The Lexus and the Human

Filed Under (Health Care) by Jeffrey Brown on Nov 1, 2011

Yesterday, I had the honor of hosting University of Illinois Jeff Margolis as a guest speaker in my Employee Benefits course here in the Illinois College of Business.  Jeff Margolis may not be a household name, but he ought to be.  Jeff just recently retired from the TriZetto Group, the company he founded in 1997 to provide the knowledge and the tools to help power “Integrated Healthcare Management.”  His company’s software touches the lives of over half the U.S. population.  He is also author of the book “The Healthcare Cure: How Sharing Information Can Make the System Work Better,” which is due out later this month (which is an extended and updated version of an earlier book, The Information Cure, from which I am drawing my post for today.)  Jeff understands the U.S. health care system inside and out.

Jeff’s talk to my classes yesterday had several really important take-aways.  Of these, perhaps the most important is the need to carefully define the problem that one is trying to solve.  I often point out to my class that there are many ways to define the problem with health care in the U.S., and each has a different set of solutions.  For example, if one simply wants to reduce health care spending, there are lots of easy ways to do that – restricting access, limiting innovation, reducing insurance coverage – but we probably would not like many of the outcomes!  Similarly, if one simply wants to increase insurance coverage, we can do that too.  But then don’t be surprised when health care spending rises as a result.

Jeff Margolis focused on a more clearly-defined problem – one with which economist like me tend to agree is right way to think about health care policy.  In my words, he was fundamentally focused on how do we get the right care to the right people at the right time?  At its essence, this is a question of resource allocation, and the only way we are going to get it right is if all the relevant actors (doctors, hospitals, patients, payers, etc.) have access to the information they need.  Information about patients.  Evidence-based information about which treatments work for which types of patients.  Information about costs.  And so on …

To give but one illustration, in Chapter 3 of Jeff’s “Information Cure” book, he has a great chapter (the title of which I have borrowed for this post), called “The Lexus and the Human.”  He starts with the provocative question – “which would you rather be: a Lexus or a human being?”  He then goes on to point out that when a Lexus is “ill” (i.e., something is wrong, and it needs to be diagnosed and fixed), there is a very rich, thorough, and transferable set of maintenance records that are easily communicated to the right mechanic in the right garage at the right time.  In contrast, when a human being shows up at a Doctor’s office, the information is often incomplete, fragmented, and out of date.

As Jeff states, “in stark contrast to Lexus’ systematic way of maintaining and repairing its cars, the U.S. healthcare system lacks the coordination to care for humans as reliably and comprehensively … the Lexus enjoys a much higher degree of precision regarding its care.  For starters, our system does not reliably enable providers and consumers to access medical records wherever and whenever we need them.”  He talks about the inefficiencies this creates in the form of duplicate tests.  But he also notes the much more serious consequences, such as being improperly treated because an emergency room doctor was unaware of your drug allergy.

He also points out that the information exists.  But the overall system needs re-engineered to optimize the use of this information.  He points out that health plan providers – such as Cigna, Aetna, United Health Care, and others – may be in the best position to help us get there because of the role they play in the health care supply chain.  Unfortunately, our politicians are so busy villain-izing these health care companies that we may be overlooking an enormous opportunity for increasing the efficiency and efficacy of our health care system.

It is worthwhile food for thought.  There is much, much more to say on this topic – and hopefully I will return to it in future posts.  But for now, thanks to Jeff Margolis for lecturing at his alma mater, and for helping to educate the next generation about ways to productively identify and solve problems.

Reframing Obamacare

Filed Under (Health Care) by Tatyana Deryugina on Oct 20, 2011

Most  economists would say there is conceptually no difference between a coffee shop (a) charging 5 cents more for each drink and giving a 5 cent discount to people who bring their own mugs and (b) charging 5 cents more to people who don’t bring their own mug. The price you pay for a drink is the same in each situation (both when you have your mug and when you don’t).

However, to some people the distinction is a big deal. Most grocery stores give customers discounts for bringing their own bags, but protest against proposed requirements to charge for each plastic bag. Apparently, judges, lawyers and much of the US population also don’t agree that the two situations are equivalent. The mandate to buy insurance has led over half the US States to challenge its constitutionality. But is that requirement really that different from some other incentives that US citizens have faced for decades?

The way the Obama administration chose to write the law is as follows. Everyone is “required” to buy insurance. But clearly, the government cannot truly force people to buy it or impose severe punishments. Thus, the “price” for not buying insurance is a tax penalty. By now, it should be clear that the Obama administration could have gone a different route. Instead of imposing a tax penalty and calling it a “mandate”, it could have simply raised taxes and introduced a tax deduction for people who buy health insurance. Same exact incentives, much less legal trouble. We don’t call the mortgage interest deduction a “mandate to buy a house” and we don’t call the deduction for dependents a “mandate to have a child”. No one has sued over them (at least that I know of). But changing something from a reward to a penalty is apparently enough to land you in the Supreme Court.

Maybe the Obama administration knew that people would react to the other framing differently? Although I don’t know of any studies that show this, people appear to dislike explicitly paying more for not doing something, like not bringing their own mug or grocery bag and not buying health care (even though you are implicitly paying more when you forget that mug). This would imply that people are much more sensitive to tax “penalties” than to tax “deductions”. I’m guessing that calling something “mandatory” and penalizing non-compliers has larger effects on take-up than simply providing a tax deduction for doing something.

Of course, it’s also possible that a proposal which raised taxes and introduced tax deductions for buying health insurance would have never passed Congress (or that the administration never imagined so many states would challenge the law).

No Mourning for Death of the CLASS Act

Filed Under (Health Care, U.S. Fiscal Policy) by Jeffrey Brown on Oct 16, 2011

Late Friday afternoon, the Obama Administration announced that it was killing the CLASS Act, a controversial provision of Obamacare (the Patient Protection and Affordability Act) that was going to create a new, government-run long-term care insurance program.  The fact that they released this news late on a Friday should not come as a surprise, as this is a time honored trick for trying to bury bad political news (although as noted by Stefano DellaVigna and Joshua Pollet, this strategy does not have long-lasting benefits, at least for publicly traded companies!)  Undoubtedly, both sides of the political divide will try to spin this issue for political benefit, as it is the first part of Obamacare that has been ended.

Politics aside, however, the death of this program is a rare victory for good economics.  And we should give credit to the Obama Administration for killing it before it grew into yet another unworkable and fiscally irresponsible government program.

Early in the life of this blog (November 2009), in a post entitled “A Solution in Search of Problem,” I wrote about why the proposed CLASS Act (which went on to become law several months after my post) was fatally flawed.  In that post, I made the point that “the government has developed a solution to a supply problem that does not exist, but has failed to address the demand problems that do exist.”

In a recently issued NBER working paper, MIT economist Amy Finkelstein and I wrote in much more detail about the flaws of the program.  Here are a few of the points we make:

“Medicaid will likely impose a large implicit tax on CLASS benefits, just as it does on private insurance policies … Overall, the Congressional Budget Office (2009) has estimated that only 4 percent of the adult population would enroll in the CLASS Act program by 2019.”

“The pricing and financing of the program are controversial. Monthly premiums are to be set by the Secretary of Health and Human Services with a goal of maintaining program solvency over 75 years.  In practice, this means that premiums will likely be below “actuarially fair” levels. To understand why, imagine that individuals start paying premiums immediately, but the average payout will not occur for, say, 25 years in the future. To oversimplify, a 75-year “solvency” calculation counts 75-years’ worth of premium payments, but only 50 years’ worth of benefit payments. Thus a program could be technically solvent even though it is being run on a negative NPV basis. The financial problem is magnified by the fact that individuals below the poverty line and full-time students would be able to participate at only $5 per month (in 2009 dollars). A number of experts have also voiced concerns about adverse selection into this voluntary program, and have suggested that the program will face a troubled fiscal future.”

This program was a clear example of a program designed by people with good intentions, but who had such a frail grasp on basic economic concepts that they designed a program destined to fail. We should all just be glad that it was killed before it developed its own constituency and grew to become “too big to fail.”

Are U.S. physicians overpaid?

Filed Under (Health Care) by Nolan Miller on Sep 28, 2011

Let me begin by stating that the answer to this question is “no.”

Now that I’ve headed off the slew of angry calls from my family, the reason why I’m writing about this question is a recent article in the journal Health Affairs by Miriam Laugesen and Sherry Glied entitled “Higher Fees Paid to US Physicians Drive Higher Spending For Physician Services Compared to Other Countries.”  The study compared fees paid to general practitioners and orthopedists in the US with those paid in Austrailia, Canada, France, Germany and the UK.  They summarize their findings as:

“Public and private payers paid somewhat higher fees to US primary care physicians for office visits (27 percent more for public, 70 percent more for private) and much higher fees to orthopedic physicians for hip replacements (70 percent more for public, 120 percent more for private) than public and private payers paid these physicians’ counterparts in other countries. US primary care and orthopedic physicians also earned higher incomes ($186,582 and $442,450, respectively) than their foreign counterparts. We conclude that the higher fees, rather than factors such as higher practice costs, volume of services, or tuition expenses, were the main drivers of higher US spending, particularly in orthopedics.”

In light of this finding, one might be tempted to conclude that physicians in the US are overpaid relative to other countries.  However, while it is true that physicians in the US make more than physicians in other countries, in order to interpret this finding it is critical to note that high earners in the US in general make more than high earners in other countries.  Relative to other countries, the US income distribution is more skewed, meaning, for example, that the highest 1% of earners in the US make more than the highest 1% of earners in other countries.

Now, we might ask ourselves, what is the relevant comparison group for a US physician?  A US college student who is deciding whether to be a physician doesn’t compare the income that could be made as a physician in the US with the income that could be made as a physician in German.  He or she compares the income to be made as a US physician with the income to be made as a US lawyer or MBA.  In other words, the right question isn’t whether US physicians are paid too much relative to German physicians, but whether US physicians are paid too much relative to others at the top end of the US skill/eduction/income distribution.  With respect to this point, the answer seems to be that incomes for US physicians are largely in line with incomes to other high earners in the US.  (Not to be too hard on Laugesen and Glied, they discuss this point at the end of their article.)

The issue of how physician incomes fit into the broader income distribution is discussed in a paper from earlier this year by David Cutler and Dan Ly that appeared in the Journal of Economic Perspectives.  They compare incomes for general practitioners and specialists to the income of “high earners,” (people in the 95th to 99th percentile of the income distribution) in 13 other OECD countries.  In the US, this ratio is about 1.37 for specialists.  In the other OECD countries, the ratio ranges from 2.56 in the Netherlands to 0.8 in the UK, with the average being 1.45.  For GPs, the ratio of income for US physicians to that of other US high earners is 0.92.  In the other countries, this GP ratio ranges from 0.68 in Norway to 1.41 in Canada, with the average being 0.98.  Thus, in both cases, US physician earnings seem to be in line with earnings of high earners, with the US being slightly below the OECD average for both GPs and specialists.

So, what do we make of the two studies?  Well, Laugesen and Glied have a point that high physician incomes appear to drive the high cost of healthcare in the US relative to other countries.  But, the reason why US physicians earn so much isn’t because “the system is broken,” they “take advantage of the system” or some other nefarious motive.  Rather, US physicians earn a lot because high earners in the US earn a lot.  So, the skewness of the US income distribution is in part responsible for the high cost of health care in the US.

This point is potentially important for understanding how we might reduce healthcare costs in the US.  Often, proposals to reduce Medicare spending focus on reducing provider payments.  However, if this reduces physician incomes we might expect that in the long run, as physician incomes drop relative to other professions, we’ll have fewer physicians and more lawyers, MBAs, etc.  As long as wages in these competing professions remain high, it will be difficult to squeeze down on physicians too much without driving them out of medicine.  If the highest-ability students are the most likely to move to a different profession, we might find that those who still choose to be doctors are not as good: the overall quality of the talent pool of young physicians might drop.  At the same time, to the extent that this reduced physician supply leads to shortages, it will put upward pressure on physician fees, and we’ll be right back where we started from.  In short, it is unclear that we can reduce healthcare costs too much by reducing payments to physicians.

In Crashes, a Grandparent Safety Factor? Not so fast!

Filed Under (Health Care) by Nolan Miller on Sep 14, 2011

The New York Times ran a story the other day on a recent study in the journal Pediatrics that “suggests that children may be safer when their grandparents are driving than when their parents are at the wheel.”  Both the article and the study are flawed.  The study is bad for statistical reasons I’ll get into in a minute, and perhaps it is better to say that the popular reading of the article makes too much of what the results actually show.  The New York Times story is flawed because there is no evidence that any thought was given to the possible merits of the study, merely summarizing without any attempt to do actual journalism.  This is all-too-often true of the coverage of health-related studies in the popular press.  As most people are unlikely to read the original study or critique its methods, reporting results without providing critical analysis runs the risk of giving people wrong, and potentially dangerous, advice.

So, what results are reported in the study?  The study finds that grandparents were the drivers in 9.5 percent of vehicle accidents involving children, but only 6.5 percent of the injuries.  Thus, children in crashes in which their grandparents were the drivers were less likely to be injured than children in crashes in which their parents were.  The result remains true even once you control for known risk factors for injuries such as use of proper restraints, the size of the vehicles involved and measures of severity of the crash.  In fact, it gets a bit stronger, so that children in grandparent-driven crashes are 50% as likely to be injured as children in parent-driven crashes.  The Times quotes one of the study’s authors as saying “But there’s something about the crashes of grandparents that we were unable to measure that was protective. It would be great to figure out what this is, because it could be protecting a lot of kids.”

OK.  So what’s the problem?  Well, the study fails to account for the fact that far more miles are driven by parents than by grandparents.  So, while it may be true that a child in an accident where a grandparent is the driver are less likely to be injured than a child in an accident where the parent is the driver, grandparents seem to have so many more accidents per mile driven that a child is far more likely to be injured in a mile driven by a grandparent than a parent.

Let’s do some back-of-the-envelope calculations.  Some facts take from the paper, which are not ideal, but the best I could find on short notice.  “Up to 33% of nonresidential drivers will make trips with children a few days per month, 14% a few days per week, and 5% almost every day. Of such nonresidential adult drivers, 42% were the child’s grandparents.”  A quick look at the original source suggests that 4% of grandparents report driving children “almost every day” and 14% report driving kids “a few days a week.” Now, I think it is reasonable to assume that a child drives with a parent every day (especially since the data in the study are limited to kids who are driven in cars and have accidents).  It is a bit of a leap, but if 4% of grandparents drive kids almost every day and we round up to account for more occasional grandparent drivers, then maybe the number of trips kids take where their parents drive is something like 20 times as large as the number of trips which grandparents drive (since 100% of parents drive their kids every day).  If the trips are of equal length, then that suggests 20 parent-driven miles for ever grandparent-driven mile.  My guess is that parents tend to make longer trips, so if anything it is more parent miles per grandparent mile, but lets go with 20:1.

Now, the study found 1143 accidents involving grandparent drivers and 10716 accidents involving parents.  But, to account for the fact that parents drive more often, you should weight these numbers, basically by multiplying the 1143 by 20 to get a number that corresponds to what would happen if grandparents drove as many miles as parents.  If we do that, we get 22860, meaning that a child is twice as likely to have an accident in a mile driven by a grandparent than in a mile driven by a parent.  Going on, the study found 113 injuries when grandparents were the drivers and 1189 injuries when parents drove.  Again multiplying the grandparent number by 20, you get 2260, meaning that children are also twice as likely to be injured in a mile driven by a grandparent than by a parent.

Given the data, in order to find that the risk of accident and injury to a child in a mile driven by a parent and grandparent were approximately the same, you’d need the child to be driven by a parent for no more than 10 times as many miles as by a grandparent.  This seems unlikely to me.

So, my reading of the data is almost exactly opposite those of the paper.  Now, to be fair, the paper usually states its conclusion as ‘if you are going to have an accident, it is better to have one with a grandparent behind the wheel.’  This is accurate, although I believe the authors are overstepping the data when they go on to suggest there is something protective about grandparent driving.  (In fact, the protective thing is that it doesn’t occur all that often!)  The more relevant question is ‘should I send my child to school with a parent driver or a grandparent driver?’  The study does not directly investigate this question (and the New York Times does not investigate anything in its write-up).  My suspicion and back of the envelope calculations suggest that the answer to this question is clearly in favor of the parents.

Note: the views expressed in this blog should not be interpreted as an indictment of any particular grandparents, especially my son’s grandparents, who are welcome to drive him around whenever they like.

More options are bad?

Filed Under (Health Care) by Nolan Miller on Aug 31, 2011

One of the bedrock principles of standard economics is that choices are good.  The simple version of the argument goes like this.  If I offer you a choice between an apple and a banana, increasing the options so that you can now choose an apple, a banana, or an orange must make you better off.  Since in the second case you can choose anything you would have chosen in the first (i.e., apple or banana), but you can now also choose an orange if you prefer it, it must be that the adding the option to choose an orange makes you better off.  Or so the simple argument goes.

While this argument has a certain intuitive appeal, recent evidence of poor decision making when faced with multiple options suggest that choices, while good in theory, may lead to poor decisions and thus be bad in practice.  One such study (written by J. Michael McWilliams, Christopher Afendulis, Thomas McGuire and Bruce Landon) just appeared in the online edition of the health policy journal Health Affairs.  The authors of the study consider seniors’ decisions about whether to enroll in Medicare Advantage, the HMO-like system that runs alongside traditional Medicare.  For seniors in the U.S., the default decision is to participate in “traditional Medicare,” the plan where seniors go to whatever doctors they like and the government pays the bills.  But, they also have the option to choose to enroll in a Medicare Advantage plans.  In exchange for accepting some limits on which providers they can see (or financial consequences for choosing out-of-network providers), Medicare Advantage plans often provide additional benefits, such as prescription drug coverage, vision or dental benefits, or else reduce their patients’ out-of-pocket costs of care.  These plans differ in their networks and in the portfolio of benefits they offer and may or may not appeal to any particular senior.

The standard “choices are good” argument would seem to apply here.  The more Medicare Advantage plans are available to a particular senior, the greater the chance that she will prefer one to traditional Medicare.  So, the likelihood of enrolling in Medicare Advantage should increase as the number of participating plans increases.  The Health Affairs study finds that this is, in fact the case, but only up to a point.  Increasing the number of plans available to a person tended to increase enrollment in Medicare Advantage, but only if there were fewer than around fifteen plans available.  When there were between fifteen and thirty plans available, additional options did not affect enrollment, and when there were more than thirty plans available, increasing the number of options available actually decreased enrollment in Medicare Advantage.  This tendency for additional options to reduce enrollment was particularly pronounced for participants with decreased cognitive abilities, consistent with the contention that when the choices facing a person become too complex, they tend to make worse decisions.

Lest you think that “choices are bad” only applies to the elderly, another such example comes from a 1995 study by Donald Redelmeier and Eldar Shafir that appeared in the Journal of the American Medical Association.  

In the study, a group of family physicians were presented with two different scenarios:

Scenario 1:  The patient is a 67-year-old farmer with chronic right hip pain. The diagnosis is osteoarthritis. You have tried several nonsteroidal anti-inflammatory agents (e.g., aspirin, naproxen, and ketoprofen) and have stopped them because of either adverse effects or lack of efficacy. You decide to refer him to an orthopedic consultant for consideration for hip replacement surgery. The patient agrees to this plan. Before sending him away, however, you check the drug formulary and find that there is one nonsteroidal medication that this patient has not tried (ibuprofen). What do you do?

The second scenario was the same as scenario 1, except the underlined phrase was replaced with “two nonsteroidal medications that this patient has not tried (ibuprofen and prixicam).”

Now, if more options are good, what we should have seen was that the doctors should be more likely to pull back the referral and try another drug in the second case than in the first.  After all, if you would pull the referral and try ibuprofen, then you should certainly pull the referral if you can try ibuprofen or prixicam.  However, the doctors behaved quite differently.  While 53% of physicians choose to refer the patient under scenario 1, 72% choose to refer the patient under scenario 2.  (The study reports similar results for other decision making scenarios.)  It seems that the doctors, faced with the additional task of thinking through whether to treat the patient with ibuprofen or prixicam after pulling the referral, simply decided to avoid the issue by continuing to refer the patient.  Just as in the case of Medicare Advantage plan choice discussed above, when faced with a difficult or complicated choice, decision makers responded by avoiding the effort of figuring out the right answer and simply going with the default option.

Studies such as these are disturbing on many levels, not the least of which because they shake our confidence in the basic principle that options are good.  If increasing a person’s available options is not good, then we need to think hard about which options we should offer them and which we should suppress.  Decisions like this, which involve making judgments about the kind of information others would find useful are value-laden and inherently difficult.  In light of this, the above studies suggest that the most likely response will be to stop offering people choices altogether.  Wait, that can’t be good either.

MacGyver goes to the hospital

Filed Under (Health Care) by Nolan Miller on Jun 30, 2011

For those of you who thought that duct tape cannot fix everything, here’s one more reason why you’re wrong.  A paper presented at the 38th Annual Educational Conference and International Meeting of the Association for Professionals in Infection Control and Epidemiology demonstrated a simple yet effective method for interacting with patients in isolation due to contagious infections.  The method saved a 504 bed health system 2700 labor hours and $110,000 over the course of a year.

To understand the beauty of the technique, you need a bit of background.  When a patient with a dangerous, contagious infection is in the hospital, special precautions are needed in order to interact with him (or her).  Typically, healthcare workers must put on gowns and gloves each time they enter the patient’s room, no matter how trivial the interaction.  This process takes time, and it expends resources, since the gloves and gowns cannot be reused.

So, how did they do it.   Basically, they took a role of red duct tape and made a box just outside of the door to the room and told the workers for simple interactions (e.g., asking if the patients needed anything), instead of suiting up, just yell to the patient from the “Red Box.”  Simple as that.  Not only did it save time and money, but two thirds of healthcare workers felt that it reduced communication barriers between themselves and their patients.

Now, usually medical research is pretty good about acknowledging potential conflicts of interest.  But, I couldn’t help noticing that the conference where this research was presented was sponsored by none other than 3M!  Coincidence, or something more?

I wonder how much money could be saved if the hospital also used bubble gum and a Swiss army knife?

I first read about this in Modern Healthcare, here (gated).  Read the press release here.

Psst … State of Illinois. Yeah, you. Here’s what to do about Health Alliance.

Filed Under (Health Care) by Nolan Miller on May 26, 2011

There has been much consternation regarding the state’s recent decision not to include Health Alliance among the HMO offerings available to state employees.  For those of you who haven’t been following it, Health Alliance is the HMO associated with the Carle Clinic system, one of the primary healthcare providers in the area.  According to the News-Gazette, the reason why the state chose Blue Cross Blue Shield over Health Alliance was that Health Alliance, whose bid was about 16 percent above BCBS, had bid too high, and was unwilling to lower their bid any further.

Because of Health Alliance being excluded from the state’s HMO offerings, many UI workers will face the choice of switching into the more-expensive Quality Care Health Plan, a plan with a very extensive network, or into one of the Open Access Plans which, based on what I can see, are pretty vaguely defined, or changing their provider.  This choice is complicated by the fact that other local healthcare providers have said that they don’t have the capacity to absorb Health Alliance’s customers.

So, what to do?  The answer is easy, and traces back to an idea health economists, especially Alain Enthoven, started talking about in the during the early 1990’s called “Managed Competition.”  The idea in a nutshell is that an employer like the State should make multiple managed care (i.e., HMO) offerings available to its employees through a competitive bidding process where the employer’s contribution to the employee’s insurance is a fixed dollar amount, often one pegged to the lowest bid. So, suppose the employer agrees to pay 80% of the lowest bid and there are two HMO options available.  HMO ABC bids $100 to cover an employee, while HMO XYZ bids $120.  In this case, the employer would pay $80 toward an employee’s insurance regardless of which plan he chose.  An employee choosing ABC would pay $20 per month, while an employee choosing XYZ would pay $40 per month.

Managed competition has several nice features.  First, by charging employees who choose the more expensive plan the difference in the plan cost, you give employees incentives to seek out and choose high value plans.  In deciding whether to choose ABC or XYZ, they ask themselves whether the extra $20 per month is worth it and “vote with their feet.”  Second, because employees are actively seeking high-value care, the HMOs now have a stronger incentive to lower their cost and provide higher quality.  In other words, by creating a type of “internal market” for health plans, the employer promotes competition to the employer and employees’ benefit.  Of course, insurers may not be too happy, but they can always choose to withdraw from the employer if they don’t feel they’re being adequately compensated.

So, what should the State of Illinois do?  Easy, offer Health Alliance to employees who want it, but charge them 16% more than they charge for the BCBS plan.  Employees who value Health Alliance’s provider network will pay the extra money, while those who do not will choose one of the other options.

In fact, the State already offers two HMO options, BCBS and HMO Illinois and charges $16 more per month for families (two adults and at least on child) choosing HMO Illinois.  So, why not just offer Health Alliance, too?  I can’t find the bid information here, but if it means charging Health Alliance members an extra $10 or $20 or $50 dollars per month, why should the state care?  Its cost is the same regardless of which plan the employee chooses?  Employees who value Health Alliance will pay the extra, those that do not won’t, and there you go.  Now, part of the problem with Health Alliance’s high bid may be that there is not enough competition in the Champaign-Urbana area, but adopting this policy will also encourage new entrants into the local market who could induce Health Alliance to lower its prices.

ADDENDUM (5/27/2010):  OK.  So the more I thought about it, the less happy I was with the post above.  While the managed competition model has a lot going for it, the system the state has set up, where there is little geographic competition among the HMO offerings, does not really give managed competition a chance to work.  A couple of additional points:

(1) Competition: if the state wants to bring down healthcare costs, it should be promoting competition.  So, it should have a goal of ensuring that there are multiple plan offerings competing in any particular geographic area.  But, this might be likely to be more successful in denser areas that can support multiple, competing healthcare systems.  Downstate areas, which are less densely populated, may need another approach.

(2) Affordability: the main problem with the argument I laid out above is one of “affordability.”  If health care is just more expensive downstate, then a managed competition approach applied uniformly across the state will mean that people living downstate pay more out of pocket for health insurance.  One solution would be to adjust the fixed payment for health plans based on local cost conditions.  However, we would want to keep in mind that healthcare costs are one part of the “cost of living.”  Downstate areas, if they have higher healthcare costs, offset that with lower housing costs, gas costs, etc.  So, it is unclear to me whether the right way to deal with this is through the employer’s system of paying for health insurance, or whether this is just another factor that contributes to differences in the cost of living that are better addressed through wage differences.

Have a policy you really want enacted (or not)? There’s a VSL for that!

Filed Under (Environmental Policy, Health Care, U.S. Fiscal Policy) by Nolan Miller on Mar 9, 2011

Last year, my colleague Don Fullerton blogged about the tradeoffs between dollars and lives saved inherent in doing public policy.  The short version: we don’t have unlimited resources.  Potential government policies differ in the cost of saving a life. We should look for the policies that save the most lives at the lowest cost.

Today, I want to discuss a slightly different take on this problem that was raised in a recent New York Times article.  The article approaches the problem from the other direction.  Namely, how much should we be willing to spend to save one life?  This question is critical for deciding whether or not a particular policy should be enacted.  For example, consider a new workplace regulation that would increase safety.  If the regulations cost $10 billion and save 100 lives, then it spends $10 million to save one life.  Government agencies take this number and compare it to a benchmark “value of a statistical life” (VSL) to decide whether the policy should be enacted.  So, if the VSL is $5 million, the policy that spends $10 million to save a life should not be implemented.  If the FSL is $15 million, it should.

This is an approach to policy making that makes a lot of sense and has been increasingly adopted in recent years.  So, what’s the problem?  Well, suppose you have a vested interest in passing a policy (or not), irrespective of the costs.  An easy way to do this while still ostensibly adopting the cost-benefit approach is to simply adjust the VSL.  If you have a policy that spends $7 million to save a life, then you need to adopt a VSL greater than $7 million in order to enact the policy.  If you don’t want the policy to be enacted, adopt a VSL less than $7 million.  In the end, we have ideology thinly masked by science, or at least math.

The Times article has a number of examples of such changes.  In particular, they discuss how the VSL used by various agencies has increased between the Bush and Obama administrations.  For example, the George W. Bush administration’s EPA put the VSL at $5 million, while the Obama EPA puts it at $9.1 million.  The result?  More EPA regulations on things like air pollution.  The Bush FDA had a VSL of $5 million, while the Obama FDA uses a value of $7.9 million.  Again, the result is more regulation.

Now, the consistently higher VSLs adopted by the current administration could be consistent with a number of stories.  In particular, they are consistent with causality running in either direction.  Either Democrats put more value on statistical lives saved and so they advocate bigger government, or they advocate bigger government and so they choose a bigger VSL in order to justify it.  Further, the same argument could be made about the Bush administration.  Perhaps the small government Republicans used a VSL that was too low to justify laissez faire policies.  (The article quotes Kip Viscusi, a leading academic thinker on the VSL as saying that he thought “agencies have been using numbers that I thought were just too low.”  Viscusi advocates a VSL around $8.7 million.)

Changes in the VSL can change policies,  as the article illustrates:

“It looks like they just cooked the books — they just doubled the numbers,” said Todd Spencer, executive vice president of the Owner-Operator Independent Drivers Association, a trade group for the trucking industry, which faces higher costs under some of the Transportation Department’s new rules. The Bush administration rejected a plan in 2005 to make car companies double the roof strength of new vehicles, which it estimated might prevent 135 deaths in rollover accidents each year.

 At the time, Transportation officials figured that the cost of the roofs would exceed the value of lives saved by almost $800 million. So the agency proposed a smaller increase in roof strength that might save 44 lives a year.

Last year, the Obama administration imposed the stricter and more expensive roof-strength standard, and it published a new set of calculations showing that the benefits outstripped the costs.

Most of the difference came from the increased value of human life. By raising that number to $6.1 million from a figure of $3.5 million in the original study, the Obama administration rendered those 135 lives — and hundreds of averted injuries — more valuable than the roofs.

In addition to problems related to changing the VSL in order to advocate a particular policy position, there are further inefficiencies raised by the fact that different government agencies are using different VSL numbers to guide policy.  So, if FDA places the VSL at $8 million and the EPA puts it at $9 million, then we may find that EPA is spending money to save lives that would save more lives if routed through EPA for environmental projects instead.

So, why not set uniform standard?  It comes back to policy advocacy (note: OMB policy quoted here  was enacted in 2004 and continues to be in effect.  So, it is bipartisan):

The Office of Management and Budget told agencies in 2004 that they should pick a number between $1 million and $10 million. That guidance remains in effect, although the office has more recently warned agencies that it would be difficult to justify the use of numbers under $5 million, two administration officials said.

Close observers of the process point to two reasons for the variation in numbers. First, they say that setting a single standard is not worth the high-stakes battle that would be required with advocates on both sides. The Obama administration, like its predecessors, has preferred to deal with the issue informally, on an agency-by-agency basis.

Second, they say the lack of a standard preserves flexibility.

Flexibility, yes.  Consistency and efficiency … not so much.

Defensive medicine may account for one third of medical imaging costs. Maybe.

Filed Under (Health Care) by Nolan Miller on Mar 3, 2011

Economists generally don’t believe in asking people about their preferences.  We prefer to let people’s behavior tell us about that.  Why? Well, we believe people respond to incentives.  If a person gets paid to tell you the sky is pink, you’re going to find that your survey provides conclusive proof that the sky is pink.  That’s why I don’t know what to think about the following study, which was presented at a recent meeting of the American Academy of Orthopaedic Surgeons.  Although the full paper is not available (or at least I could not find it), it is further discussed in this article from medpagetoday.com.  

The study, entitled “The prevalence of defensive orthopedic imaging: a prospective practice audit in Pennsylvania,” reports the result of a survey of 72 orthopedic surgeons who, for one day, anonymously recorded whether each of the 2068 scans they ordered was primarily done for “defensive” reasons.  That is, whether the test was primarily ordered to protect the doctor from a future lawsuit.

Here are the headline results.  The study finds that approximately 20% of imaging tests are ordered for defensive purposes, accounting for about 1/3 of all imaging costs.  About 40% of MRI scans were ordered for defensive purposes, and they accounted for the vast majority – 86% — of the defensive costs.

At a time when rapidly increasing health care costs are a key policy concern, these results are particularly troubling.  And yet, they suggest a way in which we might be able to significantly cut health care costs without sacrificing quality.  To take one example:

One example: a torn meniscus, a knee cartilage injury that is a leading reason for knee surgery. Studies have shown that a doctor’s judgment based on symptoms and an exam is even better than an MRI to diagnose the condition. Yet patients hardly ever go to surgery without having the imaging test, [the study’s lead author Dr. John ] Flynn said.  (quoted from AP).

So, if we can reduce procedures ordered for defensive reasons, we can reduce health care costs.  This is often taken as an argument in favor of medical malpractice / tort liability reform.  If doctors worry less about huge judgments, they won’t order as many defensive procedures.   As I’ve argued before, the weight of the evidence on this point is that there just isn’t enough money involved to significantly reduce overall health care spending.  But, this is new evidence.  So, how should it change our beliefs?

Well, it is certainly intriguing.  And, it has a certain ring of truth about it.  If I’m a doctor, and by ordering an expensive MRI I can (1) potentially help my patient, (2) at least make the patient feel better because something is being done, (3) not cost the patient much out of pocket because he/she is insured, (4) stick it to insurance companies, whom I hate with a burning passion, and (5) protect myself from liability, all by checking a box on a piece of paper, it seems like a no brainer.  But, that obvious chain of logic is a reason why these results should probably be taken with a grain of salt.  Doctors understand this, and they have a vested interest in medical malpractice reform.  So, if I were asked to complete this survey, and I understood that if defensive medicine were shown to be a big problem that would make medical malpractice reform more likely, which would put more money in my pocket, I’d answer the survey questions by saying that a lot of the procedures were done for defensive purposes.

So, in terms of the standards of scientific evidence, the study is not iron clad.  On the other hand, it supports an unscientific claim I made in my earlier blog on medical malpractice regarding the difficulties we have detecting defensive practices using conventional methods.  These methods work by comparing a treatment and control group.  However, if defensive medicine is widespread even among the control group, then we won’t be able to detect it.  The imaging study suggests that this might, indeed, be the case.