Is life expectancy a flawed measure of health care quality?
theincidentaleconomist.com
theincidentaleconomist.com
That's ok, I'm politically incorrect. I'll stick to the comparison between the US and Japan, since numbers are easy to come by[1]. On the face of it, it looks as if Japan has us beat: 82.6 years to 78.4 years.
Now lets break things down by subgroup:
People of Japanese descent: 84.5 years in the US, 82.6 years in Japan. This long lived group makes up <5% of the US (too lazy to look up exact #, but east Asians in total are about 5% of the US), and about 99.4% of Japan.
http://minorityhealth.hhs.gov/templates/browse.aspx?lvl=2...
People of European descent: 78 in the US, unknown (but irrelevant) in Japan. This group makes up about 80% of the US, and < 0.6% of Japan.
People of African descent: 72 in the US, unknown/irrelevant in Japan. This short lived group makes up about 12% of the US, and as far as I'm aware Japan only knows about this group via video games.
http://www.cdc.gov/nchs/data/nvsr/nvsr53/nvsr53_15.pdf
So if the US had the same demographics as Japan, our life expectancy would be 2 years longer.
[1] Japan is the only industrialized nation with the property that the (possibly great great) grandchildren of Japanese immigrants is classified by the census a "Japanese American".
Probably the best measure of heathcare quality you're going to get is looking at survival rates from specific treatable diseases whose incidence is largely uncorrelated with climate/genetics/lifestyle.
*i) [free] accessibility to specialist advice from medical practitioners should improve the number of people willing and able effectively managing their weight/cholesterol level etc (ii) healthcare spending can include preventative measures such as campaigns to encourage healthy behaviour (which depending on their efficacy may be be an optimal use of resources, particularly for a heathcare provider that views treatments as an overhead rather than a revenue source)
What if someone was to add something like "People of Japanese descent: $85 000 average family income in the US, $50 000 in Japan." (I've made up the numbers from personal experience) Obviously richer ppl live longer.
In any case, I agree with you. Income is a major effect, although it's far from clear that this is due to income buying health care. In fact, most of the evidence suggests health care has nothing to do with it. For instance, during the RAND experiment, richer people were healthier than the poor regardless of which category they were randomly dropped into. (Observational studies since then draw less clear conclusions, since observational studies are crappy in comparison to randomized experiments.)
Another fun fact. The shortest lived people in the US are the Lakota, who live 35 years less than Japanese Americans (not a typo, life expectancy = 50). This group gets free health care via IHS (I believe it is structured similarly to Britains NHS), so it seems that there are other factors at work.
Huh, imagine that. Life expectancy is, in fact, not a great measure of health care, since it is affected by so many other factors.
IHS is pretty much a horrorshow. It's one of the best arguments against govt-run healthcare.
It doesn't much matter if some govt somewhere can run healthcare well. The US govt runs healthcare badly. Fix that first.
Look, this is not to say that it is invalid, far from it. It’s just that I think you have to be a lot more careful. It could be that Japanese vs Japanese Americans is one of those Apples vs Oranges comparisons. Are Japanese Americans really like Japanese?
I may only be comparing lemons to oranges, which I agree is only marginally better than apples to oranges. But that's ok - I'm certainly NOT trying to make the claim that US health care is better than Japanese health care as a result [2]. I'm simply pointing out that if you include slightly better controls, the effect completely reverses itself. This strongly suggests the original comparison is crappy.
[1] Genes matter. A fact I've heard - Indians have narrower blood vessels than non-Indian Whites, leading them to be more vulnerable to hypertension. The UK NHS takes this into account when giving medical advice, though supposedly US doctors do not (that would be racist).
[2] To draw conclusions like that, you need to take groups of statistically similar people with the same disease and measure health outcomes. No one does this, it's very hard.
Why don't you go up to a Japanese American and ask, "hey, how long have you been in the US?" (Be sure to duck immediately after saying it.)
I've always felt that the life expectancy stat only works in a statist, hyper-controlled system. Once you let people make personal decisions about their health, the utility of the stat declines.
The scientists in this article try to distinguish health-related from non-health-related causes of death. But even our health-related causes of death are not exactly under the control of the medical establishment. People eat food as if they had an active death wish.
I wonder if an argument could be made that given the USs propensity towards unhealthy lifestyles that the US is in fact great at healthcare because the life expectancy is as high as it is?
Finally, the best measures I've seen to date on healthcare quality are looking at individual outcomes. For example, if you get prostate cancer what are the odds of living in the US versus other countries. I can't find the exact article now, but various countries do better at various health issues with the US leading many of them.
http://www.webmd.com/cancer/news/20080716/cancer-survival-ra...
The highest survival rates were found in the U.S. for breast and prostate cancer, in Japan for colon and rectal cancers in men, and in France for colon and rectal cancers in women, Coleman's team reports.
This is no argument to be made about the US having great health care. I don't know how a place where a random illness can wipe out your and your family's finances can even be considered first world.
To address your specific claim "we must be good because we make horrible life choices and don't die!" I would point at the UK and alcohol consumption. Their bad life choices are driving costs up to the point that they'll probably have to make some big changes to deal with them but so far the system is doing a reasonable job of keeping them from dying from it.
I didn't claim anything and simply wondered aloud. Also, the article was about life expectancy and quality of healthcare and spoke nothing about costs.
Your statement does bring up another question. Would that same hypothetical family prefer to have their finances ruined or have their loved saved from some random illness? Of course the best solution is perfect care at zero cost, but I think we can agree that it is impossible. If the US is actually getting better health outcomes than other systems perhaps the cost is worth it?
And to address your specific claim, I've always thought that the government should provide health related catastrophic fallback insurance similar to the way they do for floods. I don't think the government needs to be involved in the day to day healthcare or insurance matters, but I agree that getting cancer for example should not ruin your families finances.
False dichotomy. The US isn't so much vastly better that the choice is go there and be saved or stay in e.g. France and die. In fact all the sources I've shown seem to indicate the US isn't the best even if you can afford it. Keep in mind that doctors tend to make 3 times more than the average salary in their European "socialist" countries, just not 5 times like their US counterparts.
>but I think we can agree that it is impossible
Of course, it always costs, the question is just how much waste is going into giving huge bonuses, etc., to CEOs of insurance companies who make even more money by cutting people's coverage (insurance is a profitable business in its own right, these guys just want even more).
>If the US is actually getting better health outcomes than other systems perhaps the cost is worth it?
It it were we could discuss it. The reports from both the WHO and Berkely indicate pretty clearly that this isn't the case.
>And to address your specific claim, I've always thought that the government should provide health related catastrophic fallback insurance similar to the way they do for floods. I don't think the government needs to be involved in the day to day healthcare or insurance matters, but I agree that getting cancer for example should not ruin your families finances.
This sounds like something reasonable to look at. Just keep in mind that so many people are living paycheck to paycheck that even a surprise bill of $5k can be enough that they never recover from it.
This is relevant to this particular article, because he may be missing subtle, but important factors that a lay-person would miss. For example, see yummyfajita's post in this discussion.
Unless, of course, they agree with me, in which case they are simply wise academics trying to share their wisdom with the unwashed masses. :)
There is a logical problem here, and it's much deeper than life expectancy. It's a problem with this form of argument. The problem is that, if you use statistics to make political arguments, you're going to be arguing forever. Whatever this guy wants to do -- and god bless him I hope he gets his way -- next year the same guy will be using statistics to tell us to do something else. That's because you can slice statistics in an almost infinite amount of ways, and I can guarantee you that one of those ways will show some kind of imbalance or injustice that needs addressing.
That's all fine and dandy if the entire country could turn on a dime. I wish it could but it can't. Not reality. Instead we live in a system where it takes decades to make even one change, and it's extremely rare to stop doing things that don't work -- we just do more new stuff. I hate to rag on the IRS, but take a look at the tax code. It's so bad that I'm willing to wager nobody alive knows how to file taxes exactly correct.
This puts us in an infinite cycle of where we're always measuring things with statistics, finding some social cause worth pursuing, painfully making changes, then making more measurements. Each time through the loop the system gets more and more unwieldy and complex. It's a feedback loop that increases government dysfunction.
So instead of arguing about one cause or another, I'd like to suggest a rule for such discussions: if you want to make a case for some kind of social policy change, then make it in terms of a trade-off from whatever the existing status quo is. I realize there are a lot of problems with this suggestion, but at least it starts putting things in terms of trade-offs instead of wish-lists. Just from looking around, it seems to me that the real world is a lot more about trade-offs and compromises than moral crusades and justice. But that's just me. The only thing that I am observing is that this pattern can be repeated indefinitely, and each time through the loop conditions will continue to deteriorate. It's reactive leadership by statistics, and that almost never works.
Ok! I moved to Canada (Quebec) from the US. The healthcare system is overloaded, there are weird things like a) it's extremely hard to get a family doctor, b) you have to go to a clinic early in the morning to get a number + time of day and then come back at that time to be seen, and c) you still need to pay your own way (or get insurance) for dental / vision. You have to wait weeks/months to have a surgery, but you have to do that in the US and then promptly file for personal bankruptcy.
That's still a whole hell of a lot better than a random unexpected illness or injury sending an employed, insured person into bankruptcy. Seeing a doctor is now a possibility instead of a dangerous worst-case scenario that could ruin my life. No more hoping the pain/infection/whatever just goes away on its own.
Thus, I still think moving towards a government program / single payer / whatever you want to call it system is the best idea. Now, I based this all on personal anecdotal experience and not statistics, so I wonder if anyone will actually agree with me. :)
There are many things messed up about healthcare in the US, but I've never had to wait for it. I had an emergency appendicitis surgery and from when I went to the doctor to when I was put under was about 2 hours.
I've also had a non-emergency knee surgery (replaced ACL). I went to the doctor one day and had an MRI that same day. I had to wait 2 days for the MRI results, but when they came back positive for an ACL tear I had the option of surgery 2 days later. I opted for 2 weeks later so I could get work all set up for me to be out a couple weeks.
Now, I have insurance and paid a bit here and there out of pocket, but waiting for care is one thing I didn't do.
I think that may be key. I had insurance and appendicitis, and waited around 12 hours before surgery, but my brother was in a car accident without insurance, and was almost sent home. Luckily, a family member who is a nurse caused enough of a storm that he received the proper care.
Not having insurance is what really skews the statistics in the U.S. Cook County hospital is an example of a hospital which provides care for the uninsured, and the wait times are significant.
I haven't seen it, but I'd be interested to see the statistics on health care in the United States for the uninsured versus the insured: Wait times, costs, quality of care, quality of life, and life span.
Do you have any evidence that this occurs much in the US?
http://www.businessweek.com/bwdaily/dnflash/content/jun2009/...
Basic definitions: a bankruptcy with $1001 or more in medical spending in the previous 2 years is a "medical bankruptcy". By this criteria, Michael Vick's bankruptcy was medical, even though any sane person would realize there were other causes [1]. Also, "uncontrolled gambling", "drug abuse" and "adoption/birth" are medical causes of bankruptcy, by the definition of that study.
If we adopted the criteria of that study, we would discover that nearly every bankruptcy has "major tax cause", for example.
By the way, a slightly less misleading way to phrase the conclusions of that study would be "medical bankruptcies drop by 33%, even while medical costs increase." Luckily, the language of the study hid that from lazy reporters. See this explanation: http://www.theatlantic.com/business/archive/2009/06/elizabet...
Try again.
[1] Getting arrested for dogfighting and losing his job in the NFL, for example. While he certainly spent a few grand on doctors, he also owed about $600k in back taxes.
[edit: oops, linked to the debunking, forgot to link to the actual study. http://pnhp.org/new_bankruptcy_study/ Thanks floor(e x 10^10).]
is the biggest factor in "medical bankruptcies".
If you can't work because you're sick, you're going to have financial problems because the "cost of life" bills (food, shelter, transportation, heat/cooling) still come in. Healthcare doesn't address this, so it's absurd to bring up "medical bankruptcies" as an argument wrt healthcare.
you link to another website instead of the study >http://www.theatlantic.com/business/archive/2009/06/elizabet...
reminds me of the old Marx joke "I would never join any club that would accept me as a member"
Why do you require a control group for observational data?
Yes, personal experience and statistics are very important in setting public policy. And heck, make the government do anything you want it to do. But that's not what I'm talking about.
Let's say you run a company with ten thousand workers. But they seem unhappy. One day a VP comes in and tells you that 30% of the workers are not able to read as much as they would like to stay current with their job.
So you, wise person that you are, implement a free book program. The next month the number hasn't changed much. So querying further, you find out that folks don't have enough time to read the books -- it's not that they can't afford them. So you implement extra time off on Fridays so folks can read.
Months later, no change. This time you bring in a comittee of experts. After much investigation, they find out that folks would rather play golf than read, and the extra time has just been spent on leisure activities instead of the reading. So you implement a program where folks can read at work on Friday afternoons.
Months later, no change. This time it turns out that the books they are reading are not job-related. So, after bringing in an expensive outside consulting team, you decide to put video cameras in the office...
What's happening? Are these people just lazy or what? Is management stupid? How many times do we have to set up the conditions where more reading can take place?
Two things have happened. First, you've confused the statistic with the individuals. The 30% who want to read more aren't necessarily the same as the ones playing golf, which aren't necessarily the same as the ones reading other books. But in your mind they're all the same.
You see, all you have is a metric, an aggregation of data. It's not like there's any one universal person out there -- everybody is unique. The system is made up of great variations in motives, ability, and circumstances. Teams of people work because they are all different, not because they are all the same.
But you'll keep on measuring, and keep on poking around, and the numbers will never do what you want them to do. Why? Because top-down, statistics-based reactive management is full of shit. Doesn't work for your company, doesn't work for other companies, doesn't work for the government.
That's not a political statement -- vote for your guys and do whatever you want to with the world. Lots of things need fixing. Let's go fix them. This has nothing to do with wanting less or more government. It's a statement on how systems of people operate.
People either understand that concept or they don't. And if they don't, they're always looking at aggregate numbers instead of system constraints. It's just a futile and endlessly spiraling methodology to fix any problem of any system of interacting people. It continues to fail over and over again, yet we keep framing the discussion this way. Probably because we have a vast lack of education in economics.
The final thing that's happened -- and this is the worst of all -- is that you've lost track of the goal for the metric. Somewhere along the way you're chasing percentages of folks reading books. What you really want is people leading happy and fulfilling work lives. But the minute you put a number on that, your abstraction starts carrying you away from your real goal. I've seen this happen over and over again, and I've never seen anybody go back to the original problem.
These are just stupid people tricks. All I'm saying is that we have to be aware of them to keep from repeating yet again for the hundredth time.
In reality, if you want to make people's lives better, you need to deal with one person and the thousand parameters around his or her life. Then you can make a difference. But once you start averaging -- even in groups as small as ten or so -- you've gone off the rails and into the weeds.