Category Archives: economics

How to Eliminate $226 Billion of Overuse in Health Care

by Tom Liu

This past weekend, I had the opportunity to attend the 2013 Lown Institute conference in Boston. The Lown Institute is an organization founded by Dr. Bernard Lown in 1973 that promotes a humanistic, patient-centered practice of medicine. A major topic covered during this conference, as well as during last year’s inaugural conference, was that of overuse in health care.

Just How Much Overuse Is There in the U.S.?

Dr. Don Berwick, who gave the keynote address, estimated the cost of overuse in the U.S. to be $158-$226 billion in 2011. Interestingly, the methods of the four studies cited for the $158-226 billion figure were primarily based on macro-level economic approximations (e.g. comparison of DRG intensity between U.S. and Canada)—not on micro-level analyses of overuse in violation of widely-accepted standards.

Which makes sense, given that for many questions of what constitutes overuse, the science may simply not be clear. In 2012, Dr. Deborah Korenstein and colleagues published a review of the literature on overuse in the U.S. The study’s subtitle (“An Understudied Problem”) reveals the punch line. While they were able to document overuse rates for specific treatments that have clear standards (e.g. antibiotics for upper respiratory infection), they concluded that “the overuse literature includes relatively few procedures and diagnostic tests.” And they attributed that to the uncertainty of our science:

“The limited overuse literature is understandable given the challenges of developing standards to measure overuse. […] the process of defining appropriateness for many services remains incomplete owing to both gaps in the evidence and failure to translate evidence into appropriateness criteria.” –Korenstein et al., 2012

Which puts us as (aspiring) providers in a bit of a quandary.

Eliminating Overuse May Take More Than a Checklist

For providers (especially in a profession susceptible to paternalism), the most straightforward solution might seem to be to direct patients away from those wasteful, inappropriate treatments, shaving 7-8% off of our $2.7 trillion and growing health care expenditures. But for the majority of cases, there may not be enough evidence to clearly support a treat or don’t treat decision. And even when evidence-based recommendations exist, they are likely based on population-level analyses, which may conflict with the desires of the individual patient.

As Jessie Gruman argues, there is a coming conflict between clinicians pressured to adhere to a burgeoning number of quality measures and patients who are becoming increasingly engaged in their treatment decisions. Dr. Gruman was at the Lown conference, and she described her experience choosing a new doctor when her old one refused to give her a treatment that he deemed was “not worth it” (despite a 20% success rate). Dr. Gruman belongs to a growing chorus of advocates calling for increased patient engagement in their care decisions. Have the provider lay out the treatment options, with each option’s risk and chance of success, and let the patient decide.

At the same time, I kept hearing my former Swarthmore professor Dr. Barry Swartz whispering three words in my ear: “paradox of choice”. Presenting people with 24 varieties of jam was enough to confuse them into inaction. Present patients with too many treatment options under actual life-or-death situations, and you could create a lot of (unwarranted?) stress and anxiety.

It seems to me that while some patients may be ready to be empowered consumers choosing from among a menu of options their provider lays out, others may not be there (yet). As Dr. Ranjana Srivastava, another panelist, aptly described, “Even with a menu of options, patients expect their doctor to take charge of their treatment.”

Overcoming the Culture of Overuse

Therefore, I believe the role of providers in reducing overuse will be much more complex than simply adhering to evidence-based recommendations to root out overuse. It will require engaging with each individual patient, intuiting that patient’s preferences for autonomy vs. provider advice, and having a conversation about the value of each treatment option (see Teaching Value Project)—including being willing to argue that the most aggressive option may not always be the best (though it certainly might be for that patient).

These are all skills that are often overlooked in our current medical education system. But the alternative—sticking with an ingrained culture to overtreat—may soon become unsustainable.

P.S. If you are interested in the issue of overuse, I strongly encourage you to check out the Lown Institute’s Right Care Declaration and sign if you so choose.


Tom is a health care researcher with experience in public health and blindness prevention. Follow him on Twitter at @tliu14 or check out his blog.


What the Stock Market Crash Reveals About Medical Errors

by Tom Liu

Proponents of high-deductible health plans want to give patients more skin in the game, to solve our system’s problem of escalating costs. Should our system have more skin in the game to do right by our patients?

The Best Risk-Management Rule Ever?

It is not only economically efficient, but morally imperative, to have “skin in the game”. That’s what Nassim Taleb, author of Fooled by Randomness (2001) and The Black Swan (2007), argues in a recent paper and interview on EconTalk.

Dr. Taleb opens by recounting the “eye for an eye” philosophy of Hammurabi’s code—or, in his opinion, “the best risk-management rule ever.” Three thousand years later, Immanuel Kant posed it in a slightly less morbid way through his notion of a “categorical imperative”: “Act only according to that maxim whereby you can, at the same time, will that it should become a universal law.” Or put more simply, do unto others as you would have them do unto you.

Corporate managers, academics, predictors, warmongers, and politicians, Dr. Taleb argues, are exempt from this moral imperative. They take risks and stand to benefit from the upside of those risks, but are shielded from the downside.

The Moral Hazard of “Fat-Tailed” Phenomenon

In fact, this problem is particularly severe for phenomena Dr. Taleb defines as “fat tailed domains”. A fat-tailed phenomenon is one in which an extremely rare but high-impact event dominates the effect of all other events. Repeated instances of a fat-tailed phenomenon (such as stock market outcomes every year) might look like this (source):

fat tail

A problem exists in that the reputation of market forecasters is based on how often they correctly predict the direction of the market movement, and not by how accurately they predict the final value of the market. (More technically, they are judged by a “binary metric” for what is actually a very skewed distribution.) A forecaster who is frequently right wins widespread admiration, even as people who follow that forecaster’s predictions ultimately see their savings wiped out by that rare, “blow-up” event. The forecaster, meanwhile, is insulated from the full pain of the investment loss.

The more skewed the phenomenon, the easier it is to hide the true impact of a mistake behind a façade of “pretty good performance”.

“Forecasters with steady strings of successes become gods.” –Taleb and Sandis, 2013

Skin in the Game for Patient Safety

Medical errors are a prime example of a fat-tailed phenomenon. For 98.6-99.4% of hospitalizations in the U.S., the patient is discharged without a lethal adverse event. But for the family of the patient who falls into that 0.6-1.4% of hospitalizations, getting killed due to medical error is an extremely “high-impact event”. I would imagine that the physician and care team—if they were aware that their error had caused the patient’s death—would feel very terrible. I’m sure even the hospital administrator would feel pretty bad as he/she looks over their adverse event reports. But will their suffering come close to what the patient’s family feels from the loss?

I’ve previously written about why we haven’t eliminated medical errors. News flash: hospital errors don’t cause 44,000-98,000 deaths each year, as we previously thought. They cause 210,000-440,000 deaths per year. That makes hospital error the number three killer in the U.S., after heart disease and cancer.

Slow innovation is arguably one of the most effective ways to spur adoption of safer practices. But it is, by its very nature—well, slow. The nation’s third leading cause of death may warrant a bit more urgency. And that brings us back to the moral imperative of skin in the game. Today, individual hospitals and clinicians are rarely judged by the impact of their medical errors. When they are, they are evaluated based on the frequency of their medical errors—a binary metric (error vs. no error) for a very skewed phenomenon (the magnitude of suffering to the patient and family). Given the immense suffering caused by medical errors, it would seem that providers should share the burden in some way—perhaps not literally by Hammurabi’s standards, but, as Dr. Roberts suggests, “substitut[ing] the physical eye for the economic value of the eye”. And yet, the vast majority of public and private payers today are still paying hospitals (even rewarding them) for medical errors.

Dr. Ashish Jha recently wrote an article arguing that incentivizing hospitals for patient satisfaction more than patient safety has led them to invest in lavish amenities over patient safety improvements. To be fair, Medicare finalized a rule this August that will penalize hospitals in the lowest quartile for medical errors or hospital-acquired infections by withholding 1% of their overall payment. But that may not be a strong enough incentive to catch hospital executives’ attention, as Dr. Jha points out in this blog post. If we believe that market forecasters should invest in the same stocks they predict, and that warmongers should be subject to the draft themselves, then why shouldn’t health care providers have some skin in the game when it comes to patient safety?


Tom is a health care researcher with experience in public health and blindness prevention. Follow him on Twitter at @tliu14 or check out his blog.

Two quick plugs

by Adrianna McIntyre

It’s been a slow week for the blog, but a busy week for the bloggers. First, you should go check out Karan’s latest, over on’s Short White Coat blog.

First, do no harm: Medicine in the Information Age

Minutes from melting in the summer heat, I dumped my stuff at a table and homed in on the hospital café’s soda display for something – anything – cold. The gentleman at the next table glanced my way and said:

“Are you in the medical field?”

“Yes sir, I’m a medical student.”

He eyed my drink and asked, “Did you hear that diet soda can increase your risk of diabetes by 70%? Even just a few cans a week.”

And I have a piece over on Bloomberg View, coauthored with Austin Frakt. It takes a closer look at the issue of rate shock—which yes, is real, but may not be motivated by the reasons you think (hint: some young people get sick, too).

Obamacare Rate Shock Isn’t What You Think

So why the sleight of hand? Maybe because FreedomWorks’ premise is wrong. Recent evidence suggests that the law won’t actually raise premiums for young adults for the reason its opponents claim. That’s because income-based tax credits will disproportionately help young Americans afford coverage.

Go read!


Adrianna is a graduate student in public policy & public health at the University of Michigan. Follow her on Twitter at @onceuponA, or subscribe to the blog. 

Most young adults on the individual market today will qualify for subsidies in 2014

by Adrianna McIntyre

The Kaiser Family Foundation has a new report out that examines how people currently in the individual market will be affected by the reforms taking effect in 2014. Premiums will change for  variety of reasons. You should read the whole issue brief, or Jon Cohn’s commentary here. As Kaiser acknowledges, premiums will go up for some people and down for others. They go a step further, though, and look at how many people in the current market will benefit from the premium tax subsidies.

About half (48%) of people now buying their own insurance would be eligible for a tax credit that would offset their premium. This does not include over one million adults buying individual insurance today who will be eligible for Medicaid starting in 2014 (i.e., they have family income up to 138% of the poverty level and are living in states that have decided to expand Medicaid under the ACA).

Now, that’s all adults, regardless of age. I was curious, how does that stack up against just young adults? I looked at these numbers a few months ago, but my analysis back then was limited to childless adults. Josh Fangmeier was kind enough recently to rereun the numbers for everyone, with or without kids (more related to that here). A predictable trend holds for those currently on the individual market: younger individuals tend to have lower incomes, so they’re going to benefit more than average from the exchange subsidies offered to those below 400% of the poverty line.


Bottom line? Just over 70% of young adults that hold individual plans today could qualify for subsidies in 2014. Even if you chop off the kiddos under 26—who might not need an individual plan, since they can take advantage of extended dependent coverage—the proportion of subsidy-eligible 26- to 35-year olds is 65%, still well over the 48% attributed to the whole market.

There’s a caveat that opponents have been quick to point out in the past: in some states, young adults near 400% FPL may not receive subsidies. This isn’t a bug—the tax credits are issued to offset premiums that exceed an income-dependent threshold; in some states, premiums are actually coming in below expected rates, and below that threshold for young adults.

That’s not to say that some won’t still see an increase in their rates—many will, because insurers will be required to offer a minimum package of benefits. They’ll also be prohibited from price-discrimination against unhealthy beneficiaries. There is a valid debate that we can have about both of those provisions (it’s worth noting that two-thirds of Americans support guaranteed issue). But if you want to talk rate shock, make sure you also talk about the fact that young adults are more likely to receive subsidies than their older counterparts.

You can access a description of the methodology used to compile relevant Census data here, and a spreadsheet of the data itself here


Adrianna is a graduate student in public policy & public health at the University of Michigan. Follow her on Twitter at @onceuponA, or subscribe to the blog. 

Prominent health policy scholars weighed in: “We need the individual mandate.”

by Adrianna McIntyre

On Wednesday, the House passed bills to delay the employer as individual mandates, the latter as a matter of “fairness” (seriously, the bill is called the “Fairness for American Families Act”). I didn’t exactly veil my thoughts about this maneuver last week—I think my exact words involved “willful ignorance” and “political theatre”, but the full post is here.

I’m not the only one with thoughts threading in that direction. My attention was recently called to a letter signed by a cohort of academic heavyweights; any health wonk worth her salt should recognize about a dozen names (or more). They deftly capture both the imprudence of putting off the individual mandate and how utterly disconnected it is from the employer mandate. An excerpt of the letter follows (emphasis added, footnotes removed, full letter here).

Without the [individual] mandate, some people will choose to gamble or to free-ride, undermining the fairness and financial stability of the health insurance system … insurance reform without subsidies and mandates has consistently failed.  In the five states that have tried comprehensive insurance market reform without an individual mandate, healthy people chose to stay out of insurance, sick people took it up, and premiums increased.  Only broad participation in insurance markets can end the cycle of insecure coverage and high costs.

The Obama Administration’s recent decision to delay ACA’s requirement that large- and medium-sized employers sponsor coverage for their employees or pay a penalty is independent of the individual mandate.  The employer assessment is designed to bolster the ACA’s financing and to ensure equity between large firms who do and do not provide insurance.  This assessment will have only a very small impact on employers, since 97% of firms with more than 50 employees already offer insurance. The individual mandate stands in stark contrast, as nearly one in five non-elderly Americans is currently uninsured.

Delaying the employer assessment has almost no effect on the implementation of the ACA.  The only important effect will be to raise one fewer year of revenue from this component of the law.  In contrast, delaying the individual mandate would cut at the core of the vision of private-market based insurance market reform.

Requests to delay the individual mandate are really requests to gut the Affordable Care Act.  Millions of Americans face immediate health care needs and financial challenges addressed by health reform. They cannot wait.

 The Senate is not expected to pass either bill—and even if it did, the President has stated that he would veto both. Vetoing legislation that codifies the employer mandate delay seems counterintuitive, but Philip Klein suggests that signing such a bill “would be seen as a tacit acknowledgement that he acted outside of his authority by issuing a regulatory delay of the mandate.” I think he has a point, one that holds whether or not the Administration actually overstepped legal limits (here are the best critique and defense I’ve seen).

employermandateBut, as the letter’s authors note, delaying the employer mandate is expected to have an extremely marginal effect on implementation and costs of the Affordable Care Act. Meanwhile, the Urban Institute anticipates that delaying the individual mandate would halve the number of people gaining coverage, with insurance attracting less healthy/more expensive enrollees. This isn’t idle speculation—Wednesday also brought news that nongroup insurance rates in New York are falling because the state has maintained stringent consumer protections without a mandate, resulting in some of the country’s highest premiums.

You don’t have to listen to me. The people who signed the letter include some of the most noted scholars in health economics and policy, and they haven’t minced words: we know better than to make this mistake on the national level.


Adrianna is a graduate student in public policy & public health at the University of Michigan. Follow her on Twitter at @onceuponA, or subscribe to the blog.