« "Stabbed in the Back and Sold to Russia" | theCleverBulldog's Blog | Swords, Shields, and Stupidity »
Lies, Statistics, and the 45,000 dead from lack of insurance
The shocking headlined screamed "45,000 die each year from lack of health insurance". WOW! That sounds bad. Now let's see how they came up with that.
from :http://www.boston.com/news/health/blog/2009/09/uninsured_hold.html
"Researchers from Cambridge Health Alliance report in the American Journal of Public Health on a study that followed 9,005 adults under 65 years old who took part in a national survey conducted by the Centers for Disease Control and Prevention from 1986 through 1994. After 12 years, 351 people had died. Sixty of them were uninsured and 291 were insured. "
From this, and based on the estimate of 40 million uninsured, they extrapolated 45,000 deaths per year. Nowhere did they investigate the cause of the deaths, they simply attributed all deaths to lack of insurance to get the shocking headline number. So, using their logic, the 291 others deaths in the study must be attributed having insurance. Therefore having insurance increases your risk of death by 500%! YES! Almost a quarter of a million killed each year because they have insurance, and Obama wants EVERYONE to have insurance? Aren't enough people dying already? What kind of a monster is he?
(For the sarcasm impaired, I realize this is absurd and a complete misrepresentation of the data. The point is that the original 'study' is just as absurd, but no one bothered to look critically at the data because it supported their agenda.)
from :http://www.boston.com/news/health/blog/2009/09/uninsured_hold.html
"Researchers from Cambridge Health Alliance report in the American Journal of Public Health on a study that followed 9,005 adults under 65 years old who took part in a national survey conducted by the Centers for Disease Control and Prevention from 1986 through 1994. After 12 years, 351 people had died. Sixty of them were uninsured and 291 were insured. "
From this, and based on the estimate of 40 million uninsured, they extrapolated 45,000 deaths per year. Nowhere did they investigate the cause of the deaths, they simply attributed all deaths to lack of insurance to get the shocking headline number. So, using their logic, the 291 others deaths in the study must be attributed having insurance. Therefore having insurance increases your risk of death by 500%! YES! Almost a quarter of a million killed each year because they have insurance, and Obama wants EVERYONE to have insurance? Aren't enough people dying already? What kind of a monster is he?
(For the sarcasm impaired, I realize this is absurd and a complete misrepresentation of the data. The point is that the original 'study' is just as absurd, but no one bothered to look critically at the data because it supported their agenda.)
Advertisement
I get the sense you really didn't read the studies. Or you would have discovered they did take into account a great many factors to reach their conclusion. That those without insurance have a have a high mortality rate than those that do have insurance. Just to be clear they are saying these are excess deaths. In other words deaths that may have been avoided.
http://www.ajph.org/cgi/content/abstract/AJPH.2008.157685v1
September 21, 2009 3:25 PM | Reply | Permalink
I read their study, and statistically it is total crap. It was no different than the old line about eating carrots causes death, because 90% of all people who died had eaten carrots within 6 months of their death. I'm not arguing that lack of preventative care doesn't affect health and mortality rates, I'm just saying that the study is flawed, that it was designed to produce the result it produced, and was designed to support an agenda. It was not a real study.
September 21, 2009 3:44 PM | Reply | Permalink
Is it crap because it is or because it doesn't fit your agenda? Knowing your politics my opinion it's the latter. Funny how an agenda can cut both ways. My sense is the Harvard medical review has no axe to grind and is merely pointing out facts as they see them.
However I think we can agree that 1,000s die every year because they have no health insurance and cannot afford to go to the Dr. So whether the 44,789 is dead on accurate extrapolation of the study is a point, but not the point you should probably be taking away from the study.
September 21, 2009 4:29 PM | Reply | Permalink
Multiples more die each year from cancer, so should we spend our trillion dollars to insure 45,000 people or spend it on cancer which kills half a million people each year?
September 21, 2009 8:10 PM | Reply | Permalink
This is one of the sillier posts to appear on TPM in recent days. Epidemiology is a sophisticated science, and those who conducted the study were careful to compute the rate of excess deaths associated with insurance lack, and then to compare insured and uninsured individuals to determine whether the two populations differed in other ways as well that might explain the excess among the uninsured. These other ways are known as "confounding factors", and all plausible ones appear to have been excluded - income, pre-existing health status, etc.
Conceivably, some totally unexpected "other" variable might have explained the results, but in terms of probability, it is far more likely that the excess was almost entirely attributable to the lack of insurance, and close to 100 percent likely that at least a substantial part of the excess reflected that lack.
It takes an extraordinary ideologial bias, in my view, to try to explain away the conclusions, and to see that bias in this post is unfortunate.
September 21, 2009 3:46 PM | Reply | Permalink
Two other studies put out different figures. 18,000 from an earlier IOM study, and 101,000 annually from a London School of Public Hygeine study.
Regardless which study is used, tens of thousands of Americans die every year due to lack of health insurance... the studies all give approximately the same order of magnitude. It's solid. It also makes sense, in that the story where it makes absolutely no difference and nobody ever died from lack of health insurance is as likely as Hannity voting for Obama.
September 21, 2009 4:10 PM | Reply | Permalink
No one dies from a lack of insurance, they die of disease, old age, accidents or other causes. Does lack of insurance mean lack of treatment? No, since hospitals cannot refuse emergency treatment. Does it mean less preventative care? Maybe. Does that result in a higher death rate? How can you tell?
September 21, 2009 4:53 PM | Reply | Permalink
Bulldog, you should probably give up before you embarass yourself further. Of course the uninsured can die from lack of treatment if they delay treatment until an illness is severe enough to warrant an Emergency Room visit. In some cases, that's too late. This is not uncommon among uninsured individuals.
You should probably also go back and correct your arithmetic errors in calculating the number of insured and uninsured deaths. I realize this is an area you don't understand well, but the arithmetic part is pretty elementary.
September 21, 2009 5:04 PM | Reply | Permalink
The Dog's point being that they could have just as easily have died from a shitty diet as from lack of health insurance.
This is the same tone you took with the immunization debate with the same predictable results.
Doctor, heal thyself.
September 21, 2009 5:11 PM | Reply | Permalink
That's not correct, Jason. All reasonable consequences of dietary variation - diabetes, BMI, etc., - were included in the analysis and don't explain the results. The evidence that lack of insurance was the nearly exclusive cause of excess mortality is reasonably good. The evidence that it was the predominant cause is even more compelling, in that all plausible alternatives were excluded, but anyone interested can visit the link above to draw his or her own conclusions. There is very little likelihood that insurance lack was not a major factor, and probably more or less the entire explanation, although minor contributions from other factors an never be completely excluded in an epidemiologic study. Certainly, dietary diffrerences as an important factor can be excluded with high certainty.
September 21, 2009 5:32 PM | Reply | Permalink
I was more speaking of the way in which you disagreed rather than the specific disagreement. Your agreement with the study's findings isn't all that is needed to determine its relevance to the population at large.
I went to the link and I think they give dietary differences as short a shrift as the current medical reform efforts have given the problem - which is to say little if any at all. Not to mention the way they excluded people from the study seemed a little heavy handed.
What about the questions they didn't ask? What about the relative affluence of people who have insurance versus those who don't? How does lack of income affect their diet and hence their overall health?
What TCB was getting at, and what I agree with if that was his point, is that all numbers can be manipulated and interpreted to back up whatever preconceived notions we have of a given issue.
September 21, 2009 5:58 PM | Reply | Permalink
Numbers can are are manipulated, and it can't be a coincidence that this 'study' just happened at the time of the big health care reform debate. It is simply not valid to use this data to come to a conclusion about insurance, since the causes of death are not categorized. Pancreatic cancer kills regardless of insurance, regardless of treatment. It is just not valid to assume that people are anything other than unique, and the numbers in this analysis were too small to be significant.
September 21, 2009 7:13 PM | Reply | Permalink
Is Mr. Moolten a doctor?
September 21, 2009 8:37 PM | Reply | Permalink
http://www.youtube.com/watch?v=g1ZFNOAL-Tg
September 21, 2009 8:42 PM | Reply | Permalink
Yeah. He is a cancer researcher. Not sure if that makes him am MD or a PhD, but I am pretty sure there is a big brain tucked away in there.
September 22, 2009 6:03 AM | Reply | Permalink
A set of statements based entirely on supposition and not on fact. You may be right, but you have no analysis to prove it.
September 21, 2009 7:15 PM | Reply | Permalink
For those interested in how this type of study collects and analyzes data, here is the link to the study:
http://pnhp.org/excessdeaths/health-insurance-and-mortality-in-US-adults.pdf
September 21, 2009 4:55 PM | Reply | Permalink
You've never heard the expression "lies, damned lies and statistics"?
September 21, 2009 8:09 PM | Reply | Permalink
I won't address in detail every question raised above, but rather invite anyone interested to review the data, which should itself answer those questions.
Epidemiologic studies, by their nature, can never formally prove cause/effect relationships with 100 percent conclusiveness, but this study is robust enough, and its methodology standard enough, to conclude with very high certainty that lack of insurance was at least the predominant causal factor responsible for the excess death rate, if not the only factor. One can never exclude a minor influence of some other unsuspected variable, but the study successfully excluded a major role for baseline health status, income, smoking, alcohol, dietary and metabolic abnormalities, as well as other potentially relevant variables that might reasonably be considered.
Although there is little doubt that the data demonstrate a cause/effect relationship between insurance status and the risk of death, there will undoubtedly be room to quibble about the accuracy of the quantitation - is the excess truly 45,000, or is it 30,000, or 60,000? Given the multivariate analysis and the multiple adjustments, some imprecision is likely. However, the excess is very unlikely to be a small number, and thus remains a matter of concern.
Finally, 45,000 excess deaths annually is relevant to the risk we impose on illegal immigrants by denying them insurance coverage. Rough assumptions about their number as a total of the 46 million Americans who are uninsured implies that probably one to several thousand of these illegals die every year due to the lack of insurance. It may be a political necessity to inflict this on them, but it remains a moral failure of our society to add the death penalty to existing penalties for immigration law violation.
September 21, 2009 8:26 PM | Reply | Permalink
This is the same sort of reasoning you used for supporting wide-spread vaccinations for HPV, which affects such a ridiculously low percentage of the female population in its more virulent forms that vaccination without more long-term studies is irresponsible.
You suffer from the inverse of the Luddites - too much confidence in science by way of neglecting simple, common sense questions. Somewhere along the way from there to here, we stopped accounting for inherent preference for certain facts as a way to gauge our own objectivity.
Being suspicious of information that backs-up my existing beliefs is an automatic reaction for me.
September 22, 2009 6:34 AM | Reply | Permalink
"Two other studies put out different figures. 18,000 ... and 101,000 annually..."
Given the wide gap here, it's fair for Bulldog to question the assumptions in the studies. That's what any peer review panel does, and what Fred has defended well. Is the article hyperbolic? From Bulldog's perspective, yes. Those are "scare" numbers. Of course there is an agenda.
"Regardless which study is used, tens of thousands of Americans die every year due to lack of health insurance..."
It's also semantically correct to point out that no one is actually killed by lack of health insurance.
Bulldog, meet Big Picture:
People who have insurance are much more likely to get HEALTH CARE, care that ends up saving or prolonging their lives, than those that are not insured.
So how do you want to spend your tax dollars?
Now for MiddleClassBill:
"Multiples more die each year from cancer, so should we spend our trillion dollars to insure 45,000 people or spend it on cancer which kills half a million people each year?"
No, I'd like to spend it like this. Check out Dr. Dean Ornish' testimony in the senate:
help.senate.gov/Hearings/2009_02_26/Ornish.pdf
I'll say this over and over again. Insurance is not the goal here. It is health care. But unless this country can get to a single-payer system, insurance is going to be involved somehow. So let's make it work for as many as possible, start with prevention, and have a public option for the gaps.
As my neighbors say, what's wrong with that, eh?
September 22, 2009 12:08 AM | Reply | Permalink
I have great health insurance and haven't been to the doctor in years. The last time I tried, I sat in the hospital for six hours and still left without being treated. Not my insurance company's fault I didn't get care then either.
I think you missing the crux of his argument which is that scary numbers don't convince people - they just make it seem like there is something to hide - and this country is much too diverse to use broad-brush analysis of this sort to form our ultimate conclusions.
There was no statement about how TCB feels about health insurance reform, just that he doesn't like to be manipulated for political ends.
September 22, 2009 6:39 AM | Reply | Permalink