnitedHealth Group, has a $500 billion market cap.

Thus the poisonous legacy of Arrow’s “moral hazard” thesis has been two warring policy narratives that blame one side or the other of the doctor-patient relationship for rising health costs. It has given us a policy conversation steeped in mistrust and cynicism. You can tell if someone is a progressive or conservative merely by asking who they blame for rising health costs! 

There were credible alternative explanations for the post-Medicare cost explosion. Recall that the point of expanding health coverage in the first place was that better access to care DOES in fact improve health.  Medicare lifted tens of millions of seniors out of poverty, improving both their nutrition and living conditions. Medicaid dramatically broadened access to care for tens of millions in poverty. This expansion of coverage, and the added costs, deserve much credit for the almost nine year improvement in Americans’ life expectancy from 1965 to 2015. 

It is also worth recalling that the two most explosive periods of inflation in the post-WWII US economy were the late 1960’s, the so-called Guns and Butter economy that financed the Vietnam War, and the mid 1970’s to 1981, fueled by the Arab oil embargo. These periods of hyperinflation coincided with the coverage expansion, amplifying their cost impact.

And of course, the 1980’s also saw a flood of optimistic, high energy baby boomer physicians, the result of a dramatic federally funded expansion of physician supply begun by Congress during the 1970’s. The reason for this surge: we did not have enough physicians to meet the demands of the newly enfranchised Medicare and Medicaid populations. 

This surge in aggressive young physicians coincided with dramatic expansion in the capabilities of our care system. Non-invasive imaging technologies such as MR, CT and ultrasound, ambulatory surgery dramatically lowered both the risks and costs of surgical care. The advent of effective cancer treatments, cut the cancer death rate by one-third from its 1991 peak. The advent of statins, and less invasive heart treatment has reduced mortality from heart disease by 4% per year since 1990, despite the rise in obesity!

Medicine today is of an different order of magnitude of clinical effectiveness, technical complexity and, yes, cost, than that on offer in 1965. No one would trade that health system for the one we have today.    

However, the biggest problem with the moral hazard theses–both of them–was the assumption that the physician and the patient are primarily motivated by “maximizing their utility” in the healthcare transaction. Arrow knew better. He emphasized the role that fear and existential risk played in their interaction, given that illness, particularly serious illness, is, as he put it, “an assault on personal integrity”. Given the level of personal risk, it is easy to understand why both patients and physicians will not be obsessing about the risk/benefit relationship of every single medical decision.

By reducing the physician-patient interaction to a morally fraught mutual quest of the proverbial free lunch, economists have not only insulted both parties, but grossly oversimplified this complex interaction. Is a sick person really “consuming” medical care, like an ice cream bar or a movie? Is the physician really “selling” solutions regardless of their effectiveness, unconstrained by pesky professional ethics, or rather groping through fraught uncertainty to apply their knowledge to helping their patient recover? 

In contrast to virtually every other Western country, American health policy has been obsessed for nearly sixty years with fighting moral hazard and, in the process, saddling almost 100 million Americans with $195 billion in medical debts (the vast majority of which are uncollectible). Isn’t it ironic that those other wealthy countries that provide their citizens care free at the point of service spend between 30-50% less per capita on healthcare than we do?  And that both physician visits and hospitalization rates are far lower in the US than most of these countries.   

There is no question that healthcare in the US today is very expensive. But health costs have been dead flat as a percentage of US GDP for the past thirteen years. The explosive growth in health costs is over. Increasingly, attention is turning to the real culprit–socially determined causes of illness, and the inadequacy of our policies toward nutrition, shelter, mental health, gun violence and investment in public health. It’s time for the economists to eat some humble pie, and acknowledge that medicine will probably never fit in their cartoon universe of “Pareto optimality in perfect markets”.

Jeff Goldsmith is a veteran health care futurist, President of Health Futures Inc and regular THCB Contributor. This comes from his personal substack

By KIM BELLARD

Larry Levitt and Drew Altman have an op-ed in JAMA Network with the can’t-argue-with-that title Complexity in the US Health Care System Is the Enemy of Access and Affordability. It draws on a June 2023 Kaiser Family Foundation survey about consumer experiences with their health insurance. Long stories short: although – surprisingly – over 80% of insured adults rate their health insurance as “good” or “excellent,” most admit they have difficulty both understanding and using it. And the people in fair or poor health, who presumably use health care more, have more problems.

Health insurance is the target in this case, and it is a fair target, but I’d argue that you could pick almost any part of the healthcare system with similar results. Our healthcare system is perfect example of a Rube Goldberg machine, which Merriam Webster defines as “accomplishing by complex means what seemingly could be done simply.”   

Boy howdy.

Health insurance is many people’s favorite villain, one that many would like to do without (especially doctors), but let’s not stop there. Healthcare is full of third parties/intermediaries/middlemen, which have led to the Rube Goldberg structure.

CMS doesn’t pay any Medicare claims itself; it hires third parties – Medicare Administrative Contactors (formerly known as intermediaries and carriers). So do employers who are self-insured (which is the vast majority of private health insurance), hiring third party administrators (who may sometimes also be health insurers) to do network management, claims payment, eligibility and billing, and other tasks.

Even insurers or third party administrators may subcontract to other third parties for things like provider credentialing, utilization review, or care management (in its many forms). Take, for example, the universally reviled PBMs (pharmacy benefit managers), who have carved out a big niche providing services between payors, pharmacies, and drug companies while raising increasing questions about their actual value.

Physician practices have long outsourced billing services. Hospitals and doctors didn’t develop their own electronic medical records; they contracted with companies like Epic or Cerner. Health care entities had trouble sharing data, so along came H.I.E.s – health information exchanges – to help move some of that data (and HIEs are now transitioning to QHINs – Qualified Health Information Networks, due to TEFCA).

And now we’re seeing a veritable Cambrian explosion of digital health companies, each thinking it can take some part of the health care system, put it online, and perhaps make some part of the healthcare experience a little less bad. Or, viewed from another perspective, add even more complexity to the Rube Goldberg machine. 

On a recent THCB Gang podcast, we discussed HIEs. I agreed that HIEs had been developed for a good reason, and had done good work, but in this supposed era of interoperability they should be trying to put themselves out of business. 

HIEs identified a pain point and found a way to make it a little less painful. Not to fix it, just to make it less bad. The healthcare system is replete with intermediaries that have workarounds which allow our healthcare system to lumber along. But once in place, they stay in place. Healthcare doesn’t do sunsetting well.

Unlike a true Rube Goldberg machine, though, there is no real design for our healthcare system. It’s more like evolution, where there are no style points, no efficiency goals, just credit for survival. Sure, sometimes you get a cat through evolution, but other times you get a naked mole rat or a hagfish. Healthcare has a lot more hagfish than cats.

I’m impressed with the creativity of many of these workarounds, but I’m awfully tired of needing them. I’m awfully tired of accepting that complexity is inherent in our healthcare system.

Complexity is bad for patients, bad for the people directly giving the care, and only good for all the other people/entities who make a living in healthcare because if it. Instead of making pain points less painful, we should be getting rid of them.

If we had a magic wand, we could remake our healthcare system into something much simpler, much more effective, and much less expensive. Unfortunately, we not only don’t have such a magic wand, we don’t even agree on what that system should look like. We’ve gotten so used to the complex that we can no longer see the simple.

I don’t have a Utopian vision of a healthcare system that would solve all the problems of our system, but I do have some suggestions for all the innovators in healthcare:

• If your solution makes patients fill out one more form, log into one more portal, make one more phone call, please reconsider.
• If your solution takes time with patients away from clinicians, making them do other tasks instead, please reconsider.
• If your solution doesn’t create information that is going to be shared to help patients or clinicians, please reconsider.
• If your solution only focuses on a point-in-time, rather than helping an ongoing process, please reconsider.
• If your solution is designed to increase revenue rather than to improve health, please reconsider.
• If your solution doesn’t recognize, acknowledge, report and act on failures/mistakes/errors, please reconsider.
• If your solution can’t simply be explained to a layman, please reconsider.
• If your solution adds to the healthcare system without reducing/eliminating the need for something even bigger in the system, please reconsider.
• If your solution steers care to certain clinicians, in certain places, rather than seeking the best care for the patient in the best place, please reconsider.
• If your solution adds costs to the healthcare system without uniquely and specifically reducing even more costs, please reconsider.
• If your solution doesn’t have built-in mechanisms (e.g., use of A.I.) to be and stay current on an ongoing basis, please reconsider.

 I’m sure all those innovators think their idea is very clever, and many are, but remember: just because an idea is clever doesn’t mean it’s not Rube Goldbergian.  They need to step back and think about if they’re adding to healthcare’s Rube Goldberg machine or helping simplify it. My bet is that usually they’re adding to it.

So, yeah, I agree with Mr. Levitt and Dr. Altman that health insurance should be less complex.   Just like everything else in the healthcare system. Let’s start taking the healthcare Rube Goldberg machine apart.

Kim is a former emarketing exec at a major Blues plan, editor of the late & lamented Tincture.io, and now regular THCB contributor.

Now that the healthcare industry can work with clarity on care coordination strategies and programs, a new expansion of ACO models, trends in patient behavior and the companion issue of provider scope of practice have quickly emerged as critically-relevant spotlights. Historical perspective helps.

Simply put, even with the political tumult this fall, there is strong bipartisan support for aligning payment and care delivery models with improving quality to create a smarter and sustainable healthcare system, backed by historical precedent.

For me and my colleagues in the trenches of pursuing fiscally sound care delivery nearly a decade ago, it is well remembered that the origins of accountable care reside within a 2004 HHS document entitled “The Decade of Health Information Technology: Delivering Consumer-centric and Information-rich Health Care.” This “Framework for Strategic Action” (as it is also known) was delivered to then-HHS Secretary and GOP-appointee Tommy Thompson. And it was delivered by the nation’s first National Coordinator for Health Information Technology, Dr. David Brailer.

The document’s goals of introducing health IT solutions to clinical practices, electronically connecting clinicians, using “information tools” to personalize care and advance population health reporting followed an executive order calling for widespread adoption of interoperable EHRs within 10 years.

That core bipartisan support for these goals, also evidenced by the success of meaningful use, has weathered the political winds, and no doubt like many in health IT, I keep a copy of this foundational document at hand.

To continue to get us to where we are today, the report was followed the next year by the Physician Group Practice (PGP) demonstration, a five-year program of 10 sites pursuing early shared savings goals. This program was widely resurrected as a reference point when the current Medicare Shared Savings proposals were first issued.

A year later, Dartmouth Medical School’s Dr. Elliott Fisher began voicing the concept and vocabulary of accountable care during a Nov. 9, 2006, Medicare Payment Advisory Commission (MedPAC) meeting then put to paper by year’s end. MedPAC’s research over the past year only further supports this evolution.

Today, more than 27 state legislatures have proposed programs related to accountable and coordinated care, and there are more than 250 accountable care communities active in the vast majority of states. More than 70 of these are led by physicians, nearly double the number only eight months prior. And while closely associated with the CMS Medicare Shared Savings program (rightfully so now that an additional 10,000 Americans are becoming Medicare-eligible every day), health plan, private payer and even employer models are keeping pace.

Medicaid Models, Patients and Doctors Orders

Now following suit, state Medicaid officials are moving quickly to also understand and establish accountable care models around community ACOs, provider-led programs or hybrid models merging health plans and care providers.

This public-private initiative is being aided by organizations such as the non-profit Center for Health Care Strategies (CHCS) and the CMS Innovation Center, as all stakeholders realize the need for coordinated care for a patient population most in need of preventive and cost-efficient medicine that can build upon the Medicaid coverage expansion within ACA. Right now state Medicaid ACO pilot programs are being formed in at least seven states.

Meanwhile, many of the nation’s uninsured and elderly are increasingly taking advantage of the growth and accessibility of retail health clinics.

The number of Americans visiting these clinics for vaccinations, treatments for respiratory infections and preventive measures, for example, quadrupled – from nearly 1.5 to six million people – between 2007 and 2009, according to an August 15th Rand Corp. study published in Health Affairs. It is notable the study found that nearly 33 percent of these patients lack health insurance.

These rates will be impacted by the coverage mandate and the future of health insurance exchanges also within the Affordable Care Act, likely combining to fuel an increase in patient volumes at traditional practices as well, adding stress to our already strained delivery system in terms of the documented decline of the number of primary care physicians.

That dynamic will also continue to fuel expanding scope of practice debates on the roles of nurse practitioners (NPs) and physician assistants (PAs) moving within primary care. These issues are tied together provided that quality care can be achieved in retail settings, which I believe has been initially demonstrated and can continue to accelerate into more advanced primary care as an ambulatory option for more patients.

Steps to Accountable Care Success

Accountable care and care coordination in all of its forms is an essential building block for improved healthcare, along with EHR adoption, meaningful use and interoperability. In broad terms, this transformative journey seeks to improve patient safety and quality of care. The vehicle for that journey: further integration of care and a focus on disease management through new bundled payment models, value-based purchasing initiatives and benchmarking analysis. That’s where health information enters the picture. The robust use of data aggregation, analytics, and shared information directly support patient care coordination and population health management, which are the most critical clinical components of managing risk-based reimbursements.

For care providers and practices seeking to form or join an accountable care community, there are prerequisites to address:

1. Begin by assessing your EHR, interoperability and overall technology infrastructure, as well as your beneficiary patient volume. Then engage your peers, associations, payers, employers, and health systems in your community to identify government, private payer, or combined opportunities.
2. If your practice or organization is approached to participate in an ACO, evaluate it carefully. Consider your financial and strategic incentives for joining, data requirements, and access to bi-directional data and whether your commitment is binding or non-binding.
3. ACOs positioned for success should have three- to five-year plans that incorporate growth strategies and best practices. These include utilizing health information technology, engaging and educating patients, developing care management resources, and monitoring care delivery and follow-up.
4. It is also important to assess your own understanding of the different risk models being offered. Determine how much risk you can assume initially and over time.

The Supreme Court ruling on the ACA was a big step in this journey, and the next focal point is of course whether Shared Savings structures and the financial risk tracks succeed, causing more providers, health systems, private payers, and employers to embrace coordinated care and payment models.

We are seeing solid evidence of this already, which represents an encouraging sign of what the next several years will bring.

Justin Barnes is co-chair of the national Accountable Care Community of Practice (ACCoP), chairman emeritus of the EHR Association and a vice president at Greenway Medical Technologies. He has appeared before White House and Congressional panels on matters of health information technology on more than a dozen occasions since 2005, and has advised current and former presidential administrations on industry policy.

In a time of EHR naysayers, mean-spirited election year politics, and press misinterpretation (ONC and CMS do not intend to relax patient engagement provisions), it’s important that we all send a unified message about our progress on the national priorities we’ve developed by consensus.

1. Query-based exchange – every country in the world that I’ve advised (Japan, China, New Zealand, Scotland/UK, Norway, Sweden, Canada, and Singapore) has started with push-based exchange,replacing paper and fax machines with standards-based technology and policy. Once “push” is done and builds confidence with stakeholders, “pull” or query-response exchange is the obvious next step. Although there are gaps to be filled, we can and should make progress on this next phase of exchange. The naysayers need to realize that there is a process for advancing interoperability and we’ll all working as fast as we can. Query-based exchange will be built on top of the foundation created by Meaningful Use Stage 1 and 2.

2. Billing – although several reports have linked EHRs to billing fraud/abuse and the recent OIG survey seeks to explore the connection between EHR implementation and increased reimbursement, the real issue is that EHRs, when implemented properly, can enhance clinical documentation. The work of the next two years as we prepare for ICD-10 is to embrace emerging natural language processing technologies and structured data entry to create highly reproducible/auditable clinical documentation that supports the billing process. Meaningful Use Stage 1 and 2 have added content and vocabulary standards that will ensure future documentation is much more codified.

3. Safety – some have argued that electronic health records introduce new errors and safety concerns. Although it is true that bad software implemented badly can cause harm, the vast majority of certified EHR technology enhances workflow and reduces error. Meaningful Use Stage 1 and 2 enhance medication accuracy and create a foundation for improved decision support. The HealtheDecisions initiative will bring us guidelines/protocols that add substantial safety to today’s EHRs.

4. Privacy and Security – some have argued that EHRs reduce security by making records available in electronic form, possibly over internet connections. Efforts to enhance certification of the security of EHRs, encrypt data at rest, and create guidance for EHR modules that interoperate with built in security will further protect the data that needs to be shared for care coordination and population health.

5. Innovation – some have argued that meaningful use led to the growth of a small number of vendors and dependency/lock in with those vendors. Meaningful Use Stage 2 requires interoperability between vendors, export of data from EHRs to reduce lock in, and standards that will enable a new generation of modular “plug ins”. I’m confident that SHARP grant funded work, like the SMART initiative will lead to an ecosystem of applications from small vendors – an app store for health.

Thus, our mantra should be that Meaningful Use Stage 1 and 2 have created a foundation for query-based exchange, accurate billing, safety, security, and innovation.

Stage 3 work is already in progress and from the early thinking that I’ve seen (will post a blog about that in a few weeks), the trajectory of Meaningful Use will address all the naysayers concerns.

John D. Halamka, MD, MS, is Chief Information Officer of Beth Israel Deaconess Medical Center, Chief Information Officer at Harvard Medical School, Chairman of the New England Healthcare Exchange Network (NEHEN), Co-Chair of the HIT Standards Committee, a full Professor at Harvard Medical School, and a practicing Emergency Physician. He’s also the author of the popular Life as a Healthcare CIO blog.

Oh, that clever Center for Public Integrity.  Look what they’ve gone and done now!  My, oh my.  According to the article, doctors are much of the the problem, billing “billions” of Medicare upcharges according to the center.

But what if the medical coding game itself is flawed?  Stop for a moment and imagine what it would look like if lawyers billed like doctors.  Suddenly, we see how bizarre the world of government billing codes and chart-completion mandates has become.

Not long ago I asked readers what my time is worth on a per-hour basis.  Collectively and independently, they settled on a number of about $500/hr (see the comments).  Now look for a moment at what Medicare pays, even at its highest level of billing for a physician’s time for evlauation and management of a medical problem: for 40 minutes of a physician’s time, it’s $140 (or $210/hr) before taxes.  Again, we see another disconnect as to how doctors are valued in our current system.

Doctors are working long hours to collect these fairly low fees from Medicare while jumping more hoops than ever to do so.  They have become pseudo-experts at the coding game, trying to get as much money for their extra efforts as legally possible.  But these fees paid by Medicare do not cover payments for time spent on phone calls, e-mails, and working insurance denials.   These services are still considered by our system as gratis. To partially counteract this coding problem, doctors realized (and the government insisted) that doctors use electronic medical records.

But when independent doctors set out to implement these records they quickly discovered that the expense and long-term maintenance costs of local office-based EMRs could not compete with more sophisticated systems already in use by their neighboring large health care systems.  Because of ever-increasing cost-of-living and overhead costs, not to mention the threats of large fee cuts, doctors have migrated to large health systems faster than ever.  With the fancier electronic record at those systems (streamlined for billing, collections, and marketing) fields required for higher billing codes (but not always material to the problem at hand) are completed in less time.  So are doctors really the problem?

It depends on who’s looking.  Since every medical test and order is tied to a doctor’s name, then of course it looks like doctors are the problem.  And yet it’s the government who has mandated the codes, the requirements for chart completion, and the electronic records to which our electronic signatures are attached.  But we should ignore these facts; in the eyes of the Center for Public Integrity, of course its the doctors’ and hospitals’ fault.

And what do you think the government’s response is to all of this?

Why, get get ten times the number of billing codes, of course!

So take a moment and imagine a world without codes might look like.  A world where doctors are paid for their level of expertise, time with patients, time with communication for those patients, and time their connected to the EMR to enter codes, document, e-mail, and care for patients.  No codes, just time-based billing at a level of commensurate with their skills.  If we can track billing codes, we can track doctors’ time.  Gosh, it’s sounding sane isn’t it?

If we really want out of this coding and billing conundrum, we should stop the coding schemes.  Pay doctors for what they are worth in today’s market.  Pay doctors for their time as well as their productivity.  Throw away the codes, the consultants, the code licensure fees, and the nonsense.  Compared to current administrators of these coding schemes, people might actually discover that doctors and hospitals are the path to salvation for excessive health care costs rather than the instigators of coding fraud.

Westby G. Fisher, MD, (aka Dr. Wes) is a board certified internist, cardiologist and cardiac electrophysiologist practicing at NorthShore University HealthSystem in Evanston, IL. He is also a Clinical Associate Professor of Medicine at the University of Chicago’s Pritzker School of Medicine. He blogs at Dr.Wes, where this post originally appeared.