UHC, Denials, and Death

Posted on by under Big Data, Health Care Series, Palestine

This post has been superseded by a new and improved version written a few days later (Dec. 18, 2024). 

Let me just get straight to the point: I have a real worry about how people are reacting to the UHC killing. It’s not the usual worry that we’re being mean to Brian Thompson. It’s that a lot of what people are saying shows a misunderstanding of how the health insurance denials process works. I work in health care denials management on the provider side, and have in fact dealt with UHC‘s denials reps. I have no sympathy for them or for the insurance industry generally. UHC is the apex predator in an industry of predators. But on the whole, I don’t think it makes sense to say that insurance denials kill people. I’ll grant that excess mortality and morbidity are possible through an insurance denial, but death-through-denial is not the modal case of premature death or even close to it, and it’s a mistake to suggest otherwise.

Setting aside denials for lack of prior authorization, the vast majority of denials are denials for reimbursement after the procedure has been done and the patient has long since been discharged from the hospital. They’re not ex ante denials of service. It is prima facie mysterious how an ex post facto event can affect the mortality outcome of a procedure that is in the past, indeed, where the discharge date is probably months or even years in the past. I don’t claim it’s literally impossible. I can imagine cases where it’s possible. But it is far from usual. On the face of it, “Smith died from an insurance denial” means “Smith had a procedure done, then died because his insurance company failed to reimburse the provider a few months later.” That is not how people usually die in medical contexts. In fact, it’s not transparently obvious how you could die by such a process.

Again, setting aside prior authorization denials, we get substantively adverse mortality outcomes from people who lack insurance, not people who have it. The usual adverse effect in the case of those who have it is bankruptcy, not premature mortality (due to denial of coverage for claim submitted). It is certainly possible for bankruptcy to probabilize mortality, but no one would equate bankruptcy with death. My late wife declared bankruptcy three times. She died only once. 3 – 1 = 2. So that dog doesn’t hunt.

It’s worth remembering that the denial appeals process often takes years to complete. I do reporting for denials (on UHC, no less), and it’s standard in the denials management industry to illustrate denial rates by selecting dates of service two or three years in the past. To kill someone, a denial would have to reach back years into the past, then travel forward in time and kill them. I don’t flatly deny that that can happen, at least in a somewhat extended or metaphorical sense–I recently made a point via such an example myself*–but I deny that it’s the usual case. The correct inference is that the denial appeals process adds needless administrative costs to the health care system, not that it kills people.

Since I myself would grant that there is some elevated mortality risk from denials, you might wonder why I’m making such a big deal of this issue. It’s because I have a deep fear of any line of reasoning that turns anomalous cases into modal ones. That type of reasoning is why we currently face the prospect of the mass deportation of migrants in the United States. Americans have come to believe that if 1 in x migrants is a criminal (where x is a very large number), they’re all criminals. If they’re all criminals, why not deport them all? That is psychopathic reasoning, but if the lesson people learn from the UHC shooting is that anomalous cases of mortality through insurance denials are the typical case, we just face a different form of the same reasoning in a different context. If you can do it for insurance denials, you can do it to migrants. You can do it anywhere.

One irony here is that little attention has been paid to the Change Health hack (under UHC), which affected 100 million people, and was a likely result of our involvement in the Ukraine War. (The hack was done by a Russian gang, likely in retaliation for our involvement in Ukraine.) There is a documented mortality effect from computer hacks in health care settings. The causality is clear: when you knock a health information system offline, doing so affects clinical outcomes. The Change hack happened almost a year ago. How much of a furor has there been over the likely mortality outcomes of the Change hack? Practically none. Yet the causality is a lot clearer in that case, and more obviously attributable to UHC, than the causality arising from denials. A computer hack can kill someone by causing the clinical event that kills them. An ex post facto insurance denial can’t.

Here is another problem. How is it that Americans can be so eager to get Brian Thompson shot, but think that no one can open his mouth about Palestine without having to condemn Hamas? No one is being asked to condemn the Thompson shooting before speaking about insurance denials. The idea is: we’ve all suffered so much from UHC that we’re entitled to our schadenfreude about Thompson’s death. But decades of occupation and blockade don’t seem to yield any comparable reasoning when it comes to October 7. If Americans are entitled to schadenfreude over Thompson, why aren’t Gazans entitled to schadenfreude over killing the inhabitants of Nahal Oz? How is it that Thompson’s killing elicits unapologetic glee, but October 7 elicits a sense of existential dread and horror?

People have come to regard UHC as a murderous organization because (again, prior authorization aside) they’re confusing denials of reimbursement for services rendered with denials of service. “Thompson was killed because he denied service and got people killed.” Mostly not. A blockade can kill. Ethnic cleansing and genocide can kill as well. Prior authorization denials aside, an ex post facto denial for reimbursement mostly can’t kill you. You can’t be killed by someone’s denying payment for the medical procedure you got months ago. No matter: we live a country whose people pay zero attention to the blockades, ethnic cleansing, and genocide that their own government helps initiate, while demanding sympathy for insurance denials, up to and including assassination. These are the civilized people who insist that Palestinians are bloodthirsty savages. They themselves are holy virginal innocents, or so they want us to believe.

Look: either it’s schadenfreude for all, or it’s schadenfreude for none. Once we get to discriminatory schadenfreude, we’ve hit rock bottom. Which may be where we are.

*To be precise, the example in my Sept 29 post said this:

B loses his health insurance, and in the process, goes bankrupt. Having done so, B is unable to pay for crucial health care, and prematurely loses his life. …

For obvious reasons, there are unlikely to be statistics on the frequency of cases exactly like Lost Health Coverage, but both bankruptcy and premature death due to denial of health insurance benefits take place with startling frequency in the United States (emphasis added).

In this example, the patient previously had but then loses his health insurance coverage. The “denial of health insurance” involved is a denial of benefits due to a wholesale loss of coverage (itself due to loss of employment), not a denial of reimbursement for services rendered while being a policy holder. The denial in my Sept 29 post was meant to illustrate the culpability of the employer, not the insurer. So while the example linked bankruptcy to premature mortality (a connection I would hardly deny), it does so in a very different way than defenders of the Thompson assassination are suggesting.

The overarching lesson here is that the term “denial” is ambiguous. We might remedy this by trying to distinguish technical from non-technical senses of “denial,” but it turns out that the phrase “technical denial” is itself ambiguous. There is a technical sense of a “technical denial” (where “technical denial” contrasts with “clinical denial”), and a non-technical sense in which something that’s technically a denial contrasts with something that’s one in the informal, colloquial sense. It’s confusing, to be sure, but the blame for the failure to explain it lies with mainstream media coverage, obsessed with sensational events like the Thompson shooting, but relatively uninterested in clarifying ordinary, workaday facts that would make sense of the broader context.

I write here exclusively in my own name and on my own responsibility, not on behalf of my employer, or any other organization. 

7 thoughts on “UHC, Denials, and Death”

  1. There’s a difference :
    Most migrants aren’t guilty. Brian Thompson is guilty, at a horrendous scale, even without talking about an hypothetical increase in mortality from this actions.
    It’s the same with october 7: most of those who did weren’t guilty, and if they were, they were against at most a few familles.

    Also, the denials cause deaths in other days than increased mortality caused by poverty. What about those who won’t seek treatment, gambling with their health, because they can’t afford to gamble with their money on denials? What about those who seek treatment, but can’t find a doctor or a hospital covered by their insurance who will receive them in a timely maner? And what about those who can’t afford insurance in the first place, who will refuse treatment because UHC directly contributed to the system that made healthcare costs prohibitively expensive?

    All of these factors increase mortality, so he is guilty of a large amount of deaths, even if indirectly. And if not for deaths, he’s at the very least guilty of forcing a significant number of a
    Americans in abject poverty.

    While I don’t support death penalty, I don’t condemn the murder of a serial killer.
    If you want a better comparison, instead of killing a random Israelian, it’s like killing a member of the IDF (military group Known for their cruelty). I’d say schadenfreude is waranted for the latter

    Like

    Reply

    • “instead of killing a random Israelian, it’s like killing a member of the IDF”

      Isn’t nearly every “random Israelian” a conscript member of the IDF at some point?

      Like

      Reply

    • What I was disputing in the post was the widespread belief that you can equate an insurance claim denial with a wrongful killing. Many people have come to believe that the Brian Thompson assassination was justified because UHC has a higher denial rate than comparable insurance companies, and this fact is what implies that Thompson was a serial killer. The belief only makes sense if we assume that every denial, or most denials, kill. This is a gigantic, self-defeating over-simplification. Not only will it not achieve anything of value, but it will eventually be exploited in destructive ways that few people intend.

      Think of denials as falling on a continuum with two extreme ends. At one end, there are prior authorization denials that deny coverage that is contained in a policyholder’s policy. This, I agree, is equivalent to wrongful killing.

      At the other end are claims denials for services that have already been rendered, where the service has fully responded to the patient’s health care needs. Setting aside very unusual circumstances, this is neither a killing nor a wrongful killing.

      In between these extremes are a gigantic number of intermediate cases. Some lie closer to the first extreme, some to the second. And some are in the middle. The farther they are from the first extreme, the less clear it is that they’re wrongful killings.

      Obviously, if the patient doesn’t die, it’s not a wrongful killing. If the patient dies, but the denial is not what caused the death, it’s not a wrongful killing. Even if the patient dies because of the denial, there is a question of whose fault that is. It could be the provider’s fault. Then it’s a wrongful killing to be attributed to them, not the insurer. If the patient dies but was never contractually entitled to the treatment that would have saved her, that’s not a wrongful killing, either. If the patient needs X, but the policy doesn’t cover X, I don’t think you can justifiably accuse the payer of killing the patient.

      I am not saying that the payer is right in all these cases. What I’m saying is that these are not cases of wrongful death. If someone is dying a few yards away from me, and I don’t save him (but could), that is wrong, but it is not a wrongful death. Nor could anyone justifiably kill me over it. In fact, if they tried, I could justifiably kill them (and would).

      Someone could say, fine. Maybe Brian Thompson should simply have been killed over however many prior authorization denials he presided over that wrongfully killed someone. Even I admit that those are wrongful killings, after all.

      The problem, however, is that anyone who thinks this is missing the real implication of their view. If prior authorization denials are wrongful killings, and wrongful killings justify retaliatory killing, then the retaliation for them cannot be limited to Brian Thompson. The target has to be UHC itself. It is not enough to kill the CEO of UHC. The organization itself must be destroyed, along with everyone in it. Brian Thompson is not engaging in prior auth denials all by himself. The whole company is.

      Not just that company. In fact, while UHC’s denial rate is high, its prior authorization rate is low. If a low prior auth rate is good enough to kill the CEO of UHC, it’s good enough to kill everyone in UHC. But if it is good enough to kill all of UHC, it is good enough to kill all of Humana, too. Humana has the highest prior auth rate in the country. Actually, Medicare has a prior auth rule, too, and so does Medicare Advantage. So why not go on a rampage against all of them? Everybody engages in prior auth denials. If guesswork about denials is good enough to kill Brian Thompson, guess work should be good enough to go on a rampage against UHC, Humana, Cigna, Aetna, Medicare Advantage, and CMS, too. Once you go down this road, you’ll find that there’s no shortage of targets. If you want blood, you’ve got it (to quote AC/DC).

      https://www.kff.org/medicare/issue-brief/use-of-prior-authorization-in-medicare-advantage-exceeded-46-million-requests-in-2022/

      “Looking across insurers that are more similar, the low end of the range [for prior auth] was 0.9 requests per enrollee in both Cigna and UnitedHealthcare plans” (p. 9).

      That’s what happens if you limit the killings to prior auth denials. You seem to want to go well beyond that. Do denials cause death through avoidance of treatment? In an indirect way, yes. Do denials cause death when the patient can’t find a provider due to lack of coverage? Yes. But in these cases, the cause of death is not simply the payer. It includes the provider. The provider could after all reduce their charges or write off their expenses, or whatever, and provide life-saving treatment at a reduced cost or none. If they refuse, the provider is at least a co-cause of the patient’s death along with the payer. Should we therefore start shooting hospital CEO’s, too? How about their CFOs? Or their Chief Medical Officers? Or the best-compensated physicians themselves, like surgeons and anesthesiologists? They’re all part of the picture.

      The reason why people die due to lack of coverage is not simply because UHC engages in denials. It’s because Americans don’t want to provide universal coverage. If so, we have another co-cause. Not just payers and providers, but the people themselves. Should we shoot them?

      I understand the rage that induces people to want to shoot Brian Thompson, but it’s literally a dead end. If people want to admit that and do it, fine. Just be clear what you’re saying and why. But what people want to do is equate denials with wrongful deaths and then valorize the killing of one CEO while refusing to see that the logic of their position requires a lot more killing than that. That’s incoherent. If someone can come out and admit the logic of this view, and then say, “I’m still for it,” they’ll get my grudging respect, for honesty if for nothing else. But otherwise, it just looks like the circus has come to town.

      Like

      Reply

  2. Pingback: UHC, Denials, and Wrongful Death Revisited | Policy of Truth

Leave a comment