Competition and Insurance

I think the points Kevin Drum is making here start out in a good place but wind up heading in the wrong direction. The salient fact about competition and health insurance isn't that one can't imagine policies that would create a more effectively competitive insurance market, the problem is simply that decent people think the results of such a market would be undesirable.

Under competitive conditions, companies get better at what they do. Normally, that's good. Electronics companies make gadgets that people want, at a cost cheap enough for them to afford them. Restaurants offer tasty food, enjoyable ambiance, efficient service, etc. But what well-functioning insurance companies do is assess risk accurately. And the general premise of health care policies in most countries is that health care should be delivered to people who need health care. This is just fundamentally incompatible with well-functioning insurance companies playing a large role in the financing of health care.

Comments

This is a really important point, and is exactly what was going through my head when I read Cowen's comments. Why equate health care and health insurance? There's obviously an insurance aspect to catastrophic events, but that's only one piece of many needed in a functioning health care system.

We're stuck with an unfortunate, historic convolution of health care and health insurance. But that doesn't mean we have to be confused about it.

Posted by Ben V-L | February 8, 2008 3:45 PM

But what well-functioning insurance companies do is assess risk accurately.

So cute. Well-functioning insurance companies minimize coverage payments using strategies like cherry-picking clients, litigation, and influence of state insurance commissions. And accurately assessing risk.

Posted by SomeCallMeTim | February 8, 2008 3:45 PM

Another thing that a company might be good at is pricing different levels of coverage. That way, you can decide how important it is to pay the costs for high cost/ low value operations that you may or may not need. I think this is a good thing.

But its value is pretty much dwarfed by the fact that insurance companies will mostly just focus on the personal risk aspects of the pricing. And I agree, this is not really a service that we as a society need. Despite arguments to the contrary, lifestyle choices have little importance here. Your own personal medical history is the most important by far. And that's not something you should be deprived of health care over.

I disagree that its necessarily the case that well-functioning insurance companies must focus on this issue, though. There's always the community rating requirement approach to eliminate that consideration from their strategic options. But I'm not sure there's much value to be had in trying to get whatever benefit we can out of the existence of these private companies when we'll always be just swimming upstream on this issue.

Posted by mpowell | February 8, 2008 3:52 PM

But what well-functioning insurance companies do is assess risk accurately. And the general premise of health care policies in most countries is that health care should be delivered to people who need health care. This is just fundamentally incompatible with well-functioning insurance companies playing a large role in the financing of health care.

Er, care to explain why?

Your premises are dodgy, anyway. Insurance is obviously about far more than just assessing risk. And the premise in most countries, including the U.S., is that health care delivery is a matter of both need and ability to pay. If you're rich, you're probably going to be able to get a hip replacement faster than if you're poor regardless of where you live.

Posted by Mixner | February 8, 2008 3:53 PM

MY told us just yesterday that "there's little evidence that health care spending really helps people." Presumably he's changed his mind since yesterday, or else this post just doesn't make any sense. Why do "decent people" believe that funds should be spent to help people, if there's "little evidence" that the spending will help?

Posted by Thomas | February 8, 2008 3:55 PM

There's also another problem with this type of "competition."

With lots of things, like TV sets or tasty food, it's pretty easy for the consumer to weigh the alternatives. Basically, bigger screens and sharper pictures are easily to evaluate.

But with the extraordinary complexity of a typical health care plan---endless difference in choice, deductibles, coverage, etc.---it's really hard to weigh the trade-offs. So I'd guess that most of the decisions tend to be either random/arbitrary or based on advertising/marketing. And obviously, the contract also doesn't give any indication of just how much effort the insurance company will make to avoid paying claims, which is another huge profit center.

That's a similar problem with lots of these fancy new "consumer drugs" like Vioxx that make so much profit for the companies...but sometimes kill so many people!

Posted by RKU | February 8, 2008 3:56 PM

Despite arguments to the contrary, lifestyle choices have little importance here. Your own personal medical history is the most important by far. And that's not something you should be deprived of health care over.

Your personal medical history is likely to be largely a matter of your lifestyle. So why is it unjust to charge people who lead unhealthy lifestyles more for health insurance than people who lead healthy ones? Most insurers charge smokers higher premiums than non-smokers. You think this is wrong, do you?

Posted by Mixner | February 8, 2008 4:00 PM

Under competitive conditions, companies get better at what they do.

Not quite ... under competative conditions companies get better at that which allows them to float under such conditions. Usually and hopefully that is at least to get better at what they actually do, but not necessarily, c.f. SomeCallMeTim's point.

People refer to competative conditions as "Darwinian" and that's not for nothin': evolution is all about natural selection from diversity under competative conditions. But organisms have not evolved to be good at even what they do, but rather to be good enough to survive and pass the genes along ... that is all.

But has competition caused the evolution of a perfect being? No, evolution's allowed whatever creatures can pass their genes along under this competition to survive. Some of what results is breath-taking (synovial fluid, brain complexity) and some of what evolution lets slide is horrifically inefficient and annoying (adapting the Pentose-Phosphate pathway for photosynthetic carbon fixation, i.e. the Calvin cycle, from which one cannot evolve photorespiration free ways of carbon fixation resulting in stupid evolutionary patches being the best an organism can do; the waste pipelines running through our bodies' primary recreational facilities; butt and nose hair).

In my more paranoid moments I wonder if the whole push to teach "intelligent design" in schools is merely a way to ensure that students don't think critically of evolution's failures but rather come away thinking "wow ... competative pressure can result in systems that look intelligently designed" and thus be suckers for whatever "the magic of competition will be even better than any intelligent fix to whatever problem faces us" market-based snake oil is being sold at the moment.

Posted by DAS | February 8, 2008 4:08 PM

> Your personal medical history is likely to be
> largely a matter of your lifestyle.

You might want to tell that to Jim Fixx.

Cranky

When you see him next...

Posted by Cranky Observer | February 8, 2008 4:12 PM

Your personal medical history is likely to be largely a matter of your lifestyle.

No it isn't.

Posted by Cryptic Ned | February 8, 2008 4:15 PM

No it isn't.

Yes, it is. Behavioral factors are not just large, but the single largest contributor to premature death.

Posted by Mixner | February 8, 2008 4:20 PM

And the general premise of health care policies in most countries is that health care should be delivered to people who need health care.

And in this country we call that "socialism".

Posted by right | February 8, 2008 4:30 PM

Right,

I'm afraid you've missed the point here.

First, few people on these boards consider that the insult you do.

Second, the real question is who decides what healthcare you need. Healthcare is extraordinarily expensive, so we will always have to make decisions about how to distribute the available resources. Universal healthcare will give the goverment control of these decisions, but they will still need to be made.

Posted by heedless | February 8, 2008 4:45 PM

Insurance funadmenatlly is the transfer of risk. You join a risk pool of other individuals and pay a premium to cover the expected losses of the entire group. If you incur a fortuituos loss, then the company to whom you have transferred the risk owes you indemnity for the loss according to the contract. The pool of money from the collective whole (your risk pool) is reduced and transferred to you. And the end of the year, the company looks to see if there is any money left. If not, your premiums go up.

It works well with activities like car insurance. If you drive a higher value vehicle, in an area of statistically higher theft rates, and have prior accidents/speeding tickets/DUI's, you will pay more for liability and comprehensive coverage than the person who does not have those risk characteristics. That's just common sense. You pose a higher risk of loss to the collective pool, so you should pay more. the people in your risk pool demand it.

It doesn't work so well with health insurance.

Nobody looks to their car insurance to pay for maintenance, yet we expect health insurance to pay for every trip to the doctor and every pill we take. Likewise, it doesn't seem fair that the people who cost the most to the health insurance risk pool should cover their share of the risk. So they drain the resources and premiums go up for everybody else at renewal time. That causes the relatively healthy to re-think whether it's a good deal to stay in the pool. Each time the premium is raised, the healthy move out and the sick stay in. Guess what happens the next year? That is the "death spiral" for an insured group.

But selection goes both ways. People love to complain about the company denying coverage for pre-existing conditions. But why should the person who stayed out of the risk pool while healthy, paying no premium, suddenly be able to join when he finds out he's sick? For the cost of a premium, he buys insurance and gets the other people in the risk pool to cover his known loss. Then he drops out again after the treatment is paid for. That's a raw deal for the rest.

For car insurance, the risk of loss is fairly defined, because you know what your car is worth. With health insurance, you never know the risk of loss, because a catastrophic $1,000,000 accident could be lurking around the next corner.

Forcing everyone to buy in to one risk pool may be the only solution. But that system forces many who don't want to pay (the young and healthy) to subsidize the sick and old. It is a transfer of risk and wealth.

Depending on your level of comfort with national control of any fundamental part of the economy and people's lives, then single payer is an option. If you think those who cost more should pay more, then the current system of medical underwriting should maintain some vitality.

It just depends who you want in your pool.

Posted by Beer Here | February 8, 2008 4:48 PM

Let's say we grant that behavioral factors effect one's health more than any other variable (a premise I am only willing to grant for the furtherance of this debate). We should still consider that behavior is not solely a matter of personal choice, but is often socially ingrained in an individual or still a function of genetics. We know very little of development of taste and habits, but it is perfectly plausible that many of them are not so much a matter of free choice, as they are qualities, that while internal, have been developed by some external, difficult to control source.

I feel strongly that this is a matter that has largely been ignored in the political arena because it makes us feel uncomfortable (also the Right has succeeded in lampoon this position to death. The simple truth is that people are often not responsible for their position in life, in terms as health, as well as financially and social. The truth is that opportunity is a limited function and if it's a value we truly believe in it is critical for us to go beyond simply "leveling the playing field" to accounting for difference in players by "handicapping the playing field".

This, as it applies to something as fundamental as health insurance, is critical. People don't deserve to have less access to health care because they were unfortunately born with a proclivity to eat poorly, engage in risky activity or substance abuse than do to be put at greater risk because they were born with a congenital heart defect. The truth is these are essential the same, only the heart defect is more legible. We ought to search for inequalities rather than only helping those whose inequalities are so apparent that we cannot help but see them.

Posted by Jonathan | February 8, 2008 4:55 PM

Heedless,

few people on these boards consider that the insult you do.

I'm not trying to be insulting, but let's call it what it is. Matt is saying we should become a more socialist country because (a) all "decent people" think we should do so, and (b) "other countries" have done it. Terrific argument.

the real question is who decides what healthcare you need.

Yup. Some might prefer to individually make the decision, based on what he or she could afford. Others could enter into private contracts with insurance companies on a competitive basis to cover eventualities as outlined in the coverage agreement. Either is compatible with a free market insurance system.

Healthcare is extraordinarily expensive, so we will always have to make decisions about how to distribute the available resources.

You know what else is expensive? Houses. How will we ever figure out who lives where?

Universal healthcare will give the goverment control of these decisions, but they will still need to be made.

Why on earth do we think the government would do a good job of making those decisions? Why? Where's the evidence? What have they ever done a good job managing? The government -- both parties included here -- has never, ever, in history solved a problem without throwing bottomless pits of money at political friends. Why does that seem like an attractive option?

Posted by right | February 8, 2008 5:04 PM

Re: Your personal medical history is likely to be largely a matter of your lifestyle.

Bullshit. Obviously that plays a role, but it is drawfed by plain dumb luck, including the luck involved in what genes you inherit. I'll bet that evreyone here can attribute most of their doctor visits to things that basically being in the wrong place at the wrong time so that they picked up a bad germ, or suffered some random injury.

Re: Nobody looks to their car insurance to pay for maintenance, yet we expect health insurance to pay for every trip to the doctor and every pill we take.

That's not maintenance. Maintenance involves things like eating, drinking, sleeping and exercize. Healthcare imvolves repairs when something goes wrong (and sometimes preventative measures against something going wrong).

A comment of my own: I really wonder if peopel in these deabtes have a clue how insurers really work. I never see mentioned the single most significant way that insurers save money: by "negotiating with" (really, bullying) providers into accepting lowers fees than they would like to charge. As I have said before, check out an EOB next you have something done. The difference between what is billed and what is paid is sizable, and not accounted for by your copay. That's how insurers save money.

Posted by JonF | February 8, 2008 5:08 PM

JonF,

The single best comptetive advantage insurance companies bring to their insureds is the negotiating power they have with the providers (docs and hospitals). What do you think health care would cost without network contracts? In fact, you know-- just look at the EOB's you cite and you'll see the outrageous costs that providers try to pass on to patients. Withot insurer controls, premiums would be double what they are.

Single payer would reduce provider comp even more. Why do you think providers scream about Medicare reimbursement rates? Because the government tells them how much they'll receive. If the government takes control of all healthcare coverage, the providers are screwed.

Posted by Beer Here | February 8, 2008 5:15 PM

If you drive a higher value vehicle, in an area of statistically higher theft rates, and have prior accidents/speeding tickets/DUI's, you will pay more for liability and comprehensive coverage than the person who does not have those risk characteristics. That's just common sense. You pose a higher risk of loss to the collective pool, so you should pay more. the people in your risk pool demand it.

And if you smoke, drink heavily, use drugs, are obese, and/or play a lot of dangerous sports, you are likely to need more health care than if you don't. That's just common sense, too. And just as it isn't fair to require responsible drivers to subsidize the irresponsible behavior of bad drivers, it isn't fair to require people who take care of their health to subsidize the irresponsible behavior of people who don't.

Nobody looks to their car insurance to pay for maintenance, yet we expect health insurance to pay for every trip to the doctor and every pill we take.

If health care were limited to inexpensive routine and preventive primary care, it wouldn't make much sense to fund it through an insurance model, since the expenses would be largely predictable and the variation in demand between individuals would be relatively small. But health care is not limited to inexpensive, predictable expenses, and the variation in demand is not small. That's why we need insurance.

Posted by Mixner | February 8, 2008 5:22 PM

Mixner,

If single payer passes, I claim the right to deny the fat lady in line at BK her double whopper with cheese, since I have a vested interest in her future health costs.

That's a joke. I think.

Posted by Beer Here | February 8, 2008 5:26 PM

JonF,

Bullshit. Obviously that plays a role, but it is drawfed by plain dumb luck, including the luck involved in what genes you inherit.

No it isn't. As the piece I linked to earlier states, the evidence indicates that behavioral factors contribute more to premature death than genetic ones. Behavior is about 40% of the total, more than any other type of factor.

"Shortfalls in medical care," by the way, contribute only around 10%.

Posted by Mixner | February 8, 2008 5:29 PM

Mixner,

Behavioral factors are not just large, but the single largest contributor to premature death.

Lifestyle, including exercise, diet, and smoking, is important. This is especially so for "premature death" in seniors (see the article you linked to). By definition, seniors are the individuals who have survived for 65 years. These people, thus, have much better pool of genes than the newborn babies. Most people with serious genetic problems just don't make it to 65.

So yes, it would be a good idea to have more incentives for seniors to exercise. Yes, lifestyle is easier to change than person's genes, or (retroactively) how well they had been fed in early childhood. But no, it's not accounting for a majority of serious health problems.

And anyway, as far as cost (and cost control) is concerned, the largest expenditure (80% over lifetime?) is in the last three weeks of life. Even with perfect lifestyle, people will eventually die. The real discussion should be about these costs and their control, not over who should pay for $10 vaccination.

Posted by AO | February 8, 2008 5:29 PM

Right,

The government has fixed many problems historically on a modest government. This relates to other arguments made in the education thread as well. Take for example, the problem with the quality of meat 100 years ago. The fix was cheap and effective; the FDA. It also has allowed the US to enjoy relatively safe drugs. See, there are many ways the US government has effectively managed social problems and regulated an inefficiency very...efficiently.

Posted by freddiemac | February 8, 2008 5:34 PM

HRC's plan for mandates administered through the health insurance industry is "One Big HMO," to paraphrase the IWW's "One Big Union," meme. Insurance companies make money by maximizing revenue while minimizing payout. Has anyone thought to look at what for-profit hospitals do to Medicare? I represented them once a time, 30 years back, as a junior attorney and it wasn't pretty. The insurance companies are as bad or worse.

Posted by John Shreffler | February 8, 2008 5:37 PM

Let's say we grant that behavioral factors effect one's health more than any other variable ... We should still consider that behavior is not solely a matter of personal choice, but is often socially ingrained in an individual or still a function of genetics.

Sure, but then so is risky behavior in other contexts. The disproportionately risky driving behavior of young men, for example, is probably due in part to their raging testosterone levels and peer-pressure from friends. Does that mean insurance companies are unfair to charge higher premiums to young male drivers? I don't think so. I doubt many other people think so, either. At best, this "it's not solely a matter of personal choice" argument might mitigate their personal responsibility for their actions, but it doesn't relieve them of it altogether.

Posted by Mixner | February 8, 2008 5:39 PM

So wait, are we agreeing that Edward's plan of mandating everyone have insurance isn't a good idea?

Posted by freddiemac | February 8, 2008 5:46 PM

But no, [lifestyle is] not accounting for a majority of serious health problems.

And you know this, how?

Your claim doesn't really make sense, anyway. No one thing "accounts" for serious health problems. Instead, various influences contribute to a individual's state of health and lifespan. Obviously, everyone eventually dies, no matter how healthy their lifestyle. And some people have serious health problems that have nothing to do with their behavior. The evidence nevertheless indicates that behavior is the single biggest factor in determining health and lifespan. Even if behavior were just a large factor, rather than the biggest one, it would still be unjust to ignore it in the pricing of health insurance.

Posted by Mixner | February 8, 2008 5:49 PM

The claim that behavior determines health history has not been proven in a way relevant to this discussion.

Sure, if you exercise you will live longer. But everyone dies of something eventually. Smokers are more likely to contract lung cancer, but that's actually a pretty cheap disease to get. Its hard to detect until its pretty much a sure bet to kill you within the year. Plus those guys have been paying a tax on their cigs their whole life.

The point is there is a difference between a corrolation in health costs and a correlation in life expectancy. Studies showing the latter do not speak to the former. But if you have something like Crohn's diseas, you're looking at a lifetime of substantially elevated risk for something that's jsut bad luck.

Posted by mpowell | February 8, 2008 5:49 PM

Mixner,

It's an implication I definitely have considered. In fact, it would be an even more radical implication when considering our system of justice. If we recognize that people are often driven by forces well outside of their control it seems a bit unjust to punish them for such actions. Which is why society has conveniently pushed such thoughts aside.

The truth is that accepting such a premise would substantial alter our thinking on a lot of issues, but why is it legitimate to make such a distinction between choice and ingrained actions, if such a distinction is more about societal convenience than actual fairness. If we were to reconsider our society in light of such premises we'd probably have to move away from a risk-insurance model to simple aggregating risk for everything universally. I suspect this wouldn't be the most terrible world in which one could live, but it would be awfully painful to get there.

Posted by Jonathan | February 8, 2008 5:56 PM

mpowell,

Sure, if you exercise you will live longer. But everyone dies of something eventually. Smokers are more likely to contract lung cancer, but that's actually a pretty cheap disease to get.

No, lung cancer is not a cheap disease to get. And the health problems caused by smoking are far broader than just a greatly elevated risk of lung cancer. Smoking is a major contributor to heart disease and a host of disorders related to the circulatory and respiratory systems.

Plus those guys have been paying a tax on their cigs their whole life.

If the tax is used to subsidize non-smokers for the additional burden smokers impose on the health care system, it contradicts your position that a person's behavior should not influence how much he pays for health care. If the tax is used for some other purpose, it's irrelevant to the point we're debating.

Posted by Mixner | February 8, 2008 5:59 PM

Right,

You know what else is expensive? Houses. How will we ever figure out who lives where?

Ever looked at the prices for earthquake insurance for houses near the San Andreas fault? It's essentially unobtainable. Yet you are actually free to move elsewhere. On the other hand, people with a bad variant of BRCA2 gene cannot change that at will. Yet a perfect free-market equilibrium would be for these people to pay in full for their (or their daughters') near-imminent breast cancer treatment.

It is considered decent to have a safety net for bad luck (e.g. genetic; which is both predictive and irreversible). You can call this socialism or whatever, but frankly, "there but for the grace" strikes me as both more decent and more commonsensical approach than your free-market fundamentalism.

Posted by AO | February 8, 2008 6:02 PM

No it isn't. As the piece I linked to earlier states, the evidence indicates that behavioral factors contribute more to premature death than genetic ones. Behavior is about 40% of the total, more than any other type of factor.

Even if this is true, it still means that 60% of premature deaths (which is only one measure of health, BTW) are due to non-behavioral factors. And it's not necessarily easy to disentangle "behavioral patterns" from other factors when determing insurance premiums; for instance, given that obesity is heavily influenced by genetics, charging obese people more for health insurance is unfair to those people who are obese because of their genes.

If you want individuals to pay for the costs of unhealthy behavior, there are fairer ways to do so than allowing insurance companies to charge sicker people higher rates. We already impose taxes on tobacco & alcohol; we could tax junk food if we want people to pay for the costs of unhealthy eating.

Posted by Peter H | February 8, 2008 6:05 PM

Behavior is about 40% of the total, more than any other type of factor.

In other words, even in your cherry-picked article lifestyle does not account for the majority of health problems.

Posted by mq | February 8, 2008 6:11 PM

Take for example, the problem with the quality of meat 100 years ago. The fix was cheap and effective; the FDA. It also has allowed the US to enjoy relatively safe drugs.

Sure, the FDA is dandy, and I'm sure with narrow mandates government agencies can be more-or-less effective (the SEC is another example), if only by the law of large numbers.

However, I'm pretty sure the FDA does not (a) buy meat from meat producers, check it over for quality, and then sell it to consumers at politican determined "fair prices", or (b) subsidize steaks for everyone who can't afford to eat them.

So let me rephrase my earlier question: When in history has the government ever successfully played a key role in market economics without throwing bottomless sums of money at political friends? Education? No. Defense? No. Medicare? No.

Why will healthcare be any different? We're going to spend shitloads of money, and get terrible outcomes. It's the Dems' version of Iraq. The election in November is going to be about which one is stupider.

Posted by right | February 8, 2008 6:18 PM

Right,

Virtually every developed country spends less on health care and covers a higher percentage of people than the United States, with, at the very least, no evidence of inferior health outcomes.

Posted by Peter H | February 8, 2008 6:22 PM

Even if this is true, it still means that 60% of premature deaths (which is only one measure of health, BTW) are due to non-behavioral factors.

So what? The relevant point is that chosen behavior is a huge contributor to health and premature death, and thus a huge contributor to health care costs. It would therefore be unjust to ignore behavior in determining how much an individual should pay for health insurance, just as it would be unjust to ignore behavior in the pricing of auto insurance.

And it's not necessarily easy to disentangle "behavioral patterns" from other factors when determing insurance premiums; for instance, given that obesity is heavily influenced by genetics, charging obese people more for health insurance is unfair to those people who are obese because of their genes.

While the risk of obesity is certainly influenced by genes, they're obviously not the only factor. You can't seriously argue that for most people, chosen behavior has no significant influence on whether they're obese or not. And that argument would be even harder to make with respect to smoking, or alcohol abuse, or drug abuse.

If you want individuals to pay for the costs of unhealthy behavior, there are fairer ways to do so than allowing insurance companies to charge sicker people higher rates. We already impose taxes on tobacco & alcohol; we could tax junk food if we want people to pay for the costs of unhealthy eating.

That would depend on how the taxes are used, among other things. But, sure, I'd be open to a junk food tax, and an increase in tobacco taxes, if they were used to lower the costs of health care for people with a healthier lifestyle. It amounts to the same thing: People with unhealthy lifestyles pay more for health care than people with healthy ones.

Posted by Mixner | February 8, 2008 6:28 PM

Virtually every developed country spends less on health care and covers a higher percentage of people than the United States, with, at the very least, no evidence of inferior health outcomes.

Nonsense. U.S. cancer survival rates, for example, are the best in the world. This is almost certainly a consequence of our higher levels of spending on health care services to diagnose and treat cancer.

Posted by Mixner | February 8, 2008 6:33 PM

Mixner,

(Me:) But no, [lifestyle is] not accounting for a majority of serious health problems.

(Me:) And you know this, how?

Well, you say it yourself :) :
No one thing "accounts" for serious health problems.

Stated like that, actually, it's just plain wrong (an absolutely healthy person with one bad copy of RB1 gene will get cancer with 100% certainty. So this one thing accounts for very serious health problems).

But assuming you are actually curious: the contribution of lifestyle to the development of serious disease (e.g. effect of regular exercise on the onset of diabetes in risk groups) is statistically significant but nowhere near 100% or even 25% (on the order of 5% lower chance over 5 years -- and that for people with at-risk BMI). For population at large, the effect would be fairly small, though noticeable and important.

Again, the problem is not in how to enforce "correct lifestyle" (give an insurance premium break to people who exercise; charge smokers more). If lifestyle is overwhelmingly important, then just make premiums near zero for those who satisfy your criteria.

The real problem is the huge cost of modern healthcare and especially the irrational cost of the end-of-life care (when subsidized, willy-nilly, by the other premium payers and taxpayers). Wider risk pool is helpful.

Posted by AO | February 8, 2008 6:40 PM

I'm not one to stand up for the insurance companies, but we're missing the elephant in the room here - the insurance companies, if competitive, would get really good at finding ways to avert their clients' risks. There are many ways that well functioning insurance markets do this (incentives, watchdogging the providers, finding more efficient ways to distribute care, applying actuary tables to epi studies, etc.). Adverse selection is a problem, but it wouldn't be accelerating in the absense of gov regulation. As Obama's economist pointed out, the problem is its too expensive, and competition rarely increases price.

As a response to AO, I'm curious where you get your figures. Lifestyle is clearly the problem for an overwhelming number of conditions. Hospitals aren't filled with thin people with bad genes these days - their filled w/ fat people who have complicated their bad genetics with bad lifestyle. If BMI doesn't show that, doen't be surprised - it's a terrible standard.

Posted by John | February 8, 2008 6:51 PM

Nonsense. U.S. cancer survival rates, for example, are the best in the world. This is almost certainly a consequence of our higher levels of spending...

Meanwhile: "30-day acute myocardial infarction case-fatality rates are below 7% in Denmark, Iceland, and Switzerland, compared with almost 15% in the United States. Incidence of major amputations among diabetic patients in Finland, Australia, and Canada is less than 10 per 10 000 compared with 56 per 10 000 in the United States. And Australia, Canada, England, and New Zealand all have a better 5-year kidney transplantation survival rate than the United States." (http://www.annals.org/cgi/content/full/146/6/473)

Is that also almost certainly a consequence of our higher levels of spending, Mixner? It's great that 5-year breast cancer survival rates are better in the US than elsewhere. But cherry picking your examples doesn't make a case that the healthcare system here is "just better".

Posted by AO | February 8, 2008 6:53 PM

As I've said before, I believe the lifestyle issue is a red herring.

a) If lifestyle choices lead to sickness and death, this is it's own punishment. There is no reason to add additional punishments such as higher health care costs. (Not to mention social opprobrium.) Really, do we need more reasons to make (say) fat people unhappy with their lives?
b) The whole question of cost has not been settled because it tends to view these costs in isolation, not considering, for instance, the fact that people who die earlier receive less in social security payments.
c) We cannot, and should not, force each other to maintain "healthy" life-styles. Conservatives can't talk about "freedom" and the one hand and then call for Big Brother-style monitoring of lifestyles on the other.

Posted by Matt Stevens | February 8, 2008 7:07 PM

Stated like that, actually, it's just plain wrong (an absolutely healthy person with one bad copy of RB1 gene will get cancer with 100% certainty. So this one thing accounts for very serious health problems).

No, that one thing "accounts" for that one serious health problem, not "serious health problems" in general, which are a combination of genetic, behavioral and enviromental influences.

But assuming you are actually curious: the contribution of lifestyle to the development of serious disease (e.g. effect of regular exercise on the onset of diabetes in risk groups) is statistically significant but nowhere near 100% or even 25%

As reported in the article I linked to, the evidence indicates that behavior contributes 40% to premature deaths. I don't know where you get your figure of less than a 25% contribution from lifestyle to "the development of serious diseases." Are you just guessing?

Again, the problem is not in how to enforce "correct lifestyle" (give an insurance premium break to people who exercise; charge smokers more). If lifestyle is overwhelmingly important, then just make premiums near zero for those who satisfy your criteria.

This makes no sense, either. I'm not sure what you mean here by "overwhelmingly important." Lifestyle is certainly a very important contributor to poor health and premature death. But, obviously, even people with extremely healthy lifestyles still use health care. Just not as much health care as people with unhealthy lifestyles. So it would be absurd to charge the former group "near zero" premiums.

The real problem is the huge cost of modern healthcare and especially the irrational cost of the end-of-life care (when subsidized, willy-nilly, by the other premium payers and taxpayers).

I agree that we probably spend far too much on end-of-life care. These costs are mostly paid by Medicare. We should reform Medicare to impose much greater limits on funding for end-of-life care. But there seems to be little interest among either party, especially the Democrats, in reducing Medicare benefits at all.

Posted by Mixner | February 8, 2008 7:09 PM

AO,

It's great that 5-year breast cancer survival rates are better in the US than elsewhere.

It's not just "5-year breast cancer survival rates" for which the U.S. has superior "outcomes," it's cancer survival rates overall, and also most specific types of cancer. Considering that cancer is one of the primary causes of death, I'm not sure why you think this is "cherry-picking."

Posted by Mixner | February 8, 2008 7:16 PM

John:
the numbers (as I remember them) are from a recent talk at the Mass General Hospital. They were mentioned in passing, but I can ask around and get the references (won't happen until Monday, though). I'm not a diabetes expert, so I don't know if BMI is a good indicator. If it's a bad predictor of diabetes, then the case is actually made more general: for a random population, effect of exercise was small.

I'm not sure I agree 100% with your take on the "nature-nurture" split in disease etiology. But that discussion would be a bottomless pit.

On your main point, I think that you strongly overestimate the ability of insurance or any free market to find global optimum. This is especially the case when the societal optimum is distinct from strictly monetary.

Posted by AO | February 8, 2008 7:20 PM

AO,

I find it ironic that you complain about cherry-picking, since you obviously cherry-picked the text you just quoted from the article in which it appears. The very first line of that article is:

The U.S. health care system is among the best in the world.

The statistics you quoted are just examples the article cites in support of its claim that health care in the U.S. "isn’t always the best." Of course, since no one has claimed that health care in the U.S. is always the best, this is a strawman.

Posted by Mixner | February 8, 2008 7:25 PM

Matt Stevens,

a) If lifestyle choices lead to sickness and death, this is it's own punishment. There is no reason to add additional punishments such as higher health care costs. (Not to mention social opprobrium.) Really, do we need more reasons to make (say) fat people unhappy with their lives?

By that reasoning, we shouldn't charge people with really bad driving records any more for car insurance than people with really good driving records. According to you, we shouldn't add "additional punishments" in the form of higher insurance premiums to the "punishment" of a greater risk of injury, death or car repair bills that they have already imposed on themselves by their reckless driving. You really believe that, do you?

Posted by Mixner | February 8, 2008 7:35 PM

But selection goes both ways. People love to complain about the company denying coverage for pre-existing conditions. But why should the person who stayed out of the risk pool while healthy, paying no premium, suddenly be able to join when he finds out he's sick? For the cost of a premium, he buys insurance and gets the other people in the risk pool to cover his known loss. Then he drops out again after the treatment is paid for. That's a raw deal for the rest.

Umm, here in NJ there's this lil thing called mandatory car insurance -- solves that pesky problem.* Yet somehow this isn't seen as quasi-socialist, whereas healthcare mandates are beyond the pale.

* - Yes, I know mandates aren't infallible -- but my example was no more simplistic than that proposed by 'Beer Here.'

Posted by EpicureanQuaker | February 8, 2008 8:01 PM

Mixner,

... most specific types of cancer. [...] I'm not sure why you think this is "cherry-picking."

According to this (http://www.medscape.com/viewarticle/561737), the US-Europe difference was in solid tumors, and in that group the differences were primarily due to breast and prostate cancers 5-year survival rate. (I'm not sure that constitutes "most" specific types of cancer; and don't forget leukemias etc where there was no difference). The most glaring difference is the more aggressive early testing in the US for both these types of cancers. Good!*

Cancer is a major killer, but so are heart attacks. And the mortality rate is twice as bad in the US (see cited article in the Annals of Internal Medicine). How is your focusing on two particular cancers not cherry-picking? You are making a sweeping statement, based on squinting just right, my friend.

As for your cited paper of behavior contributing 40% to premature death in seniors -- I already made a point that it has very little to do with general population and serious disease. These are very different groups, and the measured effects have very little to do with each other. Apples and oranges. Moreover, I don't see how this is affected by the size of the risk pool (e.g. universal insurance or not).

Well, it's Friday night. Have a good one.

By the way, I didn't particularly enjoy your habit of declaring anything new to you "nonsense".

-----
*(Well, there are caveats. Basically, there is a question whether early detection leads to cancer diagnosis 5 years + 1 day before death, which would result in a better 5-year survival rate, but not a better "real" outcome. I don't know how important is that factor.)

Posted by AO | February 8, 2008 8:05 PM

AO,

According to this (http://www.medscape.com/viewarticle/561737), the US-Europe difference was in solid tumors, and in that group the differences were primarily due to breast and prostate cancers 5-year survival rate.

Your URL takes me to a subscription wall. As I said before, the superior performance in the U.S. applies to cancer overall (not just breast cancer, and not just solid tumors). You can find the data here.

Not only does the U.S. have much better 5-year survival rates than Europe for cancer overall, but it scores better for 12 of the 16 types of cancer studied. For most of these types, the performance of the U.S. is not just slightly better, but much better. The only type of cancer for which Europe has a statistically significant higher survival rate is testicular cancer, and even there the European advantage is small (97.3 vs. 95.4).

Posted by Mixner | February 8, 2008 8:31 PM

AO,

Cancer is a major killer, but so are heart attacks. And the mortality rate is twice as bad in the US (see cited article in the Annals of Internal Medicine).

I see no statement in that article comparing the mortality rates from heart disease in Europe and the U.S.

As for your cited paper of behavior contributing 40% to premature death in seniors. I already made a point that it has very little to do with general population and serious disease.

On the contrary, it has everthing to do with the general population and serious disease, since the figure refers to premature deaths among all age groups. Not just seniors, as you falsely assert.

Posted by Mixner | February 8, 2008 8:44 PM

Nonsense, Matt. Countries like Germany and the Netherlands combine a voucher system, private insurance companies, and national regulation of private insurance companies pretty well.

What's missing from both Clinton and Obama is a voucher system and a call for national regulation of health insurance.

The first? They're afraid to talk about where the money will come for it.

The second? They're thinking too small, and probably afraid of influencing donors.

Posted by SocraticGadfly | February 8, 2008 9:35 PM

So what? The relevant point is that chosen behavior is a huge contributor to health and premature death, and thus a huge contributor to health care costs. It would therefore be unjust to ignore behavior in determining how much an individual should pay for health insurance, just as it would be unjust to ignore behavior in the pricing of auto insurance.

You’re missing the point. You think it’s OK for insurance companies to charge (or avoid insuring )sicker people higher premiums. That means that people who get sick because of their “lifestyle choices” will be denied coverage and/or pay higher premiums, but so will people who get sick entirely because of chance.

And, by the way, I don’t have a problem with charging smokers (or heavy drinkers) a couple of hundred extra dollars in yearly premiums. I do have a problem with individuals having to pay higher premiums, higher deductibles, and being excluded from vital coverage simply because they’re sick.

Nonsense. U.S. cancer survival rates, for example, are the best in the world. This is almost certainly a consequence of our higher levels of spending on health care services to diagnose and treat cancer.

Looking at 5-year cancer survival rates doesn’t tell the whole story.. The United States conducts far more tests than any other country, which turn up many more cancers. So it ends up diagnosing lots of cancers that aren't lethal or are slow-moving enough to not require treatment, which results in a higher survival rate. The overall percentage of Americans who die from cancer is actually higher in the United States than the UK.

There are plenty of statistics that show that health care outcomes are worse in the United States than in other countries. For example, there was a study in Open Medicine that compares Canadian and American outcomes on 25 different measures and shows that on balance Americans do worse on more types of treatment and better on fewer here in the U.S than in Canada. There was a study in Health Affairs that found the United States, of 16 countries surveyed, had the worst rate of amenable mortality. I happen to think these studies need to be taken with a grain of salt (like the 5-year cancer survival rate statistics), because they don’t take into account factors like lifestyle and incidence rates. That’s why I said there is “no evidence of inferior health outcomes” in other countries. These studies don’t necessarily prove the United States has an inferior quality of health care, but they do make it hard to argue that the quality of care is better than the rest of the world.

By the way, you said earlier that “shortfalls in medical care” contribute only 10% to premature deaths, and now you’re saying that our higher health care spending accounts for our better cancer survival rates. Sounds to me like you’re trying to have it both ways.

Posted by Peter H | February 8, 2008 11:03 PM

A couple of my hyperlinks failed to appear, so I'll post them again:

The study on health care outcomes in Canada & the United States:
http://www.thestar.com/News/article/204163

The study on amenable mortality:
http://content.healthaffairs.org/cgi/content/abstract/27/1/58?ijkey=05uD000683MNE&keytype=ref&siteid=healthaff

Posted by Peter H | February 8, 2008 11:15 PM

A couple of my hyperlinks failed to appear, so I'll post them again:

The study on health care outcomes in Canada & the United States:
http://www.thestar.com/News/article/204163

The study on amenable mortality:
http://content.healthaffairs.org/cgi/content/abstract/27/1/58?ijkey=05uD000683MNE&keytype=ref&siteid=healthaff

Posted by Peter H | February 8, 2008 11:16 PM

Peter H,

You’re missing the point. You think it’s OK for insurance companies to charge (or avoid insuring )sicker people higher premiums.

I didn't say that. I said I think it's okay for insurance companies to charge higher premiums to people who engage in unhealthy behavior. And you just said you agree with me, at least with respect to smokers and heavy drinkers.

Looking at 5-year cancer survival rates doesn’t tell the whole story.. The United States conducts far more tests than any other country, which turn up many more cancers. So it ends up diagnosing lots of cancers that aren't lethal or are slow-moving enough to not require treatment, which results in a higher survival rate.

As I showed in a previous post, U.S. cancer survival rates are significantly better for most types of cancer, including the most dangerous types. Only a very small fraction of cancers are "slow-moving enough to not require treatment." Most cancers are life-threatening. The superior performance of the U.S. is a consequence of both better diagnosis and better treatment of cancer. The earlier cancer is diagnosed, the greater the chance of successful treatment.

The simple fact is that, in general, if you get cancer your chances of surviving at least five years are much better if you live in the United States than if you live in Europe. The U.S. health care system is simply much better at detecting and treating cancer than the health care systems of Europe.

Posted by Mixner | February 8, 2008 11:38 PM

By that reasoning, we shouldn't charge people with really bad driving records any more for car insurance than people with really good driving records. According to you, we shouldn't add "additional punishments" in the form of higher insurance premiums to the "punishment" of a greater risk of injury, death or car repair bills that they have already imposed on themselves by their reckless driving. You really believe that, do you?

Well reckless driving is different from (say) eating too much, anyway, because it's dangerous both to the driver and to people around them. In addition, in the absence of injury the consequences of accidents are purely monetary. You can't say that about obesity, for example. So I wouldn't push the comparison too far in any case.

But no, I don't think society would collapse if people who suffered serious automobile injuries weren't charged extra on their premiums. Having a punctured lung is punishment enough. Call me a softee.

Posted by Matt Stevens | February 9, 2008 12:12 AM

Peter H,

The study on amenable mortality:
http://content.healthaffairs.org/cgi/content/abstract/27/1/58?ijkey=05uD000683MNE&keytype=ref&siteid=healthaff

A nation's "amenable mortality rate," especially as defined in that study, tells you nothing about the performance of the nation's health care system. The "amenable mortality rate," just like the total mortality rate, is determined by a vast constellation of variables, from diet and exercise to accident rates and pollution levels, that differ significantly between different countries. Unless you control for the influence of these other variables on the "amenable mortality rate," you cannot isolate the effect of the health care system, and thus you cannot draw any conclusions about its performance. For a more detailed explanation of this point, see here.

By the way, you said earlier that “shortfalls in medical care” contribute only 10% to premature deaths, and now you’re saying that our higher health care spending accounts for our better cancer survival rates. Sounds to me like you’re trying to have it both ways.

Well I can only assume that's because you don't have a very good understanding of the issue. Yes, we spend more on cancer diagnosis and treatment, and as a result we have significantly better cancer survival rates. But most people do not die prematurely from cancer even in countries with relatively poor cancer survival rates. Cancer is mostly a disease of the elderly, who are at much greater risk than the general population of dying from a host of age-related diseases. So the effect of better cancer diagnosis and treatment on the rate of premature death for the population as a whole is relatively small.

I raised cancer survival rates to rebut your false claim that there is "no evidence of inferior health outcomes" from the health care systems of other countries. The much better cancer survival rates in the U.S. directly contradict your claim.

Posted by Mixner | February 9, 2008 12:12 AM

Matt Stevens,

Well reckless driving is different from (say) eating too much, anyway, because it's dangerous both to the driver and to people around them.

That's only relevant to the charges for third-party liability insurance. Yet we also charge drivers with bad driving records higher premiums for coverage of themselves and their own vehicles. Are you saying you think it is wrong to charge them more for this coverage, that the higher premiums are "additional punishment" that they should not have to pay?

In addition, in the absence of injury the consequences of accidents are purely monetary.

But, obviously, many accidents involve injury to the driver as well as damage to his vehicle. A large part of auto insurance premiums goes to cover expenses arising from injuries. Those expenses can even greatly exceed the costs of replacing or repairing the motor vehicle, especially if the injury is severe. Auto accidents are one of the leading causes of premature death, and an even larger cause of non-fatal injury.

But no, I don't think society would collapse if people who suffered serious automobile injuries weren't charged extra on their premiums.

I didn't ask you if you think society would collapse. I asked you if you think drivers with bad driving records (speeding, running red lights, drunk driving, etc.) should not have to pay any more for their insurance premiums than drivers with good driving records. In other words, do you think responsible drivers should have to subsidize the irresponsible behavior of reckless drivers? Apparently, you do. I can't imagine more than a small minority of drivers would agree with you.

Posted by Mixner | February 9, 2008 12:30 AM

"But what well-functioning insurance companies do is assess risk accurately."

Exactly, just as a casino assesses its odds of losing accurately.

Turning to insurance companies to deliver better health care is like turning to Vegas to deliver better baseball. Both gamble on the outcomes, but neither controls the action (the 1919 White Sox notwithstanding).

Efficiency is in the eye of the beholder, and US health care delivery is efficient only if you're taking a percentage of the action.

Posted by cpl | February 9, 2008 12:35 AM

Peter H,

For example, there was a study in Open Medicine that compares Canadian and American outcomes on 25 different measures and shows that on balance Americans do worse on more types of treatment and better on fewer here in the U.S than in Canada.

You're wrong about this too. Your (working) link doesn't go to the study, but only to a newspaper article about the study. If you look at the study itself, you will find that the authors report that their findings were inconclusive and that:

The only condition in which results consistently favoured one country was end-stage renal disease, in which Canadian patients fared better.

But there are also serious problems with the study. It's actually a meta-study of older epidemiological studies. Among the problems are the age of the data (most of the data it cites is around twenty years old; some is more than thirty years old), the geographical scope of the data (e.g., individual cities rather than the entire nation) and its definitions of "better" results (better only for certain groups of patients rather than patients as a whole). The authors also have a political agenda, which probably influenced their selection of studies to try and favor Canada. David Himmelstein is co-founder of Physicians for a National Health Program, a political lobbying organization promoting single-payer health care in the United States.

Posted by Mixner | February 9, 2008 1:10 AM

Again, as I said, auto insurance is different because in most cases, the "cost" of reckless driving can be rectified with money that the insurance company can provide. This is not the same with health problems such as heart disease, which are painful to deal with even if medical care is fully paid for. That's why I focused on someone who was injured, which is closer to the medical example than someone who was just speeding (which doesn't carry the social stigma that, say, obesity does).

In addition, as I've said before, everyone gets sick but not everyone has a automobile accidents. Both you and the fat man are going to require medical treatment, and there's no reason to assume that your medical treatment is going to cost less than his. (Even if they did, there's no basic to conclude that you cost more to society over all, if you're expected to live longer.) In those circumstances, I see nothing wrong with charging both of you the same amount. Insurance companies may have the right to charge the fat man more, but I don't believe that society is better off because of that.

Posted by Matt Stevens | February 9, 2008 1:20 AM

That should be "no reason to conclude that you cost less to society over all," dammit.

Posted by Matt Stevens | February 9, 2008 1:23 AM

As I showed in a previous post, U.S. cancer survival rates are significantly better for most types of cancer, including the most dangerous types.

Like I said, five-year survival rates are an misleading measure. Dr. Gilbert Welch, a Professor at Dartmouth Medical School, wrote about this in the LA Times

Most of us assume that rising five-year survival means that cancer death rates are falling. But in a study for the Journal of the American Medical Assn., my colleagues and I found no relationship between changes in five-year survival and changes in how many Americans die from cancer, as measured by the number of cancer deaths per 100,000 population.

Consider the two cancers with the most dramatic increases in five-year survival since 1950, prostate cancer (from 43% to 94%) and melanoma, or skin cancer, (from 49% to 88%). Despite all the press about recent declines, the current death rate from prostate cancer is actually a little higher than in 1950, and the death rate from melanoma is more than twice as high.

And then consider the case for one of the cancers with the worst five-year survival rates: stomach cancer. There has been only a trivial change in stomach cancer five-year survival since 1950 (from 12% to 19%), but it is America's biggest cancer success story because people die at one-fifth the rate they did in 1950.

Only a very small fraction of cancers are "slow-moving enough to not require treatment." Most cancers are life-threatening

That's not necessarily true. Prostate cancer is the best example of this: many more people have prostate cancer than will actually experience symptoms of it, much less die of it. More on this from Dr. Welch:

Tests like the P.S.A. [prostate specific antigen] are forcing doctors and patients to think about the word ''cancer'' in a radically different way. Most of us think of it as a deadly disease if left untreated. But there are also microscopic cancers that will never cause problems. We all have abnormal cells, but most of us will not die from cancer. Pseudodisease is most common in prostate and breast cancer. But it is also an issue for kidney cancer, melanoma and lung cancer. When we look for cancer early, we cast such a wide net that we catch both real and pseudodisease. Since we can't tell which is which, we treat it all. And for those for those who never needed treatment, treatment can only hurt.

http://query.nytimes.com/gst/fullpage.html?res=9C04E6DD1439F932A15755C0A9629C8B63

The superior performance of the U.S. is a consequence of both better diagnosis and better treatment of cancer. The earlier cancer is diagnosed, the greater the chance of successful treatment.

More testing for & detection of cancer is not an umitigated blessing. For some common cancers, it is not clear that early detection and treatment actually prolong patients' lives. And if detection unnecessarily exposes patients to the risks of chemotherapy, surgery and radiation, it can hurt them more than it helps.

Posted by Peter H | February 9, 2008 2:46 AM

As I showed in a previous post, U.S. cancer survival rates are significantly better for most types of cancer, including the most dangerous types.

Like I said, five-year survival rates are an misleading measure. Dr. Gilbert Welch, a Professor at Dartmouth Medical School, wrote about this in the LA Times

Most of us assume that rising five-year survival means that cancer death rates are falling. But in a study for the Journal of the American Medical Assn., my colleagues and I found no relationship between changes in five-year survival and changes in how many Americans die from cancer, as measured by the number of cancer deaths per 100,000 population.

Consider the two cancers with the most dramatic increases in five-year survival since 1950, prostate cancer (from 43% to 94%) and melanoma, or skin cancer, (from 49% to 88%). Despite all the press about recent declines, the current death rate from prostate cancer is actually a little higher than in 1950, and the death rate from melanoma is more than twice as high.

And then consider the case for one of the cancers with the worst five-year survival rates: stomach cancer. There has been only a trivial change in stomach cancer five-year survival since 1950 (from 12% to 19%), but it is America's biggest cancer success story because people die at one-fifth the rate they did in 1950.

Only a very small fraction of cancers are "slow-moving enough to not require treatment." Most cancers are life-threatening

That's not necessarily true. Prostate cancer is the best example of this: many more people have prostate cancer than will actually experience symptoms of it, much less die of it. More on this from Dr. Welch:

Tests like the P.S.A. [prostate specific antigen] are forcing doctors and patients to think about the word ''cancer'' in a radically different way. Most of us think of it as a deadly disease if left untreated. But there are also microscopic cancers that will never cause problems. We all have abnormal cells, but most of us will not die from cancer. Pseudodisease is most common in prostate and breast cancer. But it is also an issue for kidney cancer, melanoma and lung cancer. When we look for cancer early, we cast such a wide net that we catch both real and pseudodisease. Since we can't tell which is which, we treat it all. And for those for those who never needed treatment, treatment can only hurt.

http://query.nytimes.com/gst/fullpage.html?res=9C04E6DD1439F932A15755C0A9629C8B63

The superior performance of the U.S. is a consequence of both better diagnosis and better treatment of cancer. The earlier cancer is diagnosed, the greater the chance of successful treatment.

More testing for & detection of cancer is not an umitigated blessing. For some common cancers, it is not clear that early detection and treatment actually prolong patients' lives. And if detection unnecessarily exposes patients to the risks of chemotherapy, surgery and radiation, it can hurt them more than it helps.

Posted by Peter H | February 9, 2008 2:48 AM

The "amenable mortality rate," just like the total mortality rate, is determined by a vast constellation of variables, from diet and exercise to accident rates and pollution levels, that differ significantly between different countries. Unless you control for the influence of these other variables on the "amenable mortality rate," you cannot isolate the effect of the health care system, and thus you cannot draw any conclusions about its performance.

I specifically said that the study should be taken with a grain of salt, since it didn't account for lifestyle differences and other factors. What I said was America's last-place finish provides prima facie evidence that the United States does not have superior health care to the rest of the world.

And if you're going on harp on the limitations of amenable mortality, you should also acknowledge the problems with using 5-year survival rates of cancer.

Posted by Peter H | February 9, 2008 3:42 AM

Having lived in Europe for several years now, I can testify that the US culture of preventative health care and the regular prophylactic treatments we are advised to seek (think yearly cervical smears from age 18) IF WE HAVE GOOD INSURANCE are at odds to the culture over here (especially in UK - France is a little better). But the other thing to note is that in the US (as compared to UK) something like 50% more money/person is spent on healthcare. Also in Britain specifically there is a cultural norm of not seeking healthcare until one is really truly sick ('mustn't complain. stiff upper lip').

Here in Britain they are having big problems with the NHS, but those problems are directly comparable to problems in the US: lawyers and politicians (read HMO managers and financial experts) are attempting to define what makes good healthcare. Health care decisions not made by medical professionals will not be optimal no matter who pays.

That being said, my birth control is free (as opposed to $25/month - equal to two weeks of food!). I never have a co-pay and all prescriptions cost 7 pounds. Medical students' education is government subsidized and in consequence, all UK trained doctors are required to give back a certain number of years working for the NHS (as opposed to the more lucrative private sector). If the British government allocated as much funding/person to the NHS as is paid in the US AND if British people gave up their perennially stiff upper lips for a culture that valued yearly check-ups and preventative healthcare, I have no doubt that health care solutions would be miles and miles better here.

NOT providing affordable (or god forbid, free) health coverage to all americans is proof that US govt (and the culture in the USA) are content with a class-system that dooms whole swathes of the population to minimal health care and poor health solutions. We are creating a sicker, poorer, lower-class reminiscent of pre-20th century and we proudly say that their problems aren't ours and we shouldn't have to subsidize their care. Why the hell not? Call me an idealist, call me a socialist (since I am), call me naive but I think that allowing people to die for lack of treatment simply because you or I are not willing to give up a little of our own wealth is criminal neglect and undermines the fundamental Constitutional dictum that all people are equal. By denying adequate healthcare to a large proportion of the US population (we'll leave the rest of the world out of this for now) we are simply reiterating the selfish and self-serving Orwellian construct that some people are more equal than others.

Posted by Expat_UK | February 9, 2008 6:25 AM

"The government -- both parties included here -- has never, ever, in history solved a problem without throwing bottomless pits of money at political friends."

Dear Right, the government ended mass poverty among the elderly through social security and medicare. It brought electricity to rural America, soil conservation to the dust belt, and higher education to millions of World War II veterans. It ended legal segregation and guaranteed voting rights for all Americans. Under Clinton or Obama, it will bring medical insurance within the reach of everybody. And what does your crowd offer, other than inequality at home and endless war abroad?

Posted by Stan | February 9, 2008 8:35 AM

Re: Even if behavior were just a large factor, rather than the biggest one, it would still be unjust to ignore it in the pricing of health insurance.

But a better way to handle this is not to price behavioral factors into health insurance premiums but rather into the costs of the behvaior itself. We already do this with tobacco and alcohol; we could do it with other unhealthy products as well. A 5$ bag of potato chips is mor likely to deter junk food eating than higher insurance costs.

re: U.S. cancer survival rates, for example, are the best in the world.

I've seen this repeated a lot, but I'd like to know what's behind it. Earlier diagnosis? Prolongation of the terminal phase? Moroever you should also be aware that the US ranks dead last among first world nations in surviuval rates for major preventable diseases in general.

re: By that reasoning, we shouldn't charge people with really bad driving records any more for car insurance than people with really good driving records.

When a person suffers a traffic accident that is not their fault their auto premiums do not go up no matter how expensive the claim. Health premiums go up for people whose health problems are genetic or purely accidental.

Posted by JonF | February 9, 2008 8:56 AM

Mixner: Let me know when you get inoperable cancer. I'll be sure to laugh my ass off, tool.

Posted by Leon | February 9, 2008 9:15 AM

This is from a commenter at Clive Crook's blog. It explains the nicely the problem with relying on 5-year cancer survival rates

The guidelines for screening and practices are different in the US and in the UK, this is a known fact. It is not clear how much of an improvement in mortality is due to these differences.

Comparing ratio of cured to diagnosed is not an acceptable metric when one group is screened a lot more than the other; mortality rate per population is.

For example, regardless of whether PSA works or not, there is no question that it results in overdiagnosis: detections of very early and very slow growing cancers (or even non-progressive cancers) that would've never spread in one's lifetime if left undetected. Overdiagnosis is a known and accepted side effect of screening, the only question is how large it is, but it is very difficult to estimate (for example estimates for mammography vary from 5 to 40%). Since non-progressive cases are easy to cure as they wouldn't have spread anyway in one's lifetime, this increases the ratio of cured to detected by adding the same positive number to both nominator and denominator and makes the comparison of survival rates as a percentage of cured/detected invalid. If you don't understand it, consider this: let's say you have 8 cases detected in the UK and 4 cured. Let's say because of early detection you have 2 additional cases detected in the US, but as they never become apparent in person's lifetime they represent overdiagnosis. We'll get the ratio of cured/detected in the UK as 4/8 = 1/2 = 50% and the ratio of cured/detected in the US (4+2)/(8+2) = 6/10 or 60%. The US number looks better, but the number of people who die is still the same - 4. Not all additional cases detected in the US may be overdiagnosis, but this example shows how misleading the comparison between survival rates can be.

Additionally, it is ridiculous, to compare 5-year survival rates when in one case there is a lot more screening than in the other. If you don't understand it - here is a simple example. Let's say two men died at age 50 from cancer, in one case it is diagnosed early at age 43 because of a test, in another - at age 47. Who is better off? The end result is the same, but the former survived for more than 5 years. This is called lead-time bias, and this is another reason why overall mortality reduction is the only acceptable measure of a screening test effectiveness in randomised trials.

Posted by Peter H | February 9, 2008 10:32 AM

Peter H,

Like I said, five-year survival rates are an misleading measure. Dr. Gilbert Welch, a Professor at Dartmouth Medical School, wrote about this in the LA Times [stuff about cancer mortality rates]

Once again, you are confused. What we're trying to measure is the performance of the health care system, remember? Cancer mortality rates, like all other aggregate mortality rates, are meaningless as a measure of health care system performance because they depend crucially on the rate at which cancer occurs in the population. That rate is determined primarily by variables that have nothing to do with the health care system--genes, diet, exercise, smoking, exposure to environmental carcinogens, etc. The rate at which cancer occurs in the population tells you nothing about the health care system. Cancer survival rates, in contrast, are a direct measure of the effectiveness of the health care system at diagnosing and treating cancer. The significantly better cancer survival rates in the U.S. indicate that the U.S. health care system is much better than European health care systems at diagnosing and treating cancer.

That's not necessarily true. Prostate cancer is the best example of this: many more people have prostate cancer than will actually experience symptoms of it, much less die of it.

It is true. Prostate cancer is only a small minority of all cancers, and even prostate cancer is potentially fatal. For most cancers, early detection is crucial to the chances of successful treatment, and to the degree of physical and mental trauma the treatment is likely to impose on the patient. The longer cancer remains undetected and untreated, the more likely it is to spread and to become more difficult to treat, or completely untreatable.

Posted by Mixner | February 9, 2008 1:19 PM

Peter H,

I specifically said that the study should be taken with a grain of salt, since it didn't account for lifestyle differences and other factors. What I said was America's last-place finish provides prima facie evidence that the United States does not have superior health care to the rest of the world.

No, the study isn't prima facie evidence of anything about the health care system. The failure to control for any of the myriad variables that influence the rate of "amenable mortality" in different countries means that it tells you nothing whatsoever about the role of health care. It's like trying to rate the quality of a car's tires by looking at its fuel efficiency. Unless you control for all the other factors that affect the vehicle's fuel efficiency (the weight of the vehicle, the size of the engine, the drag coefficient, the mechanical efficiency of the transmission, etc.) you can't draw any conclusions whatsoever about the effect of the tires on that number. As a measure of health care system performance, "amenable mortality rate" isn't just flawed, it's worthless. In order to measure the quality or performance or effectiveness of the health care system, you have to look at data that measures "outcomes" that are actually attributable to the health care system, not data that measures overall levels of health or mortality in a population. Cancer survival rates are a prime example of data that measure outcomes attributable to the health care system, and they show a clear advantage to the U.S. health care system over the health care systems of Europe.

This is from a commenter at Clive Crook's blog. It explains the nicely the problem with relying on 5-year cancer survival rates

The comment refers specifically to screening for prostate cancer. Prostate cancer is highly unrepresentative of cancer in general because it is often slow growing, most often occurs in older men, and treatment can cause permanent disability (impotence). The government has concluded that there is insufficient evidence to recommend either for or against routine screening for prostate cancer. Physicians are advised to carefully explain the pros and cons of screening and treatment to their patients. There is no scientific consensus at present that the higher rate of prostate cancer screening in the U.S. is either beneficial or harmful. And again, prostate cancer is only one type of cancer. The U.S. cancer survival rates are significantly better than European rates for most types of cancer and for cancer overall, not just for prostate cancer.

Posted by Mixner | February 9, 2008 1:56 PM

JonF,

But a better way to handle this is not to price behavioral factors into health insurance premiums but rather into the costs of the behvaior itself. We already do this with tobacco and alcohol; we could do it with other unhealthy products as well. A 5$ bag of potato chips is mor likely to deter junk food eating than higher insurance costs.

No, we don't. Tobacco and alcohol taxes are not used to subsidize non-smokers and non-problem-drinkers for the higher health care costs caused by smokers and problem drinkers. Unless and until taxes or some other mechanism is put in place to do this, insurance companies are entirely justified in charging higher premiums to smokers and problem drinkers, commensurate with their higher level of risk.

I've seen this repeated a lot, but I'd like to know what's behind it. Earlier diagnosis?

Earlier diagnosis is a large part of it. The earlier you diagnose cancer, the better your chances of treating it successfully. According to the author of the study, the U.S. is better at both diagnosing and treating cancer because it invests more in diagnostic and treatment facilities, has better-trained, better-organized and more-highly-skilled medical professionals, and makes better use of evidence-based guidelines.

Posted by Mixner | February 9, 2008 2:16 PM

Expat,

Your discussion is also very confused. You allege, for example, that the U.S. "denies" "adequate healthcare" to a "large proportion" of its population. But what is "adequate" health care? How do you decide? And how large a proportion of the population is denied it?

Britain's National Health Service is possibly the very worst health care system in Western Europe. Does the NHS provide "adequate" healthcare to even a majority of its population? It has some of the worst cancer survival rates in Europe. Only 40% of British cancer patients ever even get to see an oncologist. Just a couple of weeks ago, I read in the Economist that around 40 new drugs are expected to be licensed in the next few years to treat cancer, and that the NHS is expected to refuse to pay for most of them. Is that "adequate" health care?

Frankly, if I had a serious health problem, I would much rather be even an uninsured American than a Briton left to take my chances with the crappy health care provided by your NHS. At least in the U.S., I would have access to a large network of public and private programs and clinics that would provide me with free or heavily-subsidized care at American standards, and as a last resort I could qualify for Medicaid.

Posted by Mixner | February 9, 2008 2:55 PM

But no, I don't think society would collapse if people who suffered serious automobile injuries weren't charged extra on their premiums. Having a punctured lung is punishment enough. Call me a softee.

It's not that you're a "softee"; it's that you don't grasp the concept of insurance. Charging higher risk people more is not "punishment." It's the very essence of insurance.

Posted by David Nieporent | February 11, 2008 8:52 PM