In the continuing debate about implementing a health insurance reform that works, much attention is paid to the issue of how many people have been, or are projected to be, uninsured under past legislation, or reform bills currently being proposed or actively under consideration in Congress; comparatively little attention is given to the issue of mortality associated with lack of health insurance. There seems to be a blind spot about discussing how many deaths have likely occurred, or likely will occur as a consequence of Congress’s past or contemplated actions.
People who care about reforming our health insurance system need to remove that blind spot by squarely facing up to what the consequences for mortality among the uninsured have been of failing to pass a Medicare for All insurance reform, with no co-pays and no deductibles, of the kind specified in the benefits section of the Expanded and Improved Medicare for All Act, HR 676, introduced into the current 115th Congress by John Conyers (D-MI). You can find some indications about what those consequences might be in articles, videos, and blog posts, but very few will face the stark reality of immoral and evil political choices made routinely by today’s “pragmatic” American politicians.
Context
During the run-up to passing the Affordable Care Act (ACA) in 2009 — 2010, a brief report on the results of an important study, authored by Andrew P. Wilper, MD, MPH, Steffie Woolhandler, MD, MPH, Karen E. Lasser, MD, MPH, Danny McCormick, MD, MPH, David H. Bor, MD,and David U. Himmelstein, MD appeared in the American Journal of Public Health in December of 2009. The report provided an estimate of fatalities in 2005 among the uninsured ages 18 – 64 of 44,789, approximately 45,000.
This finding was stated clearly in the report. But on p. 2294, in the conclusions section, the authors say:
Lack of health insurance is associated with as many as 44789 deaths per year in the United States, . . .
This suggests that the 2005 estimate applies to the years since that time or to the whole population under 65. I’m sure that the authors of the report meant to say no such thing. But that became the headline conclusion in most popular articles, blog posts, and youtube videos from December 2009 up to now, more than 7 years since the report appeared, and more than 11 years after the end of 2005.
The 45,000 number became a popular number of annual fatalities uncritical commentators could toss off during debates when arguing for Medicare for All, when pushing for the ACA, when critically evaluating the ACA, or Republican attempts to repeal that law. There have been critical analyses of that number of course, such as here, here, here, here, and here. But nevertheless the idea that there were 45,000 or so fatalities annually prior to the implementation of the ACA had legs and is still often quoted.
So what is the most likely number of additional fatalities among the uninsured compared to the insured in the United States from the end of 2005 to the end of 2016? Is it a constant? Is it varying? Is it falling? Is it growing?
Results of Some Studies
Not surprisingly (1) the number of such fatalities has varied over the years with the number of the uninsured, and (2) there have been variations across different studies in estimating the increase in the odds of dying in America when not insured. In some earlier work, I used the ratio of the fatality estimate (44,789) to the uninsured population estimate in the 18 – 64 group (34,975,000), or .0012806 in 2005 and Census Bureau data for the 2005 18 – 64 uninsured population (34,975,000) to arrive at the odds ratio, or one’s increased odds of dying if uninsured. That odds ratio is 1 additional fatality out of 781 people in the 2005 18 – 64 uninsured population.
Since the 2005 uninsured population over all age groups was 43,035,000, the mortality for the whole uninsured population, assuming the 1 in 781 odds ratio holds, was 55,111, rather than the 44,789 reported in most, if not all of the media. This revised mortality estimate for 2005 is also based on the assumption that the 44,789 number is for the 18 – 64 age group alone. In e-mail correspondence, Dr. Steffie Woolhandler, MD who supervised the project, confirmed this assumption, and, by implication, that a mortality estimate for the whole population of the uninsured wasn’t actually reported in the 2009 study.
A number of more recent studies of mortality were reviewed in a recent analysis by David Himmelstein and Steffie Woolhandler called “Trumpcare or Transformation” in the American Journal of Public Health. Their editorial in the Journal as reported by Physicians for a National Health Program (PnHP) offered the following figure:
The number needed to insure to prevent one death is substantially equivalent to the odds ratio assuming that the investigators in the studies cited did a good job of controlling for other possible causes of increased fatalities among the uninsured other than the condition of being uninsured. The table provides a range of 300 to 1239 in the number needed to insure to prevent one death, providing much doubt about what the odds ratio if one is uninsured.
I find it reasonable to exclude the finding of 1 in 1239 in the earliest 1993 study because it is based on the oldest data, and advances in medicine since then may have exacerbated the differences in mortality between insured and uninsured populations. In addition, the result from the 2009 Wilper et al study of 1 in 1094 probably reflects an error, since the number of uninsured in the 18 – 64 category in 2005 is 34,975,000 according to the Census Bureau, while the number of fatalities associated with being uninsured by Wilper et al was 44,789, an odds ratio of 1 in 781.
If these adjustments are correct, then the Wilper et al, Finkelstein et al, and Sommers et. al 2014 studies are reasonably close to one another at 1 in 781, 1 in 769, and 1 in 830, respectively. But, the Sommers 2016 study, among other things, found:
Mortality declines were closely linked to county-level changes in insurance coverage, with one death prevented annually for every 239 to 316 adults gaining health insurance, which implies a cost per life saved that compares favorably to the standards used to evaluate existing public policy interventions.
This seems to be the basis for Himmelstein and Woolhandler’s statement that the odds ratio from the Sommers study is 1 in 300. But that finding is sufficiently at variance with results of other studies to make me skeptical. If the 1 in 300 ratio annually were correct, and if the uninsured number was constant at 45,000,, then mortality would have been roughly 150,000 annually during 2005 and the years immediately following.
So, at least until further studies are done, the 1 in 300 ratio seems excessive and a range for the odds ratio of 1 in 830 to 1 in 769 seems more plausible, and I’ll continue to use the 1 in 781 ratio based on the Wilper et. al. study below.
Further Results: The Butcher’s Bill
Here’s the butcher’s bill presented to us by the Obama Administration in consequence of its decision to take HR 676, Medicare for All “off the table” in 2009. Of course, this doesn’t take into account any side effects of that decision due to the rise of the tea party and the “alt-right” in the wake of the struggle producing the ACA, and the further struggle against it by the revanchists of the right since its passage.
Table One: Estimated Fatalities Associated with Lack of Health Insurance, 2009 – 2016
Third, it provides the total of estimated fatalities associated with lack of insurance for the period 2010 – 2016. The total, 356,049, omits fatalities for 2009 because I think the Obama Administration needed a year, but no more, to pass and implement the 2009 version of the Conyers – Kucinich HR 676 bill. So, its decision to take HR 676 “off the table” leaves it partially responsible for mortality associated with lack of insurance from 2010 – 2016, but not before.
The detail in the table shows that mortality is not constant at 45,000 as the media assumed, but instead increased slightly in 2010, then decreased slightly both in 2011 and 2012, before decreasing substantially in 2013 and 2014, less substantially in 2015, and finally only slightly in 2016. This pattern was built into the Affordable Care Act, which was implemented over the period from 2010 – 2014 only very gradually to create projected “deficit neutralty” over a decade by CBO. So, deficit neutrality and faux fiscal responsibility was prioritized over fatalities, bankruptcies, family breakups, and other negative outcomes of lack of insurance facilitated by our ever faithful public servants in Congress and by a President on his way to proving that he was anything but the second coming of Franklin Delano Roosevelt and a Second New Deal.
Table Two: Projected Fatalities Associated with Lack of Health Insurance, 2017 – 2024
The projected butcher’s bill under Trump in Table Two shows that it is likely to be worse than Obama’s, assuming the following. 1) Trump finally gets an Act on the order of the recently rejected AHCA out of Congress this year. 2) CBO projections (see P. 32) about the number of uninsured are correct. 3) the odds ratio of 1 in 781 is pretty accurate and also constant over time. 4) Census Bureau population projections are pretty close to the mark. And 5) Congress stays Republican and Trump gets a second term.
Outcomes are subject to change since only 4) is highly likely. They could get much worse, depending on who wins the power struggle in the Republican Party, and the whole safety net is eliminated or privatized.
Table Two shows big jumps in the uninsured over the next 4 years and slower increases thereafter resulting in a 65% increase in mortality through 2024. This is a substantial increase over the ACA period. But also keep in mind that projected additional deaths compared to Medicare for All, no co-pays, no deductibles, for the next 8 years are 467,419, a very big number, larger than the number of American military deaths in World War II (405,000).
Mysteries and Moral Monsters
There has been evidence since the 1990s that mortality associated with being uninsured is appreciably greater than among the uninsured, and this suggests some mysteries. Among those who accepted the result of the 2009 Wilper et al study and its estimate of 45,000 annual fatalities in 2005, why was there little disposition among them to frame the fight for enhanced Medicare for all as a fight for the lives of at least 45,000 Americans going forward?
Why was there little disposition to present these results in a time series table to emphasize the price in deaths that would be paid for the next 4, 8, or 16 years unless Medicare for All was passed, and to have compared those projected deaths with the number of military deaths in World War II, or the number of deaths suffered as a consequence of the 9/11 attacks, or other comparisons that would have brought home to people the moral issue involved in failing to pass Medicare for All? Wouldn’t that have been the most powerful way to frame the campaign for health care reform?
Why was there no disposition to question the results of the study by simply noticing that the estimate was for 2005 only and that mortality associated with lack of insurance would be likely to increase along with population as the years pass? Why was there no disposition to notice that since the population of the US increased by 11 million from 2005 to 2009, the number of uninsured was likely to have increased by at least thousands of souls annually?
And in the runup to the ACA, when CBO and the Democrats were claiming that 33 million more Americans would be covered by the ACA than before, why did so few people point out that even then that the ACA would not be fully implemented until 2014 and that in 2010 – 2011 it would have almost no impact on lives saved, and that it would not have a major impact on lives saved until 2013 – 2014, and that even then many millions would still be uninsured thereafter, given that population growth would be adding to the uninsured as we moved forward in time?
In short, why was there so little interest in the many thousands of lives that in the best of scenarios would be lost if the ACA was passed in preference to Medicare for All? I remember the floor celebration staged by the House Democrats when they finally passed the ACA sellout Rube Goldberg monstrosity, that model of inequality, of different strokes for different folks, (see the listing and the links under Obamacare’s lack of universality) that is the ACA!
At the time, my immediate reaction was that they were dancing at their own funerals, and I immediately wrote a post called “What have they got to run on?” But political considerations aside, in the run-up to the ACA disaster how could the Democrats and the President simply look away from the moral issue of those many thousands of lives they were still leaving on the table in working toward passing the ACA? How can they look away now?
For that matter, how could the Republicans look away and stick resolutely to their “free market” ideology in light of warnings that doing so would kill so many Americans? And how can the Republicans now, after so many years and so many failures of private health insurance in meeting the needs of Americans including those who vote for Republicans, continue to look away from the mortality issue? Weren’t they the party that brought us “death panels”?
In blogging about the AHCA recently, Matt Bruenig recently said:
The difference between Obamacare and AHCA is 24 million uninsured people while the difference between single-payer and Obamacare is 28 million uninsured people. If you favor Obamacare over single-payer or dismiss single-payer as relatively unimportant, then you are a moral monster at least on par with the AHCA proponents you condemn.
While I agree with Matt’s implication that the problem we face in getting Medicare for all passed is politicians in both parties who are “moral monsters”, in the end the failure to pass Medicare for all is not about people being left without insurance. It’s about the deaths, bankruptcies, divorces and other social ills that result from this legislative choice. That’s why the people in Congress, and those who influence them in making these choices, are moral monsters.
Also, health insurance reform is not the only area where our problem may be the development of moral monsters lacking empathy, and the ability to recognize or care about how, when, and in what ways, their seemingly pragmatic choices will impact others.
Our elected leaders are acting in ways that either threaten or even kill us. Just think about their disposition to get into unnecessary wars; to fail to do anything significant to contain change, or to legislate Medicare for all. Just think of their repeated attempts to weaken and privatize the internet, and to weaken and threaten the survival of the American middle class.
They frequently pass legislation or promulgate executive orders that provide the means for wealthy individuals and financial organizations, and big multinational corporations to loot us, extracting our financial wealth, pollute our water supplies, poison our food chain, and liberate our large communications empires from significant regulation by agencies that might protect us from their depredations. They also are working hard to pass legislation that deprives us of our voices and our votes, ending American democracy.
So, getting back to the main question of How Many Americans Will Die Needlessly, Before Congress Finally Passes National Improved Medicare for All (NIMA); only the people and the moral monsters can answer it. For our part, it’s up to us to do the things we need to do get the moral monsters killing us out of office.
Permalink